This electronic thesis or dissertation has been downloaded from the King’s Research Portal at https://kclpure.kcl.ac.uk/portal/ Take down policy If you believe that this document breaches copyright please contact [email protected]providing details, and we will remove access to the work immediately and investigate your claim. END USER LICENCE AGREEMENT Unless another licence is stated on the immediately following page this work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ You are free to copy, distribute and transmit the work Under the following conditions: Attribution: You must attribute the work in the manner specified by the author (but not in any way that suggests that they endorse you or your use of the work). Non Commercial: You may not use this work for commercial purposes. No Derivative Works - You may not alter, transform, or build upon this work. Any of these conditions can be waived if you receive permission from the author. Your fair dealings and other rights are in no way affected by the above. The copyright of this thesis rests with the author and no quotation from it or information derived from it may be published without proper acknowledgement. Life after coca the experience of informal institutions in former coca leaf farming communities in Colombia Ostler, Sophia Awarding institution: King's College London Download date: 22. Jun. 2022
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
This electronic thesis or dissertation has been
downloaded from the King’s Research Portal at
https://kclpure.kcl.ac.uk/portal/
Take down policy
If you believe that this document breaches copyright please contact [email protected] providing
details, and we will remove access to the work immediately and investigate your claim.
END USER LICENCE AGREEMENT
Unless another licence is stated on the immediately following page this work is licensed
under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
You are free to copy, distribute and transmit the work
Under the following conditions:
Attribution: You must attribute the work in the manner specified by the author (but not in anyway that suggests that they endorse you or your use of the work).
Non Commercial: You may not use this work for commercial purposes.
No Derivative Works - You may not alter, transform, or build upon this work.
Any of these conditions can be waived if you receive permission from the author. Your fair dealings and
other rights are in no way affected by the above.
The copyright of this thesis rests with the author and no quotation from it or information derived from it
may be published without proper acknowledgement.
Life after cocathe experience of informal institutions in former coca leaf farming communities inColombia
Ostler, Sophia
Awarding institution:King's College London
Download date: 22. Jun. 2022
Life After Coca The experience of informal institutions in former coca leaf farming communities in
Colombia
By Sophia Ostler
PhD Candidate, Department of Political Economy
King’s College London
2
I would like to dedicate this thesis to my mother for giving me the motivation to follow my dream.
To my father for giving me the support to hold onto it.
And to my husband for giving me the strength and courage to see it through until the end.
3
Table of Contents
Table of Contents .................................................................................................................................................. 3
Method and Data ................................................................................................................................................. 60
1.3 Symbolic value of land titles ....................................................................................................................... 104
Brief background on land and politics in Colombia ....................................................................... 104
Symbolism of land ownership recognized by the state ................................................................... 108
But how is the symbolic value of land titles linked to AD? ........................................................ 109
Appeal of land titling to farmers in AD programmes ................................................................. 110
5
1.4 Literature Review ........................................................................................................................................ 115
Theoretical mechanisms by which land titles support AD efforts .................................................. 115
I. Transferability of land ............................................................................................................ 115
II. Increased value of land ........................................................................................................... 118
a) Increased resource allocation ............................................................................................. 118
b) Long-term commitment ..................................................................................................... 119
III. Increased agricultural innovation ....................................................................................... 120
IV. Access to credit .................................................................................................................. 121
Assumption of increased land tenure security ................................................................................. 122
Land titles vs land security .......................................................................................................... 122
Land security vs land investment ................................................................................................ 124
Pre-existence of de facto property rights to land ......................................................................... 128
Working hypothesis ......................................................................................................................... 134
Hypothesis A: Farmers with land titles have access to increased amounts of credit ...................... 139
Hypothesis B: Farmers with land titles choose to access increased levels of credit ....................... 141
Hypothesis C: Farmers’ increased access to credit results in more profitable legal agricultural
production ........................................................................................................................................ 145
Cases that confirmed the hypothesis ........................................................................................... 145
Cases that did not support the hypothesis ................................................................................... 147
Hypothesis D: Farmers’ increased productivity in legal farming reduces their dependence on illegal
drug production ............................................................................................................................... 158
No increase in land productivity ................................................................................................. 158
Return to illegal drug production ................................................................................................ 161
6
Possible counterproductive effect of land titles .......................................................................... 166
Summary of findings ....................................................................................................................... 169
2.3 Literature Review ........................................................................................................................................ 187
Theory of extra-legal governance ................................................................................................... 187
Brief historical overview of extra-legal groups and the illegal drug trade in Colombia, 1990s –
3.3 Literature Review ........................................................................................................................................ 261
About rules and norms .................................................................................................................... 261
Gender and property rights to land .................................................................................................. 264
Gender and education ...................................................................................................................... 266
8
Gender norms and economic behaviour .......................................................................................... 267
I began my research into the illegal drug trade, like many, from a place of perplexity and concern. For
several years, coca leaf production in Colombia has been increasing. It grew by 3.7 times between
2013 and 2017 (UNODC 2019a, 21). To counter this, the Colombian state is spending about $270
million USD each year on programmes to encourage farmers to voluntarily stop growing illegal drug
crops (Cruz Olivera 2019; Cuesta 2019). This is just a small fraction of what it spends annually on
supply reduction strategies including forced aerial and manual eradication campaigns and their law
enforcement component, which has been estimated as $2.1 billion USD (Cruz Olivera 2019).1 The
latter is estimated to amount to about 2.6% of Colombia’s annual national budget (Estrada 2019). I
was perplexed about the vast resources that are spent on fighting an industry that clearly has
everything going for it: the supply, the demand, and the institutions that connect these two. And I was
concerned about the huge number of victims who are caught up in this vast market, between the
threats and promises of law enforcement and of organized crime.
It was the time in the lead-up to the United Nations General Assembly’s Special Session on the World
Drug Problem, which was due to happen in 2016. I sat, like a fly on the wall, taking minutes in
meetings between parliamentarians that had a common interest in drug policy. They were from the
UK and from various Latin American countries, and they went around in circles, discussing issues
with no clear answers on how to resolve them. They passionately highlighted the need for, as well as
the harm caused by, policies seeking to reduce the supply of illegal drug crops, aiming to interdict the
trafficking of illegal drugs, and attempting to punish their consumers. There was no clear quick fix for
1 Calculating a precise figure of what the Colombian government spends on the war against drugs is complex for several reasons. There are significant discrepancies in what the Ministry of Defense and the National Council for Economic and Social Policy report as the cost of aerial eradication of one hectare of coca leaf, the former estimating it as $540 USD (only accounting for the cost of army and police support for eradication) and the latter as $19,400 USD (Cruz Olivera 2019). Other estimates calculate the cost of spraying one hectare of coca leaf with glyphosate in 2016, including the cost of airplanes, herbicide, protection, etc. as $2,400 USD, and given the effectiveness rate stands at 4.2%, the cost of actually eradicating one hectare of coca leaf with aerial spraying was $57,150 USD (Daniel Mejia 2016, 9). There is also the separation of budgets for security and for anti-drug policies, therefore the latter will not reflect increases in the former. For example, Colombia’s defence expenditure increased from 3.6% of its GDP in 2003 to 6% in 2006 due to the vast increase in security forced over the four years (Keefer, Loayza, and Soares 2010, 13). Currently, the most recent figure for what has been spent on the war on drugs, including the military component of the budget, is for the period between 2000 and 2008, which was when Plan Colombia, a large anti-drug law enforcement strategy, heavily subsidized by the USA, was being implemented. During this period, each year the Colombian government spent an average sum of $812 million USD, and the USA spent $472 USD. Together, both countries spent about $1.2 billion USD per year, about 1% of Colombia’s annual GDP (Daniel Mejia 2010; Keefer, Loayza, and Soares 2010; Daniel Mejia and Restrepo 2014; Gaviria, Mejía, and Weiskopf 2017) There is a more recent study for what the Colombian government has spent on the war on drugs between 2008 and 2015, estimating it at about $270,000,000 USD annually. This was about 0.5% of the national budget for those years (D. Rico et al. 2018, 9). However this figure excludes the military and security costs, and Colombia’s defence budget is the highest in the region, estimated to be over $10,000 million USD and accounting for 3.1% of its GDP (Aristizabal Bedoya 2019).
11
the illegal drug industry, but decriminalization was often brought up like a potential solution. One of
the strongest arguments against this innovative drug policy, was simply that we do not know enough
about what the effect of moving away from prohibition would be. And to many, it just seemed too
risky to experiment with it. There were defiant countries like Uruguay, and Portugal and some states
in the USA that were beginning to take small leaps of faith and piloting this approach in a very limited
way. However, the discussions at parliamentary level were still a long way off from addressing the
precarious situation for those at the very bottom of the pecking order of the cocaine business: coca
leaf farmers.
I wanted to know what a less punitive approach might mean for them. I thought that if I could hear
their stories first-hand, I might help build a clearer picture of how effective non-punitive drug policies
are, and most importantly why they are so. Overcoming our ignorance on the role that informal
institutions play in making coca leaf production attractive to farmers, is particularly significant to
learn more about how voluntary illegal drug crop substitution efforts, hereby referred to as Alternative
Development (AD), can last. AD policy is often devised in a top-down approach and rarely consults
the perspective of people on the receiving end of these programmes. I wanted my research to capture
people’s experiences of AD to complement our understanding of how it works at the grassroots level.
I knew that illegal drug farmers had strong reasons to farm coca leaf. Most rural farmers in Latin
America are poor and in financial need of this cash crop. And yet I most often heard politicians refer
to organized crime groups as the drivers of farmers’ motivation to participate in illegal drug
production. They were cast as responsible for keeping this illegal industry going and as culprits for
forcing farmers to continue producing illegal crops. The sentiment among those proposing a punitive
approach to drug policy, was that if only these criminals could be weeded out, the illegal drug
industry would finally wither. But this contradicted a basic premise in economics: where there is a
market, institutions will develop to keep it going. It seemed that no matter how many individuals were
forcibly removed as actors in this illegal trade, it would continue working, unless there was
institutional change at a deeper level. I believed that if I could draw lessons from places where this
institutional change had already happened, I might be able to help shed some light on why farmers
12
choose to switch from illegal crops to legal alternatives. More importantly, it might give policymakers
a better understanding of how non-punitive drug policies can be made more effective to suppress the
supply of illegal drug crops.
So, I began my quest and enrolled for a PhD in Political Economy. It was then that I understood the
important role that land played in influencing illegal drug farmers’ behaviour. After writing a long
literature review of the illegal drug trade and organized crime in Colombia during the twentieth
century, I found there are many ways in which the success of the cocaine industry can be explained.
There is the historical system of patronage, that has perpetuated inequality in how people’s homes as
well as their source of livelihoods are distributed. There is the vast number of people who have been
forcibly displaced from their land. This has created a ready supply of farmers willing to take the risk
of growing coca leaf if it helps them improve their poor livelihoods. And there is the centrally
managed state land agency, that has been tugged by diverging political interests over the years and has
lacked the capacity to enforce its rules. All this stems from the idea of inequality in land ownership
and has depleted farmers’ trust in the state as their protector and granted legitimacy to armed groups
that have sought to represent their interests. This has given steam to illegal drug farming over the
years.
I wanted to understand why farmers so often used their land for illegal crops instead of for legal
alternatives. I had hoped to explain this from the angle of institutions using Ostrom’s concept of the
tragedy of the commons as my analytical framework. I thought perhaps that the system of property
rights was encouraging people’s self-interested behaviour, thereby spoiling the legal economy, or
enhancing the illegal one. To develop this idea, I thought long and hard about how to classify the land
in which coca leaf is grown in Colombia. Is it a common pool resource i.e., rivalrous, and non-
excludable? It is a truly rivalrous good, but it is not entirely non-excludable since there has been a lot
of land grabbing and informal colonizing, so not really. Is it a private good i.e., rivalrous, and
excludable? Some land that is used to grow coca leaf is rivalrous and excludable in Colombia, but
most of it is informally owned, and is not excludable if there are armed groups that can claim
territorial control. Is it a club good i.e., non-rivalrous and excludable? As just mentioned, very little of
13
it is excludable and so it cannot be cast as a club good. Is it a public good i.e., non-rivalrous and non-
excludable? Perhaps it is non-excludable, but it is certainly highly rivalrous and therefore not a public
good. I found myself in a conceptual labyrinth, unable to classify land as any one of these goods.
At this point, I thought that I should leave my theoretical ruminations to one side and proceed with my
empirical contribution, hoping it would shed light on theory at a later stage. I wanted to find out
whether there really is a link between inequality in land ownership and illegal drug production in
Colombia, as is so often alluded to in the literature. And if so, how changing the institutions of land
ownership peacefully, might help disincentivise farmers from growing illegal crops. As I gathered
numerical data on land ownership, I realized that this too, was not a straightforward endeavour. On
the one hand, the data I was finally able to get hold of was limited. I acquired data on the annual
number of registered land plots and total number of land plots for a period of 15 years. But I could not
find data on the size of land plots, nor on the number of owners of multiple land plots, as this is not
public information. So, this data could only tell us so much. On the other hand, working out land
ownership distribution is complex, even for trained statisticians. With the data that was available to
me, I would only be able to work out the average size of land plots and the proportion of formal
landowners per municipality. However, even this calculation, which would have required
considerable training in coding and quantitative research skills on my part, would still not have given
a satisfactory answer to my question. After all, I wanted to know how addressing inequality in land
ownership might peacefully shift people away from illegal drug farming.
It seemed more appropriate to understand the issue from people’s lived experiences. How did those at
the base of the illegal drug trade believe their land ownership status had influenced their land use
behaviour? This approach would not provide the all-encompassing answer for how non-punitive drug
policies can be more effective in reducing illegal drug farming in Colombia than hard-line law
enforcement, as had been my original goal. However, it would at least convey a more accurate sense
of how, and why, efforts to redress inequality in land ownership may influence people to cease
farming illegal drug crops.
14
As I mapped out how I could make this happen with the time and resources that were available to me,
I realized that it would be more practical to carry out research in communities that had ceased to grow
illegal crops voluntarily and where the issue of land ownership inequality had been addressed. Besides
making it more feasible to interview farmers safely, I believed these places would provide greater
insight into the mechanism by which non-punitive measures may or may not change illegal drug
farmers’ behaviour and why. Surely this would be of value for drawing lessons on drug policy more
broadly.
I began my quest by contacting various policymakers in Colombia. I wanted to know how they
understood inequality in land ownership to be related to the illegal drug trade, given that this was
perceived to be the root cause of drug farming in much of the literature. I also hoped they would
facilitate some entry points for my fieldwork research where AD had been deemed successful and
where the institutions of land ownership had been transformed in some way.
After a month interviewing experts in Bogota about how land related to drugs and violence in recent
Colombian history, if at all, it was clear that most of them believed these issues were related. Their
narratives centred on how farmers were drawn to illegal drug farming because they had lacked formal
land ownership. They believed the informal status of farmers’ property rights to land meant they did
not trust the state to respect their land ownership and so were more reliant on armed groups to protect
them from being pushed from their land. In addition, their lack of land titles meant they could not
access loans to invest in alternative legal uses of their land and this was why they were lagging in
legal agriculture. By this stage, it had only been a year since the Peace Agreement between the
government and FARC, and land titling was cast as the most obvious solution to this association. It
was a timely opportunity to find out whether these policymakers’ perceptions were validated by the
farmers that were being offered land titles to stop growing coca leaf.
I came back to London to analyse the thirty odd interviews and to think about how theories on
property rights to land could help explain illegal land use behaviour. As I read over the transcripts, I
identified another explanation that policymakers often assigned to the success of non-punitive drug
policies which echoed what I had anecdotally heard parliamentarians discuss in my former job. Many
15
believed farmers were being forced to grow coca leaf by armed groups, so the mere offer of state
governance would be enough to reassure them they could safely stop farming illegal drug crops. I was
curious to learn more about the role that armed groups had played in influencing farmers’ choice of
land use. If the market emerges spontaneously, then forcing people to grow a crop against their will,
seemed just as pointless as forcing them to stop farming it, and this applied to armed groups as well as
the state. From what I had read about extra-legal governance, it seemed to make more sense that
armed groups had been complicit both in farmers’ decision to farm, and to stop farming, illegal drug
crops. I knew that that these were questions I could ask the same people I would be interviewing
about changes in land ownership and their decision to stop farming coca leaf. But as the role of armed
groups was a more sensitive topic, I decided I would focus the interviews on land ownership and if
my interviewees brought up something relevant to violence and armed groups, I would take the
conversation in that direction to learn more.
Three months later, after a 10-hour flight to Bogota, a 1.5-hour flight to Florencia, and a jumpy three-
hour ride in a chiva (a rustic bus used in rural Colombia) on untarmacked roads, I found myself in the
small community of Santiago de la Selva. I arrived on a hot and sticky day, and as I walked down the
dusty path, people observed me with interest. I bought a little bag of water from the shop and
quenched my thirst with the taste of plastic and smoky rust. It was a Monday morning, but this did not
seem to matter to the inebriated men sitting on the veranda of a bar. Reggaeton music pumped loudly
from the bar’s loudspeakers, competing with the loud evangelical chants that were being blasted from
another megaphone somewhere. I made my way to my temporary abode and reconciled with the fact
that I would be sleeping in a room with no glass on the window, and a dysfunctional bathroom, for a
week. I met my contact who took me for lunch at a local woman’s house. It smelled of urine and the
plastic tables were covered in flies. The woman watched me with a tired and expressionless face as I
ate the bits of hard meat, chicken skin and rice with stones in it. I washed it down with a mug full of
sugar water mixed with a faint hint of coffee. I was homesick but there was no phone signal, except
for under one very specific tree where there was just enough signal to send and receive text messages.
16
My contact suggested that I interview people in the village instead of in their homes because there
was a parasite in some farms that itches and gets under your skin and genitals and is hard to get rid of.
I accepted his advice and began to interview people who lived locally in the village that same
afternoon. Many people talked about the extreme violence they had witnessed in this village only a
few years ago by armed groups and pointed to the bullet holes that could still be seen on some doors.
Some mentioned that kidnapping was still happening in neighbouring communities. I felt uneasy and
did not know if I would last until the end of the week there. Unlike my entry points in other places,
my contact here was unaffiliated to the state. His lack of communications and transport equipment
and the absence of organizational infrastructure gave me a lesser sense of personal security. I spent
the night hoping no one would walk into my room through the window frame and thinking up excuses
that would help me cut my visit short.
Daylight came, and with it a slight sense of relief. As I breakfasted at the same woman’s place, I
talked to her bubbly teenage daughter-in-law, who was helping her in the kitchen. She had left school
early to marry and, while her husband worked in other villages, she helped his mother with
housework and to cook meals mainly for male day labourers who could not go home for lunch. She
spoke of how little work there was since the days of the coca leaf bonanza. Her motivation for
marrying contrasted with some theories I’d heard mentioned by some men the day before. Apparently,
girls in this community have babies and settle down young because the hormones that are injected
into chicken these days “accelerates” their sex drive.
My contact found a young woman who was willing and able to come and clean my bathroom. When
she finished her job, she shily hung around and asked me questions. Suddenly she burst into tears and
told me about her problems. She was frustrated she hadn’t finished school yet despite being 24 years
old. As a single mother of an eight and a three-year old, she could not afford to pay the fees to get her
diploma. She could not leave her children and go to find work in town and what farming jobs were
locally available did not fit easily around her childcare responsibilities. Her dream was to work in a
beauty salon in town but even for this she needed a school diploma. She said she was tempted to take
up the offer an old man had made her: to “take care of her” as her partner.
17
One evening, a schoolteacher invited me to meet some final year students at the girls’ school where
she worked. When I arrived, there were about twenty teenage girls and two teachers, thirsty for any
bit of knowledge or advice I could share with them. I asked the girls about their plans. One of them
said she had worked hard and got the top grades in her class, but these were still not high enough to
pursue her goal of studying nursing in town. She wasn’t sure what she would do now. It struck me
how it was completely beyond her control that the only school she had access to had been unable to
help her reach the same academic level playing field as the rest of the country.
These encounters began to sensitize me to the situation that women find themselves in, in places
where the local coca leaf economy has faded away. It was increasingly evident that gender norms
were another institution that deserved my careful consideration, if I wanted to explain the attraction of
coca leaf farming to women. I decided that I would develop this theme during the rest of my
fieldwork in addition to my other research questions. Furthering our understanding about how AD
policy trickles down on women’s everyday life, will help us get a fuller picture of the socio-economic
implications for communities that stop growing coca leaf crops. This is vital information if AD
policymakers expect men and women to abandon the illegal drug trade with equal enthusiasm.
As the week progressed, people’s faces and names became more familiar, and I felt safer each day.
My trust in those around me grew, as did the community’s trust in me. It was a time when former
armed groups were realigning themselves either with the state, following the Peace Agreement, or as
dissidents of such. It was a brief interlude in communities like this one, when the state’s territorial
control was uncontested, and this created a peaceful setting to conduct research. This was a unique
opportunity to be an emissary for these former coca leaf farmers and analyse the effectiveness of a
non-punitive drug policy from their perspective.
The complexity of the decision to opt out of the illegal drug trade, and our failure to analyse it fully,
may be important in explaining why coca leaf farming is proving resilient, despite governments’
efforts to stamp it out. The fact is there are still gaps in what we know about what determines people’s
behaviour at the bottom of the illegal drug supply chain in Colombia. If AD is to be made more
effective in helping farmers to opt out of the illegal drug market for good, we also need to know
18
whether these institutions influence people’s behaviours and decision-making in the way that many
policymakers believe they do. My aim with this research has been to help fill some of these gaps by
answering the following questions.
First, does the legal recognition of land ownership entice farmers to prefer legal agriculture over
illegal crops? There has been little research on the influence of land titling on farmers’ decisions to
stop or continue working in the illegal drug trade.
Second, do the armed groups that protect the illegal drug trade influence whether communities opt in
or out of coca leaf farming? Not much is known about the role that extra-legal groups play, in
persuading farmers to desist from growing coca leaf crops in places where it once thrived.
And third, we know that women play an important role in small-scale coca leaf farming (Parada
Hernandez and Marin Jaramillo 2019; UNODC 2018b), and that they tend to suffer the consequences
of an economic contraction more harshly and for longer than men (Oriana Bandiera et al. 2018). So,
what bearing does the substitution of illegal drug crops have on women’s economic opportunities?
Background
Coca leaf production in Colombia
In the last 60 years the coca leaf has evolved from being a wild medicinal herb, traditionally used by
indigenous communities in the Andean region to reduce altitude sickness, to a major export crop. In
2012 its production was worth approximately $189 million USD, accounting for about 0.5 % of
Colombia’s GDP (Gaviria, Mejía, and Weiskopf 2017, 17). Once it has been processed into cocaine
and trafficked, its value in the same year was been estimated to be approximately $5 billion USD per
year, contributing 1.2% to Colombia’s GDP (Gaviria, Mejía, and Weiskopf 2017, 9).
Colombia’s illegal drug trade began with marijuana farming and trafficking in the 1930s, mainly
along the Atlantic coast. In the mid 1960’s it expanded to other places in Colombia in response to
increased US consumer demand and to eradication programmes in other supplying countries, mainly
Mexico and Jamaica. Then the USA stepped up its own marijuana production in the 1970s, so demand
for Colombian marijuana began to wane. This is when dealers began to trade cocaine instead and
19
found that coca leaf is much more profitable than marijuana due to its higher value to weight and
volume ratio. Wealthy drug traffickers began to build cocaine laboratories, and to export it with the
help of distribution networks made up of Colombian immigrants in the USA (Thoumi 2009, 51–115).
They bought large amounts of land to launder their money and paid for armed groups to protect the
drug producing areas and trafficking routes from law enforcement.
Today Colombia is the world’s top coca leaf producer and over the past five years, the land area used
for coca leaf crops has greatly increased, surpassing its neighbouring competitors: Peru and Bolivia
(see Figure 1). In 2017, Colombia was estimated to have produced 171,000 hectares (ha) of coca leaf,
an area which is the equivalent of 410,400 football pitches. And the leaf yield harvested by coca crops
increased by 33% from 2012 to 2017 (UNODC 2018a). With this outstanding performance, in 2018,
Colombia met 70% of the global demand for cocaine (Economist 2019).
Figure 1 Land area with coca leaf crops (ha) from 2001 - 2017 (source: UNODC Country Reports)
State’s response to illegal drug farming
And yet, unless it is a limited amount for personal cultural purposes and on an indigenous reserve,
farming coca leaf is illegal in Colombia. The 1961 UN Single Convention on Narcotic Drugs listed
coca leaf as “a substance liable to abuse” and the 1988 UN Convention Against Illicit Traffic in
Narcotic Drugs and Psychotropic Substances required that its farming be made a criminal offence.
Ever since, the Colombia’s state’s drug policy has been resolute to reduce the supply of coca-leaf and
020000400006000080000
100000120000140000160000180000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Bolivia Peru Colombia
20
to interdict the supply-chain of cocaine. It has sought to supress coca leaf farming in three different
ways: with glyphosate aerial spraying campaigns to destroy coca leaf plantations; by sending officials,
protected by the army, to manually uproot coca leaf crops; and by offering subsidies and positive
incentives to convince some farmers to give up on coca leaf farming voluntarily, in what is referred to
as Alternative Development (AD).
But the state’s attempts to stamp out coca leaf production have largely been frustrated. Fresh coca leaf
crops have quickly resurfaced after aerial eradication campaigns. Farmers have confronted officials
who have tried to manually pull out the crops or they have replanted them in better hidden places.
They have also introduced varieties of coca crops that give higher yields of cocaine. Hence the effect
of prohibition has largely been the displacement of coca crop farming into new areas (Rincon-Ruiz et
al. 2016). Figure 2 shows the Colombian government and UNODC data accounting for the density of
coca leaf production over a period of ten years, from 2006 to 2016 (UNODC 2017b, 25; 2007, 9).
These maps highlight not only the intensification of coca leaf production, but its spatial displacement
(or diffusion) towards Nariño and the border with Ecuador (bottom left corner) and towards Norte de
Santander (top right corner). This trend depicts that there are clear issues with the different supply
reduction strategies. My thesis focuses on how AD can be improved as a supply reduction policy.
21
Figure 2 Coca leaf cultivation density in Colombia in 2006 (left panel) and in 2016 (right panel)
The effectiveness of AD is thought to be ambiguous by some experts (Gaviria, Mejía, and Weiskopf
2017; Daniel Mejia 2016). Nevertheless, the state continues to invest in farmers’ local communities,
in the hope that the voluntary substitution of their illegal drug crops with legal alternatives will
follow. Next, I will elaborate on the general rationale for AD and what we know so far about why its
effect is, at best, heterogeneous. I will then set out why and how I will study some specific informal
institutions that may be linked to its effectiveness.
About Alternative Development
Fundamentally, the aim of Alternative Development (AD) is “to create the economic and social
environment in which households can attain an acceptable standard of living, without the need for
drug crops cultivation” (Mansfield 1999). In other words, AD seeks to encourage communities to
abandon illegal drug farming as their means of livelihood. And the overall strategy is to make
farmers’ new forms of land use more productive, and the community’s trade opportunities in the legal
22
realm more profitable. The creation of alternative legal employment opportunities for illegal drug
farmers is attempted in a range of ways. In the short run, the state and aid agencies subsidize capacity-
building projects for farmers and donate resources to help develop other ways to use their land more
productively. Recently, land titles have been a new AD tool that is gaining traction in many illegal
drug crop producing countries. The logic of legalizing farmers’ land ownership is that it gives farmers
greater financial certainty and facilitates their long-term investment in legal alternatives to drug crops.
Whether this is the experience of farmers who have stopped growing illegal drug crops, is the subject
that I shall discuss in the first part of this thesis.
In the longer run, AD is pursued by addressing the deeper issues of poverty and marginalization that
create greater dependence on illegal drug crop farming. The state and aid agencies thus invest in these
farmers’ communities’ local infrastructure and their access to public services. This includes roads that
connect them to motorways to facilitate the transport of products; power and water to provide
refrigeration which facilitates the storage of products; and access to health and education services to
improve the skillset and efficiency of workers. These efforts differ from more general development
strategies in that they are tailored to meet the needs of small-scale farming communities, and their
success is mainly measured in terms of drug supply reduction (UNODC 2015).
One advantage of AD, over punitive ways of supressing drug supplies, is that it does not have the
same detrimental effect that forced drug crop eradication has on communities. This is because there is
less violent confrontation involved between the state and the armed groups governing the illegal drug
trade, or even with the farmers themselves. Besides causing less physical harm to farmers, it allows
the state to retain greater legitimacy and support at the local level.
However, the non-use of violence does not mean that AD is exempt from having any negative socio-
economic effects on farming communities. People may seek to improve their safety by substituting
their illegal crops at home and thereby reduce their exposure to violence, but this still comes at the
risk of reducing their long-term food security and compromising their livelihoods. And yet, we know
very little about this. Any change in livelihood where a big pay cut is involved, is bound to put people
at risk of not being able to find sufficient work to pay their bills. It is inevitable that this will have a
23
knock-on effect on farmers’ wellbeing. We know even less about how women in these communities
fare compared to men, after they stop growing coca leaf, even though women tend to be in the less
well paid and less secure jobs and perform more domestic work.
The reality is that farmers who have benefited from AD may continue working in the illegal trade,
either elsewhere or at a later stage, once they feel they can do it again safely. This happens when AD
efforts inject money in areas of extreme poverty but fail to resolve their issues around access to
agricultural markets and local public infrastructure, so end up facilitating further coca production
(Dion & Russler 2008). In the worst-case scenario, AD may provide farmers with perverse incentives,
and thus improve the local illegal economy. For example, farmers who do not receive AD subsidies
may feel neglected by the state and may start growing illegal crops just to qualify for them, or
increase their support for the armed groups that promise them a better future (UNODC 2015).
However, it is possible that AD can succeed for reasons that are external to the AD intervention itself.
This is another angle to AD effectiveness which I shall cover in the second part of my thesis. Changes
in a community’s governance scenario may create the circumstances that permit AD to succeed. This
helps to raise questions about the degree to which the design of AD programmes is responsible, or
not, for the reduction of illegal drug farming. Moreover, as the third part of my thesis will show,
understanding more about the socioeconomic effects of AD on women may help explain another
contributing factor for why it does not always result in reduced illegal crop farming.
Before discussing three institutions that I believe can help us to understand why AD succeeds or fails
to be effective, it is useful to consider the greater context of how this non-punitive approach to drug
supply reduction originated and continues to be pursued in Colombia despite its limitations.
Origins of AD
AD evolved from a simple pay-out system for opium poppy farmers in 1960s Thailand, to a hands-on
strategy to encourage alternative legal farming in 1970s South East Asia (Renard 2001). One of its
flaws is that it can create perverse incentives for farmers to initially accept the state subsidies but then
continue growing illegal crops elsewhere or return to farming illegal crops afterwards (Davalos et al.
2011; Thoumi 2012; Quimbayo-Ruiz 2008). To address this issue, AD programmes began to adopt an
24
intensive microeconomic approach in the 1980s. This included explicit investment in local
infrastructure state services and market support measures. The 1981 – 1998 Thai-German Highland
Development Programme is an example of this approach and is the star success story of AD (Renard
2001).
AD efforts in Thailand lasted for three decades, effectively transforming it from a large opium
producer in the 1970s, to an opium-free state as of this date. The success of AD in this case has been
attributed to the fact that the land where opium poppy was grown is fertile and able to produce other
profitable crops. And this was supported by a fast-growing economy and the fast development of
infrastructure. Grassroots support for AD came from state-run fear campaigns that opium
consumption was increasing in local communities, and from people’s general respect for, and
adherence to, the King’s authority (Renard 2001). However, the supply of opium poppy has since
shifted to Myanmar. Moreover, Thailand has become the main transit country for amphetamines. This
supports the theory that, like other forms of drug prohibition, AD is effective in displacing illegal crop
production elsewhere, but not necessarily in reducing net crop production, or its demand (Thoumi
2012).
AD efforts in Colombia
AD emerged in Colombia in 1985, at a time when the forced eradication of illegal drug crops was
creating huge social repercussions (UNODC 2015). Illegal drug farming communities, which were
devoid of any alternative sources of income, developed ideological opposition to the state. In addition,
punitive forms of drug law enforcement were helping armed groups gain greater political clout in
these communities. Thus, national, and international financing for AD was a way to mitigate some of
the grievance other forms of drug supply reduction were causing.
From 1998 to 2013, Colombia received 33% (the largest share) of AD assistance worldwide (UNODC
2015). The funding allocated to AD has been increased in Colombia whenever the funding for other
forced eradication initiatives has increased. For example, at the height of the military and aerial
spraying campaigns rolled out in Colombia in a programme called Plan Colombia in the 2000s,
USAID funding for AD was increased via two programmes. The first was called ADAM, which
25
started in 2001 and promoted agricultural and social infrastructure development at the municipality
level. The second programme started in 2005 and was called MIDAS and focused on building the
capacity of local governments to lower microfinance fees and assist private lenders so as to boost
investment in small rural businesses (Benkert 2008, 49). In 2003 the Colombian government also
began to increase its own AD efforts with two programmes. One was called the Forest Ranger
Families Programme (PFGB) which aimed to work with illegal crop farming families to reduce
deforestation and to substitute their crops to legal produce. It used bimonthly cash transfers as
conditional financial incentives and provided technical assistance, this time with more support for
local public infrastructure like roads. The other AD programme was called the Productive Projects
Programme (PPP). It sought to increase productivity and employment opportunities of families’
dependent on or at risk of joining illegal drug production.
Most recently, the state launched the National Programme to Substitute Illicit Crops (PNIS) in 2016.
It was one of the outcomes of the Peace Agreement that the state negotiated with FARC, the largest
extra-legal group that had governed coca leaf production for decades, especially in South West
Colombia. The deal between the state and FARC was that in exchange for lifting illegal drug farmers’
penal sentences and providing them with subsidies, small-scale coca leaf farmers, pickers and
traffickers would voluntarily abandon the illegal trade.
AD effectiveness in Colombia
From what can be measured, AD in Colombia has had very mixed results (UNODC 2015). Between
2006 and 2015 about a third (32%) of the total area where there were coca leaf crops, were subjected
to AD programmes. In 40% of this area where AD was introduced, coca leaf was abandoned for three
years. However, in about the same proportion (39%) of this area where AD was introduced, coca leaf
continued to be farmed on and off over a period of up to ten years. In 17% of the area where AD was
introduced, coca leaf never ceased to be farmed at any point. And in 4% of the area where AD was
introduced, coca leaf farming stopped briefly but resumed after three years (UNODC 2016a, 72–74).
Overall, in 2019 it was estimated that of the coca crops that were eradicated nationwide, between 50%
and 67% were replanted (Jaramillo 2019). In Puerto Asis in Putumayo, one of the municipalities
26
included in this study, cocaine seizures increased by 170% between 2017 and 2018, despite the state’s
targeted AD efforts since 2016 with the PNIS programme (Acosta 2019). The success rate of AD is
therefore not higher than that of other eradication efforts.
Measuring the effectiveness of AD is difficult for various reasons. Firstly, the effects of investment in
local development are only evident in the long term. Secondly, the effects of simultaneous forced
eradication and interdiction efforts, among other factors, makes it difficult to single out the overall
effect of AD on coca production in Colombia. AD is often entangled with other eradication efforts
such as aerial spraying or forced manual eradication efforts. This makes it difficult to attribute trends
in illegal crop farming uniquely to AD programmes, if at all. Moreover, the levels of illegal crop
production may be affected by exogenous factors that are unrelated to AD, such as an increase in
value of coca leaf, unfavourable climate conditions for growing other legal crops i.e., drought, and the
oil price crisis and its knock-on effect on unemployment in rural areas. And as the factors that lead to
illegal crop farming are context-specific, there is no set of universal indicators for AD success to be
measured in the same way across all programmes. This means that AD is tailor-made to each context
and it is not possible to make an accurate cross-comparison of AD programme success, or even get an
accurate estimate of AD effectiveness (Thoumi 2012, 991).
It seems, therefore, that the most quantifiable way of measuring AD effectiveness is therefore net
reduction in illegal crop production. However, even measuring the number of illegal crops in
Colombia presents significant methodological issues (Thoumi 2012, 991). The data for illegal crops in
Colombia comes from two reports which measure slightly different things. One set of data measures
coca leaf grown on 31st December each year while the other estimates the annual average (Velez
2016). The former is published by the Colombian Drug Observatory (ODC) and the United Nations
Office on Drugs and Crime (UNODC) through its Illicit Crop Monitoring System (known as SIMCI)
in June. This goes back to the year 2000 and is widely considered the most accurate figure. It uses
high resolution aerial photography to measure the spread and intensity of crops and illegal activities.
It supplements the analysis of these satellite images with field visits and interviews with residents to
see exactly what the coordinates of crop cultivation are and at the end of each year it flies over these
27
areas to verify the coordinates. The advantage is that its methodology is the same as that used in
Bolivia and Peru to measure the quantity of illicit crops, so the data on illegal crop production is
comparable with those countries. A major limitation of this report, however, is that the data published
up to mid-year uses photos up to 31st December from the previous year, which fails to include further
eradication efforts that may have taken place in the first semester of the following year (Velez 2016).
The other measure is carried out by the US Department of State, also once a year, and uses a sample-
based methodology (Mejia and Posada 2008). Photos are taken in certain areas that are representative
and from there the national data is calculated. However, the accuracy of the image resolution
technology used is affected by size of the area covered and cloud coverage. Coca crops are easily
camouflaged with other crops, or under large trees (Velez 2016). Because a minimum amount of
foliage is required for satellite pictures, recently planted coca crops are hard to locate and the density
of crops per hectare is also difficult to determine because of differences in soil fertility. And a
confidence interval is not given instead of a total figure because certainty is more politically strategic
for continued drug policy (Thoumi 2012).
Both reports show general agreement on the broad trend of illegal crop production, but there have
been significant discrepancies in past years. For instance, in 1999 and 2000, SIMCI’s estimates were
significantly lower than the US figures. Then since 2001, the opposite has been the case. In some
years the calculated increase and decrease have been contradictory, and this has sent mixed messages
about the effectiveness of AD, among other drug policy efforts (Mejia and Posada 2008).
Explanations for AD failure
When AD fails to be effective it is usually explained by longstanding structural factors that prevent
farmers from being able to make a better living from their land than what they could with highly
profitable illegal crops (Lupu 2004; Dion and Russler 2008; Thoumi 2012; UNODC 2015). These
issues include socio-economic and geographic marginalization, lack of infrastructure, small land plot
size and the loss of the singular advantages associated with coca leaf crops. Sometimes design
problems in the AD programmes themselves are also attributed the blame for AD failure, as it the fact
28
that some of the results of AD investment will only become evident in the long term. I shall expand
on some of these challenges next.
Illegal drug farming is at its best when it has the right climate and is remote enough to be out of easy
reach from law enforcement actors. Their geographical remoteness means that these communities are
often too marginalized from the main market towns when they decide to switch to legal alternatives.
The absence of roads that connect them to motorways means that farmers find it more difficult to
access the markets in which to sell their produce. This is a major obstacle for large-scale agricultural
production of legal alternatives to take off. The same marginalization also means these communities
tend to lack public goods and state services such as power grids, water irrigation and sewage systems.
Without these basic goods, the competitiveness of their local economy is limited. Power and water
supplies are needed for processing plants, refrigeration, and storage equipment for perishable
products. Similarly, access to health and education services are needed to ensure there is a fit and
healthy labour force and for the local economy to succeed in legal alternatives. In addition, farmers of
illegal drug crops usually have small, humble plots of land, which in themselves cannot generate high
levels of profit in the legal market (Thoumi 2012).
The attributes of the coca leaf crop itself is another way in which its resilience as the favourite form of
agriculture is explained. It provides harvests on average every three months which makes it
particularly attractive to farm as it gives farmers more rapid and regular returns compared to most
other types of crops (Thoumi 2012). Coca leaf is also a light crop, that is not as heavy to transport as
other bigger fruit and vegetables. This means that farmers to do not incur as many additional transport
costs as other crops. And finally, coca leaf is a sturdy crop that grows even in poor soil conditions.
This means it can be grown in remote areas that are unattractive for most other agricultural uses.
Farmers who switch to legal agriculture and face these challenges thus find it harder to compete with
farmers from other areas that are more accessible to the market and with better road infrastructure and
who have already established a comparative advantage in legal agricultural produce. So, when the
former lose their market share in the illegal drugs market, to other drug producing communities, they
become uncompetitive in both legal and illegal markets (Thoumi 2012). They become more
29
dependent on subsistence farming, and their income shortfall is starker. For example, in some places
in Colombia a farmer can earn from one hectare of land used for pepper COL$500,000 [about $145
USD] per year, or COL$240,000 [about $69 USD] if used for yuca (cassava). Whereas a farmer with
the same land area used for coca leaf can earn at least COL$6 million [about $1,745 USD] a year. It
should not come as a surprise then, that a field experiment conducted in 2013 found that about two
fifths of farmers growing coca leaf in Colombia thought it prohibitively costly to destroy their coca
leaf crops (Ibanez and Martinsson 2013). The income they get from farming coca leaf is not a luxury
they can forego. In some instances, farming or picking coca leaf is the only way some parents can
afford to send their children to school (Navarrette 2019b).
Besides the logistical factors so far discussed, sometimes the flaws in the design of AD programmes
are attributed the blame for failing to persuade farmers to abandon illegal drug crops. Unclear
communication and a lack of understanding of the social structure and culture of the community, can
mean that AD programmes are run without local buy in. An AD programme fails if it is carried out
when there is a lack of local support, as was the case of Chapare in Bolivia. In this case, USAID’s AD
programme since the 1980s did not engage with local government and failed to eradicate coca leaf,
whereas European Union’s AD programmes since the 1990s which did work with local governments
yielded better results (Farthing and Kohl 2005). Similarly, a lack of transparency in decision-making,
clear responsibilities and coordination between the agencies implementing AD, can cause confusion
and fail to be effective (Thoumi 2012). For example, in Cauca, Colombia, German Foreign Aid
Agency (GTZ) AD efforts conflicted with the United States Agency for International Development
(USAID) aerial eradication objectives. Farmers’ coca crops were interspersed with coffee crops, as
they were undergoing a slow transition to AD, but aerial eradication destroyed both crops and farmers
withdrew their support for the AD project (Thoumi 2012). In this case AD failed to coordinate with
other simultaneous forced eradication efforts.
And finally, the results of much of AD investment only become evident in the long-term. AD
programmes that introduce intensive capacity-building projects, to train farmers in mass processing,
marketing, financing, quality control, packing, handling competition in alternative crop production,
30
will not transform things overnight. Farmers will need to develop skills for alternative economic
activities through their own experience and as they build familiarity with the new market. Until this
happens, it is difficult for these communities to be competitive in their new choice of livelihood
(Thoumi 2012). Likewise, the money invested in roads and infrastructure and services will have a
gradual effect in transforming the local legal economy of these communities. Thus, the effectiveness
of AD efforts is diluted over a timeframe that makes it difficult to measure.
Why study role of informal institutions in AD success?
Despite the uncertainty around AD’s effectiveness, AD efforts continue to be funded in Colombia.
There are institutional incentives for the state to persevere with AD efforts, even when they are not
successful. Most states are committed to persevere with drug supply reduction instead of harm
reduction policies. One explanation for why the current consensus is to continue a hard-line approach
to the prohibition of illegal drug production is that state security agencies and the military benefit
from the increased defence budgets required for forcing farmers to stop growing illegal drug crops.
Similarly, manufacturers of military and enforcement technology profit financially from the enforced
prohibition of illegal drug farming. But because the forced eradication of illegal drug crop farming is
not only ineffective but unpopular among other groups of people who campaign for a non-punitive
approach to drug policy, AD is used to improve the state’s public image. The huge loss of life and
mass displacement caused by the confrontation between the army and rural farmers, or their extra-
legal protectors, gives the state a bad press and a poor human rights record. And the agricultural
damage caused using glyphosate in aerial spraying campaigns has devastating ecological effects
(which will be discussed in part one of this thesis). If the state is seen to be offering help, and not only
punishment, for farmers to stop growing illegal crops, it signals a gesture of goodwill which can help
to garner greater public support for its war on drugs.
In addition, the general perception of AD among policymakers is of a glass half full. AD can work
and occasionally does work and should therefore continue. Indeed in 2017, the UNODC annual drug
crop monitoring report claimed that 54,027 families in 33 municipalities in 12 different departments
31
in Colombia had effectively abandoned the illegal drug trade because of AD efforts. Why, when
illegal drug production is at an all-time high in Colombia, are AD efforts seemingly working in some
communities? Is it to do with the special approach that some of these AD programmes are taking?
Why does it work in some places but not in others? And at what cost, if any, do these changes come to
the communities?
In the spirit of new institutional economics, I propose that observing informal institutions, such as
sanctions, social customs and constraints, that inform farmer’s decision-making in communities that
have transitioned to the legal economy, can help answer these questions (Ostrom 2009; North 1990;
Mackay, Kenny, and Chappell 2010, 581). Institutions, as defined by Douglass North, are “humanly
devised constraints that structure political, economic and social interaction. They consist of both
informal constraints (sanctions, taboos, customs, traditions, and codes of conduct), and formal rules
(constitutions, laws, property rights) (North 1991, 97). The effectiveness of AD programmes is often
focused on how formal rules can nudge farmers stop producing illegal drug crops. For example, the
laws that pertain to licensing, tax, land ownership and access to education and health are believed to
influence the success of AD. Nevertheless, there are also informal constraints, like de facto property
rights, extra-legal governance, and gender norms, that can be very significant in shaping illegal drug
farmers’ choices and thus, the success of AD. In this thesis I shall seek to understand how three
informal institutions might help or obstruct farmers’ voluntary transition to the legal economy.
Puzzle #1 Role of property rights
The informal nature of property rights to land is often used to account for why farmers grow illegal
drug crops, and it has become accepted among policymakers that formalizing property rights to land
can contribute to the success of AD. Their general belief is that land titling offers economic benefits
to small private land owners who are outside the formal economy (OECD 2015, 21; Guarin 2017).
And international organizations such as the World Bank and the UNDP Commission on Legal
Empowerment of the Poor, assert that land titling helps to reduce poverty in developing countries
(Bromley 2008; World Bank 2006). As such, land titling has increasingly been used since the 1990s
32
and 2000s by some governments and international aid organizations in Afghanistan, Peru, and Bolivia
to tackle poverty in drug producing areas and thus improve the success rate of their AD efforts.
In Afghanistan, the Land Authority (ARAZI) issued landless farmers temporary titles to government
land if they committed to not grow opium poppy crops (“Afghan National Drug Action Plan 2015 -
2019” 2015, 9, 30). USAID funded land titling projects, including Land Title and Economic
Restructuring Activity and Land Reform in Afghanistan to reduce conflict in Afghanistan (Gaston and
Dang 2015). In Peru, USAID also funded land titling as part of their AD programme, which was
implemented by the Government’s Special Land Titling and Cadastre Project (PETT). However, the
effect that land titling had on illegal coca production there has not been documented (USAID 2016a,
8; 2005, 4). And in Bolivia, the USAID Integrated Alternative Development Program funded land
titling in the Cochabamba region with the dual purpose of fostering economic development and
reducing coca leaf production. It was implemented by US company Chemonics International Inc.
from 2004 to 2006 (USAID 2006). And from 2005 to 2010 it included the standardization of land
titles on over 400,00 hectares of land (USAID 2016b, 9–10). Land titling is also used by the General
Directorate for the Development of Coca Growing Regions (DIGPROCOCA) in Bolivia to indirectly
contain drug production. After it issues land titles, it monitors that each farmer does not exceed one
cato (between 1600m2 and 2500 m2 , depending on the region) of coca leaf (Ledebur and Youngers
2012; Farthing and Ledebur 2015).
Land titling in AD in Colombia
In 2015, Colombia also piloted land titling as a renewed component of AD efforts. The National Land
Agency (ANT) ring-fenced a modest budget of COP$12,000 million (about $3.2 million USD) to
subsidize land titles for farmers who had stopped growing illegal drug crops on their land (Colombian
Presidency 2016). The programme, initially set to run from 2016 to 2018, was called Formalizar para
Sustituir (“Formalize to Substitute”) and was implemented by the United Nations Office for Drugs
and Crime (UNODC). Its aim was to give land titles to 7,300 families across Colombia who ceased to
grow illegal drug crops on their land by 2018 (El Tiempo 2017b). By mid 2019 the Colombian
government reported that a modest number of 2,882 land titles had been issued in this programme,
33
benefitting 1,300 families (Colombia Government 2019). While a national newspaper gave the more
generous estimate that by April 2019, 7,000 families had benefited from this land titling programme,
at a cost of COP$18,000 million (about $4.7 million USD) as reported by the ANT (Cruz Olivera
2019). This discrepancy might be explained by the fact that this is just one of various land titling
programmes in Colombia, and that the ANT is processing a backlog of 180,000 applications for land
titles that has accumulated over time. It would therefore not be surprising if some of the figures of
beneficiaries of the Formalizar para Sustituir programme are confounded with those other land titling
beneficiaries.
Although land titling is a novel approach to AD in Colombia, it is not without precedent. In the 1970s,
along with building roads to promote cattle ranching, INCORA (the national land agency at the time)
issued land titles in areas where farmers had settled and were growing coca leaf, such as in Caquetá.
The World Bank also offered these farmers loans for cattle ranching, which was an indirect way of
incentivising farmers to switch to the legal use of their land (Diaz Parra 2019). However, the
Formalizar para Sustituir programme is unique in that it is part of the official national AD efforts, it
specifically targets former coca leaf farming communities, and it makes land titling conditional on the
voluntary substitution of illegal drug crops. At the time of this study, it had been carried out in
Antioquia, Cauca, Putumayo, and Nariño.
In practice, the process of acquiring a land title comprises a mixture of legal procedures, including
clarifying informal ownership documents or outdated paperwork. It can involve finishing the
registration process, as land must be measured by a topographer and issued with an “escritura” before
its ownership can be legally registered. In addition, it can involve adjudicating private ownership of
previously state-owned wasteland, known as baldio land, to farmers who had informally settled it. By
November 2017, the state was reported to have granted land titles for 861,000 hectares of baldio land
(El Tiempo 2017b). This is less than a quarter of the baldio land that was owned informally (estimated
as 4,000,000 hectares). The ultimate target is to issue land titles to just over half of this land area
(2,115,000 hectares) in 15 years (Melendez 2015; Las2orillas 2014).
34
Land titling and drug production in Colombia
Land titling is generally believed to enable former illegal crop farmers to privatize the financing of
their alternative land use, potentially leaving them less tied to the promise of ongoing development
aid. But so far, very little is known about the true effect of land titling on illegal drug production in
Colombia. It is a known fact that most coca leaf farming families in Colombia do not formally own
their land.2 The vast majority of coca leaf farmers grow it on land they do not formally own so a
strong correlation between informality of land ownership and drug production has been identified
(Agencia Nacional de Tierras 2018). Based on this, there has been a study that correlates land titling
with less coca leaf production in Colombia.
This quantitative study by Muñoz-Mora, Tobón-Zapata & d’Anjou (2018) found a significant
negative relationship between land titles and the area of coca leaf production at the municipality level
in Colombia between 2000 and 2009. The observational study, more specifically a retrospective
cohort, was based on a data set, provided by the Colombian Geographical Institute Agustin Codazzi
(IGAC), of land plots that had a formal title or deed registered at the local registry offices. It used a
sample of municipalities that were subject to a programme designed to keep this registry up to date
and where there had been coca leaf at some point, or which had the altitude conditions that favoured
coca leaf production i.e., 500 to 2000 meters above sea level. The sample of municipalities was
further refined to cover those on the receiving end of Plan Colombia policies to reduce coca leaf
production, including interdiction, aerial and manual eradication and AD programmes (information
provided by UNODC and the Colombian government).
The study concludes that it is improved land institutions, through land titling, that cause farmers to
quit coca leaf production. It suggests two reasons for this trend. Firstly, farmers who acquire land
titles are eligible for AD benefits if they use their land for legal farming. And secondly, that it
becomes more costly for people to farm coca leaf after the land is formalized because the state can
expropriate them from their land if they are caught. Therefore, it is less costly to grow illegal drug
2 Interviews given by Miguel Samper, director of the National Land Agency (ANT), Bogota 16/08/2017 and by Eduardo Diaz, director of the National Crop Substitution Programme (PNIS), Bogota 14/08/2017
35
crops on land without land titles. The study argues that farmers with land titles become more
interested in legal farming, which in turn encourages further AD efforts in the community. And as
these farmers reduce their illegal drug production, they also become more dependent on legal farming,
reinforcing the aim of AD.
However, these results are based on a sample of 192 municipalities, which is about 17% of the total
number of municipalities nationwide (1,123), and where the rule of law and state presence was
stronger because they were subject to Plan Colombia policies during that period, and therefore more
exposed to surveillance by law enforcement. A correlation between informal property rights and
illegal drug production in these municipalities is not very informative when about 65% of rural land in
Colombia is informally owned (El Tiempo 2016). And over 40% of small rural producers, who
account for 70% of legal agricultural produce in Colombia, do not hold land titles (Semana.Com
2012). So, it is possible that the importance of land titling in the success of legal agriculture is
overestimated by this study.
More to the point, there remains an important gap in empirical research as to how land titling is
associated with former drug farmers’ voluntary crop substitution efforts. There is a clear need for
theory development with regards to how land titling may reduce illegal drug crop farming. Given
there are no studies yet on the role that land titling plays in ensuring AD success, Part I of this thesis
seeks to fill this gap. My overall question with regards to property rights, is how, if at all, land
titles featured in farmers’ calculation to stop or return to growing illegal drug crops.
Puzzle #2 Role of extra-legal governance
Farmers choose to produce illegal crops partly out of necessity and partly out of convenience. In brief,
it is costlier to profit from legal agriculture than coca leaf, especially when there is less exposure to
law enforcement. This consideration underlies why policymakers believe AD succeeds when the state
is able to develop farmers’ market access and local public infrastructure, and to create employment in
something other than coca leaf production (Dion and Russler 2008; D. Rico 2017). Just as
unemployment increases the relative gains from entering crime, higher wages in legal trade arguably
result in lower returns from farming illegal crops (Dube and Vargas 2013). Similarly, failed AD
36
efforts are most commonly explained by insurmountable poverty and marginalization challenges
(Lupu 2004; Vanda Felbab-Brown 2014). In sum, the key to reduced illegal drug farming is generally
perceived to be in the hands of the state and whether, through its formal institutions, it can make it
more profitable for farmers to use land for legal trades through AD efforts, or it is able to make it
more costly for them not to through more punitive eradication measures.
State governance comprises formal rules that operate like public goods and can make trade more
efficient such as standardized weights and measures, which reduce the transaction costs of exchange
(Eggertsson 1990b, 247). They can also reduce the cost of coordination and communication among
entrepreneurs and make reliable information available to them (Grafton 2000). And other things being
equal, state governance is optimal for coordinating information, managing policy institutions, and
providing infrastructure that facilitates regulation, efficiency and certainty for land markets (Feder and
Feeney 1991). This is, however, assuming the state holds the monopoly of the legitimate use of
physical force, and that it has the capacity to enforce the law and tax its citizens (Daron Acemoglu
2005; Weber 1919).
By its very nature as law enforcer, the state cannot regulate illegal trade. So when the state is absent or
unable to enforce the rules, a space is created for extra-legal groups to govern.3 The extra-legal
governance these groups offer is basically a set of informal institutions that define and enforce rules,
capture the benefits of trade (by giving assurances for actors to confidently participate in trade) and
facilitate collective action for common goods outside of the state remit (Dixit, 2009; Gambetta, 1993;
The logic for AD assumes that if a coca leaf farming community finally has access to the formal
governance institutions offered by the state, it will become more productive in legal farming. This is
tied to the idea that formal institutions are inherently superior to other forms of informal governance
that may have been operating before. I shall next challenge this assumption. The resilience and
3 Some may find, the term “extra-legal” a misnomer, after all, it is not as if there are no laws, only that they are not state laws. However, by “extra-legal” I refer to “extra-state laws”.
37
success of coca leaf farming in many communities indicates that extra-legal groups not only provide
effective governance for illegal drug farming to flourish in ways that legal trade cannot, but they have
a strong incentive to keep the illegal business thriving. The persistence of illegal markets is evidence
that extra-legal governance can supersede the norms of the state (Bubb 2013). Moreover, in these
communities the state is not always perceived as the legitimate source of authority. Therefore, it is
likely that coca leaf farmers have little incentive to replace the informal institutions that have enabled
them to profit from the illegal economy with formal institutions offered by the state. But if farmers
find it harder to profit with state governance instead of extra-legal governance, why then, do they
sometimes choose to do so?
Extra-legal governance
Transaction cost theory suggests that if extra-legal groups and farmers can still make higher profits
from illegal crops, they have every reason to resist state governance. It is not like the presence of the
state gives farmers the freedom to stop doing something they do not wish to do. Contrary to common
perceptions, farmers are not generally forced to work in the illegal drug trade. A UNODC survey
conducted in Colombia identified “pressure by armed groups to cultivate illicit crops” as the reason
why farmers’ worked in the illegal trade in only 2% of the cases. The same survey identified “source
of income” as the main reason for farmers’ engagement in illegal crop farming (71%) (UNODC
2015). Moreover, distrust in the state is common in illegal drug farming communities. And the state
It is unlikely that state promises of improvement, or even actual improvements, in public
infrastructure and services in these communities will make legal trade more profitable than illegal
trade. For reasons discussed earlier in the introduction, the level of income made by growing and
selling coca leaf often exceeds alternative legal income-generating activities. And sometimes state
investment in public infrastructure and market access can even help the illegal local economy thrive
(UNODC 2015).
If a community where AD succeeded once had a successful illegal trade, it also means it had an extra-
legal group that operated like a stationary bandit and provided governance that made illegal crop
38
farming profitable (Olson 1993a). And if this was the case, the stationary bandit would not have
released this business without resistance and confrontation with the state (or at best adaptation to, or
even preying on AD efforts).4 The pertinent point is that there must be another reason why farmers in
these communities found state governance more attractive than the extra-legal governance they had
before. And this reason is unlikely to be that it makes better business sense to trade in legal
alternatives to coca leaf.
Does this mean that in the rare instances when AD succeeds, it is not because farmers choose to adopt
state governance, in lieu of extra-legal governance, but because they are effectively obliged to do so?
This cannot be the so, for two reasons. Firstly, AD is a voluntary process. AD success differs from
other drug supply reduction policies in that farmers want to substitute their illegal crops and do not
revert to illegal farming once law enforcement withdraws. For AD to be considered successful the
state cannot claim victory by force alone. Farmers who feel obliged to accept the terms of the state by
force alone tend to protest or simply refuse to cooperate with the state. Alternatively, they may face
retaliation from extra-legal groups who carry out attacks on farmers who try to substitute their coca
crops, which stops AD from progressing (Asmann 2018). Therefore, in communities where AD
succeeds, farmers should believe that substituting their illegal crops will maximise their returns.
And secondly, the use of force is not effective either. If AD success boiled down to the state stepping
up its resources to dislodge the territorial control of the extra-legal group in charge, the policy
application of this theory would be for the state to pursue a “war on drugs” wherever possible.
Admittedly, there are some successful AD stories in Colombia when prior to investing in AD, the
state increased its security and policing e.g. the Macarena Integral Consolidation Plan in 2007 (Ripoll,
Berrio, and Rubiano 2013; D Mejia, Uribe, and Ibanez 2011). Most evidence, however, has shown
such efforts to be largely ineffective as well as harmful (Coyne & Hall, 2017; Mejía & Restrepo,
2013). The “war on drugs” is a reputably ineffective approach to drug supply reduction because extra-
legal groups are so effective at protecting the illegal trade.
4 For an example of this see (Ballve 2011), who describes how between 2002-2005 an AD programme linked to Plan Colombia, financed 22,000 palm oil crops that were later found to be controlled by extra-legal groups.
39
Given there are no studies yet on how the dynamics of extra-legal governance can influence AD
success, Part II of this thesis seek to fill this gap. I propose that state-centric explanations for AD
success may underestimate the role that changes in extra-legal governance play in paving the way for
AD to succeed in communities where coca leaf had been a prosperous crop. And these changes can
happen independently from any action taken by the state. At the cost of simplifying the complex
reality of how the state and extra-legal groups interact in the Colombian context, I use the Olsonian
stationary-roving bandit model to offer a novel approach to explain successful AD, from the farmers’
point of view. The point of is to develop a theory that acknowledges that extra-legal groups’ self-
interest is not enough for an illegal trade to prosper indefinitely, and official state intervention alone
cannot prompt farmers to voluntarily embrace AD. My question with regards to extra-legal
governance, is how, if at all, it featured in farmers’ calculation to stop or continue growing coca
leaf crops.
Puzzle #3 Role of gender norms
Another informal institution that has been side-lined in how we account for the frequent failure of AD
are the effect of gender norms. This is significant because even if the new laws and rules introduced in
contexts of AD are not discriminatory, their effect may still create gender bias when they interact with
gender norms. These can play a significant role in how economic opportunities are distributed in
communities transitioning away from illegal agriculture.
It is important to analyse their effect on economic opportunities in contexts of AD for two reasons.
Firstly, understanding how one group of people may benefit at the expense of another, is intrinsically
fair and valuable, and can help inform policies that promote gender equity. After all, policymakers
have called for further research into women’s role in drug production to help make drug policy more
Secondly, from an instrumental rationale, greater insight into the customs that affect women and men
differently in the context of AD may be crucial for making AD not only more successful, but durable.
If the substitution of illegal crops creates a local economy that is legal, but that cannot offer women
equal economic opportunities as men, it is overlooking an important socio-economic effect of AD. So,
40
understanding the incentives women might have to seek to work in coca leaf production, either at a
later stage or elsewhere may help future AD efforts have more lasting results in many places.
There is an additional benefit to this line of enquiry. Lower levels of inequality and gender
discrimination is associated with greater economic growth and productive capacity (Braunstein 2008;
A. Alesina and Rodrik 1994; Duflo 2012). Therefore, greater gender equality is associated with
economic development, which is the prime objective of AD efforts. Studies have also shown that
more productive women, despite being kept out of top jobs, increase productivity of less productive
male workers (Klasen 2006). And the benefits trickle down to the whole family, as women have been
found to invest a greater proportion of their income on children and home, which is usually their main
sphere of influence (World Bank 2006). Given how economic development reduces gender inequality,
some may argue that AD policies should focus on the economic process rather than on gender
equality. However, as Duflo points out, gender equality also reduces poverty, so both factors need to
be addressed in the quest for economic prosperity (Duflo 2012).
So far, there is scarce data from women involved in illegal drug farming which means there has been
relatively little attention given to their situation. Existing studies on women formerly employed in the
drug trade have mainly focused on female drug mules and female drug sellers. This should not be a
surprise given that the data that is most readily available comes from women that are in prison
(UNODC & Colombia Ministry of Justice 2019, 20). For example, UNODC and the Colombian
Ministry of Justice published a joint report in October 2019 identifying the socioeconomic situation of
2500 women whose incarceration was “related to drug problems”. However, their analysis was based
on the women’s situation prior to conviction for a wide range of drug-related crimes. Their sample
excludes the experience of women who worked in small-scale drug farming because the state
response for this crime is seldom imprisonment, and more often other drug eradication efforts.5
Perhaps it is this lack of research why policymakers believe that when farming communities cease to
grow illegal drug crops, women’s income generating opportunities may be more limited than men’s
because they lack equality to men in land ownership or schooling and so are less able to find
5 Imprisonment tends to be the sanction for people who own between 20 and 100 drug crops, as stipulated in Article 375.
41
alternative work. These mainstream arguments about the broad challenges that women in AD contexts
face are based on evidence that comes mainly from other cultural contexts. It still ambiguous if
women who cease to work in illegal drug production in Colombia also face these are challenges.
In the third part of this thesis, I therefore analyse the ways in which women may be differently
affected after their community transitioned away from farming illegal drug crops. I observe the gender
differences in the economic opportunities available in these communities, and how they arise. My
follow-on question is how gender norms, despite there being scarce gender difference in property
rights to land and access to education, continue to shape women’s economic opportunities and
incentives. Given there are no studies yet on how gender norms influence AD success, Part III of this
thesis seek to fill this gap. My overall research question regarding gender norms is how they
affect women’s economic opportunities and incentives differently to men’s, if at all, in
communities that have substituted illegal coca leaf farming.
In sum, this thesis raises questions about, and presents new angles through which to understand, how
AD, as a non-punitive drug policy, sometimes works, sometimes does not, and how it may be
improved.
A note on normative perspectives of land
I should add that the premise for my research parts from the normative assumption that AD is
desirable but, in fact there are other ways to conceive the notion of “good use” of land. The degree of
agricultural productivity is not the only way to value soil, as Lyons describes in her ethnographies of
soil in Putumayo (Kristina Marie Lyons 2014; Kristina M. Lyons 2020). With the population
expansion in Putumayo, Amazonian soil has been turned into farms and gardens which has
stigmatized unproductive soils. Lyons notes that “capitalist agriculture has increasingly stabilized a
specific idea and treatment of soil to a mechanized and rigid timetable of production that has
generated volatile changes, destruction, and loss” (Kristina M. Lyons 2020, 170). And as a result, land
in this area is perceived by state institutions merely as a physical space that should be protected,
policed or worked (Kristina Marie Lyons 2014, 221).
42
However, campesinos in Putumayo, for instance, understand the quality of soil less in terms of
capitalist agricultural productivity and more in terms of “life propagating relations”, encapsulating
cultural practices and knowledge and not just the natural determinations that agronomists
acknowledge. Coca leaf farmers are not necessarily “problems that must be solved” by the state but
can also be regarded as legitimate actors in the determination of their present and future (Kristina
Marie Lyons 2014; Kristina M. Lyons 2020). Soil can also be defined by how farmers relate to it and
not just in terms of how it can be classified into taxonomies for agricultural use. Therefore, my
approach should not be considered the only or best way to conceive of the situation of coca leaf
farmers in relation to their land.
Research contribution to Alternative Development policy
Overall, my findings help explain the resilience of illegal drug production, despite the state’s efforts to
suppress it, and have relevance to drug policy in three ways. Firstly, they illustrate the limitations of
land titling as an AD strategy. Secondly, they present a novel explanation as to why the state
sometimes succeeds with AD, from the perspective of farmers. And thirdly, they highlight the
economic opportunities coca leaf production opens to women as an additional reason why most
communities may resist AD.
This research matters because there are many things we still do not understand about AD, especially
why and how it succeeds in some places but not in others. It makes an important contribution to
evidence-based drug policy and our understanding about how to improve the effectiveness of non-
punitive approaches to drug supply reduction. I seek to explain how, if at all, land titles have
influenced farmers’ illegal land use behaviour. Until this date, there are no studies that look
specifically at the role that land titling plays in farmers’ incentives to stop growing coca leaf in
Colombia. The association between the two is therefore still hypothetical despite there being support
for this AD policy, and this research seeks to fill this gap.
I study whether communities with ample fertile soil chose to quit illegal drug production based on
push factors, created by the declining returns offered by the illegal drug trade itself, or by pull factors
43
created by the state governance outperforming extra-legal governance. The discussion on why AD
succeeds is usually based on the premise that the state is responsible for prompting farmers’ change of
heart. By changing the locus of the analysis to the extra-legal governance of illegal crop farming, this
research shines a light into a wider set of possibilities that can pave the way AD to succeed.
And I observe the effect the removal of illegal crops has had on women’s economic opportunities and
why. Currently, there is scant knowledge about the difficulties that women face in continuing to be
financially independent when their community stops farming coca leaf. My research into their
situation is important for making AD more gender-sensitive and thereby more effective.
I capture the lived experiences of former coca leaf farmers in some of the very few places in South
West Colombia where illegal drug production has been effaced voluntarily. These communities are
exceptional, as not only do they belong to the third of coca leaf growing places where AD has been
tried, but they also form part of the 17% of places where AD is believed to have worked. Their rarity
makes their generalizability for all coca leaf growing areas less clear than a probability sample would
be. However, they presented optimum research conditions while a probability sample would not have
been feasible for qualitative research.
Even if the experiences of these people cannot be generalized to the situation of every person who
starts and stops working in illegal drug production, they offer an enriching insight into the type of
issues that people may or may not encounter when reverting to the legal economy. And they provide
rich qualitative information that illustrates the relationship between the variables of interest and the
mechanisms through which AD has, or has not, worked.
The data I use comes from fieldwork carried out in 2017-2018, at a time when these communities
were safely accessible following the Colombian Peace Agreement. In some of these places, the
security situation has since changed and the safe conditions for research would no longer be
guaranteed at this present date. It is thereofore special data because it was gathered during the small
window of opportunity that presented safe research conditions.
44
Research contribution to academic knowledge
My research contributes to the policy debate on AD, but on a theoretical level it also adds to academic
debates in political economy more generally. My subject area is political economy, which is a
multidisciplinary social science centred on studying how institutions e.g., political, cultural, and
social – historical context, and ethical values, influence the economic outcomes, distribution of
wealth, markets, trade, and vice versa. The main issues and debates in this subject area are about the
role the state, and how involved it should be in regulating behaviour or if self-governance and free
markets yield better economic outcomes. My research addresses this in different ways.
The first part of my thesis is on property rights and focuses on whether these are more effective when
enforced by the law or by social norms. Assuming that there is a tension between formal and informal
property rights to land, instead of a continuum, I question Hernando de Soto’s theory that land titles
naturally improve the land market and roll back the informal economy. I show that the reality on the
ground is different and that informal property rights can be more significant in enabling the market to
function well. This supports Ostrom’s (2009) theory that informal institutions, as patterns of
behaviour shared by most people, have more control over behaviour than do rules that are simply laid
down by an authority. I shall discuss this at greater length in the section that follows.
In addition to contributing to the discussion on property rights, in the second part of my thesis, I
present an analysis of a type of self-governance i.e., extra-legal governance, and how this operates in
relation to the state. I build on Olson’s (1993; 200) theory of stationary and roving bandits and
suggest that each type of bandit can mutate into the other, in face of rivalry. And I develop Cheng’s
(2018) theory of extra-legal governance by illustrating how an extra-legal groups operate against their
own business interests when time for extracting rents starts running out. In my thesis I suggest that
there is no clear distinction between roving and stationary bandits. Often the literature likes to
categorize specific armed groups as one or the other, but I make a case for how the same group can
take different forms in different places and at different times. Like Olson and Cheng, I assume that
45
extra-legal groups exist to sustain their own business interests and that it is in their nature to seek to
plunder resources or make trade more efficient to extract long-term rents.
And the third part of my thesis contributes to the literature on the relationship between gender norms
and other institutions and centres on how social norms can skew economic outcomes in subtle ways.
In this final part, I show that institutional changes such as AD, are fragile and without ongoing
maintenance and economic incentives, gender norms return to their previous shape. I build on Van
Staveren and Odebode’s notion that norms do not compete with one another but shape one another.
They describe how the Yoruba women in Nigeria who can buy property and can work just like men,
nevertheless still need male approval to buy property, they marry early, they cannot inherit land, they
have less time to work and they spend more of their money on children (Van Staveren and Odebode
2007). All this means that seemingly symmetric institutions continue to be shaped by asymmetric
institutions that express a power dynamic.
The overarching premise of my work is that Coasian transaction cost theory explains the nature of
institutions. That is, institutions exist to facilitate trade, and change through a natural equilibrium
designed to optimise trade outcomes. And informal institutions are just as significant as formal ones
in shaping economic outcomes. This means that both formal and informal property rights reduce
transaction costs and thereby increase the value of market transactions. It also implies that legal and
extra-legal governance institutions can make trade more efficient. And it suggests that gender
hierarchies in the household can compromise economic efficiency, despite laws that promote gender
equity.
Spatial and regional context of my fieldwork
My findings speak for individual coca leaf farmers who lived in veredas where farmers had not been
forced to start or continue to grow coca leaf by extra-legal groups, and where land titling had been
offered specifically as an AD incentive. These places were situated in municipalities where coca leaf
continues to be grown in abundance but which were, broadly speaking, relatively closer to roads,
46
compared to the average coca leaf growing vereda (although their accessibility varied between them).6
This meant that some of these veredas were not far off from roads where coca paste and cocaine is
transported at night-time by intermediary drug dealers, known as traquetos, on fast and loud
motorbikes.7 It also meant that it was relatively easier for the army to monitor their crops more
closely, which posed a big disincentive for re-growing coca leaf at home.
The experience of these farmers is valuable to those who wish to observe how, and whether, land
titling affects individuals’ decision-making when individuals are not under pressure from extra-legal
groups to farm coca leaf. The analysis of behaviour at the individual level is key to understand the
mechanism by which land titles are supposed to persuade farmers to stop growing coca leaf. This is
the aim of the first part of my research.
Analysis at the individual level is also helpful for understanding the gendered dimension of economic
opportunities in communities that farmed coca leaf and then stopped doing so. This is what I set out to
find out in the third part of my thesis.
The analysis of behaviour at the community level, be it members of a single vereda, or a cluster of
people living in similar veredas, is useful to explain why communities opt to abandon the illegal trade
when land is apt for coca leaf farming and legal alternatives are not as profitable. This is my objective
in the second part of my study.
As I discuss in the methodology chapter under the “Selection of Research Locations” section, the
veredas in which I conducted my fieldwork were unique in various ways. However, they also had
their differences between them, as the next section describes.
Is it possible to generalize?
As King, Horrocks & Brooks (2018) explain, qualitative interviews are by their nature interpretive
and cannot establish generalized patterns of behaviour. Having said that, I see generalization as a
6 In Nariño, Linares and Los Andes were municipalities that were located a three-hour drive away on an untarmacked road from Pasto, the capital of Nariño. In Cauca, Balboa is a three-hour drive from Popayan, the capital of Cauca, and Rosas is about one and a half hours drive from Popayan. Mercaderes is half an hour’s drive further south of Balboa. In Caquetá, Santiago de la Selva, the vereda I visited in Caquetá, was a four-hour drive away from Florencia, the capital of Caquetá, and a one-hour drive away the urban centre, Valparaiso. In Putumayo, La Carmelita was a two-hour drive away from Puerto Asis. And the veredas I visited in Valle del Guamuez, were all a 30 – 45-minute drive away from La Hormiga, the capital of Valle del Guamuez. 7 Traquetos are the local representatives of large-scale drug traffickers, who buy coca paste for them.
47
continuum, rather than a binary concept. I think I can generalize from my findings to make assertions
about the communities I visited as a whole, on certain points. And of other places with similar
characteristics. I cannot generalize from my data to make assertions about all coca leaf producers: my
research locations were unique in many ways; and were selected on the dependent variable; and
chain-referral sampling can lead to network overrepresentation. However, my aim is not to be
representative (small-N research can never be) but to offer an enriching insight into the type of issues
that people may or may not encounter when reverting to the legal economy. I also propose a possible
relationship between some variables of interest and the mechanisms through which AD has worked.
What I can generalize about the municipalities where this research was conducted
Although they had much in common, the context and history of the municipalities and the veredas I
visited also varied. As a way of background, Putumayo and Caquetá are in Colombia’s Amazon
region. The Colombian Amazon was first occupied by missionaries in 1558, followed by waves of
spontaneous migration in 19th and 20th centuries. Despite state presence always being marginal in this
region, people migrated there from rural areas where unemployment was high and subsistence
farming was insufficient to make a living They sought to extract rubber (in the 1930s), and timber and
oil (Kristina Marie Lyons 2014, 215; Diaz Parra 2019; Ramirez 2011). Alternatively, they sought to
escape political violence during La Violencia period (from 1948 – 1960). They were better able to
make a living from subsistence farming and logging in the jungle as the climate was warm and humid
and the land was fertile jungle terrain.
Another wave of colonizers (sometimes referred to as colonos) came to this region following the
arrival of commercial coca leaf crops in 1978, and of the extra-legal groups in the 1980s and 90s.
Land use became dominated by illegal crop production. Putumayo came to produce 40%, and Caquetá
came to produce 16% of Colombia’s total amount of coca leaf in 2000 (Kristina Marie Lyons 2014;
“Colombia Coca Cultivation Survey” 2009; Ramirez 2011). As extra-legal groups began to increase
their presence and assert their authority through a mixture of protection and coercion, communities in
this part of the country grew larger and less cohesive. Newcomers were not trusted by the longer term
colonos (Ramirez 2011, 76).
48
Although generally speaking, coca leaf tended to been grown by small landowners in Putumayo
compared to more medium and large landowners in Caquetá, the coca leaf farmers I interviewed were
mostly small landowners (Ramirez 2011, 61). The coca leaf growing communities that I visited in the
municipalities of Puerto Asis and Valle del Guamuez in Putumayo, and Valparaiso in Caquetá, were
colonizers that had spontaneously created a de facto private property rights system from open access
baldio land they claimed as their own. In theory this land belonged to the state but was vacant because
in practice the state has not been able to enforce property rights to land there. The colonos self-
regulated their de facto property rights through oral agreements and informal documents (documentos
comraventa) that were not legally binding.
Competition increased between extra-legal groups for territorial control of the coca leaf growing areas
in Amazon in the 1990s (Ramirez 2011). Homicide became the leading cause of mortality in
Putumayo, and the coca leaf farming municipalities were those with the highest rates of homicide
mortality between 1991 and 2002 (Ramirez 2011, 80). The violence escalated to a point that it caused
mass displacement of coca leaf farmers in the 2000s. Many of those who were displaced from
Caquetá and Putumayo returned to their homeland in other parts of the country such as Nariño and
Valle del Cauca (Viloria de la Hoz 2007). According to the 2005 National census, most immigrants in
Nariño came from Putumayo, as well as Cauca and Valle del Cauca.
Those who stayed behind in the veredas I visited in Putumayo and Caquetá now aspired for land titles
as they felt greater uncertainty about their future property rights now that the demographics of their
community had changed, the extra-legal groups had withdrawn, and the state was actively offering to
subsidise land titles. Some were also fearful of their land being sold to extractive companies in the
future, as they heard was happening in other municipalities in the region. In the municipalities I
visited in these two departments, land titling by the state was partly to relocate internally displaced
people, and partly to incentivise farmers to replace their illegal crops with legal land alternatives. The
process was mainly by transferring legal ownership of baldio land to private individuals.
Cauca and Nariño are more like each other than to Putumayo and Caquetá. To start with, their climate
and topography vary a lot. Nariño is in the South West corner of Colombia, right next to Ecuador and
49
between the Andes and the Pacific Ocean. It is very isolated from the rest of the country because it is
very mountainous, and in many ways its culture and economy are closer to Ecuador than Colombia. It
was only until 1970 that the Pan Americana highway was built, connecting it with the rest of the
country, but there are still very few roads and even accessing these is difficult in the rainy season as
there are frequent landslides. Cauca is just north of Nariño and has a similar topography with a range
of climates depending on the altitude. Like Nariño, it is a particularly diverse department with
significant indigenous and ethnic minority populations living near the Pacific coast.
The people I interviewed in Cauca and Nariño had either lived in the veredas all their lives, such as in
the Cauca veredas of ‘La Villa’ and ‘La Palma’ where most of the 300 inhabitants were blood related.
Or they were people who had returned home after a few years of farming coca leaf in places around
Putumayo and Caquetá. The latter had brought the coca leaf farming know-how and had tried to
replicate the coca leaf bonanza they had experienced, but they encountered two obstacles in doing
this. Firstly, the topography of their land was much hillier and with greater variety of climate zones.
So, coca leaf was not very productive in the higher parts of the hills, where most of the veredas I
visited where located. This led some communities in higher altitude points of the Andes Mountain
range to diversify into opium poppy production, while others preferred to stick to legal crops such as
coffee as the price was like that offered for their coca leaf crops. In one vereda called ‘Esperanzas del
Mayo”, where the climate was warmer, the problem was that they faced an extreme drought that
forced them to migrate to nearer the River Mayo.
And secondly, there was not enough time to build their profits and grow their production to a
comfortable level before they began to feel extorted by competing extra-legal groups, which
effectively stifled their business. FARC and ELN arrived in Nariño in the 1980s and the paramilitaries
arrived in the late 1990s (Viloria de la Hoz 2007, 68). This meant that, unlike the veredas I visited in
Putumayo and Caquetá where coca leaf farming had had more time to develop in peaceful conditions
during the 1980s and much of the 90s with just one extra-legal group (FARC), coca leaf farmers in the
veredas I visited in Cauca and Nariño had barely started off their coca leaf business in the 2000s
before they were targeted by rival extra-legal groups. In addition, the communities seemed to be more
50
risk averse than those in the veredas I visited in Putumayo and Caquetá, as many people feared the
violence would escalate again if they continued to farm coca leaf, but this time it would be in their
home villages. As a result, coca leaf production in the parts of Cauca and Nariño that I visited, was
concentrated in hubs at the bottom of the valleys, where the climate was more appropriate, in veredas
such as Samaniego and Guaitara in Nariño, or El Plateado in Cauca.
Unlike in the veredas I visited in Caquetá and Putumayo, there was not much baldio land available in
the veredas I visited in Cauca and Nariño. Instead, most land was already owned by large landholders,
but over the years they had sold off pieces of their land informally to campesinos who wanted to make
a living from subsistence farming. Communities had self-enforced these de facto property rights ever
since, and most people had not registered their purchase as the process was too complex and costly. In
most cases it required a registered topographer to measure the land, then a certificate from a local
state registry office and one from the office of public records to prove that the cadastre was updated.
Therefore, the state’s land titling efforts to encourage farmers to opt for legal alternatives to coca leaf
crops in these places, were centred on subsidising the paperwork that clarified land ownership rights,
rather than granting baldio land to private individuals. As in Putumayo and Caquetá, there was
support from the communities in Cauca and Nariño for land titles as farmers wanted to assurance their
land would not be taken away from them by the legal owners at some point in the future.
What my research does not contribute
My research does not speak for the municipalities as a whole and can certainly not be generalised to
represent the experience of everyone living in Caquetá, Putumayo, Nariño, or Cauca as a whole.
Colombia has a wide range of populations and climates, as well as ongoing fighting between extra-
legal groups, which results in very diverse contexts and different incentives for people to farm coca
leaf and to seek the legal recognition of their property rights to land.
Difference with coca leaf farmers living in collective territories
Most notably different to the situation of the people I interviewed is that of the Afro-descendant and
indigenous communities that form a significant proportion of people who live in some of these
51
departments. Cauca has the largest percentage of Afro-descendant people (22%) and is the second
department after La Guajira with the greatest indigenous population (21%). Together they own about
a third of land in Cauca (Gamarra-Vergara 2008, 40–41). The situation in Nariño is similar. In the
2005 national census, 18% of the population of Nariño identified itself as Afro-descendant and 10%
as indigenous (Viloria de la Hoz 2007).
My research focuses on private property rights, whereby individuals have exclusive rights over a plot
of land and can profit from it and sell or transfer it to others and can stop others from using it. As this
thesis argues, state recognition of these rights through land titling is not always essential for the
market to run efficiently if these rights are recognized by the rest of the community, albeit informally.
However, the situation of Afro-descendant and indigenous communities in Colombia is very different
because they have communal property rights systems. This means that their land is not individually
owned but is shared among the group members and opportunistic behaviour tends to be regulated
through social norms. In Colombia indigenous groups’ land rights have been recognized as
autonomous since colonial times, and they have the same allocation of powers as a municipality.
Afro-descendant communities were granted collective land titles in 1995, following Law 70 in 1993,
almost all of them are in the Pacific region and a very small part in the Caribbean region. Both groups
are roughly equal in size in terms of population but the indigenous reserves are roughly twice as large
as the territories owned by Afro-descendants (Gamarra-Vergara 2008, 41). Collective property has a
very strong characteristic in that it cannot be embargoed or seized by anyone in what is known as
extincion de dominio. Once the state recognizes the land as belonging to a community then this
community has ultimate authority of the land and thus must administer it.
Some of the main and most resilient coca production nuclei are in indigenous reserves and Afro-
descendant territories. Coca leaf crops entered the Pacific region in Nariño and Cauca in the mid
2000s because of glyphosate aerial spraying that took place in Putumayo and Caquetá. Together with
their coca crops, farmers were displaced from areas where there was no state regulation of property
rights and arrived in places in Nariño and Cauca where land was communally owned. And with them
came extra-legal groups. The arrival of coca leaf farmers in these “indigenous collective properties”
52
and “Afro collective properties” has created serious land issues that are very different to the situation
of coca leaf farmers analysed in this thesis.
As the director of the Ethnic and Campesino Territory Observatory of the Pontificia Javeriana
University explained in one of my elite interviews, one of the main issues the influx of farmers in
these territories has triggered, is the beginning of individual capitalist rationality with regards to land
access. Coca leaf farmers seek to demarcate individual plots of land to grow their own illegal cash
crops in forests where people have been used to hunting, foraging, and transiting with a collective
rationality. In doing so, the coca leaf trade has imposed a private-individual ownership and a free-
market economy on communities that depend on social cooperation and equal distribution of wealth
to survive.
The arrival of coca leaf has been damaging to their culture. On the one hand, it has confined these
ethnic groups to a corner of their territory. Those unwilling to participate in the coca leaf trade are
unable to move around freely on the land that belongs to them and are forced to stop or change their
usual activities and habits as they can no longer access the forest. In cases of indigenous communities,
whose collective identity is inextricably linked to their land, the damage is irreparable. For example,
the Kogi, the Arhuaco, the Kankuamo and the Wiwa indigenous communities living in the Sierra
Nevada of Santa Marta in the Caribbean region, have a territorial identity that is linked to this specific
land, and which has been built over a long time. They consider the Sierra Nevada to have been
desecrated by people who trespassed their land to farm coca leaf and who buried people there.
Alternatively indigenous communities are left with no option but to move elsewhere, adding to the
issue of mass internal displacement (UNHCR 2009).
On the other hand, it has spread values of individual competition that run contrary to the values of
unity and respect for their traditional authority of the council or the “cabildo”. This creates tension
that can be best resolved with state enforcement of their property rights. There are villages like
Timbiqui, in Puerto Saija, Cauca, that are populated by indigenous and Afro-descendant communities
in the Pacific, where the illegal drug trade has made things more expensive even than in Bogota. This
puts strong pressure on local Afro-descendant people, who would not normally farm coca leaf, to
53
participate in the illegal drug trade. As they do so, their local economies based on natural resources,
like molluscs and special fibres, and their forest licences to trade valuable timber, are destroyed
because they require too much effort for too little money compared to what farming coca leaf entails.
Similarly, the local population in the Eperara Siapidara (Embera) indigenous community, does not
associate itself with coca, and very rarely can an indigenous person from Nariño and coastal Cauca,
be seen working as a raspachin or a coca leaf farmer. In that area, coca leaf is a wild crop so farming
it is not part of their traditional knowledge-base or local ecological customs. Coca leaf is not even
used medicinally there, like in the Nasa communities in Cauca for instance. This has meant that when
there is coca leaf money circulating in the community, their lifestyle is affected as indigenous
individuals find incentives to participate in the illegal coca leaf economy.
Another issue that stems from this, is that extra-legal groups can forcibly recruit members of these
communities to work in the illegal drug trade. Unlike the communities I focused on in my fieldwork,
where individuals joined the coca leaf trade voluntarily, in some indigenous territories people have
been forced to participate in coca leaf farming and processing. For example in the Awa community,
close to a village called Policarpa in Nariño, extra-legal groups killed 50 Awa people between 2008
and 2009 by, when the Awa leaders, who were opposed to the coca leaf business, tried to ban the
young members of their community from participating in the business (Amnesty International 2009).
The Awa were eventually supported by the state to be able to regain access to their land freely, but
many extra-legal groups remain present in ethnic territories.
In contexts like these, the cost of clarifying a nebulous land ownership situation or defending a claim
to use it for something other than coca leaf, is high, and can result in death. When there is a clear
clash of property rights systems and the levels of violence against those wanting to give up coca is
high, then the state’s legal recognition and enforcement of land ownership can be crucial for resolving
the land disputes and to prevent mass displacement. Unlike the low significance of land titling for
many of the campesinos in the places I carried out my fieldwork, clarifying and protecting official
land ownership plays a very significant role for Afro-descendant and indigenous people where coca
leaf is produced.
54
These are some issues that collective communities face over land ownership and the illegal drug trade.
However, I do not cover these in my research because they are not directly related to individual
decision-making or to AD efforts, and so were beyond the scope of my fieldwork.
Difference with campesinos living in areas targeted for oil extraction
My research objective has not been to describe the experience and significance of land titles for
people living in places where the land is sought after for resource extraction. For example, Putumayo
and Caquetá have areas of great interest for oil exploration. Since 2006, oil blocks have been allocated
to the oil industry in the Amazon on land that people informally owned but who were displaced by the
conflict. In 2019, there were nine oil companies present in Caquetá with 25 oil blocks allocated to
them, 22 of them in exploration phase and one in production. At that time there were still 15 oil
blocks available for allocation (Diaz Parra 2019). This is a strong issue of contention.
According to Colombian law, for a land to be issued with a title it must not be located within a
hydrocarbon extraction area. So, in areas like these, where access to land is sought for oil extraction,
land titling is very significant to individual farmers. The tensions with the local population are not
only about who is the legal owner of land, but how this land can be used, as there are reports of
damaging effects of hydrocarbon exploration (Finer et al. 2008). For example, around 2001 farmers in
a vereda called ‘Campoalegre’ in the Orito municipality in Putumayo, complained that oil refineries
on land neighbouring land had polluted their water sources and damaged the quality of their soil.
Therefore, land titles are of crucial importance for people in this position, who feel strongly that the
land they colonized belongs to them, and who fear big oil companies might be issued warrants to use
their land.
Difference where there are real alternatives to coca leaf
Likewise, land titling is more likely to be significant for those living in places where there is a real
alternative for substituting coca leaf farming with legal alternatives. I found this was only the case for
farmers who lived in a handful of veredas, such as ‘Esperanzas de Mayo” in the Mercaderes
municipality in Cauca, where people could make a better living from lemons than from coca leaf.
They had previously farmed coca leaf in a vereda called Altos del Mayo, up the hill, but had been
55
displaced to ‘Esperanzas del Mayo’ by the extra-legal groups. ‘Esperanzas del Mayo’ was right next
to the Pan Americana highway which meant that not only was it easier to transport the fruit but
growing coca leaf there would be impossible as they could be easily accessed by law enforcement.
However, they needed land titles to be able to take out loans to invest in lemon trees. It was also the
case for farmers in ‘Santiago de la Selva’ in Caquetá, who were able to make a living from milk
farming because of Nestle, and for some farmers in Linares, Nariño, who wanted to invest in sugar
cane crops now that there was milling machine in their reach which made sugar cane a less labour-
intensive crop.
However, most of my fieldwork took place in veredas where poverty predominates, and land is not
very productive for commercial ends. Therefore, this influenced the conclusion I drew from my
findings that the lack of land titles is not the real obstacle that keeps former coca leaf farmers’
alternative legal business from thriving. Even in the veredas that I visited such as ‘Lo Alto’ and ‘Los
Andes’ in Los Andes, Nariño, where coffee crops grow relatively well because they have the right
altitude and are not too far from roads and there is a market for it, farmers felt it was not as profitable
as coca leaf and was much more labour intensive. Farmers still wanted to access loans to invest in
their coffee crops, as they felt they could not grow coca leaf given there was greater state surveillance
of how they used their land. However, they described that those who were serious about making a
greater profit from coffee farming, had preferred to work in Huila, Antioquia, and Valle del Cauca,
where the harvests lasted up to two months, instead of two weeks, and where coffee pickers were paid
more for their labour, and given food and housing. Those who did well from coffee in Cauca, such as
the vereda of ‘El Porvenir’, which was not too far from a vast water source called El Macizo
Colombiano, still preferred to combine their coffee crops with coca leaf crops to keep a steady
income.
Difference where farmers are forced to grow coca leaf
Finally, the focus of my research is on land titling that is specifically offered by the state to encourage
farmers to stop growing coca leaf. However, there are other state efforts of land titling for other
purposes, which I do not cover in this research. For example, the Land Restitution Unit (URT) issues
56
land titles to victims of land dispossession, so that they can return to land they were forcibly displaced
from during the conflict. Alternatively, if it is still not safe to return to their original land, they are
granted ownership of land somewhere else where it is safe to live.
This is the case for many people living in current conflict zones. In Uraba, extra-legal groups have
bought up land and dispossessed many campesinos from their land (Memoria Historica 2009; El
Espectador 2019). Their objective in doing this has been to maintain an illegal trade smuggling
corridor (Ballvé 2020; 2012, 610; Grajales 2013). Farmers who have been displaced from this area
have then found themselves unable to return because of the menacing presence of extra-legal groups.
A similar situation arises in the Catatumbo region, in the department of Norte de Santander. As it is
near the border with Venezuela, it is not only a hotspot for coca leaf, but a key route for cocaine to be
exported out of Colombia. The local population has been subjected to ongoing violence as they are
caught in between extra-legal groups and the army who are constantly fighting to control this area.
Many campesinos wish the state would recognize and enforce their claim for joint land ownership, in
what is called a campesino reserve area (ZRC), so that they can freely make a living in alternatives to
coca leaf. Farmers feel that without their land tenure security protected by law, they are unable stop
farming coca leaf (Amnesty International 2020). However, as this area is also rich in minerals and
resources and therefore of interest to extractive industries, as well as to extra-legal groups as an illegal
smuggling route, their claims to this land have not been validated by the state (Verdad Abierta 2017).
Since the Peace Agreement, coca leaf production has continued to increase in Catatumbo, and no AD
programmes have been implemented this area.
It should be clear then, that the situation of the coca leaf farmers I interviewed is a long way off from
the situation coca leaf farmers face in Catatumbo and Uraba, where many people have been forcibly
displaced from their land and then unable to return to it. The focus of my research was on the situation
for farmers in what are now peaceful, rural areas, that have ceased to grow coca leaf. These farmers’
lack of legal ownership has not been an impediment for substituting their coca leaf crops, because
they were protected by the state from extra-legal group aggression and potential land grabbing.
57
In sum, my unit of analysis, the fieldwork sites, and the specific types of land titling I focused on have
all shaped the conclusion I reached in my quest for understanding the role of land titling in
influencing AD efforts. That is, that land titling is secondary to other factors such as violence, in
causing individuals to give up coca leaf farming. As has been discussed in this section, land titling
remains problematic in many other parts of the country and relates strongly to ongoing tensions. My
findings therefore cannot be generalized to explain the situation of many other coca leaf farmers
living in other parts of Colombia with a very different situation regarding their property rights to land.
Thesis structure
Following this introduction, I present the method and data I used to answer my three research
questions. I draw on 87 fieldwork interviews conducted with farmers in several former coca leaf
farming communities in south west Colombia between 2017 and 2018 where land titling has been
used by the state to dissuade them from working in the illegal trade. They focus on the three areas of
ignorance I mention above: informal vs formal property rights to land; extra-legal vs state governance
in their community; and gender norms and economic opportunities during and after the coca leaf
bonanza.
In the first part of the thesis, I investigate whether the use of land titling for AD introduced by the
Colombian state in former coca leaf farming communities, did incentivise farmers to stop growing
illegal crops. To begin with, I examine policymakers’ rationale for using land titling as a tool to
support farmers who have substituted their illegal drug crops in Colombia. I assess the symbolic value
of land titles in the context of Colombia’s history of land ownership. I review the literature on
property rights to develop four theoretical mechanisms by which land titling could conceivably
support AD. I then devise a working hypothesis based on the most relevant mechanism: that land titles
offered farmers increased access to credit, thereby helping them to rely less on illegal drug crops to
earn a living. I then present my findings. Most farmers told me they already had informal property
rights. This meant that their existing strong de facto land tenure gave them access to some credit.
Although they could access bigger loans with land titles, their land productivity in legal alternatives
58
had not improved as a result. The value of their land was low, so many farmers were risk averse, and
therefore cautious of taking out larger loans, or found they were unable to turn them to profit. Others
did not know how to handle big loans. Some farmers continued working in the illegal trade elsewhere,
despite being given a land title. And sometimes they did so to pay off loans they had accessed using
their land titles. Finally, I argue that land titling was popular for its symbolic value; but, contrary to its
intentions, this policy can displace coca leaf production to other areas and may conceivably reduce
farmers’ land tenure security.
In the second part of this thesis, I review the literature on extra-legal governance and present a
contextual overview of two major extra-legal groups that emerged in Colombia to govern the illegal
drug trade. I offer a working hypothesis for how to analyse whether AD succeeded in three
communities, where coca leaf had once abounded, due to official state intervention or to changes in
the behaviour of extra-legal groups. I go on to present my findings based on examination of the
experiences of a sub-sample of 32 farmers I interviewed who were living in three communities in
Colombia where coca leaf used to be abundant and profitable. I end by discussing the role of extra-
legal governance in AD success. And I offer a novel explanation for why some communities, where
coca leaf was once highly profitable, decided to abandon the illegal trade.
Conventional theories attribute communities’ switch to the legal trade to a successful mix of state
deterrence strategies and positive incentives. However, from the point of view of people living in two
of the three communities I studied, it was clear that this was not what happened there. Farmers in
these communities told me that the armed group that had been in charge had become increasingly
violent when another one entered the scene. This reflects how the extra-legal group in charge ceased
to operate as a stationary bandit when it faced competition from incoming rivals. The rivalry between
the groups meant the behaviour of those that had once provided effective governance for the illegal
trade, became like that of roving bandits, seeking only to extract immediate profit, even at the cost of
the illegal trade. Their changed behaviour meant that the success of the illegal trade had been
compromised before the state became involved in these communities. Farmers thus found it difficult
to profit from coca leaf farming even before the state intervened. I argue that the state “succeeded” in
59
these communities, not because it provided superior governance to the stationary bandits, but because
it filled the governance vacuum left behind by a stationary bandit (Olson 2000; 1993a).
In the third part of this thesis, I review the literature on how gender norms influence other institutions
linked to economic development, and how policymakers have so far addressed gender equality in
contexts of AD. I present the themes I observed in my analysis of the experiences of women living in
AD contexts in Colombia. I then draw on desk research and a sub-sample of 28 interviews with
women, 49 interviews with men, and 5 interviews with heterosexual couples, to examine the
experiences of women living in these communities, during and after the coca leaf bonanza. Most
women who were still living in these places told me that when their communities stopped farming
coca leaf, they had experienced a greater decline in the economic opportunities that were available to
them, compared to their male counterparts. The illegal trade had generated a wealth of paid jobs for
women mainly, but not exclusively, in cooking and cleaning. As a result, women had been able to
challenge conventional gender role attitudes in a variety of ways: they were able to work as more
equal business partners with men, they could branch out into better paid and non-domestic work, and
they were less financially dependent on men. But when AD succeeded and the local coca leaf
economy contracted, gender norms that assigned women a greater burden of unpaid work, by
specializing them in housework and subsistence farming, were reinforced. As employment
opportunities shifted away from services and towards agriculture, gender norms created additional
barriers for women to compete with men for these jobs, despite there being very similar levels of
education and respect for their property rights to land between the genders. Women in these
communities had less time available for paid work and were less flexible to work away from home,
which was what some short-term farming jobs required. In addition, the physical demand, and the
form of payment in cash crop farming, meant that men had an advantage for earning more in these
jobs. I end Part III with a discussion where I argue that these women’s economic security was
adversely affected by AD, that addressing this implication is essential for improving AD policy
effectiveness.
60
Method and Data
Introduction
This chapter will describe the methods and data I used to answer my three research questions. It will
present how the research locations and the interviewees were selected, the nature of the interviews
and the observational component of the study. It will end by discussing the strengths and limitations
of this type of research.
The three research questions of this study are:
1. How is land titling conducive to the replacement of coca leaf with legal alternatives?
The answer to this question presents a valuable evaluation of land titling as a policy to
encourage illegal drug crop substitution by increasing farmers’ access to credit.
2. How do extra-legal groups become unable to govern illegal drug farming despite having
an economic incentive to do so?
The answer to this question offers an analysis of extra-legal group behaviour which may be
correlated with AD success and raises important questions about the limited role of the state.
3. How are women’s economic opportunities and incentives affected in contexts of
successful Alternative Development?
The answer to this question opens a discussion about the significant role that gender norms
play on women’s economic opportunities in some places where AD has worked in Colombia.
I collected evidence from 35 elite interviews with policymakers and 87 in-depth interviews with
farmers in Colombia to answer these questions and applied thematic analysis on them to generalize
from a pattern or trend. I also noted my direct observations of the different environments the
interviewees lived in, including the social and the geographical features of the community, such as
quality of soil, climate, topography and road proximity, and the history of drug production and
eradication in the community. These observations were then triangulated with different sources of
information from desk-based research and informal conversations I had with other people working in
61
the communities including drivers, shopkeepers and local experts in coca leaf eradication,
agroforestry, and land titling.
I picked the specific rural communities based on where coca leaf had officially been substituted and
where land titling had been offered to incentivise farmers to stop growing coca leaf crops. My
additional selection criteria were that they had reliable gatekeepers and were accessible places that
offered safe research conditions for both me and my interviewees. 87% of the farmers I interviewed
had been coca leaf farmers, either growing their own crops on land they rented or owned or working
on a coca leaf plantation as coca leaf pickers, known as raspachines. The rest of the farmers
interviewed had worked around coca leaf production either by growing food crops for coca leaf
farmers, often at inflated prices, or by working exclusively in domestic services for coca leaf farmers,
which was more often the case of women. 80% of the farmers I interviewed who had farmed coca
leaf, had substituted their crops more than three years before the study. About a year after doing so, in
2015, the state had introduced land titling as part of a wider AD package that was offered to them.
The main objective of the interviews was to work out what farmers believed had influenced their
decision to substitute coca leaf crops with legal alternatives and how their lives had changed since. In
cases where they had not been direct coca leaf producers, their views on why their community had
opted to stop producing coca leaf was the object of interest. Although the interviews were initially
designed to primarily analyse the role of land titling, they included a large unstructured component
that allowed farmers to express their views on other factors they believed had been important in
influencing their decision to stop farming coca leaf in their community. This gave me the opportunity
to identify two other themes, besides land titling, that add to our current understanding of why AD
succeeds in some places and possibly fails to in others: changes in extra-legal governance, and effects
on women’s economic opportunities. As these themes became increasingly evident, I was then able to
re-analyse all the data I collected from the start to provide an answer to my other two research
questions.
62
For the first research question, I thematically analysed 87 of the 90 interviews I conducted as I
dismissed a few interviewees that showed signs of incoherence or mental illness. The interviewees
included people who still had not acquired a land title (43%) or were in the process of applying for
one (31%), so it was the prospect of land titles on farmers’ behaviour, which I analysed. For those
who had already acquired a land title (26%), it was the acquisition of land titles on their attitude to
returning to coca leaf production, that I assessed. I then examined the data based on my working
hypothesis.
For the second research question, I analysed the themes on extra-legal governance that had emerged
from the elite interviews, and cross analysed them with the fieldwork interviews and the notes and
observations I made during my stay in the different research locations. I selected a sub-sample of 32
interviews based on farmers living in three communities with shared characteristics. These were the
communities with the same extra-legal group in charge, and with conditions that were highly
favourable for coca leaf. The sample was comprised of 11 interviews in Santiago de la Selva, 9 in La
Carmelita and 12 in Valle del Guamuez. I then clustered the data at the community level and
organized it in summative narratives for each community. Finally, I examined this sub-sample based
on my working hypothesis.
When I set out to explore the theme of land titling, I came across recurring references to the hardship
that women were facing in these communities. I only had one chance to collect data, however the
interview style allowed me to gather data on gender issues interviewees found important and relevant.
Therefore, this is what I focused on for the third research question. I re-analysed the unstructured part
of the original set of interviews I had conducted for the first research question. These included 28
interviews with women, 49 interviews with men and 5 interviews with heterosexual couples. About
85% of the women interviewed cohabited in heterosexual couples or were married. The rest were
either widowed, separated or single. For this analysis, I coded the transcripts on the theme of gender
norms and the impact that the contraction of the local coca leaf economy had on women. I cross-
checked my findings with desk research.
63
Selection of research locations
Colombia is the largest producer of coca leaf and one of the few states where land titling has formed
part of its AD efforts. The experiences of farmers in former coca leaf producing communities in
Colombia thus offer a unique opportunity to deepen our understanding of how informal institutions,
including property rights to land, and extra-legal governance, may have influenced their decision to
abandon coca leaf production. They also shed light on the situation of women after the local coca leaf
economy subsided in these communities. This helps to raise new questions about why most
communities continually decide against substituting their illegal drug crops.
The aim of this study was to observe the robustness, degree, and commitment levels of farmers to
these substitution efforts; to ascertain how land titling and extra-legal governance affected farmers’
land use behaviour; and to observe gender differences in how farmers changed their economic
behaviour after the coca leaf bonanza. For these purposes, the ideal research would have been to visit
many former and current coca leaf producing communities at random to compare the reasons why
some farmers said they had opted to stop and others to continue growing illegal crops. This would
have provided variation on the dependent variable to better understand the significance of different
independent variables influencing farmers’ decision-making. This is because case selection based on
the independent variable of interest makes the causal mechanism, in this case of land titles or extra-
legal groups influencing coca leaf production, more evident (Seawright 2016, 85–97; G. King,
Keohane, and Verba 1994, 109; Mahoney 2007, 129).
However, there were ethical, practical, time and budget constraints which impeded this type of
research. Firstly, land titling was being offered in communities that had already stopped growing coca
leaf. Secondly, communities where most farmers were visibly dependent on illegal drug crops on their
land, are known to be protected by extra-legal groups. Their presence would very likely have
prevented most farmers from speaking openly about their experiences, besides presenting a personal
security risk to me. Research in these places would have been unable to meet basic ethical standards
and the response bias may have significantly skewed the results. So, I opted for remarkable
64
communities that were accessible to me, which some might refer to as a convenience sample. A
degree of selection bias was the trade-off for avoiding hostile conditions for conducting this research
and ensuring something could still be learnt (Mahoney 2007, 129; Achen and Snidal 1989; G. King,
Keohane, and Verba 1994, 130; Geddes 2003; Stake 2005, 451).
Besides this, as is often the case in qualitative and mixed-methods research projects, there were
additional practical considerations and logistical reasons for selecting the research locations (Koivu
and Hinze 2017). The level of risk I undertook during my research was influenced by my personal
circumstances. I am a self-funded student and a new mother of a young child. These constraints meant
that I took a more cautious approach to my fieldwork.
Given that the emphasis of qualitative research is on understanding peoples’ lives in context, it is
imperative to narrow the scope of a study so that it is not too broad (N. King, Horrocks, and Brooks
2018). Therefore, I focused on smaller, more homogenous communities, which I selected based on a
pilot enquiry I carried out in advance of the data collection. During the pilot enquiry, which lasted one
month, I carried out elite interviews with 35 people with expert knowledge of illegal drug production
and AD in Colombia. I selected these key informants based on desk-based research. They included
academics, policymakers, and public officials (see Appendix III). These key informants provided me
with an understanding of the issues I was interested in, from the point of view of those who decide on
AD policy. This was important, as it helped to highlight any discrepancies between why AD is
believed to be successful by policymakers, compared to the reasons that people at grassroots level
give. They also suggested places where coca leaf had already been substituted, where they perceived
extra-legal group presence to be absent or dormant, and where land titles had been offered by the state
as part of AD efforts. I followed up these suggestions by contacting local organizations that could
facilitate my research in these places in the following months. The remaining fieldwork data
collection period lasted three months.
Colombia is divided into 32 departments that are divided into municipalities, a bit like counties in
other countries. And these municipalities are again sub-divided into veredas (in rural areas) and
barrios (in urban areas), which are perhaps best translated as boroughs.
65
Figure 3 Location of municipalities included in the study in relation to general coca leaf farming in Colombia 2017 (Source: Colombian Observatory of Organized Crime)
The United Nations Office for Drugs and Crime (UNODC), and ANT (Agencia Nacional de Tierras -
the National Land Agency), were able to offer safe research conditions and reliable access to specific
veredas, in the departments of Putumayo, Nariño and Cauca (see Figure 3). Likewise,
COOPERCAMBIO, a local cooperative with links to the Pontificia Universidad Javeriana, offered to
facilitate my access to a specific vereda in Caquetá. A local think tank called Fundación Ideas para la
Paz (FIP) offered safe research conditions in additional communities in Antioquia, but funding and
time constraints prevented them from being included in this study.
The dots in Figure 3 indicate the general location of the veredas that were included in this study. They
were in the South West of Colombia, in the municipalities of Valle del Guamuez, and Puerto Asis, in
the Putumayo department; of Los Andes and Linares, in the Nariño department; of Balboa,
Cauca Department
Nariño Department
Putumayo Department
Caquetá Department
Rosas
Valle del GuamuezPuerto Asis
Valparaiso
Balboa
MercaderesLos Andes
Linares
66
Mercaderes and Rosas, in the Cauca department; and of Valparaiso in the Caquetá department. There
was an additional interview with an individual from a vereda in Puerto Guzman municipality in
Putumayo, although the interview took place in Puerto Asis.
At the time of this study, land titling as a component of AD had been introduced in 123 veredas in 11
municipalities (see Table 1).
Table 1 Veredas included in this study in relation to the 2015 Formalizar Para Sustituir AD programme
Department Municipality Veredas targeted in 2015 AD programme
Veredas included in this study
ANTIOQUIA CAUCASIA 52 0
CAUCA BALBOA 8 2
MERCADERES 10 3 ROSAS 7 2
PUTUMAYO
ORITO 17 0 VALLE DEL GUAMUEZ 22 7
VILLAGARZON 14 0 PUERTO GUZMAN 0 1
PUERTO ASIS 0 3
CAUCA BUENOS AIRES 2 0
CALDONO 1 0 MIRANDA 1 0
NARIÑO LOS ANDES 19 4 LINARES 22 3
CAQUETA VALPARAISO 0 1
This research included interviews with farmers living in 19 of these veredas and in 6 of these
municipalities. Farmers’ experiences in these places provided most of the insight into the potential
influence of land titling on their recent decision to stop or continue farming illegal crops. In addition,
the study included interviews with farmers living in five more veredas where coca leaf had previously
dominated the local economy and where land titling had also been introduced by the state but under a
different programme to the Formalizar Para Sustituir AD programme. Of these, three veredas were in
Puerto Asis where land titling was officially being introduced as part of the Espacios Territoriales de
Capacitación y Reincorporación programme, which was aimed at facilitating the local development
of the areas close to training and reintegration camps for FARC ex-combatants (Colombian
Presidency 2017). One vereda was in Puerto Guzman, where land titles had recently been introduced
as part of a targeted agrarian reform and land restitution effort from the state to support people who
had been displaced by the violence (Colombia Government 2015). These were programmes that ran in
67
parallel to the Formalizar Para Sustituir programme so the experiences of farmers living in these
areas was as recent as the rest. The difference is that families had not specifically been offered land
titles at the time of substituting their illegal crops but had been offered other AD benefits with land
titles offered afterwards simply as an optional follow-on subsidy.
The experiences of farmers in these veredas gave this study additional insight into the rationale for
substituting illegal crops, other than because of land titling, as well as for opting to return or not to
growing coca leaf. The remaining vereda was in Valparaiso in Caquetá, where land titles had been
introduced three decades ago by the state land agency called INCORA at the time. Farmers’
experiences in this vereda gave additional insight into the effect of land titling on their behaviour in a
longer period than two years.
The veredas included in this study were not ‘average’ or representative of coca leaf farming
communities across the country. Firstly, they were exceptional from the norm in that they had
received some AD subsidies. In 2016 there were an estimated 106,900 households involved in illegal
drug crop production and 17,970 households had received AD state subsidies in exchange for
voluntarily substituting their coca leaf crops. Recipients included households that had already stopped
farming coca leaf, so it was less than 16.8% of those farming illegal crops that had received AD
benefits.
Secondly, these veredas were unusual for discontinuing coca leaf farming. They had stopped growing
coca leaf voluntarily, unlike most other coca leaf producing communities in the country. Of the
171,000 hectares of coca leaf reported in 2018, only 29,393 had been eradicated voluntarily by
January 2019 (UNODC 2017a). This means that the farmers interviewed were living within 17% of
the land area that had formerly been used for coca leaf crops.
Thirdly, these veredas were unique in having had land titling introduced by the state. At the time of
this study, the state had only given 755 households in former drug producing areas land titles since
2015, and it aimed to give 3000 more in 2017 (UNODC 2017a, 20). This meant that only about 4.2%
of those households who were in receipt of AD benefits had been granted land titles.
68
And finally, these veredas were in municipalities where coca leaf was still being grown in abundance
and where there was looming extra-legal group presence. The arrows in Figure 4 point to the areas
where coca production was currently most concentrated: Nariño, Putumayo, and Norte de Santander,
on the border with Venezuela.
Figure 4 Coca leaf farming density in Colombia 2017 (Source: Colombia Reports Data, September 24, 2018)
69
Figure 5 Coca leaf and extra-legal group presence (Source: Colombian Observatory of Organized Crime, 2018)
The black star markings in Figure 5 show the areas where there was still extra-legal group presence in
nearby boroughs at the time of this study. So, farmers’ choice to substitute their illegal crops in these
veredas made them anomalies among their neighbouring communities.
All this meant the research locations were remarkable, compared to the wider population of drug
producing veredas. However, exceptional cases can illustrate well a unique phenomenon (Small 2009,
18). And these exceptional veredas made it possible to safely observe the phenomenon of illegal crop
substitution in the context of the prospect of, or provision of, land titles. They are extremely useful in
that they are the only empirical illustration of how the process of land titling is connected, if at all, to
AD efforts in Colombia. In addition, they offer the opportunity to observe in detail the decision-
making of farmers who have opted to substitute coca leaf at home and compare their explanations of
this phenomenon with the more generic explanations of policymakers for AD success.
It should be noted that this data was not suitable for comparing the influence of land titling against the
influence of other factors, such as extra-legal groups, on farmer’s choice of land use. This is a crucial
point because, although land titles may have played a role in illegal crop substitution, this study
cannot discern its relative weight compared to other factors. For example, the data suggests that the
behaviour of extra-legal group also played a role in farmers’ decision-making, but because there are
70
no veredas where land titling was not offered, it remains a matter of my own subjective interpretation
how influential this was.
On the other hand, the veredas displayed a range of climatic conditions including altitude,
temperature, humidity levels, and soil quality. The veredas located in Putumayo and Caquetá
departments exhibited warmer, flatter and more humid jungle conditions, while those in Nariño and
Cauca departments had cooler, more mountainous and drier conditions. This variation enhanced the
comparative analysis of the factors that were considered to have motivated farmers to disengage with
the illegal drug economy, relative to land titling.
All the research locations mentioned above were used in the analysis for the first and the third
research questions, that is, to study the role of land tilting in AD and the effect of AD on women’s
economic incentives. For the second research question, I selected a sub-sample of three communities
that had common characteristics that would make them comparable in my analysis i.e., they were
those where coca leaf had once grown abundantly prior to AD success. This meant they would answer
my research question more accurately because they would highlight how extra-legal governance
ceased to protect the profitability of the illegal drug trade.
The small sample size of this study should not be an issue given that this study does not seek to
remove bias in the same way that quantitative methods do. The point of case studies is not to be
representative but to carry out an in-depth description of a social phenomenon that can be traced over
time and which takes context into account (Gerring 2007, 19, 20; Yin 2002). As Small argues, there is
never an average community, and rare situations can nonetheless make meaningful contributions to
our knowledge of mechanisms (Small 2009).
Selection of interviewees
The sampling method I used to determine who to interview in each vereda was criterion-based
sampling, also known as purposive sampling. By using this sampling method, I “selected on the basis
of known characteristics, which might be socio-demographic or might relate to factors such as
experience, behaviour, roles” relevant to my study (Ritchie et al. 2013, 144). This ensured that the
71
data I used was that of people who had previously worked with coca leaf and / or had received AD
subsidies. This included individuals who had their illegal crops eradicated, either manually or through
aerial aspersion, as well as those who uprooted them voluntarily, although not all needed to have
owned the land where they had produced or processed illegal drug crops. It also included farmers who
had access to land where coca leaf (or opium poppy) used to be grown, who had applied for or had
already received AD benefits such as a land title. It further included individuals who had stopped
working for illegal crop farmers because they were displaced from the land where they used to do so,
or because they went out of business. This included coca leaf and opium poppy crop growers/
farmers, cocaine processors, and individuals who worked for coca leaf farmers, such as domestic
workers and coca leaf pickers working as day labourers (referred to as raspachines). These broad
criteria for interviewees helped inform me of various perspectives on the issues of interest, even if the
sample is not generalizable.
There were additional criteria such as that the individuals interviewed could feel comfortable speaking
honestly about any potential intentions to replant their illegal drug crops afresh or prune those that
were not properly eradicated or continue with the legal use of their land or return to illegal drug
production in some other form. All of them had to be over 18 years old, with no known mental
disability that prevented them from communicating with me, and able to speak Spanish so that I could
interview them directly without the need for translation.
Insiders, such as local officials working for the organizations that facilitated the data collection
introduced me to allies in the community for the initial round of interviews in each research location.
They vouched for me and helped me identify research participants. This facilitated my access to
people who would otherwise have been difficult to approach, even if it reduced my initial control over
who participated. I then identified additional interviewees independently through cold calling on
people’s front doors. I would introduce myself and ask them an initial set of questions to check if they
met the criteria of interviewees I wanted to include in my study. This allowed me to remain open to
discovery and include interviewees who met the criteria but who did not necessarily have a direct
relationship with the organizations that were working in the community. Some interviewees then
72
suggested other people to include in the study. Like with snowball sampling, interviewee chain-
referral from known acquaintances can result in more honest conversations because people feel they
can trust the interviewer more if they know other people who have also volunteered information and
recommended them to participate in the study without prejudice (Small 2009, 14). To minimise the
risk of network over-representation, I used various entry points for identifying interviewees. I also
selected individuals in similar and different contexts to make ongoing meaningful cross-case
comparisons and to minimise the risk of endogeneity and of anecdotalism i.e. deriving conclusions
only from a few well-chosen examples (Silverman 2005).
The nuances in how land titles were understood by different members of the community made it
difficult to include all farmers that I approached in the study. I excluded people who had no notion of
a what legal land title was. Their experience is therefore not represented in this study.
As the study progressed, each interview shaped the questions for the next one, as well as the broader
research questions. The sample size could not be determined prior to the study because I did not know
how many cases it would require for the theory to emerge from the data (Rudestam and Newton 2007,
108). I considered the study to have reached saturation when the new data collected from other
sources confirmed previous findings (Morse 1998), when emerging theories were validated, and the
depth and breadth of the phenomenon was reflected in the data (Yin 2002; Small 2009). This meant
that a different number of farmers were interviewed in each vereda (see Table 2).
Table 2 Number of interviewees in each vereda
Department Municipality Vereda No of Interviewees Caquetá Valparaiso Santiago de la Selva 11 Cauca Mercaderes Carboneros 3 Cauca Mercaderes Esperanzas del Mayo 6 Cauca Mercaderes Sombrerillo 3 Cauca Rosas El Porvenir 6 Cauca Rosas La Soledad 4 Cauca Balboa La Palma 7 Cauca Balboa La Villa 6 Nariño Los Andes San Vicente 1 Nariño Los Andes El Alto 4 Nariño Los Andes Huilque 2 Nariño Los Andes Sotomayor 3
73
Nariño Linares La Tola 5 Nariño Linares Linares 2 Nariño Linares Llano Grande Alto 1 Putumayo Puerto Asis La Carmelita 5 Putumayo Puerto Asis La Pradera 4 Putumayo Puerto Asis Agua Negra 1 Putumayo Puerto Guzman Santa Lucia 1 Putumayo Valle del Guamuez Alto Palmira 2 Putumayo Valle del Guamuez Brisas del Palmar 3 Putumayo Valle del Guamuez EL Cairo 1 Putumayo Valle del Guamuez El Placer 1 Putumayo Valle del Guamuez Los Guaduales 2 Putumayo Valle del Guamuez Miravalles 2 Putumayo Valle del Guamuez El Oasis 1
Having said this, total saturation is difficult to complete because new questions always emerge from
new data that is collected (Josselson and Lieblich 2003). The one-week timeframe that was available
for data collection in each department limited the number of veredas that could be included in the
study. Sometimes there was just one individual available to give an account of their own vereda,
sometimes there were several.
In total, besides the 35 key informants interviewed in the pilot enquiry, 90 individuals were
interviewed in the research locations and 87 of them were included in the analysis (see Appendix II).
This was because in two occasions there were doubts about the interviewees’ mental coherence and so
they were excluded from the analysis. In addition, one interview was with a bank manager, which did
not count as a case because she was not a farmer (but she offered valuable information on access to
credit to local farmers). There were eight interviewees who gave joint interviews, in couples, because
they felt more comfortable that way. The rest were one-to-one interviews all held in private so that
they felt at greater ease to talk. 36 out of the 90 cases were women and the median age of interviewees
was 48.
Rationale for qualitative research
The aim of this study was to understand the world from the point of view of individual experiences
and to produce detailed descriptions of farmers’ motivation for substituting their illegal crops.
74
Therefore, qualitative analysis was well suited to this research and for formulating further questions to
be explored and developed in the research process. It helped me to build a theory of the chain of
causality in farmers’ efforts to substitute coca leaf crops, whether and how land titling and extra-legal
groups’ behaviour influenced them, and how women had been financially affected in the process. And
it considered farmers’ own local meanings and views on what mattered in the context of former drug
producing communities, which helped to avoid measurement error (Mahoney 2007, 123; Stake 2005,
450).
In addition to assessing the influence of land titles from the outset, qualitative research made it
possible to develop ideas that came up during the research and discover other hidden and abstract
variables that were not evident from the start (Small 2009). For instance, the role that changes in
extra-legal governance played in influencing farmers’ behaviour was a theme that emerged from the
pilot enquiry and which I was able to incorporate into the study. Similarly, the effect of coca leaf
production and substitution on women’s economic opportunities was developed as a theme during the
fieldwork interviews because it became more evident as the unstructured part of the interviews
progressed when people could talk about issues without any prompting. Both research areas would
otherwise not have been easily identifiable with quantitative research. The semi-structured nature of
the interviews gave interviewees the opportunity to talk openly about their story and share their points
of view on what was important with no time restriction. This enabled me to then reanalyse the earlier
interviews in the light of themes that I realised were important later. As a result, this study provides
useful information for further empirical work on topics such as the effectiveness of land titling as a
component of AD, changes in the behaviour of extra-legal groups that might facilitate AD, and the
impact of AD on women’s economic opportunities. Future surveys on these topics can be more
accurately designed with measurable and relevant indicators based on the variables that farmers
themselves considered important in influencing their behaviour.
Most of the methodological critiques of qualitative studies centre on how generalizable they are. This
includes whether the sample is representative, whether interview data can indicate generalizable
patterns of behaviour and whether the conclusions reached from one case study can be applied to
75
more cases. This study did not aim to be statistically representative, as small-N research can never be
so (Small 2009, 8). The conclusions of this study are only generalizable to places that coincide with
the exceptional circumstances of the people interviewed. In qualitative research, everything is
ultimately context-dependent and the generalizability of research findings to other settings is
ultimately “the task of the reader rather than the author” (Rudestam and Newton 2007, 113).
However, sometimes “we gain better understanding of the whole by focusing on a key part” (Gerring
2007, 1, 76). Even if not widely generalizable, this study is reliable and offers strong internal validity
because it is focused on the clear objective of identifying mechanisms and took a great deal of care to
collect and interpret data from specific contexts to present a valid interpretation of the phenomenon
(Rudestam and Newton 2007; Mitchell 1983; Mahoney 2007, 128).
Not all the stories and accounts heard were interpreted as true. Farmers’ experiences were interpreted
as true if interviewees brought up ideas spontaneously without being led, and what they said was
clearly and logically argued, was consistent with the rest of the discussion, and with what other
interviewees were saying. This is what I took as “evidence” to support a point. For those adhering to a
more objectivist approach, they will inevitably prefer quantitative data to support the interpretation of
the assertions farmers made as factual. However, what farmers believed to be ‘true’ is also a form of
reality which influences how they behave. This research is subjectivist in that it parts from an
understanding that there is no unique and objective truth (Rossman and Rallis 2003, 32–34).
There were also reasons not to use a mixed-methods approach. With regards to my first research
question, it was very early days to carry out a quantitative assessment of the influence of land titling
on AD for two reasons. Firstly, land titling was only beginning to be introduced in former coca leaf
producing communities. It had been rolled out for about two years and was still being introduced in
all the research locations, except for one where land titles had been introduced considerably earlier
(back in the 1970s). In most research locations, it was being used as an explicit tool for AD because it
was granted on condition of communities’ illegal crops substitution. In three veredas included in the
study, it was being used as an implicit tool for AD because the official objective of its use there was to
boost the local development of the places where armed groups had recently ceased to control.
76
Secondly, farmers’ understanding of land titles was rudimentary and most were only beginning to
apply for increased access to credit. The interviews revealed that farmers did not understand land
titles as a binary concept, but on a spectrum of land tenure formality. Their local meaning ranged from
legal documents that formally qualify land plots as belonging to an individual, to partial land titles
which were outdated escrituras or those lacking a registro i.e., the land registration document (a type
of ‘Torrens title’ that is the record of land plot) to the informal documento compra (or carta) venta,
which were the documents acknowledging the informal purchase of a land plot, a record of informal
agreement by the community.
Units of analysis
Inferential, or case study, logic was well suited to this research because my questions were mainly
about mechanisms, in this case how land titles and armed groups (and/ or other variables), influenced
farmers’ decisions to substitute their illegal crops, and how the latter influenced women’s economic
opportunities (Small 2009, 25). Unlike surveys or field experiments, case studies do not separate the
phenomenon from the context, and they consider different independent variables to better compare the
phenomenon.
My aim, like that of case study analysis, was ultimately to analyse how reality is understood from
within a case (Stake 2005, 452). For the first and third research questions I used the individual farmer
‘a case’ or the unit of analysis. Each farmer selected for interview counted as a new case that would
help to unravel a sequence of events within a context (Yin 2002; Mahoney 2007). The logic behind
the individual as my unit of analysis, is that I part from a rational choice position that stresses
individual incentives over social structures. In short, I assume that the market functions as a sum of
the behaviour at the individual level. Therefore, whether a legal or illegal market dominates in a
community, will depend on all the individual transactions that function outside the law. And in this
case, I will observe whether and how land titles reduce transaction costs for everyone at the individual
level. Similarly, I will observe how AD affect women’s economic opportunities at the individual
level.
77
For the second research question I clustered individual interviews into communities, so that each
community was my unit of analysis. By community, I mean either one cohesive vereda i.e., Santiago
de la Selva and La Carmelita, or in the case of Valle del Guamuez, a group of veredas in a single area,
namely Alto Palmira, Brisas del Palmar, El Cairo, El Placer, Los Guaduales, Miravalles and El Oasis.
See Figure C in section 2.4 for a map of their location in relation to Valle del Guamuez.
As Gerring (2007) describes, a case is a “spatially delimited phenomenon…[that] has identifiable
boundaries and comprises the primary object of an inference” (Gerring 2007, 19). The reason why I
use different levels in my unit of analysis for this research specific question is that the spatial
boundaries of the communities I seek to study differ. In the cases of Santiago de la Selva and La
Carmelita, they are long-standing communities, where the individuals interviewed are connected to
the village and feel a sense of belonging to the vereda.
I clustered the farmers I interviewed in the seven veredas in the Valle del Guamuez as one case
because they share a context that is similar enough to demarcate identifiable boundaries. Although
they officially reside in different veredas, their plots of land are remote and disperse, so they do not
live in a close-knit community linked to a village, as did the farmers I interviewed in La Carmelita
and Santiago de la Selva. Therefore, for the purpose of this research question, the vereda jurisdiction,
or political unit, is not a relevant description of their context. Their context is better described by the
attributes of the environment they live in. Firstly, they have similar levels of geographic remoteness
and infrastructure marginalization. They are relatively close to a main road and are a 30 – 45 minute
drive away from the town, La Hormiga, a town in Valle del Guamuez which owes its urban
development to the revenue of the illegal drug trade and to its proximity to Ecuador which facilitates
the smuggling of goods and the “cheap dollar” (Ramirez 2011, 65). This makes these places unique in
a similar way, compared to the rest of the municipality, which comprises 80 veredas in total.
Secondly, they have a similar climate and soil characteristics as they are all located within 19 km
from each other. These attributes differentiate them as a unit from which one can make meaningful
comparisons with the other cases of La Carmelita and Santiago de la Selva. And finally, their
experiences concerning the timing of coca leaf production, AD and the behaviour of FARC coincide
78
in the details, as did their reticence to talk about the past, compared to the other two cases. All this
made their joint experience a phenomenon of interest worthy of analysis.
Interview methodology
Semi-structured in-depth interviews were best suited to this research because they could be targeted
on the first research question as well as provide original and insightful data which helped shape my
existing knowledge on the second and third research questions. As King, Horrocks & Brooks (2018)
explain, qualitative interviews are by their nature interpretive and cannot establish generalized
patterns of behaviour. They can, however, focus on meaning and experience related to a specific
group. As such, they focus on perceptions of causality from the perspective of research participants, in
a way that a structured interview or a survey would be unable to do.
The average duration of the interviews was between half an hour and 45 minutes, although some
interviews were much longer. The flexible component of the interview enabled a range of
unanticipated perspectives and insights to emerge, and allowed for interesting digressions, better
capturing the voices of the people interviewed. This variation also prevented me from fitting the data
around any preconceived theory and impose my own views on the interviewee (Barbour 2013, 120).
And for the more structured part of the interview, the interview script ensured that important topics
were not missed as probes and prompts allowed people to clarify and repeat their answers (King,
Horrocks, and Brooks 2018). It also helped to address under/over communication during interviews,
when interviewees shared too much or too little information.
One-on-one interviews provided farmers privacy and confidentiality so that they would respond more
honestly. I also assured them of anonymity. This was especially important for talking about sensitive
topics such as land ownership, extra-legal groups, household issues and illegal drug crop production
and substitution efforts. Many farmers had felt intimidated by extra-legal groups in the past for talking
about them with outsiders and were reticent to own up about their current use of land and sources of
income. It is likely many of them may have had an unconscious fear of being “grassed” by others for
79
things they expressed. Therefore, I built trust with the interviewees, as I discuss below (in the
minimizing interview bias section).
Other qualitative methodologies would not have offered the same “safe space” for farmers to talk
openly. For instance, focus groups would not have guaranteed confidentiality for individuals to
express themselves openly without feeling vulnerable in front of each other (Krueger 1994). A focus
group is about getting people to talk to each other and listen to their conversation. In the interaction of
focus groups, some interviewees could be influenced by other peoples’ views and select what to share
based on this (Whynn 1999). Therefore, private interviews enabled me to build trust more effectively
with interviewees and reassure them that what they said would be anonymised and not shared with
anyone else in the community.
Interview protocol
There is a trade-off between questions that are too broad and unanswerable and too narrow and not
very useful (King, Horrocks, and Brooks 2018). The questions asked were focused on interviewees’
experiences, behaviour, opinions, values, feelings, and knowledge. Following Rennie’s (1998) advice
on not starting interviews with a set of questions, to allow each interviewee to feel relaxed and for
categories to emerge from the data and not vice versa, at the beginning of each interview, I introduced
the study and explained that data would be anonymized. To build rapport with interviewees, to elicit
high quality responses and to encourage them to participate to the best of their ability, the interviews
began as open-ended discussions. Interviewees were usually prompted with an initial question of
discovery, inviting them to share a story about themselves, usually about how they came to live on
their land (Mishler 1991; Kvale 1996; Weiss 1994; Rubin and Rubin 2005).
I then introduced more structure in the follow-up discussion asking more objective research questions
about their background and demographic. This was to help clarify the unstructured part of the
interview. A list of predetermined questions (see Appendix I) helped to enable prompt thoughtful
answers from interviewees in their own words (Braun and Clarke 2013). The predetermined questions
were modified and asked in different order according to the flow of the interview. A mix of short and
clear questions interspersed with open-ended interval questions gave interviewees the chance to give
80
more nuanced insights (Barbour 2013, 129). Any new and relevant issues that might have arisen in the
first unstructured part of the interview were followed up with more direct questions in the second part
of the interview. The discussions kept the following overarching questions in mind:
1. What is your understanding of land ownership? How is this enforced?
2. What is your credit history and what are your credit aspirations, if any?
3. Have there been tensions in the community over time and how have they been resolved? Have
you been affected by violence and how?
4. What was your life like before and after coca leaf was farmed in the community?
5. Why did you work in coca leaf production? How do legal alternatives compare with coca leaf
production?
The aim was to open interviews with soft questions, that build up rapport, to be clear and to ask
simple questions, and not several things in one question, and to leave difficult questions to the end.
Minimizing interview bias
Interviews present a danger of bias so the strength of using the interview method is usually
determined by the skill of the researcher as an effective and flexible interviewer (Starr 2014). It is
possible that most people might have had an incentive to underestimate the amount of coca they once
produced, as those who had, in the past, owned up to the state authorities about having grown more
than four hectares of coca leaf, had been classified and penalized as drug dealers. A potential
challenge was that interviewees would find it difficult to trust me. To ensure the data gathered from
the interviews was as reliable and valid as possible, it was crucial to develop trust and rapport with
interviewees. Introducing the aim of my research clearly, listening patiently and sensitively, giving
interviewees enough time to answer (and ask) questions and using silence before prompting, helped
interviewees to not feel pressured. It was important for me to not to ask them leading questions, and to
keep a distance from the interviewees by avoiding non-verbal cues or by making assumptions about
interviewees (King, Horrocks, and Brooks 2018). Silence and non-verbal communication can also be
informative.
81
To reduce inaccuracy caused by participants’ poor recall, similar follow-up questions were asked with
different wording and triangulated this with other sources of data (e.g., additional participants,
observations, and published studies). This helped to validate the timing of events, cross-check
emerging themes, and corroborate evidence (Stake 2005). Sometimes scenario situations were used to
invite interviewees to imagine how they would behave which helped to compare responses to the
same material and consider different interpretations (Bloor and Wood 2006, 183). Interviews were
closed with “is there anything I didn’t ask or that you’d like to add?”, which helped to ensure the
interviewee had the opportunity to provide any additional information that they thought was relevant.
The interview settings were chosen so that both the interviewees and I felt safe and comfortable.
Private and quiet settings in public spaces were favoured and face-to-face over a table setting were
avoided (King, Horrocks, and Brooks 2018). And finally, the interviews were recorded with a digital
and password-protected audio recorder and backed them up with written notes, to ensure no details
were missed and I recalled information accurately.
Ethical considerations
This project’s fieldwork research was registered with the King’s College London Research Ethics
Office (Research Ethics Number is KCL Ethics Ref: HR-17/18-5300). All gatekeepers and
interviewees were informed about the nature of the study, the research protocol, and the academic
body to which it belongs, prior to being invited to participate. Even in informal or casual
conversations with me, people were made aware of my research intentions.
The ethical concerns of using interview data in this study include issues of privacy and
confidentiality. There was a risk that participants could feel uncomfortable talking about tensions in
the community. This was addressed by making them aware the interviews would be fully confidential
and that in no way would their testimony be identifiable as theirs. They were informed in advance that
their data would be protected (stored in an encrypted file in a password-protected laptop during the
study), anonymized by default (using pseudonyms so as not to be identifiable in future and giving a
very general geographical location to the communities in the thesis), stored for seven years after the
study and then destroyed. Personal data will be destroyed after one year of the thesis being completed.
82
With interviews, there is also the possibility that potential interviewees may feel pressured to
participate or misled into believing they will be rewarded with something for doing so. It was also
possible that research participants might associate the research with the gatekeeper organisations
increasing their trust, and then feel abandoned, and in some way “used” by me after the study. This
was addressed by stressing that participation was entirely optional and not linked to any type of
benefits. It was emphasised that my role as the researcher was separate from the community
organisations, and that they could not provide participants with any additional benefits that the
organisation they had accessed them through, may have provided. However, they were also made
aware that this research might be able to help improve policies that address illegal drug supply, as
academic studies can do.
All research participants were given an information sheet with information about the research and my
contact details in advance of conducting an interview and were asked to provide informed consent
before being interviewed to ensure they understood this. Often this consent was given orally because
many of the community members may have felt suspicious about signing a piece of paper, they either
could not read, or thought was a trick. It also helped keep the conversation more informal and increase
trust levels for the interview. The participants were made aware that they could withdraw from the
study up to two weeks after the interview had taken place by contacting me. They were also told that
they could stop the interview at any time.
Talking directly to farmers about their past personal illegal activities might present issues of trust
between the interviewees and me. Therefore, only farmers whom the state had already agreed not to
penalize in exchange for their voluntary illegal crop substitution were included in the study. They
were identified with the help of gatekeeper organizations (including community leaders, university
research groups, local NGOs, and state agencies) that already knew the communities well. This also
meant I was under no obligation to report any disclosures of illegal behaviour. Anyone who felt
compromised by talking openly about their past was automatically excluded from the study because
the research had to be carried out free from pressure so that the data could be trusted and not be
skewed by interviewees’ fear.
83
Observations
A small element of participant observation was used to complement the data gathered through
interviews. This helped me build a picture of the context in which farmers lived and to make
meaningful comparisons about land use behaviour at the community level. I spent one week in each of
the municipalities selected for the study and spent time in various places, in between carrying out the
interviews, such as local cafeterias, public buses, and public spaces. I recorded my own first
impressions and made notes on as much useful contextual data from the vereda as was visible.
I followed up what was significant and surprising about each vereda, and what interviewees treated as
significant and frequently brought up, by talking to other locals, UNODC topographers, agronomists,
lawyers, social workers, and officials who were familiar with the veredas. This included asking the
local agronomists hired by UNODC about the quality of land in terms of soil fertility and natural
resources; asking the lawyers hired by UNODC about the value of land and the general size of the
land plots; asking locals about the distance of the community from urban centres and their access to
road infrastructure; asking officials hired to implement AD programmes about their perceptions of the
local social capital including wealth, literacy, recent migration and homogeneity; asking UNODC
officials about the proximity of extra-legal groups and coca hotspots and drug trafficking corridors;
and asking local about the degree of state presence including schools, health and law enforcement.
These discussions provided me with more specialised information and helped me to check for
distortions and devise more trustworthy interpretations of the interviews.
The information I gathered this way was handwritten daily in a project journal. Recorded in it, were
also my impressions, reactions and other significant events that occurred during the day. It was thus a
reflective document in which observations, analytic comments and interpretive comments were all
stored together (Bazeley and Jackson 2013).
Data analysis
Analysis is about identifying larger significance and meaning, finding connections between events
and segmenting and coding data (Bazeley and Jackson 2013, 113). I used thematic analysis to create
84
themes from the data in an inductive way i.e., generalizing from a pattern or trend. This flexible
analytic method allowed themes to emerge from the content of the data instead of forcing me to fit the
data into pre-determined categories.
For the coding part there were three phases as is often the case in qualitative data analysis (Miles and
Huberman 1994; Mahoney 2007, 131). The first level of coding started during the fieldwork period.
After each interview, I sorted information into descriptive themes or common topics. In this way I
began to analyse the data simultaneously as it was being gathered (Charmaz 2005, 507; Glaser and
Strauss 1967; Strauss and Corbin 1994, 273). This helped me to inform further questions I could ask
and to identify further interviewees, until each of the themes were saturated. For example, the role of
violence in motivating some people to stop producing illegal drug crops came up early on. This led to
subsequent interviews focusing on this subject and prompting people to share their experiences of
extra-legal groups. Similarly, issues that women faced came up early in the interviews when they
were asked about life after coca leaf production, and so this led me to investigate their economic
opportunities further.
The second phase of coding took place after all the interviews were completed, and professionally
transcribed (in Spanish to avoid data being lost in translation). Each transcription (and three
interviews that only had handwritten notes because those interviewees had preferred not to be audio
recorded) was summarized in detail in English. This helped me build a sense of the whole, before
breaking down the detail within each interview. Key quotes in the summaries were translated and
abbreviated forms of what had literally been said by the interviewees. I then re-read each of these
summaries and made marginal notes on the recurring themes that crept up. I uploaded the raw
interview transcripts and over 200 pages of interview summaries onto NVivo 11 software. I re-read
the data and sorted it into a preliminary interpretive coding scheme, in what is known as open coding.
By the end of this coding phase, there were 22 themes for the elite interviews and 65 themes for the
fieldwork interviews (see Appendices IV and V for layout of coding scheme and record of coding
decisions).
85
The third stage of analysis was a rigorous checking to see if the data still fitted those themes to avoid
forcing it into categories. The aim at this stage was no longer to identify and label (open coding) but
to refine and interpret (focused coding) (Bazeley and Jackson 2013, 126; Miles and Huberman 1994).
During this stage I focused on capturing explanations, relationships, and theoretical constructs. For
example, the frequency of child labour and larger family size in the past partly explained the increased
difficulty for farmers to revert to subsistence farming after abandoning illegal drug production. The
irrelevant themes in the preliminary coding scheme were discarded. For example, the role of mining
and extraction of natural resources had appeared to be a cause of land tensions in the communities at
one stage, but there was insufficient data to justify this as a recurring theme, so this line of enquiry
was abandoned. The remaining themes were then analysed in the context of the wider literature, to see
if they answered the first research question and in what way. The hypothesis for the chain of causality
in how land titling might influence farmers’ illegal crop substitution efforts was developed.
With regards to the second research question, the vereda was the unit of analysis so I used cluster
analysis for analysis at the community level. This meant there was an additional layer of analysis to
the three phases described above. I typed-up and uploaded my observational data onto NVivo for the
veredas included in my subsample of the three communities I focused on. This data included notes on
variation in the quality of land, how marginalized the borough was from urban centres, how
homogeneous the local population was, the levels of social capital in terms of literacy and education,
how close the community was to coca leaf hotspots or drug trafficking corridors, how present the state
was in each farmer’s community and the overall behaviour of extra-legal groups. I looked for patterns
and irregularities, paradoxes, and contrasts in the summaries of the interviews corresponding to each
vereda. I then clustered my findings to produce a summary for each vereda, enriched with quotes.
And I cross examined them in the context of the academic literature in a way that explained why
extra-legal groups occasionally fail to protect an illegal trade.
Given the limited set of interviews in the small sample I used for my second research question, they
are not generalisable or representative of a broader pattern. To complement the inferences, I made and
further explore my insights on potential correlations between extra-legal groups and successful AD, I
86
therefore analysed some datasets on extra-legal group presence, coca leaf production and violence. I
secured these datasets as part of my engagement with local institutions and government in Colombia.
These include:
• Annual number of hectares of coca crops by municipality from 1999 to 2014 facilitated by
UNODC Colombia.
• Annual reports of acts of terror, threats, homicide, and extra-legal group presence by
municipality from 1993 to 2014 facilitated by CEDE Research Centre (Centro de Estudios
sobre Desarrollo Economico) at University of Los Andes.
• Annual reports of victims and displacement intensity by municipality from 1984 to 2015
collected by the Colombian Unique Register of Victims.
These datasets were cleaned and processed for consistency and then merged using the unique
identification code by municipality as coded by the Colombian National Department of Statistics
(DANE). All the analysis and data handling were performed with R programming language and
RStudio interface for R software.
I chose descriptive as opposed to statistical analysis for this quantitative data, because my original
question sought a qualitative explanation. I had not originally planned a quantitative study and data
collection for a statistical analysis. My interview sample was inductive and not randomly selected nor
designed to eliminate any biases, so it would not be robust to analyse data in that way. The aim was
simply to generate a new hypothetical mechanism for how AD succeeds in some communities without
the state being at the centre of it. And the graphical data provided a summary of some general trends
that complement my insights.
During the third stage of the formal thematic analysis for the first research question, the patterns in
gender came up as an additional theme. So, for the third research question, I re-analysed all the
interviews and created a new table that compiled all the experiences that were particular to women. I
re-organized the data into sub-themes on a word document and created a survey table for observing
the experiences of 28 women. I also sifted through the coding that emerged from previous analysis
87
with regards to property rights and education to identify any patterns in gender using text searches in
NVivo. In this analysis, I inferred a general pattern on how women’s economic opportunities and
experience of gender norms changed during and after the coca leaf bonanza.
Reflexivity
One of the consequences of the type of analysis I used in this study is that there is an inevitable degree
of reflexivity because I played an active role in determining the themes and the consequent results
(Braun and Clarke 2006; Boyatzis 1998). High quality interview material is usually co-created
between the interviewer and interviewee as there is a two-way influence between them. This means
that cultural background, age, class hierarchy, race and gender can influence what the researcher sees
and what people say in interviews. To add to this, it should be noted that no researcher can remain
rigorously detached from the research subjects (Schatz 2009; Burawoy 1998). Similarly, any field
notes the researcher takes will be exclusively theirs and are inevitably shaped by their own
perceptions of the facts.
It is thus important to reflect on one’s role as the researcher in qualitative a qualitative study. I should
note that I was part insider and part outsider in this study. On the one hand, I understood the language
and wider Colombian society, having lived in Bogota for nine years during my childhood. On the
other hand, I lacked detailed knowledge of people living in rural Colombia. I feel that the fact that I
grew up for some time in Bogota, made me more approachable to many interviewees, as I spoke
fluent Spanish with a Colombian accent, and understood their rural idioms and humour. This meant
that the interviews flowed freely, without the need for me to constantly interrupt them with questions
and clarifications. However, my physical appearance made me stand out from the local population,
which also created a degree of separateness from the community, so they were initially curious about
whether they could trust me. I sought to resolve this by not rushing the introduction part of the
interviews and giving interviewees enough time to ask me all sorts of questions before agreeing to
participate in the study. As my knowledge of rural Colombia before this study was mainly theoretical,
I believe my standpoint when analysing the data was still as an outsider.
88
It is not usual for hypotheses to be conceived separately from “the contaminating influences of the
data” (Gerring 2007, 72). Because of this risk, I took steps to reduce my own personal bias such as
cross-checking findings with primary sources including administrative records, and local people, and
secondary sources such as media clippings, news reports and field notes. In addition, I made sure the
field journal includes a detailed record of the process of the study and my thought trail so that if others
wish to replicate the study under similar conditions, they will potentially reach similar conclusions. It
has been argued that any questions about the reliability of the coding used, given the reflexivity of
coding, can be offset by the strength of the argument made and the clarity and comprehensiveness of
the evidence presented (Bazeley and Jackson 2013, 151).
My intention with this research has therefore been to be as transparent as I possibly can about the way
I have acquired and analysed the lived experiences of coca leaf farmers. It is not usual for the
experiences of people living in communities where AD has been successful to form part of how AD is
designed, and how its success and impact is explained. As the elite interviews and desk research
revealed, policymakers have preconceived ideas about the way in which AD succeeds. For instance,
that land titles work to stop people from farming coca leaf crops by expanding farmers’ access to
credit for investing in legal agriculture. Or that the state always plays the active role in persuading
communities to substitute their illegal crops. Or that efforts to minimize gender inequality in farmers’
access to education and land is the best way to support women in AD communities. This research
provides a kernel of valuable data that helps to raise questions about some preconceived ideas that
policymakers have about the role of land titles and the state in AD success, and its impact on women.
And this may further help to develop AD effectiveness. Finally, the fact that the data was gathered in
a small window of opportunity when the security conditions were right, and that not all these
communities are accessible for research today, gives additional significance to this research.
89
Part I: Property Rights to Land & Alternative Development
1.1 Abstract
How is land titling conducive to the replacement of coca leaf with legal alternatives? In the first part
of this chapter, I examine the rationale for using land titling as a policy tool to support farmers who
have substituted their illegal drug crops in Colombia. From a pilot enquiry with 35 policymakers in
Colombia I find that they justify the use land titling for Alternative Development (AD) partly for its
political symbolism, but mainly for increasing farmers’ access to credit. The assumption is that by
having more access to capital, farmers’ land productivity increases in legal alternatives. This then
makes farmers less dependent on illegal crop farming. I then present a review of the different
theoretical mechanisms by which land titling could potentially change illegal drug farmers’ behaviour
and devise a working hypothesis based on the strongest theory i.e., that land titles increase farmers’
access to capital and thereby their productivity in legal alternatives to coca leaf.
To observe if this theory holds true in the experience of communities that stopped coca leaf farming, I
draw on 87 interviews with farmers in South West Colombia. I observe how the acquisition and
prospect of land titles featured in these farmers’ credit history, plans to access bigger loans, land
productivity, and how all this influenced their decision to stop farming illegal crops. I find that
although land titling was locally popular for its symbolic value and did increase farmers’ access to
credit, many farmers preferred not to take out loans. Smaller informal loans were already available to
farmers, and farmers were very often risk averse and wary of taking out larger loans, or found they
were unable to turn them to profit. Only very few of those who did, experienced improvements in
their alternative land productivity.
My empirical exercise shows that for the areas I examined, land titling did not diminish the attraction
of growing coca leaf given its rate of return compared to other crops. And counterintuitively, it
presented the additional risk of increasing farmers’ debt and need to resort to coca leaf farming again.
Some farmers with land titles felt motivated to relocate illegal drug farming elsewhere. They saw this
as a more effective way to raise capital than through loans, or to pay back their bank loans on time.
90
I argue that in the context of these places in Colombia, even if there is value in the political
symbolism land titles were associated with, they had an insignificant effect on farmers’ behaviour.
This was because farmers already had land tenure security, and because the value of their land assets
was low. Policymakers should thus play down the supposed effect of land titles on AD efforts and be
aware that it can be conducive to the displacement of coca production elsewhere.
91
1.2 Introduction
Land titles are the legal recognition of informal land ownership. They transform de facto property
rights, which are institutions recognized and enforced by the community, into de jure property rights,
or institutions recognized and enforced by state law. De jure property rights, are regarded as formal
institutions because the state is the ultimate authority that recognizes and monitors them. If they are
not respected, there are explicit sanctions, normally delivered by the state. It is important to note that
although de facto property rights are not written law, and are thus classified as informal rights, they
are nonetheless a functioning institution and not only a norm. Norms are shared concepts of what
must, must not, or may be appropriate actions in particular types of situations e.g. littering (Ostrom
2009). While rules are recognized and monitored by an authority, norms are self-enforcing through
ostracism or simply by individuals themselves. In this respect, although de facto property rights are
not enforced by the law, they are more than norms because they are not simply patterns of behaviour
followed by most of the population. De facto property rights to land are norms as well as rules that are
considered informal because they are recognized and enforced by the community but not by state law.
In more abstract terms, a land title is “an instrument of thought, representing assets in such a way that
people’s minds can work on them to generate surplus value” (de Soto 2000, 218). According to de
Soto, land titles unlock the economic potential of a land asset. He argues that this is because formal
property has six effects: it describes and registers the potential value of the asset; it integrates isolated
data about the asset into one system; it makes the owners accountable (e.g. they can be identified and
sanctioned if they disrespect contracts and disobey the law); it makes assets convertible into other
resources; it ties citizens with the government and the private sector; and it ensures their asset’s
transaction is protected over time and space (de Soto 2000, 49–62). De Soto goes onto argue that land
titling is “the indispensable process that provides people with the tools to focus their thinking on those
aspects of their resources from which they can extract capital” (de Soto 2000, 218). This is essentially
a development of Coase’s theory that property rights reduce transaction costs and thereby increase the
value of market transactions (Coase 1937). By bringing everyone into the same contract, de Soto
92
argues that land titles are essential for helping people to cooperate to increase their community’s
productivity.
But the effect of land titling on people’s economic behaviour is debated. Ostrom (2009) argues that
norms represent the intrinsic benefits and costs of doing something. As such, norms have more control
over behaviour than rules. Applying this idea to property rights to land, it means that de facto property
rights to land may be better enforced than those with land titles, because not only does the community
enforce these rules, but there is also widespread social acceptability for who owns what land and how
they can use and transfer it, etcetera. In other words, de facto property rights are necessarily, doubly
reinforced by social norms. Without this common social agreement on what counts as land ownership,
it is very difficult to enforce a property right effectively, even with a land title. Therefore, it is often
argued that land tenure security counts more than a land title.
Indeed, North (1981) presented a historical discussion on the importance of well enforced property
rights for economic growth. Since then, there has been much research into the consequences of secure
land tenure on economic development (De Long and Shleifer 1993; Acemoglu, Johnson, and
Robinson 2001; Johnson, McMillan, and Woodruff 2002; de Soto 2000; Brasselle, Gaspart, and
Platteau 2002; Lanjouw and Levy 1998; Libecap 2007; Besley and Ghatak 2009; Leblang 1996;
Knack and Keefer 1995). Secure property rights offer non-economic benefits, namely liberty, privacy,
equality and community (Ellickson 1993). They further create the conditions for trust, having a
knock-on effect on social capital and cooperative behaviour, which in turn has been positively
associated with increased economic performance (Knack and Keefer 1995; D. Rodrik 2000). But
secure property rights are further associated with direct material economic benefits too, because they
shape the incentives for land use (Mueller et al. 1994).
A practical application of the above discussion is the focus of this chapter. USAID stresses that land
tenure insecurity is a driving cause for farmers in Afghanistan to grow illegal opium poppy crops
(Giampaoli and Aggarwal 2010). Policymakers frequently interpret the absence of land titles as a sign
of land tenure insecurity and thus, land titling is used as a strategy to make land tenure more secure.
This was illustrated at an international Alternative Development (AD) conference organized in 2002
93
by the German Foundation for International Development (DSE), the Deutche Gesellschaft fur
Technische Zusammenarbeit (GTZ), and the now United Nations Office on Drugs and Crime
(UNODC). One of their outcome recommendations was for AD projects in the Andean region to
“distinguish between land title holders and those without” (BMZ, DSE, and GTZ 2002, 18). This
initiative is based on a broad generalization that the existence of land titles indicates land tenure
security. The Deutsche Gesellschaft fur Internationale Zusammenarbeit (GIZ) equated land security
with the formalization of land rights and stated that “land use planning [away from the production of
coca and opium poppy] must go hand in hand with the legalization/ securing of land rights (titles)”
(Wehrmann 2011, 86). And a GIZ study that concluded that unequal and insecure land rights is what
drug crop producing communities in Afghanistan, Bolivia, Colombia and Myanmar, all have in
common, also conflated land insecurity with a lack of land titles (GIZ 2014, 1).
As such, land titling has also been used as a tool to support AD efforts in Colombia, in the hope that it
will change former drug farmers land use behaviour. The elite interviews I carried out with a range of
policymakers revealed that there were two main rationales for this intervention. One was in response
to the recent Peace Agreement between the government and the main armed group that had been
controlling the illegal drug trade, FARC. This rationale hoped land titling would help resolve the long
history of inequality in land ownership, as well as suppress the influence of armed groups and harness
the state’s authority. The other rationale was that land titles would deliver secure land tenure and as
such, change people’s land use behaviour. It assumed that land titling would increase farmers’ access
to credit, perceived as a key part for rural development, and would thus incentivise farmers to stop
producing illegal crops.
The following extracts from the elite interviews illustrate policymakers’ dual rationale for land titling
as a strategy for AD:
Land formalisation and the Agrarian Reform is the first point of the Peace
Agreement... Land is the cause of conflict…. Land titling is not the only solution,
but it is probably the most important part of the solution… Now the guerrilla
[FARC armed group] is almost on the government’s side and the Peace
94
Agreement enables the state to enter some territories to reform the land ….[part
of] these development plans contemplate land ownership … but also the credit
system… land titling is central for transforming those areas…if a campesinos
wants to use their land as collateral to guarantee a loan, they should be able to.
We are all equal [Director of the National Programme for the Substitution of
Illegal Crops - PNIS]
Land issues are central in point one and point four of the Peace Agreement. The
Ministry of Justice is trying out different variables of what works and doesn’t
work in terms of land and AD efforts…. Land titles are needed for accessing
credit … Land titles provide an incentive that is readily available. They send a
message of trust and state investment in the community [Policy Advisor,
Colombian Drugs Observatory - ODC, Ministry of Justice]
Formal land titles mean that producers can focus on the long-term effects on land
and can access credit [Policy Advisor, GIZ, German development aid agency]
In Colombia working in coca growing communities means you are working with
people who are deeply traumatized by many years of war or civil conflict, where
there is a very high degree of lack of trust… There is also a great deal of distrust
in the state because the state was never there for them, or offering health,
education, security, etc. One of the important steps in all this is trying to rebuild
the social fabric in Colombia…[so] we are regularly handing out land titles and
it’s increasingly picking up momentum. It’s a way to build trust with the
communities…they feel that now there really is something that will make a
difference in the communities as they know that now they own their land, where
they have lived for generations. Being on the national registry of land titles in
Colombia is a real accomplishment. The idea is that this will complement the AD
efforts, it’s like the cherry on the pie… If they are not the owner of their own land,
95
they don’t have the same attachment and commitment to the land… We estimate
that at least 90% of coca growing is happening on land without clear land
titling… If you are a landowner, you can get access to credit, you can share that
land, or you can give that land to your children as inheritance. So socially you are
somebody. If you are not a landowner and you are living on a piece of land that is
not yours, you’re nobody and you have no guarantees [Director, UNODC
Colombia Office]
There are many problems relating to land. The main one is the widespread
informality of land transactions. When there is land informality, campesinos are
much more vulnerable to violence like forced dispossession. It also is a barrier to
rural development. When a campesino is looking for bank credit or a subsidy…
the state is unable to make public investments in plots of land that lack titles. So
formal land tenure is not just about the legal security of the land, but also about
access to rural development measures. From this issue, many others arise, like
the intense conflict over land access in Colombia, even if it is formalized…
dispossession, abandonment and lack of opportunities is the result… The taxi
theory is that a driver who does not own the car will not look after it as well as a
driver who owns the car. Same with land. So, land ownership is the strongest
incentive for farmers to look after it and use it for legal production… There has
been no resistance to formalization [Director of the National Land Agency – ANT]
These quotes reflect a consensus that land titles work in the way that de Soto describes, leaving little
room for the subtleties that Ostrom raises. In this study I will analyse how land titles are perceived by
farmers living in communities in Colombia where the state has offered them to support their illegal
crop substitution efforts. My aim is to compare whether the reasoning of policymakers rings true at
the grassroots level, and whether land titling is conducive to the replacement of coca leaf with legal
alternatives.
96
I will start with a contextual overview of how land ownership has developed over time in Colombia
and will argue that there is symbolic significance for policymakers to roll out land titling programmes
in communities such as these. I then present some former illegal crop farmers’ perceptions of de jure
land ownership and reveal how they were appealing for the said symbolic significance.
However, this rationale falls short of answering whether land titling has any influence over former
illegal drug farmers’ land use behaviour. I will therefore go on to review the literature on property
rights to deduce four theoretical mechanisms by which it might convert farmers to the legal use of
their land: 1) by facilitating land market transactions, 2) by increasing the value of land; 3) by
encouraging agricultural innovation; and 4) by increasing access to credit thereby increasing land
productivity. I will argue that the first three mechanisms assume that land titling increases land tenure
security and this is an ambiguous assumption for three reasons. Firstly, land titling does not always
increase land security. Secondly, there are caveats in the evidence that links land security to increased
investment (and therefore to increased land productivity in legal alternatives). And thirdly, and most
importantly, the farmers I interviewed confirmed that they already had land security thanks to the de
facto property rights to land they had developed prior to any opportunity for land titling.
I will then argue that the fourth theoretical mechanism by which land titles could transform farmers’
land use from illegal to legal agriculture, does not assume land titles increase land security. Moreover,
it is the rationale used by policymakers for using land titling in AD. It is therefore worthy of empirical
verification. I will go on to assess whether my quadripartite hypothesis of how land titles could
encourage farmers to stop growing coca leaf by increasing their access to credit, is supported by the
experience of 87 farmers who ceased to farm coca leaf crops in South West Colombia where land
titles formed part of AD efforts.
From my analysis, I will conclude that even if they increased access to credit, land titles did not
diminish farmers’ attraction to growing coca leaf given its rate of return compared to other crops. In
addition, they presented the additional risk of increasing farmers' debt and need to resort to jobs
linked to coca leaf production. Therefore, land titling programs may appeal to policymakers because
they are measurable, and it reflects positively on the state given the country’s social and political
97
context, but they are not a game changer in the context of AD. Policymakers should thus play down
this effect and watch out for the displacement of coca production elsewhere that land titling can have.
Further insight into the value of informal property rights
In a nutshell, this part of my research is a rebuke of Hernando de Soto’s idea that land titles are an all-
encompassing solution to incentivising coca leaf farmers to switch to legal crops. For de Soto, land
titles are like bargaining chips in that they enable people to use their material assets as collateral for
credit, which in turn generates surplus value by creating ‘capital’. Like Adam Smith, de Soto proposes
that ‘capital’ is not just “the accumulated stock of assets but the economic worth it holds to deploy
new production” (de Soto 2001, 30). He argues that generating capital is the ultimate way by which
people can break out of poverty, because a sum of money savings per se is not enough. He argues that
formalizing the existing property rights of people working in the informal and illegal economy,
including through land titling, is what enables them to unleash their entrepreneurial potential. How
so? De Soto affirms that the formal property system allows people to generate capital by describing
and registering the value of assets, by standardising and unifying information into one system, by
reducing people’s dependence on relationships with people they know and trust, such as family
members, to respect their assets, by making assets adaptable to any transaction, by improving the flow
of communications and by protecting transactions through time and space. In this sense, land titles are
like passports, in that they are a standardized and reliable piece of paper, that makes land assets
sufficiently trustworthy for land owners to unlock their credit potential and thus maximize their value
(H. De Soto 2012, 62). According to de Soto, the marginalization of the poor in developing countries
stems from their inability to access the formal property system and thus the six benefits described
above (de Soto 2001). In short, legal institutions, such as land titles, are paramount for development,
and more so than technical assistance, market integration, strong foreign trade, and access to
technology (De Soto 1994; de Soto 2014, 33).
In practical terms, de Soto’s argument suggests that if only farmers growing illegal crops, such as
coca leaf, had land titles, they would be able to take out loans and maximize the productivity of their
land to produce legal alternatives. The investment would improve their soil and reduce erosion to the
98
point that alternatives to coca leaf would be, at best, more attractive or, at worst, as attractive to grow.
In his words, a “formal title gives the poor of the Amazon basin legal alternatives to selling coca
leaves to drug traffickers. As long as the farmers who grow coca remain informal, the government
cannot find them, identify them or reach an enforceable crop-substitution agreement with them” (De
Soto 1994, 21).Therefore the answer to illegal drug crop farming is to simplify the law so that small
legal producers have incentives to expand their business.
Inspired by de Soto, the Peruvian government undertook a large-scale land titling programme in the
early 1990s for coca leaf farmers in the Amazon region to help them to turn their back on the terrorist
organization, Shining Path, and thus also on coca leaf (Brooke 1990). Then in the early 2000s, more
land titling programmes were introduced for coca leaf farmers in Peru as part of the Special Land
Titling Project, Proyecto Especial de Titulacion de Tierras (PETT) which had been running for more
general recipients since the early 1990s, funded by USAID. However, the effect that land titling has
had on illegal coca production in Peru is not clear (USAID 2016a, 8; 2005, 4). The eventual defeat of
the Shining Path in the 1990s by no means guaranteed by coca leaf reduction, for Peru went on to
become the number one coca leaf producer in 2013, surpassing even Colombia. The UNODC
website’s accounts of successful AD in Peru makes no mention of land titles and assigns AD success
to the triumphant work of cooperatives.8 And more to the point, the fast-tracking land titling only for
some people, who meet special criteria for specific AD development projects, is not the same as
simplifying the process for everyone. And de Soto’s point is that land titles must be made more
accessible for everyone so that the capital and property market functions better for all.
On a theoretical level, de Soto’s argument can be contested for assuming that credit is the answer to
prosperity. As David Harvey argues, the benefits of credit are overestimated and the financial crisis is
but one example of its shortcomings, for investing in asset values rather than relying on labour or
industry, creates a financial debt (Harvey 2011). But my research does not seek to challenge de Soto’s
8See for instance: https://www.unodc.org/unodc/en/alternative-development/peru---success-
stories.html or https://www.unodc.org/unodc/en/alternative-development/peru.html
99
theory on these grounds. Regardless of the pitfalls of capitalism, I too share his assumption that coca
leaf farmers are motivated by materialistic goals. Admittedly, the options that are available to many of
them are very limited. But like the rest of the population, they seek to improve their livelihoods
through the market. Therefore, I sometimes refer to the farmers I interviewed as entrepreneurs.
My critique of de Soto’s advocacy for land titling is that it is an oversimplistic approach when applied
to AD. De Soto wrote his acclaimed book “The Other Path” in 1989, at the end of two decades of
many land invasions in Peru, and mass migration of the rural landless to urban areas around Lima
because of a lack of land and work (U.S. Congress 1993). Thus, his theory makes great emphasis on
informal settlements outside towns and cities. In this context, the value formalising property rights is
clear and straightforward. Bringing informal housing, transport and street vendors into the legal
economy is a no brainer if it means the governance for the market used by the vast majority of the
population can run more efficiently (de Soto 1989). But extending his theory for the informal
economy, to the illegal one seems to underestimate the attraction of growing coca leaf. Whether it’s
grown on legal or illegal land, the illegality of the crop makes its rate of return higher compared to
other legal crops. Therefore, land titles cannot be generalized as the solution for all coca leaf farmers
to stop farming illegal crops just because they operate in an informal economy.
My research on property rights reveals that the situation of former coca leaf farmers in Colombia is
different to that of the Peruvian coca leaf farmers de Soto refers to in support his theory. De Soto was
addressing a situation in Peru where even farmers with land titles found it difficult to find credit
because commercial banks had ceased to give agricultural loans after the 1969 land reform (Tammen
1991). And those who were lucky enough to be offered short- and medium-term credit by the
Agrarian Bank, as part of a USAID funded development project in Alto Huallaga (PEAH) in 1981,
were also required to have a land title (Morales 1990). Therefore, a land title in this context may have
made a huge difference in farmers’ chances for accessing credit. As I will later discuss, the coca leaf
farmers I interviewed in Colombia already had access to credit through informal channels and even
banks issued small loans with informal property rights to land. And, of those who wanted land titles,
only around half wanted them to be eligible for bigger loans. The rest could make do with informal
100
ownership for land transactions because they already enjoyed land tenure security. Therefore, as a
2005 summary of USAID reports on Bolivia, Colombia, Ecuador, and Peru, pointed out, there are
different factors associated with coca leaf production in each country. While coca leaf farming was
epitomised by the lack of guarantees for property rights in Peru, in Colombia it was more a question
of poverty (Rocha 2005, vi).
The empirical evidence I present challenges de Soto’s theory in two ways. Firstly, it illustrates how
formalising property rights is not a blanket solution for development, and therefore for successful AD.
The obstacle that most of the farmers in my study faced was not a lack of legal integration into
market, as all of them had been able to access some credit without land titles, even from the Agrarian
Bank, and were able to sell their legal produce in agricultural markets. The real problem was the lack
of roads and the poor quality of land that most of them had to work with. This meant that any
investment in their land because of a title-based loan was unlikely to make their alternatives to coca
leaf faming as profitable. This suggests that the barrier to their success in the legal economy was
structural, and often beyond their possibility to overcome, even with land titles. What use is credit if
the material land assets have no economic potential superior to coca leaf farming? This point has been
argued previously by Bromley (2008).
Therefore, the structural causes of underdevelopment should not be underestimated and can surpass
the benefits that land titling may bring to former coca leaf growing communities. This was already
seen in Peru as most of the title-based loan applicants participating in the PEAH project mentioned
above, used their loans to plant coca leaf, since their chances of “complying with their financial
obligations to the lending institution could not be met with cash returns from their crops of corn, rice
and cacao” (Morales 1990, 96). And the example of where land titling is deemed to have worked as a
policy to disincentivise coca leaf farming in Shambillo, Peru, happens to be where the land and
location is adequate for a profitable alternative product: palm oil trees (GRADE 2004). Nevertheless,
de Soto does not test alternative hypotheses for poverty, nor does he observe the relative importance
of land titling compared to other barriers to economic growth.
101
Secondly, my research supports Ostrom’s argument that informal institutions can be just as effective
as formal ones, and so the value of the formal property system is not always evident. De Soto states
that “a piece of land without such a title to specify its ownership at low cost is extremely hard to
market … [without titles and] and exchange will be restricted to closed circles of trading partners who
trust one another”, and land titles mean that people respect each other’s land more and invest more in
it with greater access to credit (De Soto 1994, 20–21). But the cases I studied showed that where there
is social acceptance of the informal system of property rights, the added value of land titling is lower
than in communities such as those which inspired de Soto’s theory. The cases I studied were in places
where there was little commercial interest in the land. This meant that farmers were embedded in their
local community for land transactions and their contracts had been effectively enforced by social
norms even when they were not legitimized by the law. In fact, journalists found that the coca leaf
farmers included in a land titling project in Peru in the early 2000s, did not feel that AD initiative had
gone well because their land titles did not match their existing informal property rights (The
Economist 2003, 58). All this goes to show that de facto property rights can ultimately be more
significant than land titles, and de Soto’s advocacy for land titles is not the golden bullet solution to
coca leaf farming.
My research further challenges de Soto’s view that legal property rights have the added benefit of
reducing violence, and even stopping wars (de Soto 2017). De Soto basis this idea parting from the
assumption that economic disparity and a clash of economic interests is the cause of insurgency and
therefore of violence. He argues that ultimately, informal property rights are to blame for this because
they create the perception that “opportunities, property and power are distributed arbitrarily” and this
gives rise to rebellion (de Soto 1989, 233). He proposes that formalising property rights is the key to
reducing violence because this is what gives the state social relevance and legitimacy in the eyes of all
citizens, especially when there is a frequent government turnaround, like there was in Peru around that
time. The logic follows that reforms aimed at integrating people into the formal economy, such as by
giving them land titles, means that they no longer have a source of discontent and so the rhetoric of
insurgent groups loses its appeal.
102
De Soto’s ideas have been credited with turning people in Peru against the Shining Path (IMF 2003).
And maybe rightly so, but only in the context of Peru. The Shining Path was a Marxist organization
which promoted the collectivization of property, and then implemented their own land titling process.
So farmers who wanted to protect their property felt beholden to the Shining Path not only to enforce
their system of property rights to land, but also to avoid further bloodshed (de Soto 2017). It therefore
makes sense that when rural people were offered land titles and state protection, the government was
able to win their loyalty against the Shining Path. This was because these farmers felt supported by
the state and felt their property was now safe without the need to resort to violence (de Soto 2017).
My thesis also shows that violence cannot always be explained in such simple terms. On the one hand,
de Soto’s understanding is that violence is the result of citizens’ frustration with “the system”, and
subversion and aggression are the “human response to frustration” (de Soto 1989, 234). He affirms
that this social frustration is not so much the result of poverty but of a perception of widespread
injustice, which gives rise violent rhetoric and terrorism. So, to prevent citizens from migrating and
revolting de Soto’s remedy is to formalize property rights, because this counters people’s perception
of structural inequality and reduces their resentment towards the state.
However, violence in coca leaf farming communities cannot always be explained simply as people’s
“response to frustration”. And formalising property rights does not always counter the authority of
extra-legal groups among coca leaf farmers or deter farmers from growing coca leaf again. From my
fieldwork in formerly bountiful coca leaf farming communities in Colombia, it transpired that the
motivation for violence in these contexts, stemmed from rivalry between extra-legal groups rather
than from coca leaf farmers’ frustration with the state. In addition, the offer of land titles did not
automatically mean that landowners felt they could unconditionally side with the law. Many of those
who had or were applying for land titles, were still willing or considering returning to growing coca
leaf elsewhere. And the coca leaf farmers’ loyalty to the state emerged less from the government’s
land titling efforts, and more from the farmers’ aversion to the roving banditry caused by inter-group
fighting.
103
And finally, part of de Soto’s argument is that the young, who are the most enterprising are also “the
most aggressive and belligerent”, and therefore “the easiest prey to the rhetoric of violence” (de Soto
1989, 234). On this basis, it would follow that the young have the most to gain from formalising
property rights, as they will be appeased once they are given a levelled playing field in the legal
economy. However, from my fieldwork it was clear that most of the farmers who were landowners,
and who were thus able to apply for land titles, were older people. It was the younger farmers who
had migrated elsewhere once the coca leaf farming had subsided in these communities. Therefore, my
research shows, albeit implicitly, that there are contexts where land titling is not directly conducive to
peace, because the benefits of AD are reaped by older generations, which in de Soto’s logic, are the
least likely to become violent.
In sum, it seems that de Soto’s association between legalising institutions and severing citizen’s ties to
armed groups is heavily biased by the context of Peru, and therefore his solution is simplistic and by
no means universal. As my research shows, there are various context specific barriers to economic
development and to sources of violence.
104
1.3 Symbolic value of land titles
It was evident from the elite interviews I carried out during the pilot study, that policymakers
considered the promise of land titling was perceived as a reward at the grassroots level. They believed
that for rural farmers, without formal land ownership, a land title represented a transfer of ownership
from the state to them as individuals. For example, in one interview the head of the UNODC
Colombia office, referred to land titling to build trust in the communities where the state has entered
with AD programmes. He said registering farmers’ land on the national registry of land titles was a
useful way of making farmers feel that they had really accomplished something, “like the cherry on
the pie”. Similarly, in another elite interview, the director of the ANT, the national land agency,
described land titling as a reward for farmers that have given up coca leaf production. And the
director of PNIS, said that land titling was a way of carrying out the Agrarian Reform which has never
been fully resolved. Land titling was a form of gift from the state to rural dweller in two ways. For
farmers who had de facto rights over private land it was a gift of subsidies to expedite their
paperwork. And for farmers who had de facto rights over baldio state waste land, it was effectively a
legal acknowledgement of that state land ownership was now private ownership.
To better understand why legally acknowledging farmers’ de facto claims over land in these
communities is politically significant in Colombia, I shall next summarise Colombia’s history of land
grievance. This will help to highlight the symbolic value of land titles.
Brief background on land and politics in Colombia
Colombia’s first land tenure system began in the 16th century with the establishment of 300 titles of
collective ownership to land for indigenous groups, known as encomiendas. The Spanish monarch
granted specific people the responsibility to protect these communities, and instruct them in the
Christian faith, in exchange for their agricultural goods. Individual use-rights and land grants were
also given to conquerors and settlers, resulting in what is known as the latifundio system, the
105
concentration of land ownership. By the late 18th century, the encomiendas had all been broken up and
sold.
Land titling of state-owned land was first introduced in Colombia in the 1800s to encourage migration
to different parts of the country. For example, private land titling of former state-owned land in
remote wastelands, known as baldios, was introduced in order to populate frontier areas and alleviate
land pressure (Albertus and Kaplan 2013; Le Grand 1986). Then increased pressure from liberal
governments began to break up the very large land holdings, and to expropriate many of the rural land
that was owned by the Church. In 1936, the “Law of Lands” (Ley 200) is famous for attempting to
reduce conflict by giving land to dissatisfied rural farmers (Adams 1966). It entitled rural residents to
gain ownership of land if they made productive “social use” of it. Landowners began to feel
threatened by the rural landless, commonly referred to as campesinos, who they feared would chip
away on their property with small parcels of small-scale farming. Landowners resented campesinos
for this and for forming the “Agrarian Leagues” that resisted their obligation at the time to work one
day a week on their to earn their access to their parcels of land (Steele 2017, 62–63). Likewise, the
campesinos resented the landed elite for these unequal policies and increased the pressure on the state
for further land redistribution policies to reduce land concentration, known as an agrarian reform. This
triggered the formation of armed groups in the mid 20th century that opposed the state, such as FARC-
EP (Brittain 2010). The landed elites have opposed them since (Albertus and Kaplan 2013, 204;
Bushnell 2007, 329; Ferrell 2019).
In 1961 the government passed the Social Agrarian Reform Act, Law 135, which promised to give
land titles to hundreds of thousands of acres of public land and established the land reform agency,
Comite Nacional Agrario, to implement land reform through regional project zones. These were
elected based on demographic pressure, landholding patterns, potential for land reclamation and
proximity to markets. However, the committee board was comprised mainly by landed elites with
only two of the 15 members representing campesino interests. The rest were members of the
Federation of Colombian Cattle Ranchers (FEDEGAN – Federacion Colombiana de Ganaderos),
members of the armed forces and elected congressmen. The intensity, geographical scope and nature
106
of the reform process was thus skewed in the interest of the landed elites, who shifted the focus from
expropriation and redistribution to and titling and contained the reforms in area of new colonization
like the Magdalena Medio and the Llanos (Albertus and Kaplan 2013, 204; Bushnell 2007, 329;
Steele 2017).
In the 1970s, 1980s and 1990s the government issued some sporadic land titling of state-owned land
for individual and communal owners, with social and economic development objectives, including the
allocation of large haciendas to workers of those lands. A public land agency called INCORA
(Instituto Colombiano de Reforma Agraria) was established in 1961 because of Law 135. Its purpose
was to reallocate land to displaced people, to redistribute state-owned land, to buy land from private
owners and redistribute it to others, and to legalize informal private ownership of state-owned land.
Some of the land in which the latter happened, was in illegal drug producing areas which had the
effect of incentivizing farmers to switch to legal farming from illegal drug production.
But INCORA was subjected to immense pressure by landed interest groups lobbying the Conservative
President Misael Pastrana’s government, by fraudsters and by illegal land grabbers (Grajales 2011;
2013). The National Peasant Association (ANUC – Asociacion Nacional de Usuarios Campesinos)
was founded by a decree in 1968 by the Liberal President Carlos Lleras Restrepo to represent
campesinos’ land interests and keep land grievance on the agenda. This got the attention of INCORA
and the Conservative party who acted to weaken ANUC by stalling redistribution and introducing
repressive policies. For instance, in 1970 Conservative President Misael Pastrana led the Chicoral
Agreement in which large landowner interests trumped previous land reform efforts and instead
introduced a settlement policy to colonize unexploited Southern regions (Grajales 2011).
In 1972 the ANUC leaders rejected this agreement so the minister of Agriculture accused ANUC of
communist inclinations and paramilitary groups began to repress even moderate factions of
campesino associations (Grajales 2011). Then in the mid-1970s Liberal President Lopez Michelsen
sought to develop areas where there was armed group presence, while FEDEGAN and the agricultural
producers’ association (SAC – Sociedad de Agricultores de Colombia) opposed this. As tensions over
107
land tenure inequality escalated in the late 1960s and 70s, a series of US funded development
programmes were introduced to reduce the attraction of communism in rural Colombia, known as
Alianza para el Progreso (Bushnell 2007, 327). Similarly, there were a series of development
initiatives to promote the formation of rural cooperatives and campesino associations (Ballvé 2013).
There was some sporadic land titling in the late 1980s but the 1990s the market-led agrarian reform
was still stifled by the agrarian elite and efforts to subsidize market-based land purchases had little
impact in distributing land equitably (Grajales 2011; Albertus and Kaplan 2013).
The national land agency has been dissolved twice in the past decade; INCORA replaced by
INCODER in 2007 and again by ANT in 2015. This has compromised the existing institutional
memory concerning the allocation of state-owned land. It has created a backlog of claims and resulted
in sketchy land titling in some places, with only part of the paperwork for land titles being carried out.
Meanwhile, many campesinos have informally colonized virgin baldios and claim to be the informal
owners.
Many in Colombia believe that, unlike other countries in the region, the state has never implemented a
thorough agrarian reform and this has consolidated a perception of inequality in rural property
(Grajales 2011). Today on average, two thirds of all farms own less than five hectares and account for
only 5% of all agricultural land (Economist 2015). On the one hand, conflict has created mass internal
displacement and land-grabbing which has increased the concentration of land ownership over the
past three decades (Reyes 2009). Over the period of 1964 to 2009 there was a huge shift of land
tenure, with tens of thousands of small and medium sized farmers being displaced by armed actors
(Palacios 2012). It is estimated that about four million hectares of land has been grabbed from
campesinos by armed actors. On the other hand, large-scale commercial investors have assembled
larger holdings of vacant land which have now been declared illegal and land transactions have been
suspended, such as in los llanos. The result has been increased tension between landless farmers and
(2011) quantitative review of the literature on land rights and investment incentives in West Africa,
found that studies with small sample sizes tend to not to identify a relationship between land tenure
and investment. Moreover, investment is difficult to measure without bias as it varies, is not binary
and is not always visible in the forms of fencing and trees (Fenske 2011; F Place 2009). For instance,
the length of time that land is left fallow is a measure of investment, even if the land is not actively in
use. Investment is also subject to recall bias and severe measurement error. Land size may influence
the density of investment (Frank Place and Otsuka 2002). And data is not always accurate enough for
rigorous econometric tests (Fenske 2011). For example, data on investment can focus on innovation
rather than intensity. Or it varies too much, so very common and very rare forms of investment can
each skew the samples on which generalizations are made. And different regions will have data with
very different sample sizes making it difficult to carry out cross-case comparisons.
Therefore, isolating and measuring a causal relationship between secure land tenure and increased
investment is not a straightforward endeavour. The inherent complexity in studying this relationship
makes it difficult to see how land tenure security increases investment in land, even if we could see a
clear link between land titles and increased land tenure security.
Context-dependent studies
It is also worth noting that much of the research on the causal relationship between land tenure
security and investment has been focused on Africa, where the property rights system is often
communal, with very specific circumstances. In these contexts, communal land ownership may be
127
enough to attract investment even in the absence of secure individual land rights (Fenske 2011).
Because land tenure security may depend on occupation rather than land titles, it may be the security
of land use rather than land titles which attracts investment. In some cases, the risk of expropriation in
African indigenous land tenure systems is too small to have a measurable impact on levels of
investment (Fenske 2011, 139).
Moreover, measuring the effect of land titling in the African indigenous property rights system may
be biased by the fact that land titles can “increase insecurity through rent-seeking, failure to recognize
informal rights, and competition over registration” (Fenske 2011, 141; Atwood 1990; Gebremedhin
and Swinton 2005). Africa presents circumstances that make it difficult to extrapolate from studies
carried out there. For example, the difficulties involved in land registration (Soludo 2000), the lack of
state capacity to enforce property rights (S. Smith 2004, 245–68) and landowners’ limited knowledge
of their rights (Deininger and Jin 2006) are factors that contribute to the futility of land titles in Africa.
Therefore, the evidence on whether land tenure security is linked to higher investment should be
regarded as context-dependent and not to be generalized.
Counter evidence
And finally, the effect of land tenure security on economic productivity may be overstated. Studies in
Africa and historical England have found only a modest link between forms of land tenure and
agricultural investment (Clark 1998; Allen 2009; Fenske 2011). Increased productivity, because of
land tenure security, may only be the case for medium income borrowers (Besley and Ghatak 2009).
A large-scale study in West Africa further found a further link between land tenure and investment
but only for some types of long-term investment such as tree planting and fallow land, and not for
land improvement, nor for short-term investments such as labour, fertilizer, manure and insecticide
(Fenske 2011). So, clear, and irrefutable evidence for the link between land tenure security and
investment is still missing.
What is more, some studies suggest that land tenure insecurity does not discourage investment
because non-landowners intensify their production to make up for the produce that will be lost
through insecure property rights (Ellickson 1993). Enforceable sharecropping contracts or tree
128
planting have been found even in insecure land tenure (Fenske 2011). It is argued by Fenske (2011)
that altruism and communal sentiment can encourage investment independently from land tenure
security, particularly for short term profits i.e. for fertilizers rather than trees and irrigation. Other
studies in Ghana, Kenya and Rwanda have found that increased investment in land productivity is not
connected to the existence of land titles (Place and Hazell 1993). Therefore, even if land titles
increase land tenure security and this increases investment, the evidence that links increased tenure
security to increased land productivity is still disputed.
Pre-existence of de facto property rights to land
So far, I have discussed two major issues with the first three theoretical mechanisms I presented for
how land titles can aid AD. These are that they assume land titles increase land tenure security, and
that land tenure security increases investment. The third, and perhaps the biggest, problem with the
first three theoretical mechanisms presented earlier is that they assume that the state is present to
enforce property rights to land more effectively. After all, for land titles to increase land tenure
security, the state must be present in the community and hold a monopoly of violence so as to be in a
stronger position to enforce the law than the status quo enforcer of property rights (Snyder and Duran-
Martinez 2009). With these optimum conditions for law enforcement, the state will pose a credible
threat and deter individuals from disobeying the law i.e., the rules of property rights. Only if the state
is in a stronger position to enforce property rights than the community is able to do unaided, will land
titles improve land transactions (Eggertsson 1990b, 35).
However, in the coca leaf farming communities I studied, property rights to land already existed and
were sufficiently well enforced without recourse to the law. Farmers had developed de facto property
rights to land with informal, but standardised contracts. Most of their land transactions had once relied
an oral agreement between sellers and buyers, and later, on a documento compra (or carta) venta.
This is a standard letter issued by the community that recognizes someone as the owner but with no
value before the law. It is an informally issued document setting out the conditions for the transaction
and which assigns sanctions for disobedience, not by the law but by social ostracism.
129
These de facto property rights to land, in the communities included in this study, had been enforced
by social norms and by armed groups at different times. The farmers that I interviewed in Putumayo
and Caquetá, recounted that a de facto private property rights system had emerged from an open
access property rights system. In other words, the state owned the vacant forest wasteland, known as
baldios but was not able to enforce its own de jure rights. As the land was effectively open access
fertile jungle terrain, individuals from various other areas began to claim this land as their own in the
1960s and 1970s. These people are known as the colonos (colonisers) because they colonized this
land for the first time in centuries, possibly since indigenous communities had been able to live freely
outside of their communal indigenous reserves. These farmers migrated to the jungle from rural areas
where unemployment was high and subsistence farming was insufficient to make a living, or where
there was political violence. They found they were better able to make a living from subsistence
farming and logging in the jungle as the land was more fertile. And so, they spontaneously established
a de facto private property rights system which they regulated through oral agreements and
documentos compraventa. These farmers illustrate how this occurred:
Land colonizers demarcated their land boundaries by opening corridors (called
trochas) that were 15 meters wide. We would cut down the thinner trees and leave
the thicker ones, and work our way form a corner, up diagonally. Our land plots
were respected but they had to be tended to because if trees grew back, someone
else could come and take that land. Disagreements [over land] would mostly be
settled peacefully because people were sometimes armed… only once did a
foreigner, presumably working for an oil company, try to force me to sell my land.
But it didn’t work as these disagreements were resolved among neighbours and
there was little the inspector could do when he got involved (Luis Antonio, 62)
By the 1980s there were many people arriving from Huila and Tolima. And my
family would give pieces of land to newcomers who would settle with large
families of eight or ten children. There were no land titles, everything was divided
informally, but it was well respected. The word of the elders was the same as an
130
escritura. We would offer them food for the first year while they settled in.
Neighbours would help each other out “trochando” i.e., clearing corridors and
cutting down trees with machetes (Marta E, 42)
The farmers that I interviewed that were in Cauca and Nariño, had a slightly different story. In these
veredas, property rights to land existed as a mix of de facto and de jure rights. The de facto ownership
in these areas emerged from vast land plots, that although privately owned by someone decades ago,
had been informally parcellated over the last century. The cost of dividing up these land holdings
through legal means, was prohibitive for most of the campesinos, so large land holders had sold off
pieces of their land informally to them. They issued documentos compraventa or gave them the
escritura but without officially updating it, so it was effectively an informal transaction. This had
happened a very long time ago for several generations and these de facto property rights had been
well enforced through self-governance ever since, in what became very close-knit communities.
What all the communities in this study had in common, was a very scarce recourse to the state. And
although they had initially relied solely on self-governance to enforce their de facto property rights,
over time, land use became dominated by illegal crop production. As a result, there was an influx of
armed groups who began to assert their authority through a mixture of protection and coercion. Their
extra-legal governance extended to the enforcement of property rights to land, but they respected the
established de facto property rights that the communities were already self-enforcing. This shall be
further evidenced in the findings to hypothesis A in Part II of this thesis. These farmers describe how
it worked:
People don’t trust the police as much as they did FARC. FARC used to say that
people should only worry about their own property, and everyone knew where the
land boundaries were. But they would not interfere in resolving land disputes
(Sandra Mireya, 35)
An old lady once came out with a machete insisting, she wasn’t going to allow a
shared pathway to cut across her side of the boundary, but people generally
131
respect each other’s land property. During the times of coca, any issues would be
resolved at gunpoint, so people were very respectful. Now issues tend to be
around pathways and boundaries. I just leave my vehicle on the inside of my
property to avoid trouble with a neighbour (Jose Gerardo, 60)
In this sense, the communities had long enforced property rights to land, and this self-governance was
the equivalent to the law (Alchian 1977: 129-130; Eggertsson 1990). By the time the state entered
these communities and began to offer land titles, on condition that they remain in, or return to, the
legal use of their land, most farmers felt that their land ownership was sufficiently recognized within
the community to be able to make many land transactions. In the words of these farmers:
There are no land tensions in the community. Neighbours respect each other’s
property. It has always been this way (Gilberto, 73)
There are no serious land tensions there and people are happy with the way that
land has been distributed. Even with a “compraventa” people respect land
ownership here. There are some disputes about land plot boundaries. The JAC
tend to resolve these disputes locally (Evelio, 39)
There are no land tensions in the area as everyone respects each other’s property.
All land has owners here (Seneida, 42)
There are no land tensions in the area, sometimes a few disputes over borders. But
the Conciliation Committee of the JAC resolve these disputes according to the
versions of each neighbour (Magula, 51)
De facto land ownership did not impede their own choice on how to use it, to sell it, albeit informally,
or take out loans. For example, land titles were not a prerequisite to join agricultural association, and
farmers were landowners even without land titles:
Land titles are not obligatory to be a member of an association. You just must
have access to land, and a few pepper crops, even as few as 200. Many
132
associations that are formed end up only existing on paper. They register and
receive benefits, but they never take off….Land ownership helped me to succeed
[in farming] as I did not have to pay rent. Most people are owners in the sense
that they don’t pay rent, even if they don’t have land titles (Pedro, 42)
Some farmers feared the de jure property rights might come to clash with the pre-existing de facto
property rights, and so wanted land titles to stay on the safe side. However, they did not feel their land
tenure would become completely secure in every way because of acquiring land titles. As these
farmers describe, it was simply a different trade-off of risk:
We got a land title later to be sure we had our papers in order and would not risk
our land being taken away. There is a rumour that the state comes to check what
land has a land title, and whatever land doesn’t will be taken back by the state…
The benefit of a land title is that you can get a loan but if you are not careful and
can’t pay it back, the bank can also embargo your land (Pedro Pablo, 55)
I am applying for a land title because I want to feel like I am the owner and not
the state. But there is a trade-off. With a land title, I will have a heightened fear
that banks can take my land away if I am unable to pay back the loan as they can
embargo my land and they cannot do this if you don’t have a land title (Pedro, 42)
Other farmers asserted their de facto rights were more valid than any land title. They did not feel that
their lack of land titles could give the state a right to take their land away, even if they were caught
farming coca leaf. As these farmers illustrated:
In a borough meeting we were told by some officials that those who needed land
and couldn’t access any, might benefit from the “land bank” whereby land will be
redistributed. But people here are not convinced about this as there is no free land
and if they were to be given the land of someone who had been expropriated from
their [informally owned] land, there is no way they would take up the offer. If they
133
did, they would be threatened. No one would leave their [informally owned] land
without a fight, even if found with coca on their land (Carmen, 43)
People do not fear having their land taken away because they do not have land
titles. There is no property to be confiscated. People are more afraid of being
arrested. They think (perhaps because they lacked education) that the police will
only target their [illegal] crops and do not run the risk of being expropriated from
their [informally owned] land (Ligia, 33)
Some farmers even described how in neighbouring communities, people had come together to defend
their de facto property rights against big oil companies, and succeeded:
When oil companies have approached communities to start exploration, they buy
land from individuals and start placing a kind of wiring…and a lot of violence has
resulted from this in 2015. The local people of Curbinata near Valparaiso
opposed them, they pulled out the wiring that the company had placed, even
through lakes. They feared their water sources would be negatively affected and
the soil would become infertile. People from our vereda came on their motorbikes
to join forces and support this local community, my brother was among them.
They took farming tools, machetes, and sacksful of stones to fight against the
ESMAD [mobile anti-disturbance squadron] and there were deadly physical
confrontations. They came back home injured and starving, but the oil company
had to withdraw (Maria Lucia, 62)
Moreover, the interviews revealed that many farmers understood land titles on a sliding scale. Some
believed their land was formally owned if they had an escritura even if it was outdated or not
registered on the national database. Others believed their land was formally owned because they had a
document compraventa. It was only when the state introduced the concept of land titling in the
community, that some farmers began to understand the difference. As this farmer explained:
134
I had an escritura for my land, but I was still not sure if this counted as a land
title, so when I saw that the UN were helping with the process, I asked them to
check my title status. It turned out to be adequate, I just was not sure about this
(Carlos Arturo, 50)
However, not everyone shared this concern. One farmer could not tell the difference between a land
title and an informal document compraventa but was not worried about this because she was sure she
and her husband had inherited their land from her father-in-law. Farmers like this had never
experienced problems for not having a land title as people had always been respectful of each other’s
land in these communities. And given that previous landowners tended to live nearby, and could
therefore vouch for the land transaction, land titling was seldom required to resolve current land
disputes.
In sum, the former coca leaf farmers I interviewed did not need property rights to be stipulated in the
law to operate well. They already had well-functioning de facto property rights, and as such, a great
deal of land tenure security. Even the Agrarian Bank recognized de facto land ownership to issue
small loans. So, overall, the data does not support the first three theoretical mechanisms by which land
titling is meant to significantly improve legal production for farmers, and thus dissuade them from
growing coca leaf.
Working hypothesis
The fourth theoretical mechanism I presented earlier does not rely on the premise that land titles
increase land tenure security. Nor does it presume that increased land tenure security promotes
investment in land. It simply implies that land titles provide an opportunity for farmers to access
commercial loans. Therefore, it is easier to observe if this mechanism operates in former drug farming
communities in Colombia.
In addition, as presented in the beginning of this chapter, it is the main rationale used by high-level
policymakers and government agencies in Colombia to justify the use of land titles as a component of
AD, besides their symbolic value. The 2015 UN World Drug Report identified farmers’ access to
135
credit for investment as a necessary component for AD to work (UNODC 2015). It is believed by
policymakers worldwide that credit is essential for purchasing initial livestock or seeds, and without
it, farmers’ efforts to substitute drug crops can never take off. Some AD programmes have tried to
address this issue with micro-credit schemes, to make capital more accessible to farmers (UNODC
2015). While other AD programmes, such as Formalizar Para Sustituir in Colombia, have used land
titles as a new way to increase farmers’ access to commercial loans (Colombia Government 2019;
UNODC 2016b). Anecdotal evidence (conveyed to me through informal chats with officials on the
ground, rather than formal interviews) also suggests the main benefit officials refer to, when
promoting land titling among farmers, is the opportunity they will have to access bank loans.
It is therefore relevant as well as important to question the widely held belief that land titling
generates good outcomes because it increases access to credit. To do so, I went to former coca leaf
producing communities to ask farmers if this mechanism rang true in their experience. To evaluate
whether land titling might have enhanced farmers’ voluntary coca crop substitution efforts, there were
four parts to my hypothesis.
The first is that land titles come hand in hand with increased access to credit. Evidence for this
increased access to credit following land titling is, so far, ambiguous (F Place and Migot-Adholla
1998; Carter and Olinto 2003; A. Field, Field, and Torero 2006; Johnson, McMillan, and Woodruff
2002). Access to credit is also dependent on the credit that is available. For example, a study in rural
Peru found that land titling does not lead to increased access to credit because titles do not always
present sufficient collateral to secure a loan (Kerekes and Williamson 2010). The Galiani &
Schargrodsky (2010) experiment found that land titling did not lead to increased access to credit in
Argentina because people had very little access to credit cards, banking accounts and formal credit, so
instead they relied on informal credit from relatives and friends, based on trust networks. Moreover,
the quality of land, in terms of the topography, soil fertility and natural resources available, has a
direct effect on the value of land, as will the size of the land plots (Galiani and Schargrodsky 2010).
And this has been found to affect the level of credit available to farmers (Besley and Ghatak 2009).
So, if the cost of the legal paperwork to acquire a land title exceeds the market value of the land plot,
136
then possession of a land title may be insufficient criteria to qualify for formal credit. This may be the
case especially in places where land value is low, and the credit market is thin.
The second part of my hypothesis is that landowners will automatically want to apply for credit.
Doing so implies using their land as collateral and risking the loss of their only asset, and possibly
their entire savings, for the sake of a business project. However, poor landowners may be less willing
to take this risk (Galiani and Schargrodsky 2010). Moreover, there is no empirical evidence to suggest
that landowners will automatically want to apply for credit and use their land as collateral. This was
found to be the case in Besley’s (1995) study in rural Ghana. There may be additional barriers to the
use of land as collateral (Fenske 2011). On the other hand, there may be ways reasons why this might
not apply. For example, credit markets in developing countries may be precarious and unable to
repossess land in poor areas (Deininger and Feder 2009).
The third part of my hypothesis is that increased access to credit increases land productivity. Even
when there is increased access to credit, this may not translate into increased land productivity, or
even reduced poverty, because the credit applied for may be insufficient for someone to overcome the
barriers to economic development. If the amounts borrowed are small, they might not level out
consumption, as happens in West Africa (Fenske 2011). And if the amounts they borrow are large,
they still might still not resolve the issue of unemployment. The point is that increased access to credit
may not overcome the underlying issues causing low productivity, such as poor soil quality or poor
boroughs, and may only make their land tenure less secure (Bromley 2008, 22–26). Land fertility and
climate can be more important factors than land titling in promoting land productivity, as was found
in Burkina Faso (Ouedragogo et al. 1996, 232). Land assets that are low in value do not attract
investment. In these cases, individuals may use their loans for home improvements or projects that are
not profitable (Galiani and Schargrodsky 2010).
In addition, rational choice theory assumes that farmers will use their increased access to credit to
invest in land productivity, but the reality may be different. People can behave in ways that do not
always maximise their economic returns due to their bounded rationality. It is therefore possible that
137
farmers may choose to spend their increased access to credit in things other than in agricultural
productivity, or that are not an investment.
In urban settings, increasing property owners’ access to credit works in the interest of the mortgage
lenders, not the borrowers, as the former have greater control over borrowers (Bromley 2008).
Kronman (1985) compares banks to kidnappers, as they take hostages through titles. This effectively
locks borrowers into long-term contracts with lenders, and makes it impossible for landowners to use
the collateral associated with the title in another credit transaction without the permission of the
lending firm (Bromley 2008, 22).
The fourth part of the hypothesis is that increased legal economic productivity is enough for
landowners to disinvest in illegal drug farming. Sometimes, increased economic productivity in the
legal realm is not enough to persuade individuals to give up illegal drug production. There are other
significant variables. Perhaps the most significant of these is the territorial control of extra-legal
groups. Their presence affects farmers’ ability to switch to, and even preference for, the legal
economy. Extra-legal groups enforce contracts and oversee the supply of crops and of required
chemicals, as well as the financing, money laundering, and export opportunities for drugs (Skaperdas
2001, 181–82). In Colombia, they hold territorial control over coca leaf production and can
monopolize the trade by setting prices with relative independence from the supply (D. M. Rico and
Gallego 2012). This means that the illegal drug trade can adapt to the growth of the legal local
economy, so that coca leaf farming remains most profitable regardless of small improvements in the
profitability of legal alternatives.
Another possibility is that landowners are unable to make profitable investments and need to pay off
their loans with profits from coca leaf production elsewhere. In Afghanistan, opium poppy farmers
have been found to grow illegal crops to pay off their mortgage and debts (McEwen and Whitty
2006). This begs the question of how landowners will pay back their loans if they are unable to make
their land more productive with legal alternatives.
138
In sum, if land titling is conducive to the replacement of coca leaf crops with legal alternatives by
increasing farmers’ access to credit, the following quadripartite hypotheses must be true:
A. Farmers with land titles have access to increased amounts of credit
B. Farmers with land titles choose to access increased levels of credit
C. Farmers’ increased access to credit results in more profitable legal agricultural production
D. Farmers’ increased productivity in legal farming reduces their dependence on illegal drug
production
139
1.5 Findings
Hypothesis A: Farmers with land titles have access to increased amounts of credit
De Soto, among others, suggests that when they acquire land titles, landowners, who were once
locked out of the legal banking system, can use their land as collateral to take out commercial loans
(de Soto 2000; Feder and Feeney 1991; Feder and Onchan 1987; Islam 1995; Cole and Ostrom 2012;
Brasselle, Gaspart, and Platteau 2002). With greater capital at their disposal, community members can
invest in their land productivity in the legal economy, start a business, and improve their living
conditions, making the profit gap with the illegal economy less marked.
My findings show that, farmers did not perceive land titles to be essential for them to access credit
because they had other ways to access loans, besides bank loans. About 41% of farmers interviewed
had had access to credit without the need for a land title. Alternate informal sources of credit included
informal loans from private individuals, loan companies, private organizations, and even extra-legal
groups who offered credit to coca leaf day labourers. It was not uncommon during the coca bonanza
for farmers to access private loans, as they often invested in each other’s businesses. A farmer
explained how a private loan helped him out, despite not having a land title:
A woman from a nearby vereda offered me a COL$2 million [about $629 USD]
loan. Now I own a small beer tavern where I sell gas and my wife has a small
shop (Gelbert, 40).
Farmers who had resorted to loan companies and private organizations for credit, had not been asked
to present proof of having a land title. They had simply borrowed money at a higher interest rate. For
example, farmers who had turned to a charitable foundation (called “Fundación Mundo Mujer”) for
loans, were only asked for personal references from guarantors who had land titles, or proof of a good
track record with their bank. Some farmers had borrowed credit from a local loan company (called
“Contactar”) which asked them for proof of a documento compraventa. This was a document issued
and used by farmers to vouch for their informal land transactions. The companies had warned farmers
140
that if they failed to pay their loan back on time, they would seize their land. Some farmers also
borrowed money from coffee producers’ committees who issued loans based on trust and a document
signed before witnesses. Some farmers had been offered credit and land by extra-legal groups after
passing a three-month trial as day labourers, so they could invest in their own coca leaf plantations.
Farmers had also accessed loans without having land titles from both state subsidized (e.g., Agrarian
Bank), and commercial banks (e.g., “Davivienda”). A local bank loan manager from the Agrarian
Bank, confirmed that banks had considerable discretion in the way they ascertained evidence of
farmers’ savings. Farmers said that prior to the common use of land titles, they had been issued bank
loans if they could prove a good credit history and present a reference letter from their local Mayor
testifying to their informal land ownership (documento de sana posesion and / or a documento
compraventa). They might also be asked for a detailed budget of how they would use the loan, and
evidence of being a state employee such as a schoolteacher, or a member of a recognized cooperative.
These were the various means by which farmers had used their de facto property rights to land to
access credit to invest in their land before land titles became the means to secure loans.
However, the introduction of land titles as part of AD efforts in 2014 – 2015 changed this in the
places I visited in Cauca, Nariño, and Putumayo. A recurring theme in the interviews in these research
locations, was that possession of land titles had since become the only means by which farmers could
access large amounts of credit. Farmers noted that they were now asked to present proof of land titles
to local mayors and to banks whenever they needed financial loans. They also perceived banks to
issue much larger loans to those that had land titles, as this farmer explained:
With a COL$12 million [about $3,775 USD] loan I took out from the Agrarian
Bank I began to produce pepper. That is the maximum they will lend you without a
land title. With a land title, you can re-mortgage your land to access a higher
credit, up to COL$50 million [about $15,730 USD]. I did not need a land title
because it wasn’t a big loan I was after (Pedro, 42).
141
In sum, farmers had had access to some credit without land titles. But land titles had become essential
to access larger amounts of credit from banks. It was evident that recent land titling efforts had
created a new culture of bank loans, which were only issued on condition of having a land title. This
alludes to the concept of performativity in language theory, in that language effects change. For
example, when a promise is made, or a judge announces a verdict, or a marriage ceremony is
concluded, people will change their behaviour. Similarly, when property rights are questioned and
land titling is introduced to validate ownership, people’s attitude to land ownership shifts and so does
their behaviour. The implication is that as more people apply for land titles, more people without land
titles believe they need to get one too.
Hypothesis B: Farmers with land titles choose to access increased levels of credit
Hypothesis A focused on whether the lack of land titles was a barrier to accessing loans. As outlined
above, the interviews revealed that land titles were becoming necessary for farmers to access bigger
loans. The literature assumes that farmers with land titles will naturally want to risk their land asset in
exchange for a loan (Besley 1995; Galiani and Schargrodsky 2010). So, hypothesis B seeks to
ascertain whether possession of land titles makes farmers more likely to take out bigger loans because
they know they can.
Before describing my findings, it should also be noted that farmers’ responses on their willingness to
take out loans may have been influenced by the time at which the research was carried out. As the
bank manager explained, farmers’ optimism about land productivity can vary with the season:
Most people who take out loans, apply for another one. However, the number of
[loan] applications vary according to the weather. In summer [drought season]
people get scared and in winter [rainy season] they find it easier to grow crops, so
they apply for more loans during the rainy seasons. Also, the rainy season is when
land needs to be fertilized and manual labour is needed for this, so people apply
for loans to pay for their wages.
142
My findings only partially support this assumption. Farmers willing to take out loans were almost as
common as those who simply wanted the option of loans but had no intention of taking one out. 46%
of the farmers interviewed had no desire for future access to credit, while 54% did.
Farmers who were more open to the idea of applying for loans, and who said that this was their main
reason for wanting a land title, were generally located in places that were well suited for alternative
legal agriculture, like coffee production. This farmer explained what many others in the same
situation echoed:
I want an Agrarian Bank loan because I want to produce coffee. I have the know-
how from some local technical advisors, the land, and the seeds. But I lack the
fertiliser and money to pay for the day labourers and their lunch. You need at
least COL$500,000 or COL$1 million [about $157 or $314 USD] to get started,
to pay for the day labourers’ food. Especially as coffee only has one harvest per
year around here (Daniel, 46).
However, among the farmers who had applied for loans or intended to do so, only 33% of them
wanted them to have more credit to invest in their land.9 The rest had other spending priorities. So,
while there were a few cases that did seem to support the hypothesis, there was a further complication
in that most of them were not using, or did not intend to use, the funds for AD. Previous studies have
found that land titles do not always increase investment in land use (Fenske 2011). This was reflected
by the many farmers who intended to spend, or had already spent, their loans on things such as home
improvement, food, healthcare, or children’s education, as this farmer explained:
9 It should be noted that in these communities, many farmers had attended talks given by officials who had promoted land titling as part of the local AD programme, and who they said had mentioned access to loans as the main benefit of land titles. Some farmers also mentioned that there had been local campaigns, led by private banks, to raise awareness among farmers about their credit services. Therefore, it is possible that some farmers’ motivation for borrowing large amounts of money was influenced by these efforts. Some farmers who were keen to take out loans, referred to these talks to justify their confidence that their investment in alternative legal forms of land use would thrive. As mentioned in the introduction, previous studies have shown that grassroots support for AD in Thailand came from state-run fear campaigns that opium consumption was increasing in local communities, and from respect for the King’s authority (Renard 2001). Therefore, the powers of persuasion by AD officials for AD effectiveness should not be underestimated.
143
I became very indebted as I used much of my credit to pay for my healthcare and
for my daughters’ university education. They now live in Pasto, one of them is an
accountant and the other one a businesswoman but is unemployed. Last year I
sold my land to pay my debts (Sabalon, 60).
Rational choice theory holds that people will rationally invest in the projects with the best returns. As
these findings show, farmers did not always consider agriculture to be the most productive
investment. Investment in a shop business or in their children’s education was seen to give higher
returns. And sometimes they saw more value in spending their money on price-inflated food,
consumption goods and entertainment, including prostitution and alcohol. Sometimes their bounded
rationality led them to believe that fraudulent investment pyramids was the best way to invest their
loans, which could leave them heavily indebted, as this farmer described:
I could access credit from the Popular Bank to invest in a fraudulent pyramid
scheme. I lost COL$22 million [about $6,900 USD], and I am still paying it back.
My monthly instalments are of COL$500,000 [about $157 USD] and with the
interest I still owe COL$40 million [about $12,500 USD] (Gilberto, 73).
Farmers who were more debt averse, tended to be in the more geographically marginalized
communities and in places with low quality land. In these places, there were farmers keen to acquire
land titles to be eligible for larger loans, but who also said that they were not intending to take out
these loans themselves. When I asked them why, their general answer was that they lacked confidence
that they could turn their land to profit. They were concerned they would be unable to pay back their
loan, and have their land repossessed by banks as a result. Previous research has shown that low
income households are generally averse to long-term debt (Datta and Jones 2001). This farmer’s
reluctance to get into debt, due to the poor soil conditions, was echoed by various others:
I have never taken out a loan because I do not like being in debt. We live in such
uncertainty with regards to employment and water that any income we have will
144
go towards our children’s education. We cannot afford debts… we have serious
water shortage problems (Mirta M, 42).
However, many farmers wanted the legal ownership of their land to be eligible for credit, as this
farmer explained:
I was offered a land title for free, among many others from around here. I’d never
had any land problems or had to defend my right to land from neighbours or
others. But I was happy to get the land title even if I still owed money to the
owner, who lives locally…the title helps me access credit, but I haven’t wanted to
get a loan as I’d prefer to not get into debt (Arturo, 55).
These farmers believed that land titles would increase the value to their otherwise valueless land,
because future buyers, who would presumably be less risk averse, would have the option of accessing
significantly larger bank loans.
Older farmers were also among those who were more risk averse, when it came to be taking out loans.
This included farmers above the age of 61, who had had lower levels of schooling but more
agricultural experience than younger farmers. As one farmer explained:
Older people are not interested in loans, they just want enough money for food.
They don’t want to sell their property either (Hubeimar, 32).
To put this in context, it should be noted that the average life expectancy for people born in 1960 in
Colombia is 56.75 (“Datosmacro.Com” 2017). And a local Agrarian Bank credit manager confirmed
that loan applicants had to be between the age of 18 and 69, because life insurance only covers losses
when people die before they turn 70. Therefore, farmers aged between 70 and 75 had to apply for
loans with a co-debtor, possibly another relative, to sign their life insurance and continue repaying
their debt if they died. After the age of 75, no one had access to credit, and this may explain debt-
aversion among the older generations.
145
In sum, the interviews suggest that farmers willing to take out loans were almost as common as those
who mainly wanted the option of bigger loans. Those willing to take out loans, were either in places
that had more promising condition for legal alternative agriculture or had other plans on how to spend
their credit, besides their land productivity. Those who were more debt averse were in marginalized
communities or were older farmers.
Hypothesis C: Farmers’ increased access to credit results in more profitable legal
agricultural production
As discussed in the literature review, the de Soto-inspired theoretical mechanism, that land titles
increase the availability of credit for investment and thereby makes land more productive, rests on the
premise that that investment makes the land asset more valuable (De Long and Shleifer 1993; Daron
Acemoglu, Johnson, and Robinson 2001; Johnson, McMillan, and Woodruff 2002; de Soto 2000;
Brasselle, Gaspart, and Platteau 2002; Lanjouw and Levy 1998; Libecap 2007; Tim Besley and
Ghatak 2009; Leblang 1996; Knack and Keefer 1995). The follow-on hypothesis is therefore that
increased access to credit increases farmers’ land productivity. My findings for this hypothesis reveal
that that this was not always the case. In instances where credit investment was well suited to the land,
farmers considered their loans had helped them boost their productivity in legal alternatives. But in
other cases, farmers did not consider that increased access to credit had led, or would lead, to greater
land productivity. This reflects the findings of previous studies that identify the limitations of
microfinance for stimulating economic development (Buckley 1997).
Cases that confirmed the hypothesis
Farmers in places where the soil and climate were well suited to agriculture, and with road access,
were generally more positive that the loans they had, or would have, access to, with land titles, would
make their land more profitable in something other than coca leaf. For example, farmers in the
‘Esperanzas de Mayo’ vereda, located in ‘Mercaderes’ municipality, Cauca department, felt that
access to larger loans, had enabled them to make a living from lemon crops. In this vereda, the
146
climate was dry and labour costs were high, and lemon trees grow well in these conditions and require
relatively little care. Moreover, this vereda was located very close to a main road which meant that
transporting this heavy fruit was not an issue. Therefore, investment in lemon crops worked out well
for farmers to earn a living without coca crops.
Farmers in places where there had been previous large-scale investment from private enterprise, felt
their title-based loans had helped them become more competitive in the legal market. For example,
farmers in veredas around ‘Linares’ town in the Nariño department believed that the new access to
bigger loans offered by land titles was helping them invest in sugar cane crops. Farmers there felt this
was proving profitable as their exclusive crop because a sugar cane company had invested in milling
machines that were not too far from tarmacked roads, making sugar cane production less laborious.
Similarly, farmers in ‘Santiago de la Selva’ vereda in the Caquetá department, confirmed that larger
bank loans helped them make their milk cattle profitable enough to replace coca leaf as their source of
livelihood, despite living in such a remote place. Their confidence stemmed from the recent
investment Nestle had made in their local transport and storage infrastructure for their milk.
And farmers in places with the right climate conditions and where there had been investment in the
infrastructure by AD programmes for coffee production, felt that investing their loans in their land
was going to be, or was already, profitable enough to stop growing coca leaf. This was the case for
farmers in ‘La Villa’ vereda, in Balboa municipality, in the Cauca department. It was also the case for
farmers in ‘Lo Alto’ vereda, in ‘Los Andes’ municipality in the Nariño department, where farmers felt
confident that a South Korean company would buy their coffee at a good price, come what may.
Ironically, many of the farmers that said they were doing well thanks to the loans they could apply for
with their land titles, were those who had already done well from coca leaf. In the interviews it
transpired that these farmers had used the capital they had saved from coca leaf production to secure
larger land plots in places with better soil and better housing. During the coca bonanza, they had been
able to produce their own food crops, instead of having to buy them from neighbours who sold food at
inflated prices. This meant that after the coca bonanza, they could continue to grow enough food to
147
feed themselves, without having to rent extra land, and were less likely use their loans for
consumption. It also meant that as they already had larger and more fertile plots of land, their loan
investment was more likely to make a profit in commercial farming. One farmers’ explanation of her
investment, illustrates the financial complexity verging between legal and illegal assets, and is an
experience that was shared by many others:
With our savings from coca production, we invested in the DMG financial
pyramid and made enough money to buy land and a house. We bought 19 hectares
of baldio land an old man owned. The old man owned 200 to 300 hectares in total.
Then in 2012 the national land agency [previously INCODER] came with a land
titling project and gave us a land title (por sana posesion). We paid COL$360,000
[about $113 USD] to a former Incoder official who acted as their agent. With our
land title, we could take out three loans from the Agrarian Bank (Lilian, 34).
Cases that did not support the hypothesis
As already came up in my findings for hypothesis B, many farmers had no interest in accessing loans
to invest in their land because they did not consider it would be a profitable investment. They felt
subsistence farming was the only realistic alternative to coca leaf. Their perception supports the idea
that having access to bigger loans does not always make land more productive because the credit is
still insufficient to overcome the larger barriers to economic development, such as poor soil quality
(Bromley 2008, 22; Buckley 1997). Macro-barriers to local development and limited soil productivity
were the two most significant barriers that farmers identified as hindering their investment in land.
Other factors they identified for being confined to subsistence farming, included poorly designed land
investments, small land plot size, high production costs, especially labour, and other investment risks
such as crop price volatility, unexpected external events, and changes in farmers’ personal
circumstances.
148
Macro-barriers to development
Farmers interviewed in Putumayo and in Caquetá, who were unwilling to take out loans to invest in
their land, identified their geographic and infrastructure marginalization as their prime reasons.
Farmers living in these areas had previously benefited from their limited road access when they had
produced coca leaf, as they were less detectable by law enforcement. After stopping coca leaf
farming, they found it difficult for their substitute legal agricultural produce to take off. This was even
the case for those with good soil, because they were too far from the towns where the markets were
located. These were often two-hour journeys away on untarmacked roads. Farmers believed this
distance made their farm produce more expensive, and therefore uncompetitive with that of farmers
from places closer to town, as they had to pay for extra transport costs. Some farmers found their
plantain, often their strongest crop, would sell in town at the same price as what it would cost them to
transport it into town. One farmer explained the difficulty with prohibitive transportation costs:
I used the loan for plantain and yucca crops and some cocoa bean. You can sell a
bunch of plantain in the village for COL$5000 to COL$6000 [about $1.5 to $1.8
USD] [and transport for one bunch of plantain costs about a third of the cost] but
your journey to get there costs COL$4000 [about $1.2 USD] each way so there is
no profit whatsoever for me and my crops and chickens are mostly for self-
subsistence (Marianeli, 46).
Another farmer explained how this barrier was a disincentive to apply for land titles. As he
understood it, the transport costs would make any alternative land use uncompetitive:
Some of the hidden places where coca is currently produced have land titles. You
need to get there by boat or motorboat. I think that alternative development in
these far-flung places would only help people grow other crops if they tackled the
transport issues and the marketing barriers producers face. People would
willingly abandon coca in those circumstances… I didn’t get a land title for my
149
coca plantation because it is too far away from town, and anything else it could
produce would be too expensive. So, it was not worth a costly land title. For
example, a lb of rice sells here for COL$1,400 [about $0.44 USD] whereas in
town it sells for COL$3,000 [about $0.94 USD]. But to get yourself into town
[Popayan] you would need to spend COL$5,000 [about $1.57 USD] in transport
costs, and between COL$30,000 and COL$50,000 [between about $9.44 and
$15.7 USD] to transport your rice sacks (Meregildo, 38).
In one vereda, I interviewed farmers who had tried to overcome this barrier by forming an association
to buy a vehicle for the entire community and thereby share the transport costs. However, this strategy
had so far not worked because there had been significant coordination issues.
Farmers in these remote places, also had limited communication infrastructure, which exacerbated the
issue. They pointed to the lack of phone and internet signal, bridges and electricity, as general
impediments for them to communicate effectively with buyers. This finding is consistent with studies
that have found mobile phone use to be correlated with financial development in Africa (Asongu
2015).
Climate and soil quality
Inadequate climate and soil quality are additional barriers that explain why there were many cases that
did not support the hypothesis that increased access to credit would lead to more profitable legal
agricultural production. Many farmers located in veredas that were better connected, still found other
obstacles for making their land more productive. They believed their crops were less competitive
because of the type of soil and the climatic conditions they had. Some farmers said they had a lot of
clay in their soil, which was unsuitable for anything beyond subsistence farming or coca leaf, which,
like weeds, grows even in hostile conditions. Many farmers felt unable to pay their loans back due to
these suboptimal climate and soil conditions. In some places in Linares and Los Andes in Nariño with
less light and cold weather, coca leaf harvests were less frequent and never very abundant. Other
150
farmers believed their soil productivity was lower than in other places, because of drought or
flooding. As one farmer described in his own words:
The sudden bout of rain following drought last year destroyed part of my crop as
there was a partial land slide. The mud then affected my plantain and maize crops,
so they were also damaged. I am the only one to produce cocoa beans in the area
and it is not proving a very successful crop because of the intense summers and
winters causing drought and flooding. I am still repaying my loan (seven years
later) with my earnings as a day labourer on other people’s land (Marco, 48).
Some farmers identified changing weather conditions as a cause of frustration. They felt that the
drastic temperature and climate changes in recent years, made their harvest times less predictable,
resulted in lower yields of produce, and made them more reliant on expensive fertilizer. This farmer
summarizes how these hostile and unpredictable weather conditions were to blame for his neighbours’
debt:
I know of neighbours who have taken out loans to start fruit or coffee production
and have failed to make a profit, only to be left with a debt instead. This is due to
the climate making land less productive (Luis Alberto, 45).
And in one community, Esperanzas del Mayo, farmers believed these climate changes had led to the
displacement of the entire vereda. Changing climate conditions affected not just farmers’ crops, but
their cattle too. Many experienced their cattle had died from dehydration. One farmer explained the
devastating effect of drought on his cattle:
I borrowed COL$20 million [about $6,340 USD] from the Agrarian Bank to buy
five cows and grass and have just finished paying it off this year (five years later)
including COL$8 million [about $2,530 USD] in interest. The loan was not very
productive as when I put the cattle on my new grassland most of it died. One out
of 15 cattle head survived (Alvaro, 55).
151
Farmers in some veredas in Putumayo and Caquetá, where soil had once been highly fertile, believed
coca production, as well as its aerial eradication, had contributed to their soil erosion. They associated
the repeated aerial spraying of glyphosate with reduced vegetable matter and microorganisms in their
soil. They described how it had taken several years for their soil to become fertile again, although they
now needed more expensive fertilizers to produce the yields required for commercial farming. They
felt these higher production costs made them uncompetitive in legal farming, even for staple crops
such as yuca and plantain. As one farmer explained, it made better business sense to work as a day
labourer in a coca plantation or cutting down trees:
Aerial eradication poisoned my legal crops, so I was not able to grow yuca or
plantain either. Our papaya crops were poisoned and animals that would feed on
them died. So, I had to work as a raspachin for COL$30,000 [about $9.6 USD]
per day to earn enough money to feed my family. It was hard on my hands. So
sometimes I would alternate this with cutting down trees to sell wood for which I
could earn up to COL$100,000 [about $32 USD] per day (Jose P, 70).
There are studies that support these farmers’ experiences with regards to the effects of glyphosate on
their environment (Kristina M. Lyons 2020, 180). Some suggest that glyphosate has no effect on soil
microbial communities, as any changes in this regard are solely to do with the place and time of year
(Busse et al. 2001). Admittedly, farmers all around the world use glyphosate as a common herbicide
and report no tangible detrimental effect on their soil. However, they usually apply it directly to their
soil and in small quantities. The indiscriminate use of glyphosate in high concentrations is a different
matter. This is the case when it is aerially sprayed by law enforcement in Colombia, over vast areas of
coca leaf as well as other surrounding crops and vegetation.
The additives and surfactants in the formulations of glyphosate, have prompted research into the
effect of its long-term, and high intensity, use on soil health. These studies have found that when
glyphosate is used in high concentrations, the recovery of soil in temperate climates is slow, and may
even exceed the duration of the vegetation period (Sihtmäe et al. 2013). This is because the use of
152
high doses of glyphosate affects microbial community functions and the bacterial composition of soil
(Ratcliff, Busse, and Shestak 2006; Zabaloy et al. 2012). It affects the equilibrium of endophytic
bacterial communities which colonize the interior tissues of healthy plants and help them supress
diseases, reduce contaminants and promote their growth (Kuklinsky-Sobral et al. 2005). Long-term
use of herbicides in general, are associated with depleted soil microbial biomass and its biochemical
activities (Sofo et al. 2012). The long-term use of glyphosate, specifically, is associated with
micronutrient deficiency (Lane and Dick 2012). The use of glyphosate has further been associated
with antibiotic resistance in fungi and bacteria, with chronic effects on animals and humans, with
changes in the microbial composition in soil plants and animal guts and with toxicity for some
amphibian species and crustaceans and water contamination (Van Bruggen and Morris 2018; Howe et
al. 2009; Tsui and Chu 2003; Amoros and Carrasco 2007).
Moreover, Amazonian soils are generally described as “thin, acidic and poor” and with a dominant
clay base by agrologists and farmers in Putumayo (Kristina Marie Lyons 2014). So compared to soils
in other parts of Colombia that tend to be a meter deep and get 90% of nutrients from the rip layers,
the land that many of these farmers have is deficient of the minerals needed for agricultural use.
Despite these challenges, some farmers had attempted to invest in new crops which they had no
previous experience with but hoped would be resilient to their climate and soil difficulties. They had
acted on advice given by AD programmes, only to be frustrated by the lack of results. They blamed
AD projects for influencing their choice of land use in ways that had so far not been productive. One
farmer recalled how he took out a loan ten years ago to buy rubber trees, on the advice from a German
aid organization working with INCORA, the land agency at the time. INCORA had been offering land
titles at the same time as these AD efforts took place. His investment had not given him the expected
returns on his loan:
I took out a loan to grow two hectares of rubber. I thought it would be easier
work, as they are trees. But in ten years they have only been drained once for
rubber which barely earnt me COL$800,000 [about $250 USD] (Cornelio, 55).
153
Some farmers said they had been given cattle by AD programmes but, as it came from warmer
climates, it didn’t adapt to their cooler environments. As a result, the cattle became stubborn, hard to
feed and to control, and had to be slaughtered before it could be sold. Others claimed that AD projects
had introduced transgenic maize and bean seeds that did not grow well on their land unless they
applied lots of fertilizer. This proved to be too expensive to be profitable. Others still, asserted AD
projects had given them sick chickens and pigs that were too expensive to feed. Overall, farmers
identified these shortcomings as reasons for why their investment had failed to be profitable. Their
experience echoes previous studies that have found AD programmes to fail because the alternatives
they disregard the local experience (Farthing and Kohl 2005).
Small land plot size
Another reason why it is possible that greater access to credit did not lead to more profitable farming
was that half of the farmers interviewed owned very small plots of land, of less than two hectares.
About a third owned or rented plots of land of more than ten hectares. The rest owned something in
between. This may have been an advantage for coca leaf, as small land plot size helps farmers remain
low profile and undetected by law enforcement and means that extra-legal groups may only be able to
ask for smaller cuts of their profits. However, when these farmers substituted their coca crops with
legal alternatives, their small land plot size may have made them unable to compete with
agroindustry. Farmers with small plots of land confirmed they could only get small loans, which
barely helped them get their subsistence farming back on track. Some of them believed that they
would benefit from economies of scale by attracting more commercial buyers. But there were
additional challenges that prevented them from coordinating their crops and working effectively in
associations.
This suggests that smaller land plots could have been another obstacle encountered by these farmers
as they tried to make their land commercially profitable. This proposition is consistent with previous
studies that have found that the quality and size of land have a direct effect on the value of land, and
the amount of credit available to owners (Galiani and Schargrodsky 2010; Tim Besley and Ghatak
154
2009; Ouedragogo et al. 1996, 232). If the amounts borrowed by farmers are small, they might not
level out consumption (Fenske 2011).
High production costs
Sometimes farmers who did not face the barriers discussed above still found that investment in their
land had not been or would not be profitable because the cost of the overheads was too high to make
their business profitable. When they had attempted to grow legal alternatives on land that was well
suited to these crops and with which they had had a lot of prior experience with, they still encountered
insurmountably high production costs. For instance, the cost of food, of day labourers, and of keeping
the soil fertile and free from pests, were regarded as new barriers to making their land productive in
legal agriculture. One farmer, for example, explained how the drastic drop in the price of coffee,
relative to the increased costs of production, had prevented him from making a profit from his land:
Coffee production is costly, and I am new to coffee production. One sack of
fertilizer costs COL$110,000 [about $34.50 USD] and one arroba of coffee costs
COL$70,000 [about $22 USD]. You need two sacks of fertilizer to yield three
arrobas of coffee. I’ve taken out a loan and had to sell some parts of my land as
the price of coffee is too low to pay for production supplies. If the price of coffee
does not rise, I would have to find another way to repay my loan… I am losing my
morale as a coffee producer. I think the solution is for the state to increase the
price of coffee (Fernando, 48).
In addition to the increased cost and need for fertilizers and pesticides, some farmers believed that
production costs had become high, in great part, due to a reduced rural labour force after the coca
bonanza. In the period prior to coca leaf production, farmers had had bigger families. Most of those
interviewed, had more than three siblings. As children, they had usually taken it in turns with their
siblings to attend school and to help with farming at home. However, most farmers now had smaller
families themselves as 64% of them had two children or fewer. Besides a decrease in the size of
155
families, farmers experienced their children had more access to education than they had had, and the
younger generation was now moving away from agricultural labour. This suggests that legal farming
could have been more profitable in years past because there were numerous children available to work
the land, at little extra cost.
In addition to this, farmers experienced a shortage in labour supply after coca leaf stopped being
grown in their vereda. They perceived many farmers to have left, as they followed coca leaf
production elsewhere to secure higher wages. Those who had stayed and substituted their coca leaf,
found they needed to pay for day labourers, unlike before the coca bonanza. However, paid labour
was too expensive. One farmer explained that people tried to redress this by helping each other out for
free:
People who did not own land and had been working as coca day laborers just
switched to other types of crops and were paid the same (about COL$25,000
[about $7.8 USD] per day). But as there are fewer day laborers around, the cost
of wages has increased. So instead of relying so heavily on day laborers, some
farmers try to help each other out in turns (Evelio, 39).
This included working on each other’s land periodically when more hands were needed to clear the
soil, to prune crops, to spread fertilizer or during harvest seasons.
A recurring theme that emerged in the data was the local urbanization that many farmers perceived
following the coca leaf bonanza. They shared their concerns about the phasing out of farming jobs, as
succinctly expressed by one farmer:
Many people have abandoned the countryside and moved to urban areas. So,
there are fewer farmers (Carlos Arturo, 50).
156
This decreased reliance on agricultural resources for rural livelihoods has previously been identified
as a general patter in the Rural South (Rigg 2005). Agricultural labour competed with construction
and cleaning jobs, and this seemingly further increased agricultural labour costs.
Investment risk factors
Finally, another factor that undermines hypothesis C is the uncertainty of investment. There were
farmers who felt their investment had failed to make their land more productive because of risk
factors that were beyond their control. Crop prices could drop or fluctuate unpredictably, sometimes
daily, causing investment loss. For example, the price of panela, processed sugar cane, has
significantly decreased since 2016 (Bloomberg 2018, cited 6 April 2019). One sugar cane farmer
claimed the local price had dropped at some point in 2017 from COL$40,000 [about $12.5 USD] to
COL$25,000 [about $7.8 USD] in the space of a week.
Some farmers assigned smuggled produce the blame, for the falling prices for their produce. They felt
they could not compete while these smuggled agricultural goods were flooding their market. A farmer
described this issue relating to chicken and eggs:
There is a lot of smuggled produce coming from Ecuador via Pasto e.g., chicken
and eggs, and they flood the campesino subsistence markets (as the big cities are
already saturated). For instance, in our village they were selling a box of eggs for
COL$7,500 [about $2.3 USD], while the local cost was COL$9,000 [about $2.8
USD] because they fed the chickens with corn, which is more expensive (Carmen,
43)
Another farmer recalled a similar experience with pepper:
There is a lot of pepper smuggled in from Ecuador. When there were no illegal
imports from Ecuador, and legal imports from India, we would sell one kg for
COL$22,000 to COL$23,000 [about $6.9 to $7.2 USD] and now it’s gone down to
157
COL$6000 to COL$7000 [about$1.8 to $2.1 USD]. One kg of white pepper would
sell for COL$28,000 to COL$29,000 [about $9.1 to $8.8 USD], but now it’s down
to COL$12,000 [about $3.8 USD]. Buyers used to come to the vereda to bulk buy
the pepper, and now producers must go all the way to town to sell it (Rito, 53).
A few farmers’ circumstances had also unexpectedly changed due to illness and theft. One farmer
recalled how the cattle she bought with a loan was stolen and became ill:
In 2012, with our land title we took out a first loan of COL$12 million [about
$3762 USD]. We bought 10 beef cattle head, prepared the grass for grazing, and
fenced the paddock. But some of our cattle was stolen and the rest became ill with
a virus called el carbon, that turns their blood black and they had to be put down.
So, we lost it all and still had to pay the loan back including the interest (totalling
COL$16 million [about $5013 USD]) without having made any profit (Lilian, 34).
Another farmer described how the money she had borrowed was stolen by a relative:
I have taken out a loan of COL$3 million [ about $940 USD] from the Agrarian
Bank since I received the land title. It was to buy some pigs and sugar cane crops.
However, I was taken for a ride by my daughter’s boyfriend who tricked me into
giving him COL$1.5 million [about $470 USD] to get started. So, I used the loan
to give him that money. And then he disappeared. I am still paying it in biannual
instalments (Maria I, 55).
These were the unforeseen ways, often beyond farmers’ control, in which some farmers accounted for
failing to make profits on their loans. This finding coincides with previous studies which identify
poverty and marginalization as explanations for AD failure (Lupu 2004).
158
Hypothesis D: Farmers’ increased productivity in legal farming reduces their
dependence on illegal drug production
As discussed earlier, policymakers have introduced land titling as an AD measure on the assumption
that farmers with land titles can use their land as collateral to access bigger loans from banks. The
policy is based on de Soto’s assumption that access to loans increases productivity. The logic follows,
that as productivity in legal farming increases, farmers have fewer incentives to return to illegal drug
production. Hypothesis D seeks to understand whether this was the case in the communities included
in this research. As already came up in hypothesis C, my findings revealed that farmers did not
perceive the new availability of credit to sufficiently improve land productivity in legal alternatives to
make coca leaf production less attractive. Not only was this the experience of a great number of
farmers, but it was evidenced by the fact that some farmers admitted they continued to work with coca
leaf elsewhere. In addition, a few farmers felt increased dependence on illegal crop farming because
of taking out loans. This was because they struggled to repay them with the proceeds of legal
agriculture. I will now discuss these findings in more detail.
No increase in land productivity
Most cases did not support the hypothesis that farmers were less attracted to, or reliant on, coca leaf
farming because of their increased access to loans. Of those farmers who were considering producing
coca leaf again, half had no desire for future access to credit, and only about a quarter wanted to apply
for a land title specifically to access a loan. Only 14% of those farmers who were struggling to make
ends meet (i.e., had visibly humble lodgings, dishevelled clothes and gave narrative evidence
confirming they lived on the breadline), were applying for land titles specifically to access credit.
About half of them had already taken out loans, with and without titles, and felt they were not in a
better position because of having done so. And crucially, about half of them were considering
returning to coca leaf production. This suggests that on a balance, farmers hoping for a better future
159
were not overwhelmingly convinced that a loan was the answer. In the interviews, coca leaf
production was usually perceived to be a better and faster way of raising money than through loans.
As many as three quarters of those farmers who had used a land title to access larger loans, appeared
to live above the breadline. Half of these had been large scale coca leaf producers i.e., farmed it on
more than four hectares of land, which suggests that they were also better off because they had
accumulated more capital from illegal crops than others had been able to. Three quarters of the title-
based loan applicants also had larger plots of land, of more than 10 hectares. These circumstances
made title-based credit applicants better off and more likely to turn the larger loans to profit.
Nevertheless, they remained uncertain as to whether the loans could continue increase their land
productivity as only 10% of them planned to take out additional loans at some point in the future.
This gives us an idea of the proportion of my broad findings; however, the thematic analysis of the
interviews provides a much clearer dimension of the issues farmers faced. Farmers had not always
expected to be in a better situation than when they had produced coca leaf. Most farmers who chose to
substitute their illegal crops and apply for land titles were prepared to take a pay cut when they chose
to transition to a trade in which they would be comparatively disadvantaged. Indeed, in some places
coca leaf farming had not always been significantly more profitable than legal farming. For some
farmers, their experience of coca leaf crops was that they were investment heavy (for pesticide, day
labourers and fertilizer) and not always profitable, also because of the fluctuations in the price of coca
leaf. When this was the case, these farmers had spread the risk of price fluctuation by growing a
variety of other crops among their coca leaf crops.
But in general, farmers interviewed in marginalised places with poor soil, considered that earning a
living from legal crops was harder and more expensive than coca leaf. Their experience was that coca
leaf provided harvests from three to six times per year, it required little infrastructure, and was suited
to cheap transport, with no need for good roads. They considered coca leaf to be a hardier crop that
would grow on poor-quality soil, steeply sloped terrain and, like a weed, resist climate variations and
160
pests. This confirms what has been found in previous studies about the attraction of coca leaf farming
(Rojas 2003; Dion and Russler 2008, 405).
When farmers had stopped growing coca leaf in these research locations, they had generally expected
to earn enough to cover their basic living expenses with the support of AD subsidies and their own
increased access to credit. They had all been approached by AD programme officials who encouraged
them to substitute and reassured them they would be able to make a living from an alternative legal
use of their land. Many farmers had trusted these assurances. However, since then, most felt unable to
overcome the obstacles to making a profit from legal farming, for reasons discussed in hypothesis C.
For some farmers, the macro-barriers to land productivity in legal farming that had existed prior to the
coca leaf bonanza, did not disappear when they had substituted coca leaf. They were uncertain
whether bank loans would help them overcome the factors that had led them to start producing coca
leaf in the first place. Farmers with poor quality land as well as low levels of education and a limited
skill set, felt particularly vulnerable. They did not have the option of replacing, or supplementing,
their income from subsistence farming, with non-agriculture related employment.
All this meant that farmers in places where agricultural production was challenging anyway, largely
did not consider private bank loans to be the answer to their poverty. Many of these farmers who
wanted titles to take out loans, thought it more practical to use them to improve their house or to pay
for their children’s education, than to invest in their land. Others wanted titles for loans because they
had been encouraged to do so by more literate and persuasive AD programme officials. However,
when I questioned their new business projections many revealed low levels of financial literacy. This
suggests their enthusiasm for loans was not always based on sharp economic calculations or well-
developed business plans. Farmers who had taken out loans, both through informal channels and with
the help of land titles, had not improved their land productivity. Despite some cases where investment
had enabled farmers to make a comfortable living from their land, their profits from legal alternatives
had not outperformed their previous income from coca leaf production. This suggests that at the time
161
of this study, farmers were unconvinced their increased access to credit, facilitated by land titles,
would increase their land productivity.
It should be noted that those who were able to move to urban areas to find alternative livelihoods by
becoming construction workers, cooks, shopkeepers, motorbike-taxi drivers or even to become coca
leaf manual eradicators, were not interviewed in this study. Therefore, the experiences reflected here
are only those who relied on land to earn a living.
Return to illegal drug production
In addition to title-based loans failing to increase farmer’s agricultural productivity in most cases, I
found that since the introduction of land titling through AD programmes, a fraction of farmers
admitted that their dependence on illegal drug production had not ceased, while many others stated
they were open to producing coca again if AD didn’t work out. Coca leaf continued to be the answer
for those seeking to supplement their income at home. This finding echoes those of a monitoring
survey which estimated 0.6% of farmers in the most recent AD programmes returned to the coca leaf
business (UNODC 2019b). However, from my sample, I would argue that the figure is much higher
than 0.6%. This official figure may underestimate how many farmers return to coca leaf, because it
excludes those who are considering returning, and only reports those who are “caught” doing it.
Perhaps fewer people own up to doing it when surveyed by the UN, than in my fieldwork study. Even
so, it should be noted that the sample of farmers in this study did not include those who may have
been working away on coca leaf plantations at the time the interviews were conducted. It is therefore
possible that my study also underestimates the trend of farmers returning to drug production.
Furthermore, the 0.6% figure calculated by UNODC does not reflect geographical differences in the
likelihood that farmers will return to coca leaf production. Farmers in the communities I visited, who
failed to earn enough from subsistence farming at home, felt they could either seek employment away
from their land, probably without success, or return to illegal drug production to supplement their
income. I found this trend was most salient in the municipality of Rosas in the department of Cauca.
162
A relatively high number of farmers in these veredas admitted they had returned to working in coca
leaf production elsewhere, after having substituted their illegal crops at home because since they
ceased to grow coca leaf at home, they had become unemployed and underemployed. In some
veredas, the substitution of illegal crops caused unemployment even for people who did not work in
agriculture. The abandonment of coca leaf production contracted the local economy and so even non-
land related businesses went bust.
These farmers casually explained how and why the coca leaf business continued to be their main
source of income:
Some people have moved their coca production further to the border with
Ecuador. They only go to plant their crops and then to pick the coca once every
three months and, in the meantime, they live here where they own shops and
businesses (Sandra M, 35).
Coca used to be my main source of income and I am a single mother and the sole
breadwinner. I spend about COL$50,000 [about $16 USD] on shopping per week.
I also used to have a job in a local restaurant but as the coca production shifted,
the business went broke. So, I try to earn money selling yoghurt locally but mostly
I have continued working as a raspachin (Hortencia, 45).
With no agricultural opportunities at home, what other kind of work can we
realistically get? [I work as a raspachin] and go for two weeks at a time and come
back for a one-week break. I go to various places depending on where there is
work … My monthly outgoings are about COL$800,000 [about $255 USD]
including my children’s education. …. I don’t work out of greed, buy out of
need…. At home, I would be paid COL$15,000 to COL$20,000 [about $4.8 to
$6.3 USD] as a day labourer whereas as a raspachin I earn from around
COL$100,000 to COL$120,000 [about $31.8 to $38.2 USD] a day. That’s why I
163
have gotten used to doing this work… I’m also producing coffee and sugar cane at
home, but the fertilizers are very expensive. Sugar cane earns me about
COL$300,000 [about $95.6 USD] per month, for half the year. The rest of the
year it makes no profit. One arroba of panela costs COL$17,000 [about $5.4
USD] to produce and is sold for COL$35,000 [about $11 USD] and it takes a
whole year to process that arroba, so it’s not worth the effort when coca leaf
grows even on stones (Meregildo, 38).
Farmers who opted to return to coca leaf as their means of livelihood, felt they only had two options:
to grow coca leaf on land elsewhere, or to work as day labourers in coca leaf plantations, what was
commonly referred to as raspachin work. They explained that these jobs were still risky but paid more
than legal farming did in their home community, possibly because one result of crop substitution is
that there is too much supply of the legal alternatives. These two farmers illustrated the significant pay
cut involved in ceasing to work with coca leaf:
I have a brother who has coca crops in a small land plot he rents further down the
mountain where the climate is warmer. He earns good money about COL$2
million [about $635 USD] every four to six months… with our three hectares of
coffee we earn COL$3 million [about $952 USD] a year in coffee production …
With our plantain, we also earn about COL$60,000 [about $19 USD] every two
weeks. We have 200kg which sell at COL$300 [about $0.09 USD] each, but it can
go up to COL$800 [about $0.25 USD] per kg. Plantain prices have dropped in the
past two years because there is too much supply (Daniel A, 33).
In my borough, I would be paid COL$20,000 [about $6.3 USD] per day as a day
labourer, not including lunch. So, it ends up being something like COL$12,000
[about $3.8 USD] after paying for lunch and transport. But in coca production
people get paid between COL$300,000 and COL$500,000 [about $95 and $159
164
USD] per week. The difference is immense. But with armed groups in control of
the coca plantations, it is risky work too (Jesus Efren, 37).
Farmers who chose to grow their own coca crops in land far away from the place they lived, either
bought land cheaply, as it was of low value, or “crop shared” with the landowners. This meant they
would pay for the maintenance of the coca crops and give the landowners half their earnings. Farmers
returning to coca leaf production in this way chose to do so for various reasons. Firstly, in places
where land titling was conditional on there being no coca leaf crops in the entire vereda (a policy
called “zero coca vereda”). If AD officials found coca leaf anywhere in the vereda, then the culprits’
neighbours would not be eligible for land titles and thus for bank loans, which created an element of
peer pressure to not grow coca leaf at home. This is where land titling, along with other factors, might
have played the largest role in influencing farmers to substitute their illegal drug crops at home.
Secondly, these farmers felt it was safer to grow coca leaf separately from any of their food crops to
avoid the latter being destroyed by aerial eradication campaigns. Farmers who wanted to re-plant coca
leaf were also encouraged by their neighbours to do it in plots of land further away for this reason.
Coca leaf farmers would be held responsible by the community for any damage that would come to
their crops or cattle. Sometimes farmers opted to grow coca leaf in land further down the hills because
there were more favourable climate conditions so they could yield more coca leaf than they would on
their homeland.
Thirdly, extra-legal groups were no longer present in these communities and farmers wanted to keep it
that way. So, they chose to separate their illegal drug production from their home crops to protect
their families from being exposed to violence and extortion, like they had been in the past.
Some farmers chose to continue working in illegal drug production by becoming raspachines.
Raspachin work was less financially risky than producing their own coca crops as they did not have to
invest in fertilizers and fungicide for the crops, and they were paid daily. On the other hand, it was
less profitable than owning coca leaf crops and much more arduous work. It involved walking
165
between three and eight hours with their luggage to get to the coca plantations to work as temporary
labourers on coca leaf plantations hand picking coca leaves, then spraying the crops with pesticide
and placing some fertilizer by the stem. Sometimes it also involved cocaine processing in cocaine
labs, or cooking meals for the other raspachines. People who did this were constantly exposed to
headache-causing chemicals and had to wrap their fingers in cloth to protect them from bleeding and
blistering. They were not allowed to use gloves or machetes to avoid harming the crops. The living
conditions were hard, their sleeping arrangements were on the floor or in shared rooms, they had to
bathe in the river, there was no phone signal or electricity, there was poor ventilation, the food is basic
e.g., cooked bananas, a spoonful of rice and some broth and in small portions. They also had to work
under other people’s command, with the surveillance of armed groups so the atmosphere was tense.
Farmers said they needed community leaders, or JAC (Community Action Board) presidents, to give
them supporting letters of recommendation to vouch for their honesty and reliability, as they were not
easily trusted by the coca plantation owners. This farmer’s testimony gives us an insight into the risks
involved in raspachin work:
In the last five years, I feel my health has been affected by the chemicals I touch
there all the time… coca is sprayed a lot. I feel my bones and muscles are weak
and I have joint pain and headaches. I used to go for two to three months at a time
but now I cannot last more than two weeks working there. I would prefer to work
in a sugar cane plantation as it is as much work as coca, but you are close to your
family and don’t feel vulnerable all the time. You can enjoy your life (Jesus Efrein,
37).
Overall, farmers believed that if people could afford to buy or rent land elsewhere, producing their
own coca leaf crops was less time consuming than raspachin work, which gave farmers more time to
work on other legal crops and jobs at home. On the other hand, illegal crop owning producers were
more exposed to losing their investment than raspachines. Farmers also believed it was easier for
younger people to move away from home and become raspachines. They said younger people were
166
more efficient picking coca leaf, so their earnings would make their move away from home worth the
effort.
Possible counterproductive effect of land titles
Most farmers who substituted their coca leaf at home, but continued working in the coca leaf business
elsewhere, did so because they believed it was the only way to supplement their livelihoods at home,
or they that wanted to pay for their own or their children’s education. Without this supplement income
from drug production elsewhere, they felt they were unable to cover these costs due to the
considerable pay cut involved in their switch to legal farming at home. What was more surprising,
however, was that some did it to pay off their debts, which in a few cases had been accrued from
loans they had accessed with the help of land titles.
In stark contrast to what I expected, it was the loans which they had taken out that made a handful of
farmers feel equally or more reliant on coca production. They saw it as the best way to pay back their
non-performing loans. Some of these farmers even got certificates from their banks for being a good
payer (i.e., buen pagador) as they had paid their loans back in time, despite having done so with
income from their illegal crops. This suggests that increased access to credit has a potential effect of
augmenting some farmers’ dependence on illegal drug production. These farmers’ accounts reflect
this pattern:
INCORA offered me a land title. They said that this would help me access loans
from banks… So, I took out a loan for cattle, as this was the only thing, they would
lend money for. I remember it was the World Bank that loaned me 14 beef cattle
head and 1 bull. Then I started growing coca in three hectares of my land. With
this I quickly paid the loan back (Hipolito, 72).
Those that remain in the vereda feel strongly tempted to start growing coca again,
about five or six already do. About 20 young people continue to work as
raspachines and domestic workers in temporary placements. One of my sons took
167
out some loans and has a small internet café. The other one works in a
supermarket. Every two to three months they go to pick coca to earn some extra
money to pay back their loan. People [from coca leaf plantations] sometimes
come here looking for workers, it is not always people who go out looking for jobs
(Carmen, 43).
As my findings for hypothesis A illustrated, farmers had different sources of credit at their disposal.
Bank loans were available to them without land titles but only up to a point, as these farmers could
only borrow small amounts of credit from banks. The lack of formal financing left some of them prey
to informal financiers and informal loans companies, as they did not require as much paperwork, as
formal banks did, to disburse bigger loans. Farmers who accessed loans this way, and failed to see any
returns to their investment, became tied to very high interest rates, and spiralled into bigger debt. It
could take them up to eight years to pay back the interest on their loan. Because informal financiers
and loans companies work without insurance, they tend to issue threats to farmers who fail to pay on
time, sometimes forcing them to sell their house, crops and cattle and creating anxiety that they will
repossess their land (Nunez 2018). These threats may have pushed some farmers to resort to illegal
drug production again to pay back their loans.
The government encourages farmers to get land titles to then take out formal bank loans. It is feasible
that this helps to steer farmers away from informal financiers and loans companies. Moreover,
because bank loans work with insurance, they discourage farmers from borrowing more money than
they will realistically be able to pay back given the size and value of their land. The Agrarian Bank, a
bank heavily subsidised by the Ministry of Agriculture, issues loans, with very low interest rates, to
small scale farmers (such as those interviewed in this study) for agricultural purposes. This bank does
not lure applicants with other products to get them into more debt than they need to be. Besides being
covered by insurance, private banks are more flexible than informal financiers and loan companies in
how they recover their money. The Agrarian Bank tends to use positive incentives and other means
before it attempts to repossess farmers’ land. For instance, a programme called AgroListo rewards
168
small scale farmers who pay back their loans on time. It basically facilitates further loans for them and
minimizes the paperwork required for it. It promotes association and cooperatives. Moreover,
commercial banks only require farmers to use their land as collateral for loans over COL$20 million
[about $6,400 USD]. All this suggests that farmers who use their land titles to access bank loans but
who fail to pay them back, are less exposed to losing all their assets.
The legal process to repossess land, for those who fail to repay their loans, is long. First, a lawyer
must issue a formal request for repayment with a deadline, of about 180 days. If guarantors aren’t
available to pay, then a bank representative will talk to the debtor and offer to cancel the loan interest,
so all that is required is the repayment of the sum that was borrowed. If debtors are still unable to pay,
a judge issues a sentence to repossess their land, which means the land cannot be sold and the
farmer’s credit history is blacklisted. This sanction is lifted as soon as the farmer repays. If the land is
of some value, banks may proceed to sell the land in a public auction. Alternatively, they may give the
details of the debtor to third parties who may offer to buy their land for a very low price. If debtors die
before they repay, their life insurance covers the cost of their debt.
However, land repossession of small farmers is scandalous and highly politicised, and draws
significant media attention given historical tensions around inequality in land ownership (Serrano
2015; IPC Press Agency 2010). Therefore land repossession is only applied by banks in extreme
cases, and as a last resort, when the debtor does not want to or cannot pay back their loan (Nunez
2018). I was unable to identify any cases of land repossession in the areas where I carried out my
research.
Given the rarity of land repossession by banks, one would think that debtors with land titles, might
feel under less pressure to repay banks, than those who owe money to informal financiers and loans
companies. Although no one could point to an experience of either themselves, or a neighbour, having
their land repossessed by a bank, the perception that they could, was palpable. Based on the interview
evidence, most farmers did not seem aware of the above information that makes it highly unlikely
their land would be repossessed if they defaulted on their loans. Many farmers applying for land titles,
169
still felt anxious that banks, or the state, might repossess their land if they failed to pay back loans,
they accessed by using their land as collateral. It was this fear of losing their land which was enough
to drive some farmers to pay back their bank loans by resorting to the coca leaf business. One farmer
described his experience of having to sell his land to pay off his debt due to this threat of
repossession:
With a land title, the (Bic) bank loaned me COL$2 million [about $639 USD],
then COL$5 [about $1598 USD] million and then in COL$9 million [about $2876
USD] to buy 105 cattle head. But at this point, I was unable to repay the last loan
as all my cattle was stolen. So, they threatened to repossess my land and I ended
up selling it all off, to pay the debt with interest. This amounted to COL$16
million [about $5113 USD] (Arcadio, 67).
It is not possible to deduce from this research the direct effect of land titling on the displacement of
coca leaf farming. However, the interviews suggest a plausible way in which the two may be
associated to one another. Farmers who titled their land and then took out loans they could not pay
back with the proceeds from their land, had an underlying fear that banks would repossess their land.
And some of these farmers sought to resolve this by returning to the coca leaf business elsewhere.
Summary of findings
Farmers’ experiences supported hypothesis A. Increased amounts of credit were available to new land
title holders. Farmers felt that land titles had not always been essential for them to access credit
because borrowing money informally was common practice prior to land titling in their area. In
addition, smaller commercial loans, at the banks’ discretion, had been available to farmers without
land titles. However, when land titles were introduced in their vicinity, farmers perceived that the
credit market and culture changed. They felt that possession of land titles had become essential for
them to access larger bank loans.
170
The interviews only partially supported hypotheses B, C and D. Farmers interviewed were roughly
split evenly between those who wished to use their greater availability of credit to invest in their land,
and those who did not wish to use their land as collateral to take out loans. Those who did not want to
take out loans, were aware of the insurmountable barriers they would need to overcome to make their
land more productive. They tended to be older farmers or located in places that were geographically
marginalised or with low quality soil.
Of those who did access credit because of land titles, only a few experienced improvements in their
land productivity. They were most often in places well suited to agricultural alternatives, or where the
climate was not ideal for coca leaf crops. They were also farmers who had accumulated capital from
their former illegal drug production, as they generally possessed better and larger plots of land.
The rest had been unable to make their legal alternatives profitable. The lack of road access and
infrastructure, poor quality of the soil, and drastic climate conditions were the main reasons farmers
identified for investment in land not rendering profitable. In addition, it transpired that farmers’ small
land plot size, high production costs and usual investment risks such as fluctuations in alternative crop
prices, theft, and illness, were other factors that contributed to sentencing farmers’ land to subsistence
farming. These factors are what many farmers said had led them to start producing coca leaf in the
first place. Some believed these challenges had become worse over time due to climate change, the
effects of aerial eradication on the soil, the increased cost of labour and urbanization.
Overall, farmers felt land titles and the bigger loans they offered, had not improved their
circumstances and many had returned to subsistence farming after eradicating coca leaf. The fact that
hypothesis C was not supported by the data, meant that hypothesis D was also flawed.
Counterintuitively, the income shortfall combined with increased debt, and a culture of fear that that
their land will be repossessed by banks if they defaulted on their loans, had made a fraction of farmers
with title-based loans return to the coca business to pay them back on time.
171
1.6 Discussion & Policy Implications
Although in some cases land titling helped AD, more often it did not. Why this heterogeneous effect?
Amidst the list of reasons already identified, three broad explanations can be identified. One is with
regards to the agency of the individuals. Farmers’ investment decisions were sometimes unwise or
uninformed. This was reflected in their choice of crops, and how well suited it was to their soil.
Sometimes farmers felt ill advised on their choice of how to spend their loans. The examples of
people spending their credit on fraudulent pyramid schemes, is a case in point.
Another explanation is that the benefits of land titles in AD are largely associated with increased land
tenure security. And yet, land tenure security already existed in these communities as most farmers
felt they already owned their land with de facto property rights. In this sense, there was no real
dichotomy with legal rules. So, even if land titling is becoming an increasingly popular policy in AD
efforts, informal property rights mattered a lot in these communities.
A third explanation for the heterogeneous effect of land titling is that the value of the land assets was
often too low. This meant that the land for which farmers were being issued with land titles was often
not easy to sell even with land titles. Moreover, farmers felt it was difficult to make the land more
productive than it already was, even with greater access to personal loans.
Farmers’ belief that land titles could not resolve the issues of unemployment and their lack of land
productivity, echoes Bromley’s (2009) findings. He found that individual investment is insufficient to
increase productivity in geographically and economically marginalized farming communities. These
communities’ geographic isolation from the larger national economy limits their ability to invest in
land productively. For example, a lack of road access restricts farmers’ access to the market because
they encounter “large and asymmetrically distributed transaction costs” such as additional transport
costs (Bromley 2008, 24–25). Without state investment in local infrastructure, individuals will spend
their private capital acquired from using their land as collateral for loans trying to create the missing
172
public good by private means. This leaves them with fewer funds available to invest in their land asset
(Bromley 2008, 25). In this sense, loans from land titles are not the answer to poverty.
This research reveals that for many of the farmers interviewed, land tenure was already secure.
Therefore, the benefits of land titling boiled down to their symbolic meaning more than their function
as collateral. It is therefore over-simplistic for policymakers to think that land titles can sustainably
curb illegal drug crop production by increasing farmers’ access to loans.
There are two policy implications that come into view. First, access to credit should not be promoted
as the pathway out of illegal farming. Second, policymakers should be aware that land titling can
displace coca leaf production elsewhere.
Play down access to credit as the solution to productivity
Like many other small-scale farmers in rural Colombia, the farmers interviewed in this study faced
numerous challenges to making a comfortable living from their land. It is usual for farmers like these
to rely on credit to tie them over until they have a decent harvest, or their cattle is ready to be sold. It
is also usual for anyone seeking a different agricultural use of their land to rely on credit if they do not
have capital at their disposal. This means that many small-scale farmers who fail to profit from their
land, and aspire to something more than subsistence farming, live in debt.
Bromley (2008) suggests that increasing landowners’ access to credit works in the interests of the
mortgage lenders rather than the borrowers. If land titling serves the purpose of commodifying land
for the credit economy and promoting AD with private loans, instead of exclusively through rural
state subsidies, it is farmers, instead of the state, who assume the cost of owning unproductive land. It
is feasible that banks might benefit from the illegal drug trade by issuing loans to farmers who pay
them back with their revenues from illegal coca production. This may be a question for further
research.
173
Bromley further suggests that a more effective way for policymakers to increase land productivity in
agricultural settings, is not through land titling, but through tackling the underlying factors that cause
land to be marginalized from the wider economy in the first place. The creation of jobs and
investment in road access are more suitable policy responses to increasing land productivity than land
titling. Farmers are better able to devote their resources to increasing their net returns only once these
macro-barriers have been overcome (Bromley and Chavas 1989). This validates the argument that
“investment [in infrastructure] is necessary to obtain [land] security” not the other way around
(Bromley 2008, 24; Sjaastad and Bromley 1997, 553). Investment in public infrastructure and market
integration should thus take a higher priority in AD efforts than land titling (Dion and Russler 2008).
Moreover, AD policymakers should be aware that promoting increased access to credit may not in
itself resolve the obstacles to making land more productive, when it is done on the cheap, and can lead
to disappointment among farmers. Policymakers should not assume or lead farmers to believe, that
individual farmers’ investment of their loans in their land will enable them able to overcome the
macro-barriers to development. The interviews revealed that there had been some mis-selling of the
alternatives. And even if farmers adopt a particular alternative en masse, and succeed, this may
undermine the profitability of the crop and hence their livelihoods. For this reason, they should not
encourage farmers to endorse land titling simply on the grounds of being eligible for increased access
to credit.
Previous research in Africa has found that the proselytising of microfinance can offer low income
entrepreneurs the illusion of a quick fix without addressing the more profound socioeconomic issues
that need to be tackled for anyone to succeed (Buckley 1997). Similarly, giving unrealistic hope to
former coca leaf farmers that their land can become productive in legal agriculture, beyond
subsistence farming, with the help of a private loan is misleading and can cause disappointment,
ultimately lessening the credibility of other AD efforts in these communities. This is especially the
case in contexts where farmers’ land productivity is challenged by macro-barriers to development,
unsuitable climate, and poor soil quality of their land. If they fail to earn enough from their legal
174
agriculture to meet their basic needs, farmers may be even worse off if besides suffering an income
shortfall they have additional loans to pay.
Risk of spatial crime displacement
The spatial displacement of the coca leaf business because of drug supply reduction strategies, is
commonly referred to as the “balloon effect”. It is defined as the geographical displacement of either
drug production or of drug trafficking routes, through the movement of people (Barret 2010). And it
happens in response to drug supply reduction policies causing a “labour force pre-equipped with coca
[or other illegal drug crop] production know-how” to spread to places where there were previously no
illegal crops (Dion and Russler 2008, 404). Alternatively, it intensifies drug crop production in areas
with less state infrastructure, and in more dispersed plots. Although it is usually associated with
forced manual and aerial drug crop eradication efforts, as this study illustrates, it can also occur
because of AD efforts. And this phenomenon may be more common than the 0.6% that is reported in
official UNODC surveys.
Some argue that even when drug production is spatially displaced as a result of drug enforcement or
AD, farmers who have been sanctioned or subsidised by the state before, will perceive greater
uncertainty regarding the risk of being caught by law enforcement again (Windle and Farrell 2012,
872). So the displacement of drug production will never be 100% (Windle and Farrell 2012).
Moreover, even if drug supply reduction policies may not reduce the overall net drug production, they
can have a “diffusion effect”. This is when they make it costly for illegal economies to reorganize,
thus making them more vulnerable to law enforcement. This disruption then deters some farmers from
entering the drug trade because they anticipate increased confrontation with law enforcement (Windle
and Farrell 2012, 870).
However, the profits made by the illegal drug trade are so high, that the illegal drug business is highly
tolerant and resilient to risk. It retains exceptionally high returns despite having to shift its operational
strategies and alter its locations in response to drug supply reduction policies (Rouse and Arce 2006,
544). But the real problem with the balloon effect is that it pushes drug production to more socially
175
and environmentally vulnerable areas (Rincon-Ruiz et al. 2016). It can then be harder to wean farmers
in these places off the proceeds of illegal drug farming. It can also result in there being even more
drug production elsewhere because the shift of location causes a temporary reduction of drug crops,
and this increases their value (Laffiteau 2011).
Some argue that Colombia became the top coca leaf producer precisely because of aggressive coca
leaf eradication campaigns in Bolivia and Peru (Dion and Russler 2008, 400; Rouse and Arce 2006;
Mora 1996; Ramirez 2011, 56). Similarly, the production of opium poppy in Afghanistan is believed
to have increased in response to opium supply-reduction campaigns in the 1970s in Pakistan, and
opium production in Myanmar increased because of supply reduction efforts in Thailand (Seecombe
1995).
Therefore, it should be of concern to policymakers that land titling in AD can contribute to the
balloon effect. Some farmers who had acquired, or were in the process of acquiring, land titles
stopped coca leaf production in the land plot for which the land title was applied for, but then
continued to work in coca leaf production elsewhere. This means that the policy may not only have
little bearing on the net reduction of illegal drug crop production, but it can spread the know-how to
new places or intensify it in others, which only exacerbates the problem in the long run.
Farmers did this when they faced significant obstacles to making their land more productive with
legal farming and could not supplement their income with non-land-based work in an urban context.
And in a fraction of cases, the pressure to pay back bigger title-based loans, combined with their fear
that their land would be repossessed (even though in practice land repossession of this kind is
unrealistic), contributed to the pressure they felt to raise money quickly and to their temptation to
return to illegal drug production elsewhere. Policymakers could address these causes by realising that
asking farmers to give up the coca business, at the same time as encouraging them to take out loans
when their prospects of making their land more productive is uncertain, is risky.
176
AD officials who promote land titles by encouraging farmers to use their land as collateral for bank
loans, should therefore make it clearer to them that they will not automatically have their land
repossessed should they face difficulty in paying their loan back. This could help redress farmers’
anxiety to avoid land repossession if they are unable to pay back their loans on time and reduce the
pressure, they can feel to return to coca leaf farming elsewhere.
Another reason why farmers recommenced their work in coca leaf farming elsewhere was because
they did not want to obstruct their community from qualifying for AD benefits. At the time of this
study, the land titling for AD was conditional on the entire vereda being free of coca leaf crops.
Perhaps a compromise for policymakers would be to rethink whether this is the best strategy, given
the perverse incentives this creates for farmers to further push drug production to more marginalized
and vulnerable areas.
177
1.7 Conclusion
My research shows that, in the areas I examined in Colombia, increased amounts of credit were
available to new land title holders. However, just over half of the farmers who had the new
opportunity to take out bigger bank loans, opted to do so. There were a few cases of farmers who
experienced improved land productivity in legal alternatives because of their increased access to
credit with land titles, but this effect was, at best, heterogeneous. In most of the cases I came across,
farmers’ increased availability of bank loans from using their land as collateral, did not improve their
circumstances.
Why was this so often the case? Amidst the list of reasons identified, three broad explanations can be
identified. One is with regards to the agency of the individuals, as they often believed they could
make better investments in things other than legal alternative crops. Another reason is that land tenure
security already existed, so a land title did not make a huge difference in other types of land
transactions. The benefits of land titles are often associated with increased land tenure security which
most farmers felt they had. And thirdly, the value of the land assets was too low. Farmers felt it was
difficult to make the land for which they were being issued with land titles more productive than it
already was, even with greater access to personal loans. They found that individual investment was
insufficient to increase productivity.
Nevertheless, the use of land titling as part of the state’s AD efforts was a popular policy among
former coca leaf farmers. Perhaps most importantly, in the context of Colombia’s history of land and
politics, the symbolic value of land ownership played a strong role in creating satisfaction among
farmers. The interviews revealed that most felt land titles gave them status as “proper owners” of their
land, which is linked to a long history of social and political tensions among the landless poor and
their previous lack of recognition by the state. Policymakers were aware of this “reward” effect of
land titling and its historic symbolism, so the political significance of their land titling efforts, should
not be underestimated. In addition, many farmers felt it became increasingly necessary to legalize
178
their land ownership as more people began to do so and that this would prevent future, often
intergenerational, land disputes.
This study highlights that for many of these farmers land is more than an economic asset, and that a
land title has symbolic meaning. In this respect, it makes sense for land titling to form part of
Colombia’s AD efforts. It is not the first time that policymakers justify land titling for AD by drawing
on local peoples’ perceptions of what is needed. For instance, in Myanmar opium poppy farmers
stated that a lack of legal land ownership and government recognition of customary land tenure rights
is one of the top reasons why they are driven to cultivate illegal drug crops (UNODC 2017c; 2015;
TNI 2016). There is value in granting people what they believe is important for them.
However, it is equally important that policymakers do not overplay the expected effect of land titles
on people’s financial circumstances. The rural communities I visited had already developed informal
land institutions, and ways of enforcing them, in opposition to the state. Therefore, while
policymakers assert that land titles are effective in changing coca leaf farmers’ economic behaviour,
this may not be the case for two reasons. Firstly, land titles were already aligned with existing de facto
property rights to land, so land tenure security was strong enough for a wide range of transactions to
be effective. Secondly, even if they increased their access to credit, land titles did not diminish the
attraction of growing coca leaf given its rate of return compared to other crops.
Land titling programs may appeal to policymakers because they are measurable, and it makes them
look good in the country’s social and political context. The state’s belated recognition of the existing
land titles may be welcomed by communities as a sort of peace offering to the officially landless.
However, land titling in AD efforts is not an economic game changer. And actively encouraging
former illegal drug producing farmers to apply for land titles by promoting their function for increased
access to credit, has ethical implications and may cause farmers to have unrealistic expectations about
their options beyond subsistence farming. Policymakers should also watch out for the possible
displacement of coca leaf production by land titling in communities undergoing AD. Failing to do so
will ultimately weakens the credibility of AD programmes.
179
More broadly, this research shows that informal property rights matter a lot and formal property rights
can bring large potential costs. Land tenure was secure in these communities despite the lack of de
jure property rights. This undermines de Soto’s theory that the lack of legal property explains why
people are locked out of the “bell jar” of capitalist apartheid and into “the grubby basement of the pre-
capitalist world” (de Soto 2000, 47, 56). He may be right that titles are not “mere paper”, and that they
serve as mediating devices that give us useful knowledge about the value of assets that is not
manifestly present (de Soto 2000, 219). But as this study highlights, the value of land assets is not
limited to the credit that a title unlocks for improving farmers’ livelihoods. In great part, it lies in the
symbolic value that farmers and policymakers assign to land titles. Overall this chapter shows, that the
simple dichotomy between formal property rights or nothing is not true, has never been true, and is a
problematic assumption to base any policy on (Ostrom 2009).
180
Part II: Extra-legal Governance & Alternative Development
2.1 Abstract
Why do some communities in Colombia abandon coca leaf farming when it once thrived? In this
chapter I use the Olsonian stationary and roving bandit framework to present a novel explanation for
how Alternative Development (AD) comes to succeed in some communities without the state making
the first overt move (Olson 2000; 1993a). In the theory section, I review literature on the nature of
extra-legal governance and present a brief historical overview of how two major extra-legal groups
governed the drug trade in Colombia. I argue that, in the context of Colombia, state incentive and
deterrence strategies do not always explain why farming communities stop growing illegal drug crops,
and AD succeeds. I hypothesise that this can happen when a stationary bandit ceases to provide
adequate governance for the illegal trade prior to any official state intervention. As the extra-legal
group becomes unable to govern the business interests of local coca leaf farmers, a governance
vacuum emerges for another stationary bandit to fill i.e., the state.
I then analyse 32 interviews with former coca leaf farmers and use some descriptive analysis of
timeseries datasets of coca leaf production and conflict in Colombia, to test my hypothesis. The
interviews are drawn from a sample of three communities in Colombia, where AD is currently
considered to have been successful and where coca leaf had once been bountiful. I find that two of the
three communities I studied, abandoned coca prior to any official state involvement. The extra-legal
group that had been acting as a stationary bandit turned into a roving bandit when it was confronted
by a rival group. Instead of joining forces with its rival, to protect the long-term business interests of
coca leaf farming, the stationary bandit changed its behaviour and began to expropriate farmers. It did
this by looting and extorting farmers to demand their loyalty by force. In the face of weakened
governance for the illegal trade, coca leaf production became less attractive and eventually drove
farmers to move away or to welcome state governance instead. I argue that it was the breakdown of
extra-legal governance which created a more receptive space for army intervention and for AD
programmes in these communities. I conclude by highlighting the importance of this research.
181
2.2 Introduction
Extra-legal groups emerge when the state cannot govern trade opportunities. Their function is to
provide extra-legal governance to create stable commercial environments to further their business
interests and to extract long-term rents by maximising their tax base. This has been the case for illegal
crop farmers in Colombia. While coffee growers rely on the stock exchange to regulate the price, on
trade associations to facilitate collective action, on the law to stipulate their rights and responsibilities,
and on the police to enforce them, coca leaf farmers rely on extra-legal groups to provide these
institutions informally. In return for a cut in the business, extra-legal groups in Colombia, commonly
referred to as “armed groups”, shelter coca leaf farmers from law enforcement and make sure the
illegal trade thrives. They do this by offering extra-legal governance i.e. a set of informal constraints
that guarantee buyers for coca leaf and coca paste; set market prices; and enforce codes of behaviour
that facilitate collective action for public goods, enforce contracts and protects farmers from
opportunistic thieves (Hough 2011; Rosenau et al. 2014; Castro Caicedo 2014; Pachico and
McDermott 2011; Arjona and Kalyvas 2012; Jentzsch et al. 2015).
Some argue that extra-legal groups can become so powerful and effective in the way they govern that
they actually pave the way for state formation (Cheng 2018; Olson 2000). Based on his study of
paramilitary group violence in Uraba, Ballvé argues that even extra-legal groups that cause an
immense deal of suffering among civilians, and are seemingly self-serving, are state builders. Despite
being responsible for the lion’s share of forced displacement in Colombia and inflicting gross human
rights abuses, the paramilitaries also gained a degree of support from civilian communities because
they were not opposed to development projects and they enforced property rights to some degree,
albeit for an exclusive group of people. Ultimately these extra-legal groups have thrived from the
widespread perception of statelessness, and which has legitimized their existence as regimes of
accumulation and rule. Ballvé stresses that state-building is not the same thing as peace-building, and
that authority and state formation can still exist despite violence and chaos for some (Ballvé 2020).
But extra-legal groups can get it wrong too. In the words of Ballvé, “an armed group’s territorial
hegemony is never simply imposed but is always to some degree a negotiated process” (Ballvé 2020,
182
8). And so, sometimes a regime that has had the support of a local population, can lose its authority.
The ascending of one actor, be it the state or another extra-legal group, is the demise of another. This
study will illustrate how this has happened in some coca leaf farming communities in Colombia, even
when business had been doing well. These cases are especially peculiar because they were in places
where the climate was warm and humid, which ensured coca leaf could thrive.
I want to know why these farmers willingly abandoned an illegal trade that had been flourishing. The
most common explanation for this phenomenon is that the army either forces an extra-legal group out,
and/or the state provides superior governance services for legal trade so farmers can still earn a decent
living. My elite interviews revealed a general understanding among high level AD policymakers, that
when given the chance to break free from the control of extra-legal groups, farmers prefer the
alternative livelihood promoted by the state. They choose state governance because they are fed up
with the violence associated with coca leaf farming, and they are exploited by extra-legal groups
anyway. In short, they need rescuing by the state and the income shortfall they must assume when
transitioning to legal alternatives is not that stark anyway. The director of UNODC in Colombia sums
up this sentiment:
The real rival [of AD efforts] is organized crime. They have no interest in these
programmes... It’s clear that you can’t work in an area where you are going to put
your people at risk. So, if there is a security issue… those places are subject to
forced eradication by the army and by the police. But sometimes you can
[peacefully] access … the marginalized areas, or the fringes of the coca areas
controlled by armed groups. This has been very much the aim: to work on the
fringes and slowly ring them in, forcing them to move out.
You are still in some sort of competition, so you need to find out some arguments
why the community would prefer to work with AD than to stick with what they
have. There are several arguments that come up. Sometimes farmers are fed up
with living in areas where these groups are present because these groups make
the areas high risk for them. There’s a high degree of violence, higher rates of
183
homicide, they don’t feel safe, and they don’t feel it’s a good place for them and
their children to live. They don’t see a future in having these organized crime
groups present. So, the whole security aspect weighs a lot and it’s not something
you can monetize in that equation. It’s just a matter of being able to live at peace.
Then if there are solutions that are better than, or as good as illicit crops, it’s a no
brainer.
The other thing is that the income from illicit crops is low. The exploitation of
coca growers is extraordinary. They are paid less [in Colombia] than in Bolivia
and Peru, but they are not really given an option. They are in a context where they
frankly can’t make a free decision, they are forced to grow coca because there are
no other options. If you are in a context where you can’t choose what to grow,
and you are forced to grow coca because they buy it at your front gate. If you
don’t and you refuse you are forced to leave your house and they will take your
land. That situation of coercion is difficult to measure.
This statement highlights the possibility that the institutions the state has to offer are more attractive
than what coca leaf farmers already have, and that extra-legal groups force farmers to produce coca
leaf, underpay them for it, and subject them to a hard life in violent conditions. This may accurately
picture how AD programmes come to succeed in some parts of Colombia. As I shall discuss below,
there is no doubt that extra-legal groups can be exploitative.
However, what the above explanation fails to answer is why, when coca leaf is so profitable, would an
extra-legal group neglect its long-term business interests. Why would it opt to force farmers to grow
nothing other than coca leaf, rather than incentivise them to produce better and more coca leaf? Why
would it make the atmosphere high risk for coca leaf farmers instead of protecting them and
maintaining a stable commercial environment for the illegal trade to thrive? And why would it offer
an inferior set of institutions to the state, making it unattractive for farmers to produce coca leaf?
184
Policymakers’ standard explanation for AD success may give us part of the answer, but it
underestimates the embeddedness of extra-legal governance. When a well-established extra-legal
group behaves like a stationary bandit, it provides optimum governance for an illegal business, and
this makes it very difficult for the state to dislodge it. It also overlooks the possibility that extra-legal
governance can dwindle prior to, and not because of, official state intervention.10 Like any
organization, extra-legal groups can have fatal flaws in their internal management, which will
inevitably weaken the governance they offer (Cheng 2018, 16).
I want to feed in these ideas of extra-legal governance to help build a fuller picture of how AD efforts
come to succeed in places where the coca leaf business had once thrived. From the point of view of
the farmers, was it the state’s offer to improve their conditions for legal agriculture that ultimately
persuaded them to substitute their coca crops? Did the state increase its threats and sanctions to a
point that farmers no longer felt protected by the extra-legal group that had been in charge? Or did
something happen to the extra-legal group which meant it could no longer govern the illegal trade
effectively? And if so, why?
To answer these questions, I interviewed 32 former coca leaf farmers living in what I qualify as three
rural communities in Colombia, where coca production was abundant before they opted to substitute
their illegal crops. I then used thematic analysis to deduce whether there had been a) an extra-legal
group behaving like an Olsonian stationary bandit by providing governance that offered a stable
commercial environment for coca leaf, and b) a change in the extra-legal group’s behaviour prior to
official state intervention, which affected the governance they provided.
I found that in all three communities there had been an original extra-legal group that behaved like a
stationary bandit. And in two of these communities, its behaviour had changed when faced with
competition from another extra-legal group, prior to any official state intervention. The rivalry
shortened the time horizons that the stationary bandit could foresee in those communities and thereby
reduced the value it assigned to its assets. Instead of merging with its rival or ceding territory, to
continue extracting long-term profits from the illegal trade, it sought short-term gains. Thus, the
10Although as shall be discussed further on, the state may be covertly intervening through its alliances with some extra-legal groups.
185
stationary bandit replaced the governance it had once provided for coca leaf farming to flourish, with
indiscriminate taking of goods by force.
Farmers in these communities described these changes as disadvantageous to their business. The
violence and extortion from both competing extra-legal groups impeded profitable coca leaf farming.
As such, many people fled, and many others desisted with coca leaf even before the state gained
territorial control in their community. Farmers’ need for governance and a stable environment to live
in, lowered the entry barriers for the next stationary bandit: the state. So, when the state came offering
basic local governance functions, the communities accepted, despite the income shortfall they would
have to endure by no longer being able to farm coca leaf.
Contrary to the usual explanations for AD success, these cases suggest that farmers’ decisions to stop
working with illegal crops can be taken irrespective of the promises of the state. When an extra-legal
stationary bandit leaves a governance vacuum behind, the offer of state governance on condition of
ceasing to farm coca leaf, becomes more attractive to a community even with the pay cut it entails.
One policy implication of this research is that the state could consider targeting its AD efforts in
communities where there are fighting extra-legal groups, because there is likely to be a governance
vacuum it can fill more easily. Arguably, it also implies that targeting a “war on drugs” in places
where extra-legal groups’ time horizons are contracting i.e., when they start preying on their
protegees, can help reduce the supply of coca leaf. However, this approach may be more successful if
accompanied by AD measures that help local populations to trust in the state.
More broadly, this research complements existing theories about extra-legal group behaviour. It
provides fieldwork examples of how extra-legal governance weakens at the micro-level when an
extra-legal group is faced with competition. It is a reminder that although extra-legal groups can
operate like stationary bandits that effectively foster trade, they can also miscalculate and become
roving bandits when faced with competition from other extra-legal groups. And I argue that this may
be a cyclical phenomenon in Colombia.
186
In the theory part of this chapter, I will explain how extra-legal groups emerge to govern illegal trade
and illustrate how this happened in Colombia. I will put forward the idea that two major extra-legal
groups have protected the illegal drug trade in Colombia by governing like stationary bandits but at
times have also failed to do so by behaving as roving bandits. And both have changed their behaviour
over time. I will go on to present conventional explanations for how the state succeeds in replacing
extra-legal governance in the context of AD. Based on these theories, I will devise a working
hypothesis for why AD may have succeeded in the three communities where coca leaf thrived in
Colombia.
187
2.3 Literature Review
Theory of extra-legal governance
Extra-legal groups are organizations external to the state, such as armed groups or mafias, that may, or
may not, be endorsed by community members, and that impose informal rules to structure human
behaviour (Varese 2011). From a political economy viewpoint, they exist to sustain their own
business interests (Cheng 2018). They do this through repetitive and structured interactions, that
determine the benefits individuals can access or be excluded from, and thus shape their behaviour.
When the outcomes of their behaviour are positive, individuals increase their commitment to the
structure to maximise productive outcomes (Skarbek 2012).
When the state, or another legal institution, is absent or unable to sanction those who disobey rules, a
space for extra-legal groups is created (Skaperdas 2001; Vélez-Torres 2014; Revelo Rebolledo and
Garcia Villegas 2010; Skarbek 2014; Gambetta 1993; Cheng 2018). This tends to happen in situations
where the community is socially and, or, geographically isolated (Acemoglu, Robinson, and Santos
2009; Skaperdas 2001; Matloff 2017). Or when a community has recently settled and is deficient of
trust and cooperation, making it less likely to self-organize and work cooperatively (Ripoll, Berrio,
and Rubiano 2013; Jansson 2008; V Felbab-Brown 2005). And it is almost always the case for
individuals operating in illegal economies (Skaperdas 2001).
Trade relies on well-functioning governance to prosper so when it concerns illegal goods and there is
no state to provide this set of institutions, entrepreneurs require extra-legal groups to provide them
(Skaperdas 2001; Kleemans 2013; Cheng 2018). In doing so, extra-legal groups define and enforce
property rights, capture the gains of trade, and shape people’s behaviour so that they act collectively
in these underground economies (Skarbek 2014, 4–6; Shortland and Varese 2014). For instance, they
can prevent business fraud, and enforce quality control, informal contracts, and standard practices
where necessary. They can provide a conflict resolution mechanism to settle disputes among
entrepreneurs and issue informal equivalents of licenses, insurance, and information to match supply
188
and demand for the illegal goods (Gambetta 1993; Schelling 1984). In short, they are a “substitute for
trust” in the quality and delivery of goods and services (Gambetta, 1993, p. 2; Cheng 2018 p. 81).
Sometimes there is demand for these governance services among entrepreneurs so that their trade can
flourish. So, an extra-legal group will provide them in exchange for a fee. Other times, there is simply
space for an extra-legal group to impose itself on a community to deliberately create demand for its
protection services by using the threat of violence (Gambetta 1993; Shortland and Varese 2016;
Skarbek 2014). Examples of this include two of the largest gangs in Latin America, the MS-13 and
the 18th Street, which prey on El Salvador’s economic development (Melnikov, Schmidt-Padilla, and
Sviatschi 2019).
There are those argue that not all extra-legal groups are the same as they can comprise a combination
of activists and opportunists. This suggests their motivation may not always be economically driven
but political and so should not be lumped into the same analytical category (Gutiérrez Sanín 2008).
However, parting from the political economy understanding that extra-legal groups exist to optimise
trade outcomes, the Olsonian stationary-roving bandit framework seems best placed to depict in broad
brushstrokes some fundamental differences in how extra-legal groups can behave.
Mancur Olson (1993; 2000) made a distinction between “stationary” and “roving” bandits. A
stationary bandit enforces the set of institutions that encourage some economic success among
entrepreneurs, and a peaceful environment for commerce to work (Shortland & Varese, 2014; Cheng
2018). In doing so, it protects entrepreneurs from being doubly taxed or robbed and ensures the
group’s own long-term interest. In return for this governance, the stationary bandit maintains a
monopoly over the resources and demands only part of entrepreneurs’ income on a regular basis. This
means the stationary bandit can accrue higher earnings than if it plundered everything in one go (as a
roving bandit does). Echoing optimal tax theory, a stationary bandit leaves entrepreneurs with part of
the profit so they retain an incentive to produce as much as they can. In sum, the Olsonian stationary
bandit encourages the economic success of entrepreneurs to ensure its own long-term benefits (Olson
2000, 14; 1993b). The state is a classic example of a stationary bandit, as even an autocracy will have
some incentive to maximize the outputs of the society it rules so it can accrue greater gains (Olson
189
1993b). And an extra-legal group behaves like a stationary bandit when trade is prosperous so it can
pocket a greater share through taxation (Olson 2000; 1993a, 87:568).
A roving bandit, on the other hand, imposes itself on a community and only seeks short-term gains
from entrepreneurs. Essentially, a roving bandit aims to loot opportunistically before exiting a
community. An extra-legal group is more likely to behave like a roving bandit when it only foresees a
short-term horizon in which it will be able to extract rents (Cheng 2018, 62).
Building on this, it follows that there are two factors that distinguish a stationary from a roving bandit.
The first is that a stationary bandit holds legitimacy in the community. Because it provides
governance services and it helps entrepreneurs keep some of the profit, it is accepted by the
community. A roving bandit, on the other hand, is spurned by entrepreneurs as it essentially destroys
their livelihoods.
The second criterion that separates a stationary from a roving bandit, is the use of force. To a large
extent, a stationary bandit keeps the peace while a roving bandit is indiscriminate in its use of
violence, if it serves to extract more resources. The stationary bandit prevents violence from
escalating because violence is detrimental to local economic productivity in three ways. Firstly, the
threat of terror diverts social capital away from productive activities and towards further extortion and
violence. Secondly, the ensuing uncertainty, incomplete information, and property destruction affect
trade efficiency and make it difficult for businesses to operate (Skaperdas 2001, 188). And thirdly, it
deters fair competition in trade, because fear, instead of the price mechanism, determines the value of
goods. So, when trade rests at the mercy of an extra-legal group’s will, rather than on what is optimal
for the market, there is monopolistic competition (Skaperdas 2001).
Kalyvas’ (2006) theory of violence is that its use is something that must be motivated, because too
much of it is costly and labour intensive. After all, violence disrupts the flow of information, increases
distrust, and attracts unwanted attention from the law enforcement and the media, which in turn is
costly to illegal business (Gambetta 1993, 44; 1994; Peter Reuter 2009; A. Smith and Varese 2001;
Andreas 2011, 423). Therefore, for a roving bandit, the returns from pillage are high enough in the
short term, for the use high levels of violence to make business sense.
190
Both types of bandits exist but cannot permanently coexist, because for a bandit to continue behaving
like a stationary one and provide effective governance, it must maintain a monopoly over trade. And
to retain it, a stationary bandit must fend off roving bandits that present competition and threaten their
control over trade (Snyder and Duran-Martinez 2009; Daniel Mejia and Restrepo 2013; Bailey and
Taylor 2009; Osorio, Weintraub, and Schubiger 2016). Confrontation creates a more violent scenario
because both bandits have to adopt more terrorising forms of violence to communicate the fiercest
reputation and command authority in the community (Schelling 1984; Gambetta 1996, 251–54). The
indiscriminate and more brutal use of violence only subsides when one bandit (extra-legal group or
the state) gains complete control of a given area (Kalyvas 2006; Bailey and Taylor 2009).
I propose that the Olsonian stationary-roving bandit paradigm is a useful typology for describing how
different groups come to hold a localized monopoly of force in different coca leaf farming
communities. However, extra-legal groups and the state cannot be categorized as either stationary or
roving bandits because the stationary - roving bandit transformation is cyclical. The situation of
stationary bandits turning into roving bandits can present itself in some communities in Colombia, and
occasionally pave the way for AD to succeed. Faced with a certain degree of competition that leads to
the breakdown of armed control in a locale, an extra-legal group that has been operating as a
stationary bandit, inevitably degenerates into a roving bandit when it ceases to provide effective
governance for the drug trade in each community. Likewise, the state, or a roving bandit, that has
been attacking a community for its own profit, may change its tactics once it is able to entrench itself
in the community and assert a monopoly over trade and provide some governance services. As its
time horizon expands and ambition develops, it may seek ways to gain legitimacy in the community
by governing trade, legal or illegal.
The following brief historical overview of some extra-legal groups that governed the illegal drug trade
in Colombia between 1990s and 2010s, and their linkages with the state, will help to illustrate the
cyclical transformation of roving to stationary banditry and vice versa. It will focus on giving some
context to the paramilitary – FARC war of the early 2000s, and the situation up until the Peace
191
Agreement in 2016, and not on the realignment of armed groups after the Peace Agreement between
the state and FARC, because that is the period of violence my fieldwork interviews refer to.
Brief historical overview of extra-legal groups and the illegal drug trade in Colombia,
1990s – 2010s
The original two largest extra-legal groups that emerged at the turn of this century in Colombia are
commonly categorized as ‘guerrilla’ and ‘paramilitaries’. I make the case that despite their different
origins and development, both groups profited from the illegal drug trade and behaved like stationary
as well as roving bandits. As such, they interchanged between the two Olsonian bandit categories.
Next, I shall describe the origins of these two major extra-legal groups.
Origins of FARC
The largest extra-legal group with political motivation that has existed in Colombia is the Colombian
Revolutionary Armed Forces (FARC). Founded in 1964 by Manuel Marulanda Velez, (a.k.a.
Tirofijo), it emerged from the Colombian Communist Party with a strong socialist ideology (Bushnell
2007; Ugarriza and Craig 2013, 57:452).11 Extra-legal groups can harness popular support and erode
the credibility of the state by using ideology and propaganda (H. Campbell 2014; Cardenas, Eslava,
and Ramirez 2016; Skaperdas 2001; Bonilla 2012). This was certainly applicable to FARC, which
developed strong links to the people in places it helped to colonize (Steele 2017). FARC’s raison
d’etre was to establish a communist-agrarian state. During the 1970s and 1980s their main demand
was for a thorough agrarian reform (Grajales 2011). And in the 1990s and 2000s its website was
loaded with anti-capitalist and nationalist language.12
11 Other guerrilla groups that emerged at this time include the Popular Liberation Army (EPL), the 19th of April Movement (M-19), and the National Liberation Army (ELN) which then splintered into the People’s Revolutionary Army (ERP) and the Guevarist Revolutionary Army (ERG). 12 FARC provided ideological training for its members. Its protocol manuals asked for new recruits to undergo military training and lessons on FARC’s agrarian agenda. The political indoctrination covered Marxism, Leninism, socialism (understood as a redefined role of the state in economic development), Bolivarianism (associated with ideas of Pan-American nationalism and anti-imperialism) and texts inspired by Simon Bolivar. Recruits were required to join the Clandestine Communist Party of Colombia (PCCC) and their promotion to command level was conditional on this political training (Ugarriza & Craig 2013: 454). Most members came from rural campesino backgrounds and the very small percentage that held a university degree went on to assume leadership positions (Rosenau el al 2014).
192
Despite its ideological origins and rationale, FARC profited from illegal as well as legal trade, and
grew in strength in rural places where there was little or no state presence (understood here as public
services and law enforcement) as a result. It levied taxes on everything including cattle and beer
(Economist 2015, 7). In the early 1970s and 1980s, FARC groups moved into regions where there was
gold, namely in Antioquia and Bolivar, and by the 1990s it was earning millions in gold revenue
(Dube and Vargas 2013, 1412; Echandia 1997; Kline 1999).
FARC also turned to the illegal drug trade as an alternate source of funding and set out to grow and
refine coca leaf to this end (Steele 2017). It began to regulate coca leaf farming, in exchange for a cut.
It protected coca leaf farmers in many regions by investing in volunteer work assignments (mingas) to
build local infrastructure such as bridges, schools, and public clinics for these communities. It
regulated the price of coca leaf, introduced coca leaf production quotas and coca paste processing, and
stipulated a living wage to all coca leaf harvesters (raspachines) (Hough 2011). And it offered useful
advice to make coca leaf farming sustainable e.g. that farmers should use three acres of their land for
food subsistence for every seven acres used for coca growing (Brittain 2010).
In the mid 1990s, there was a disease that destroyed 30% of coca leaf crops in the Huallaga Valley in
Peru, and drug traffickers transferred their crops to Colombia, experimenting with crops that gave a
higher cocaine yield. This increased Colombia’s coca leaf production by five-fold in five years (Pardo
2000, 69 –70). As a result, coca leaf farming spread in the territories that FARC controlled and there
was a massive influx of funds for FARC.
This gave FARC the capacity to actively confront the state and paramilitary armed groups, which
heightened the conflict in the late 1990s. By 1995, FARC maintained presence in 622 out of 1000
municipalities (Echandia Castilla 2000). And by the end of the 1990s, FARC was estimated to have
16,000 to 20,000 paid fighters and a yearly income of $300 million USD (Hough 2011). FARC was
the largest and wealthiest guerrilla army in the world until it agreed peace with the government which
was ratified in 2016.
193
Origins of the Paramilitaries
Another large extra-legal group that existed at the time of FARC’s glory, are commonly known as
“the paramilitaries”. They are a loose coalition of non-state armed organizations which, between 1997
and 2002, acted under the umbrella organization of the United Self-Defence Forces of Colombia
(AUC - Autodefensas Unidas de Colombia). There is a consensus that their motivation at the
individual level was primarily economic, and they were less trusted by the communities they operated
in because they mostly had no ambition to develop legitimacy at the local level.
The paramilitaries emerged from landholders’, army officers’, and drug traffickers’ joint demand for
protection from guerrilla groups (Grajales 2011; Ballvé 2012).13 Rural landed elites sought protection
from extortion and kidnappers, and they felt threatened by the Conservative government’s efforts to
negotiate land reform with the guerrilla groups in the 1980s. They began to enlist private armies for
protection from the guerrilla, sometimes with crossovers with police officers (Hristov 2009). As the
wealth and land tenure of illegal drug traffickers grew in the 1980s, their interests began to merge
with those of traditional landowners (Brittain 2010; Ballvé 2012, 610).14
The largest paramilitary group, called the Autodefensas Campesinas de Cordoba y Uraba (ACCU),
was formed by drug traffickers in 1994 to protect wealthy landowners from kidnapping by FARC
(Siegel 2013; Grajales 2011). And throughout the 1980s paramilitary groups were joined by
campesino movements in the Caribbean region who felt harassed by FARC (Grajales 2011, 778).
Then, in 1994 Liberal President Ernesto Samper allowed the creation of national neighbourhood
watch groups called CONVIVIR (Cooperativas de Vigilancia). These private rural security firms
promoted the expansion of paramilitaries and were armed with equipment from the Colombian army
(Hristov 2009, 69). The Federation of Cattle Ranchers (FEDEGAN) lobbied for these groups to be
trained by the military and by 1997 the ACCU began to merge with these militias as well as the other
private armies created in 1980s (Daron Acemoglu, Robinson, and Santos 2013). This gave rise to the
13 During the 1970s wealthy drug traffickers, including Pablo Escobar and Rodriguez Gacha bought large amounts of resource rich land to launder their money and garnered support from landowners and local communities (Siegel 2013). The most efficient syndicates of the cocaine trade at that time were the Medellin and the Cali cartels. They needed protection for the drug trafficking routes and began to protect the local elites’ economic interests by forming self-defence paramilitary groups that turned into protection racket gangs (Machado and Meertens 2010; Grajales 2011; Ballvé 2012, 610; Angrist and Kugler 2008). 14 For instance, drug traffickers hired private mercenaries to train these groups. In the late 1980s Israeli, British and Australian mercenaries were hired to train paramilitaries in techniques for building bombs and explosives (Castro Caycedo 2014: 328).
194
AUC in 1997, whose membership peaked in the early 2000s with about 30,000 fighters (Dube and
Naidu 2010; Dube and Vargas 2013).
The AUC’s mission was to look for local solutions to contain the influence of FARC in the illegal
drug trade, among other guerrilla groups, and to alter the trust levels between the guerrillas and the
local population (Hough 2011). To achieve this, it used extremely high levels of violence and terror.
The Centre for Historical memory calculates that between 1958 and 2012, a minimum of 220,000
people died in the conflict and about 80% of them were civilians. Colombia had the second highest
number of internally displaced people of any country in the world (Jentzsch et al. 2015). FARC
groups retreated from many parts of the country because of these confrontations.
Differences between the extra-legal groups
It is widely believed that FARC emerged with political motivation while paramilitary groups were
economically driven (Arjona and Kalyvas 2012; Jentzsch et al. 2015). FARC was unique in that it
operated like an army. It had a formally established chain of command and standardized units (like
regiments in state armies) and a clear division of labour. Its members lived in camps, wore uniforms,
and were not allowed to interact with civilians. It had harsh and systematic forms of discipline. This
helped it maintain structural integrity and fighting capacity. Whereas the paramilitaries were an
“unwieldy assortment of localistic units” continually suffered centrifugal blows (Gutiérrez-Sanín
2018, 640).
FARC fighters were exposed to daily ideology and discipline lessons such as how to treat civilians,
while the paramilitaries recruited children through the promise of money and a better way of life
(Pachico and McDermott 2011; HRW 2003). FARC did not provide salaries to low level fighters, and
it encouraged austerity, while paramilitaries did pay salaries and sometimes gave access to their
leaders to dispossessed land (Gutierrez Sanin and Rincon 2008; Gutiérrez-Sanín 2018, 641). FARC
did not allow individuals to profit personally via looting, while paramilitary groups did (Gutierrez
Sanin 2004, 270). Their social composition also differed in that FARC had younger and less educated
members, and a higher fraction of campesinos and women, while the paramilitaries had fewer women
recruits and more from common delinquency and the army (Gutiérrez Sanín 2008).
195
Despite these differences, both extra-legal groups adapted their behaviour in different ways over time.
To start with, FARC’s repertoire of violence was more army-like and with tougher discipline and a
structured chain of command. The paramilitaries, on the other hand had a more flexible, network-
structure and used more indiscriminate forms of violence as they had fewer mechanisms of self-
restraint (Gutiérrez Sanín 2008). However, in the 1990s, FARC became more indiscriminate in its use
of violence since its origins, profiting from kidnapping, the drug trade and from blowing up oil
pipelines to extort multinational companies. In many places this compromised the trust they had built
up at the local level.
The paramilitaries, on the other hand were more likely to commit massacres than FARC and less
likely to kidnap (Steele 2017, 111). But they began to replace their flurry of massacres, mutilations,
torture theft and destruction of property, with more selective violence including, by expanding their
extortion networks and targeting individuals (Gutiérrez Sanín 2008, 15).
What triggered this change in behaviour? One way of understanding is by each group’s organizational
dynamics. Gutierrez Sanin argues, for instance, that FARC entered a period of hardship around 2002,
not because their financing now depended on the drug trade (that was the case for all armed groups)15
but because of the way it operated. Its militaristic blueprint meant that it became isolated and lost
legitimacy among the general Colombian population. It also meant that the revenues it accrued were
put back in to funding fighting and not to pay off its members, which may have contributed to its
Returning to the Olsonian stationary-roving bandit framework, I shall argue that this was due to their
pursuit of profit and the changing timeframe that was available for them to extract rents. FARC felt
threatened by other armed groups, seeking to extract rents from the territories they controlled. The
paramilitaries on the other hand, sought to broaden their influence on political institutions to pursue of
longer-term rents. As such, both groups exhibited behaviour that resembled that of stationary as well
as roving bandits at different points in time.
15 FARC were invested in protecting the coca leaf growing while the paramilitaries where more interested in controlling the cocaine trade.
196
Use of violence by FARC: stationary-cum-roving bandit?
When extra-legal groups operate like stationary bandits, they have legitimacy among the
entrepreneurs they protect, and thus command respect for their authority without having to use force
too often (Shortland and Varese 2016; Gambetta 1993; Arias 2006; Daron Acemoglu, Robinson, and
Santos 2013; Friman 2009). FARC’s origins and behaviour with coca leaf farmers meant that it held
legitimacy at the local level, and it seldom used unpredictable and terrorizing violence in the
communities it governed. Like other extra-legal groups, it weakened citizens’ trust in the state and
weakened their willingness to cooperate with the state (Cardenas, Eslava, and Ramirez 2016, 376–77).
Its ideological foundations were arguably a strategy to retain loyalty in the communities it profited
from, without having to use costly violence (Ugarriza and Craig 2013, 57:56, 68, 448; Jentzsch et al.
2015, 21). It formed a political party called Union Patriótica (UP) in 1985 to gain tract in the political
institutions too. All this made it a stationary bandit at heart.
Some argue that FARC parted with its ideological and political roots when it began targeting farmers
and causing their mass displacement (Hough 2011, 404). As the drug cartels were dismantled in the
1990s, smaller drug trafficking networks began to take control of the illegal drug market. These
sought protection from guerrilla and paramilitary groups alike. The paramilitaries reorganized and
tried to gain control of many coca leaf producing areas. In doing so, they strived to dislodge FARC’s
territorial control of these places.16 This caused mass displacement of civilians (Grajales 2013). Rural
farmers were systematically murdered for sympathising with the FARC, as were journalists and
politicians.17
In response to this onset of confrontation by the paramilitaries and the state, FARC changed its tactics
as “it focused on economic extraction rather than building networks” (Steele 2017, 96). FARC did not
stand a chance of winning in direct confrontation with the paramilitaries because of the latter had the
support of the state. So instead of relying on its ties with the communities it controlled, FARC began
16 As the fight between these groups escalated, the level of violence was heightened by the army which had been strengthened by Plan Colombia, a US-funded anti-narcotic military offensive (Daniel Mejia and Restrepo 2013; Osorio, Weintraub, and Schubiger 2016; Bushnell 2007; Dube and Naidu 2010). The conflict soon became multilateral between the state, paramilitaries and guerrilla (Isaacson 2005; Bowden 2001; Castro Caycedo 2014). 17 Even popular satirist Jaime Garzon, who helped as a peace negotiator in the release of FARC hostages, is believed to have been murdered by paramilitaries in 1999.
197
to weaponize civilian communities to put up a fight. It thus became more indiscriminate in its use of
violence and less popular in the communities (Hough 2011; Steele 2017, 90, 93). From 1997 to 2006
the scale of victims as well as frequency of attacks increased. Entire villages were forcibly displaced
as a result. Figure 6 shows official data on the number of victims of conflict during the period: 1990 –
2015 from Colombia’s official Victims Registry. It shows how the rate of violence in the departments
of Caquetá and Putumayo, where FARC held control, was about three times higher than the national
rate. This is not to say that other departments where other extra-legal groups held territorial control
were not violent too, but these two departments are of particular interest to us as it is where the
interview data used in this study comes from.
Figure 6 Rate of reported victims per 100k habitants in Caquetá and Putumayo departments, and the national rate, for each five-year period between 1990 - 2015
To assess how violence during this period related to the presence of different extra-legal groups, I
performed an exploration using a box plot to visualise the distribution of homicides at the
municipality level categorized by the presence of different extra-legal groups from 1999 to 2014. The
panel data on extra-legal group presence was facilitated by CEDE Research Centre at the University
198
of Los Andes in Bogota, which collects time series data by year reporting presence of different armed
groups and number of crimes disaggregated by municipality. The demographic data on population
rate is publicly available from the Colombian National Administrative Department of Statistics
(DANE 2018).
Homicides peaked in 2004 to 2005 (Steele 2017). Figure 7 shows the rate of homicides reported per
1000 population by presence of extra-legal groups and shows that the rate of homicides was highest in
places where there were various extra-legal groups, shown in green (for multiple armed group
presence). This suggests the rate of victims is highest where there is contestation between different
extra-legal groups.
Figure 7 Distribution of homicides at the municipality level categorized by the presence of different extra-legal groups from 1999 to 2014 (namely FARC, paramilitaries (labelled as AUC for convenience), ELN and multi – meaning various). The median rate of homicides is represented by the solid line inside the box. The height of the box reflects the interquartile range and whiskers the extreme values.
Internal displacement peaked in 2002, the year the paramilitaries began to demobilize (Steele 2017,
120). Figure 8 is another box plot that shows the displacement pressure rate per 1000 population
(number of people each municipality receives and expels). It also draws on panel data provided by
CEDE Research Centre at the University of Los Andes in Bogota, and demographic data on
population from the Colombian National Administrative Department of Statistics (DANE 2018). Like
Figure 7, it shows that in the same period of 1999 to 2014, the rate of people that were displaced was
highest in places where there were various extra-legal groups. This suggests that in municipalities
where there is extra-legal group rivalry the rate of displacement is higher.
199
The inconsistency in FARC’s behaviour and loss of legitimacy among some rural communities, raises
questions as to whether its behaviour continued to be like that of a stationary or a roving bandit. The
evolution described above, suggests it was both, depending on the time and place. New war authors
argue that resource-driven conflicts leave no room for political agendas (Collier 2000; Kaldor 1999).
They might also argue, then, that underneath its ideological façade, FARC had no political ambition,
it simply sought to exploit resources as much as was possible before exiting. Others might argue that
the ideological motivation of an extra-legal group does not rule out that need, creed, and greed can
coexist (Ugarriza and Craig 2013). I propose that this evolution is best explained by the Olsonian
stationary/roving bandit framework. FARC was both, depending on what it needed at the time and
place, to protect its business.
Pursuit of legitimacy by the paramilitaries: roving-cum stationary bandit?
Without going as far as claiming that paramilitary groups were merely roving bandits during the
period of interest here, after all, they also provided governance for illegal drug trafficking, they did
behave in ways that are/were similar to roving bandits when they came into contact with FARC (Tate
2001). They would carry out public acts of terror, leaving very little room for winning the support of
local communities. This was especially the case following the 1991 constitution when the multiple
and dispersed paramilitary groups began to coordinate and to expand. Romero argues that during this
Figure 8 Distribution of displacement at the municipality level categorized by the presence of different extra-legal groups from 1999 to 2014 namely FARC, paramilitary groups (labelled AUC for convenience), ELN and multi – meaning various). The median rate of displacement is represented by the solid line inside the box. The height of the box reflects the interquartile range and whiskers the extreme values.
200
period the paramilitaries set out to isolate FARC to delegitimise its efforts for a political reform that
threatened the economic elites. FARC supporters threatened to join a political debate that might lead
to policies for wealth redistribution. At the same time, regional economic elites sought to maintain
certain social groups excluded from participating in politics. Therefore they backed the emboldening
of paramilitaries, who deliberately resorted to carrying out massacres in places that FARC controlled,
in an effort to uproot local support for FARC and other guerrilla groups (Romero 2003). Besides
terrorizing local populations in many places controlled by FARC and other guerrilla groups, the
paramilitaries extracted all they could from farmers by force and in a short space of time. In doing so
they also disrupted the coca leaf trade. Because their aim was to dislodge FARC’s territorial control
and to draw out short-term profits, the paramilitaries showed little will to assert their authority in these
places through peaceful means. This behaviour illustrates the paramilitaries’ manifestation as roving
bandits.
Then, over time, the paramilitaries developed characteristics of stationary bandits. They repopulated
land they had captured by force and expanded their economic initiatives. They sought state and
private investments to legitimize the new order (Romero 2003; Ballvé 2020). In the early 2000s,
paramilitary leaders sought to influence political candidates at the local level through terror and
coercion (Ballvé 2012, 614). In some places they collectively targeted civilians whose political
loyalties did not side with them and forced their displacement, to establish territorial and local
institutional control. This is what has been referred to as “political cleansing” (Steele 2017). They also
set out to co-opt politicians at the highest levels in Congress (Lopez Hernandez 2010; Steele and
Schubiger 2018). This led to the “para-politics” scandal, known as parapolitica, whereby paramilitary
groups succeeded in influencing the 2002 and 2006 national elections including through the electronic
manipulation of election results. After scrutiny in 2005 of the 2002 election results, big changes in
voting patterns were found in some municipalities, which suggest paramilitary groups were involved
in the 2002 elections. Then, in 2009, a third of Colombia’s legislators went under investigation and 32
parliamentarians and five regional governors were convicted by the Supreme Court for having links
with paramilitary forces (Acemoglu, Robinson, and Santos 2013, 5–44; Economist 2015, 12). Most of
the imprisoned politicians have been replaced by their close friends or relatives.
201
There is ongoing debate about why this shift in behaviour came about. One point of view is that
political elites used the paramilitaries to protect their own interests and to fend off the challenges that
FARC, among other guerrilla groups, were making to the status quo distribution of wealth and power.
This is what is referred to as “inverse capture”. It suggests that it was not the extra-legal groups that
seized the state institutions to further their own greed and to reduce their risk of punishment by law
enforcement, but that it was legal entities embodied by economic elites, which co-opted the extra-
legal groups to instrumentalize violence and drug trafficking to their own advantage (Romero 2003;
Lopez Hernandez 2010). In other words, the paramilitaries may have been used by those in power to
prevent FARC, and other guerrilla groups, from claiming a share of that political power.
However, returning to the Olsonian framework, even if the paramilitaries were not always the
protagonists of co-optation, they still profited from this scenario. Even if ultimately the state was in
the driving seat of the paramilitaries’ change of behaviour and strategy, this idea is not at odds with
the interpretation that they were transformed into stationary bandits. The paramilitaries may have
forced themselves into, or been allowed to enter, or even invited to entrench themselves into, the
political and legal system by the state. But either way, this extra-legal group benefited from an
expanded time-horizon in which they could extract rents. And this was ultimately to the
paramilitaries’ advantage. By turning from warfare to large-scale corruption, violent extortion and
robbery, at this point the paramilitaries aimed for longer-term rents than what they had been able to
extract from causing havoc (Dugas 2012; Reyes Posada 2009). And by securing more stability at the
political level, they could continue extracting rents from entrepreneurs without using as much
violence. This change of tack illustrates how, depending on the place and time that was available to
them, the paramilitaries operated both like roving bandits and like stationary ones too.
State’s response to extra-legal groups: 1990s - 2010s
To illustrate my hypothesis later, and to provide some background context for the events that will be
described by some of the farmers I interviewed, it is important to offer a wider picture of how the two
extra-legal groups I discuss above, related to the state, understood here as another bandit. It will
remind the reader that extra-legal groups do not operate in a complete state vacuum and will offer a
202
broader understanding of the nature and development of the state and of these extra-legal groups
through the state (Arias 2017; 2010a).
In general terms, during the timeframe of the 1990s until the 2010s, the state supported the
paramilitaries, or at least failed to confront them with the same force that it confronted the guerrilla.
For a long time, the state ignored, or was co-opted by, paramilitary groups, while it chose to openly
confront the guerrilla. While the state attempted various peace negotiations with FARC, it also
entered full blown conflict with it (Ugarriza and Craig 2013; Ballvé 2013; Hough 2011; Albertus and
Kaplan 2013, 206; Steele 2017).
State and FARC
In 1984 the state took steps to offer FARC a ceasefire and President Betancur began peace talks with
FARC (Ugarriza and Craig 2013). Betancur and FARC jointly established the Caguan Colonization
Committee (CCC) to boost economic and social investment in the region (Hough 2011). President
Betancur also proposed to allow direct mayoral elections, which was eventually approved by
Congress after pushing back the start date to 1988 (Eaton 2006, 452). FARC established a political
party called Union Patriotica (UP) as a response to the government’s peace efforts and secured some
of the region’s key political positions in 1986 (Ballvé 2013). But FARC pulled out of that peace
agreement after its demobilised members and innocent members of the UP were systematically
targeted and murdered by paramilitaries. There have now been 4,153 reported victims of this deadly
violence, 3,621 of which were homicides and 542 were forced abductions (CNMH 2018, 433). The
next government then took a harder line towards the peace negotiations and withdrew government
participation from the CCC. The peace talks broke down by 1987 (Hough 2011). FARC was given the
option to participate in the Constitutional Assembly talks if it demobilized, but it didn’t. Land reform
increased again in the late 1980s until the mid-1990s (Albertus and Kaplan 2013, 206). The next
president, Gaviria, attempted to hold peace negotiations again with FARC in the early 1990s in
Venezuela (Caracas) and Mexico (Tlaxcala), but these were unsuccessful. The constitution was
reformed in 1991 and further political decentralization was granted as a concession to FARC so that
203
the president was no longer solely responsible for appointing governors, and mayors and local
authorities were given more political and fiscal discretion (Ballvé 2012, 606–7).
These democratic reforms favoured FARC but were difficult to implement in the Atlantic coast and
central highlands where the paramilitaries held control. Nevertheless, the government felt it could
now legitimize its war-making activities (Hough 2011). Some argue that both paramilitaries and
FARC took advantage of these reforms to transfer funds from central government to municipalities
and use these funds to expand their drug production (Thoumi 2009).
By the late 1990s President Pastrana became increasingly concerned with the state vacuums, or rural
areas where the state lacked territorial control, across the country and relaunched peace talks with
FARC. He assigned a distention territorial zone for the FARC to remain in while peace negotiations
took place. This effectively granted FARC de facto sovereignty over 42,000 km2 of land, an area
known as “El Caguan” and roughly the size of Switzerland. This facilitated the expansion of coca
crops in the region, which boosted FARC’s competitiveness to participate in the illegal drug industry
(Thoumi 2009; Steele 2017). Pastrana’s government also accepted US support to strengthen and
increase its military presence in areas controlled by the guerrilla (Thoumi 2009; Steele 2017). The
USA devised a bold military and police security intervention programme worth $1.2 billion USD. It
was called Plan Colombia. Its aim was to reduce the production, refinement, and distribution of
illegal drugs by 50%. In exchange, Colombia was offered duty-free access to some exports to
compensate for the cost of their drug enforcement efforts. Japan, Canada, and Europe made modest
contributions for AD programmes to be introduced. President Uribe then took a more hard line
approach in confronting FARC, introduced a tax for the wealthy to strengthen the military and
increased police presence in rural areas in the early 2000s (Steele 2017, 105). Plan Colombia was
originally planned in 1999 to last for six years, but it was then extended until 2007 when President
Uribe introduced the Strategy for Strengthening Democracy and Social Development (EFDDS) to
phase out Plan Colombia by 2013.
204
In 2000, Plan Colombia rolled out a mass Roundup aerial spraying programme over Putumayo,
Nariño, and Caquetá (see blue dots on Figure 9). It was funded and managed by the Narcotics Affairs
Section of the US Embassy. It used a concentration of
glyphosate up to 26 times higher than what is normally
used. It also contained a surfactant called polyoxy-
ethylenamine (POEA), which is more toxic than glyphosate,
to increase the adherence of the chemical to plant surfaces.
This sparked a lot of criticism from the EU and other
national bodies, including the Ombudsman for human
rights (Defensoria de Pueblo), who demanded the
suspension of the aerial spraying programme due to the
unknown impact of the new mixture on human health and
the environment. Despite the unpopularity of aerial
spraying, the USA doubled its funding for it from 2001 to 2003. And from 2003 to 2004 the area
sprayed increased by 84%. President Uribe allowed aerial spraying to expand to natural parks and
demilitarized areas.
When President Santos was elected in 2010 the peace negotiations with FARC were reintroduced in
Havana, culminating in a Peace Agreement in 2016. FARC agreed to demobilize in return for various
state concessions: broadening campesino access to land by issuing titles to all rural landowners;
including FARC in the political system by guaranteeing seats for them in parliament; replacing aerial
with manual drug crop eradication strategies; legalizing coca production for traditional indigenous
uses; and giving compensation to victims of violence and lower sentences to FARC members who
confessed to their crimes (Economist 2015). These efforts were accompanied by greater investment in
infrastructure development in many rural areas. Some unintended effects of the ceasefire have been
the increased contestation for FARC’s former territorial strongholds among other extra-legal groups
that were excluded from the negotiations (Prem et al. 2018). In addition, since the Peace Agreement
former FARC members and deserter factions have joined FARC dissidence groups such as Columna
Daniel Aldana and Columna Miller Perdomo.
Figure 9 Departments where Plan Colombia's aerial glyphosate campaign was concentrated (indicated by blue dots)
205
State and the Paramilitaries
As discussed in the section on the origins of the paramilitary groups, some argue that the state took
the first initiative to form paramilitary groups.18 After attracting attention by carrying out massacres in
the Uraba region, the paramilitaries were outlawed in 1989. However, paramilitary groups began to
argue they were a political organization to justify their existence and were given the legal status of
self-defence groups in 1994 and 1997 with the formation of CONVIVIR (Lopez Hernandez 2010;
Steele 2017, 101). From 1996 to 2002 the paramilitaries expanded their presence, initially intensifying
the violence to expel FARC from its territorial strongholds, and then by forging ties with the state
(Steele 2017, 101). As discussed in the section on how the paramilitaries changed their behaviour over
time, the state’s involvement was central in the development of the paramilitaries strategy. The
“parapolitica” phenomenon can be interpreted not only as the paramilitaries hijacking congress and
other political institutions, but also as elites of different regions, the military and the executive, giving
rise to strengthened paramilitaries (Romero 2003; Lopez Hernandez 2010).
In addition, military-paramilitary ties in 2001 have been documented revealing that the army provided
paramilitaries with intelligence, weapons and transport equipment, training, and direct and indirect
operations support. For instance, the armed forces were ordered to stay out of a massacre undertaken
by the paramilitaries in the Meta Department, and provided intelligence and logistics for a
paramilitary group called the Calima Front (HRW 2000; 1996; Hough 2011; Dube and Naidu 2010).
And in 2006 the highest-ranking officer of the army, General Mario Montoya, was charged with
supplying weapons to paramilitaries from a military base in Medellin and in 2008 high level officials
were indicted by the supreme court for collusion with paramilitaries (Dube and Naidu 2010).
The close ties between the paramilitaries and state institutions were arguably not just driven by the
extra-legal groups’ ambition to secure more power and wealth, and to reduce their own risk of
18 In 1946 the Colombian Petroleum Workers Union went on protest in response to the rising cost of living in towns where Shell Oil and Tropical Oil operated. Instead of sending reinforcements, the Conservative government suggested Shell replicated the practice of hiring off-duty policemen to provide security as other companies had done (Siegel 2013). Then there was a military effort in the 1960s called Plan Lazo in which the Ministry of Defence armed civil patrols (Hristov 2000). In 1965 the government allowed self-defence associations to form and then in 1968 it passed the Law 48 which allowed private citizens to create self-defence militias to protect themselves from the guerrilla groups. In the 1970s the army tolerated vigilante groups such as “the Black hand” and “Death to Kidnappers and Communists” in the Caribbean plains (Machado and Meertens 2010; Grajales 2011; Steele 2017, 10).
206
punishment for their actions, but also a result of being co-opted by regional economic elites. It is
argued that it was the latter which sought to embolden the paramilitaries so as to further their own
interests and keep political modernization at bay (Romero 2003; Lopez Hernandez 2010).
This patronage of paramilitary groups also extended to foreign states. For instance, in 2005 two US
army officers were arrested near Melgar, Tolima for arming paramilitary groups with ammunition
provided by the US government (Forero 2005; Dube and Naidu 2010, 26). And some academic
studies have shown that “increases in US military aid increased attacks by paramilitary groups
differentially in municipalities containing military bases”, which also suggest a degree of collusion
between the army and paramilitaries in fighting guerrilla groups (Dube and Naidu 2010, 26).
State partisanship was extended to the media coverage of extra-legal groups. A study of the language
used by the Colombian press to represent the actions of armed actors considered how guerrillas and
paramilitaries were named, what participant roles were assigned to them and how they were and were
not introduced in descriptions of violence. It concluded that guerrilla groups were frequently referred
to in differentiated terms that allowed for the identification of those responsible for the attacks. On the
other hand, paramilitaries were more frequently reported in undifferentiated forms, denoting low
certainty and casting doubt over the authorship of the crimes. This has meant that guerrillas have
frequently been depicted as more dynamic, and paramilitaries as minor agents of violence (Garcia-
Marrugo 2013, 440).
In the early 2000s, President Uribe offered amnesty (including psychological counselling, a stipend,
medical care, food, new clothes and protection) to AUC paramilitaries who demobilized (Rosenau et
al. 2014; Ugarriza and Craig 2013; Jentzsch et al. 2015; Thoumi 2009). Some argue that demobilized
paramilitaries made use of this armistice to launder their assets (Thoumi 2009). And those that did not
demobilise gave rise to a new generation of paramilitary groups, known as Bacrim (criminal gangs)
(Dube and Naidu 2010, 6; Steele 2017, 109).19 Many of these have since reorganized into new extra-
19 In 2015 they had about 5000 to 6000 members (Economist 2015, 9). Larger paramilitary groups such as Clan del Golfo (a.k.a. Urabenos or Auto Defensas Gaitanistas) have since formed and continue to operate in similar territories as the AUC. Paramilitary predation today is documented to be strong in the oil region (Dube & Vargas 2013:1388-9). In addition, they have used gold mines and refineries to launder their money and extract taxes resulting in what is known today as paramineria (Siegel 2013). Besides control of the drug trade routes in certain parts of the country, paramilitary groups have several agribusiness and mining projects such as palm crops to launder their illegal profits (Grajales 2011; Ballve 2012:603; Ballve 2011; Grajales 2013).
207
legal groups such as Clan del Golfo, Águilas Negras, ERPAC (Colombian Revolutionary Popular
Antiterrorist Army), Urabeños and Rastrojos, among many others.
Theories for success in Alternative Development (AD)
Let us now consider how the ideas about extra-legal governance so far discussed, fit in with existing
theories of AD success. Coca leaf farming thrives in Colombian communities that have climatic and
geographic conditions that are favourable for the crop, but also where there is an extra-legal group
that operates like a stationary bandit, because it provides governance and thus a stable environment
for the illegal trade. Why then, would a community decide to replace the set of institutions that helps
its illegal business work so well, with state governance that caters only for subsistence farming? And
why would a stationary bandit neglect its long-term business interests by seeking short-term gains?
Established theory for AD success
A rational choice approach assumes that the institutions that facilitate trade, change through a natural
equilibrium designed to optimise trade outcomes (Coase 1988, 5–7, 13). This suggests that
governance institutions only last if the costs and benefits they bring to trade are not outperformed by
another system of governance. If we apply this approach to the situation of farmers who give up trade
in illegal crops and embrace state governance, we should expect that these farmers see the prospect of
higher returns in legal farming than those offered by coca leaf production. But what could provoke
this change? If it makes better business sense to farm a lucrative illegal crop, that is easy to grow, easy
to transport and easy to sell, thanks to the stationary bandit that governs the trade, what is it that
shifts?
Conventional wisdom attributes this change of heart to something in the agency of the state. The
assumption is that if the state provides the right balance of new incentives and opportunities, the
equilibrium changes so that legal trade becomes preferable to illegal trade. A combination of pull-
factors, such as AD benefits, and push-factors, such as disincentives and deterrence efforts to force
farmers to comply with the law, can make it less costly to farm legal crops and more costly to farm
coca leaf. This is when it is believed that farmers decide to stop farming coca leaf and AD succeeds.
208
This view holds that the state’s investment in local development improves legal agriculture
sufficiently to make it more profitable to farm alternatives to illegal drug crops. In addition, it could
be argued that the state’s mere promise and first steps to provide services and investment wins over
the hearts and minds of entrepreneurs in an illegal trade. For instance, some development projects to
improve material conditions in conflict zones in Iraq helped to reduce insurgent violence (Berman,
Shapiro, & Felter, 2011; Berman, Felter, Shapiro & Troland 2013). The suggestion here is that when
the state becomes more attentive to communities that have long been governed by an extra-legal
group and makes gestures to invest in their borough, it enhances social order and builds trust in the
state. As a result, farmers shift their loyalty from the extra-legal group towards the state.
A complementary argument for AD success is that the state triumphs by force. When it is unlikely
that investment alone can make legal trade profitable enough to make farmers want to abandon illegal
crops, then a degree of deterrence is needed. If the state gains enough territorial control with the army,
it can confront the extra-legal stationary bandit to a point that it must surrender. This makes it
impossible for the illegal trade to run smoothly and farmers have little choice but to accept the terms
of the state.
Consider the incentives facing the farmer in a coca leaf producing community. Without a stationary
bandit to shelter her illegal crops, she will feel more vulnerable to surveillance by law enforcement.
There is a higher chance that her illegal crops could be stolen, and she will feel the threat of state
sanctions likely enough to want to desist with coca leaf. The risk outweighs any gains she could make
in the illegal trade.
The point here is that it is possible that farmers voluntarily accept the terms of AD, perhaps not as a
first choice, but as a consolation package for a loss they are going to incur anyway. Without an extra-
legal group to provide effective governance at their doorstep, it is no longer profitable to produce
illegal crops. Thus, farmers switch to subsidised legal alternatives just to cut their losses. In addition,
as the jurisdiction of law enforcement broadens through increased police presence, courts of justice,
prison and tax collectors, farmers become hopeful that trading conditions for legal alternatives will
one day improve (Revelo Rebolledo and Garcia Villegas 2010).
209
At its core, this theory assumes that regardless of the circumstances, state governance is the optimal
institution for capturing the gains from trade (Hobbes 1651). Everyone gives up some liberty in
exchange for common security. In short, the set of formal institutions enforced by the state are
inherently superior to the extra-legal governance farmers once depended on to produce illegal crops.
So, if the state sends signals of being present, through investment and subsidies, and/or through law
enforcement, AD will succeed. This is what it takes for communities with thriving illegal crops to
accept to replace them with legal agriculture.
New theory for AD success
There is no doubt that when the state invests more in local infrastructure, and increases its military
presence, it can improve, albeit marginally, the efficiency of legal agriculture for rural farmers.
However, important as these two factors are, they do not provide the whole picture of why farmers
choose to abandon coca leaf production in their home community. After all, Coasian theory holds that
farmers would only accept the institutions governed by the state if this is how they could optimise
trade outcomes. But the farmers I interviewed were never forced to grow coca leaf, they were simply
keen to profit from it. So, if they remained unconvinced that legal alternatives were as good a
business, then why did they submit to state governance and accept AD? This suggests that AD success
is not simply the result of improved state governance. We are therefore left to consider another
explanation for AD success.
A recurring theme that transpired from the accounts of former coca leaf farmers (scattered across
Cauca, Nariño, Putumayo, and Caquetá) interviewed in this study, was their disenchantment with the
coca leaf trade due to the intolerable levels of violence and extortion. Of those farmers who gave a
single reason for why they had chosen to substitute their coca leaf crops at home, 25% revealed that
they had been driven by fear that extra-legal groups would continue to inflict violence on them. In
addition, they did not consider the extra profits they could make with coca leaf crops, worth the
unsupportable brutality that seemed to accompany the business. 32% admitted they had quit farming
coca leaf at home as they no longer felt protected from law enforcement, while 16% said coca leaf
was no longer profitable to them. Only 10% said that they had quit coca leaf farming at home because
210
they felt they could do better with legal alternatives, thanks to the positive AD incentives offered by
the state. Together, these figures suggest that farmers in communities where AD was successful, had
felt exposed to violence and to law enforcement and had been struggling to profit from trading coca
leaf, prior to any official state intervention. Furthermore, the promise made by AD officials of
improved commercial opportunities in legal alternatives, appeared to be circumstantial, and did not
feature heavily, in farmers’ decision-making.
This suggests that something must have been going wrong with the coca leaf trade before any
alternatives were on offer. I use the Olsonian stationary – roving bandit model to hypothesize that AD
succeeded due to the weakening of the extra-legal governance, once provided by the local extra-legal
group that had been operating as a stationary bandit. Instead of governing with the minimum use of
force, so that trade could prosper, the extra-legal group changed its behaviour and inflicted
unpredictable violence on farmers. Instead of demanding measured protection fees from farmers so
that they could keep an incentive to produce, it began to make unreasonable demands. So, instead of
feeling supported by extra-legal governance, and sheltered by the stationary bandit from other armed
groups, perhaps seeking to prey on their wealth or to sanction them, farmers felt doubly exposed.
Farmers were effectively chastised by their former extra-legal stationary bandit and bereft of the
informal institutions on which they had depended on to prosper. All this meant that the illegal trade
became less profitable and less attractive, and since people prefer the governance provider with the
longer time horizon, this created a space for agents of the state the to enter.
Why use the Olsonian stationary – roving bandit model?
In political economy the state is classically conceived as a sovereign territorial unit with a government
and a population. However, in this analysis I propose it might be appropriate to term agents of the
state as “bandits”. A bandit in the Olsonian sense is a set of individuals who provide governance
functions for a society. The nature of governance they provide is entirely self-interested because they
react to the different incentives that are determined by their time horizons in control. Thus, it is not
impossible for the political group that holds power in an autocracy or even a democracy, to behave as
a roving bandit if its time horizon for being in power is too short. This means they will have no
211
interest in the future output of society nor in who succeeds them in holding power. In cases like these,
it makes more sense for state agents to appropriate “assets whose tax yield over [their] tenure is less
than their total value” (Olson 1993a, 87:567). Nevertheless, Olson argues that both in autocracies and
democracies, state agents tend to provide governance with longer time horizons, so it is usually in an
autocrat’s and a government’s interest to maximise the productivity of a society. This basically,
maximises the surplus it can cream off through taxation.
Therefore, for the purpose of my analysis, I interpret agents of the state to behave like bandits, in the
sense that that they provide governance functions for a society. In the former coca leaf farming
communities I study, “the state” is understood to be the governance provider with the longer time
horizon. This meant that it was the group that could offer a peaceful order and public goods. Like a
stationary bandit, it had become the preferred governance provider that could maximise productivity
for the community during a period of anarchy as the extra-legal groups had confronted each other.
I use the dichotomous Olsonian stationary-roving bandit model because it helps us to understand the
self-interested nature of governance provision. When there is fighting between bandits, their time
horizon for power shrinks. This generates a space for a new governance provider, to establish itself,
albeit temporarily. In this case, this governance vacuum was filled by agents of the state. In other
words, the rolling back of coca leaf farming is not always because the strength of the army or
legitimacy of the state institutions. Nor is it always because what the new governance provider can
offer is superior to what the previous one had offered farmers. AD can succeed because the time
frame in which the extra-legal group can provide governance, suddenly shrinks. This means it
degenerates into a roving bandit and creates an atmosphere analogous to anarchy, which is the least
efficient situation for any trade.
The question now is, why would the local provider of extra-legal governance in a community with a
flourishing illegal trade, become violent and jeopardize its business? Metaphorically speaking, why
would it kill the goose that lays golden eggs and make way for another stationary bandit to succeed it?
As discussed above, extra-legal groups that operate like stationary bandits, use violence and
intimidation as a conflict resolution mechanism, as a communication tool to enforce their rules and to
212
protect the illegal trade from theft by rivals and from state destruction (Gambetta 2009). These can all
be useful services for an illegal trade to prosper. However, when an extra-legal group encounters
rivals and can no longer rely on their reputation to govern a trade, uncertainty increases and trade then
becomes inefficient (Leeson 2006, 43). This is most likely why the organization starts to behave like a
roving bandit. It begins to use violence to assert its authority in the community, and to extract
unsustainable amounts of protection fees from farmers. Its governance services start to deteriorate,
and its reputation becomes sufficiently blemished for farmers to see fewer returns in the illegal trade.
In doing so, the stationary bandit loses farmers’ trust and prompts them to adhere to a new stationary
bandit that offers to restore a stable commercial environment in which they can rebuild their
livelihoods, albeit less profitably than when they had farmed coca leaf peacefully.
It is feasible that AD succeeds in scenarios like these because at that specific point in time, agents of
the state have a better chance of gaining farmers’ trust and of governing like a stationary bandit. Even
if farmers’ earnings in legal agriculture are negligible compared to what they had once been able to
make with coca leaf, they will be glad of the peace. Therefore, if farmers felt the illegal trade was
declining, before there was any prospect of AD in their community, it is possible that agents of the
state simply managed to fill a governance vacuum that was left behind by the roving bandits. This is a
different explanation to the conventional assumption that AD success is the result of one provider of
governance i.e., the state, outperforming armed groups in the returns it can bring to a society.
Limitations of the Olsonian stationary-roving bandit model in the Colombian context
At this point I should add that the Olsonian stationary-roving bandit is not a typology that perfectly
describes the interaction between agents of the state and extra-legal groups in Colombia. The reality
in Colombia is more nuanced as agents of the state are not always the antagonists of the illegal trade
but are intertwined with extra-legal groups resulting in criminal networks (Arias 2006, 296; 2017).
There are copious examples of linkages between paramilitaries, guerrilla groups and ‘legitimate’
sectors of the state and society. These include political officers who protect, and even participate in
the drug trade, collusion between the army and the paramilitaries, and the use of aerial fumigation of
coca leaf by law enforcement as a form of aggression against FARC (Ramirez 2010; Arias 2010a;
213
Ramirez 2011; Eaton 2006; Steele and Schubiger 2018; Daron Acemoglu, Robinson, and Santos
2013; Arias 2017). Therefore, like in the Latin American region more generally, in Colombia extra-
legal groups do not operate in parallel to the state but through it, and vice-versa, and violence is not
the result of anarchy or state absence, but the way the state operates to maintain the status quo (Arias
2010b; Arias and Goldstein 2010; Arias 2010a; Pearce 2010).
In my own fieldwork farmers accounts pointed to examples of how agents of the state worked in
parallel with extra-legal groups and not always in opposition to the illegal drug trade. There were
policemen who would offer to turn a blind eye to farmers’ coca leaf crops in exchange for a bribe,
there were community leaders, or presidents of the JACs (Community Action Board) who provided
reference letters for individuals to be given employment as raspachines in coca leaf plantations
elsewhere. There were also interviewees who described their incredulity that paramilitary attacks were
not condoned by the state as they were not stopped by the army despite the latter having made it
presence known to the community and promised to protect them from armed groups.
Let it be clear then that there are drawbacks in using the ‘stationary’ and ‘roving’ bandit model in
relation to the armed groups in Colombia. The definition of the state as the governance provider with
the longer time horizon, or a ‘stationary bandit’, does not capture the complete picture of how agents
of the state operate in Colombia. Similarly, the binary abstraction of ‘stationary’ and ‘roving’ bandits
does not wholly capture the behaviour of the extra-legal groups in the Colombian context because of
their ties to agents of the state. The fact that there are links between extra-legal groups and agents of
the state and that these can co-exist to maintain mutually beneficial peace in places where an illegal
trade dominates, highlights the complexity in rationalising why AD succeeds in some places.
Therefore, I stress it is the “officially declared” state intervention that cannot always explain AD. It is
likely that agents of the state still intervene, deliberately and covertly, through their networks with
paramilitaries, to cause the extra-legal group that had so far behaved like a stationary bandit, to
become a roving bandit.
However, the focus of my research was not to identify criminal infiltration into state institutions, or
the collusion between extra-legal groups and agents of the state. I did not gather detailed data on how
214
FARC and the paramilitaries governed through state institutions or civil groups, and vice versa,
because this was beyond the scope of my research. Nevertheless, I will allude to how the paramilitary
groups seem to have served as the advance guard of the state in some of these cases.
So, although there is hybridity between agents of the state and extra-legal, or criminal, groups in
Colombia, my quest was to help explain why some coca leaf farming communities decided to stop
farming coca leaf before there was official state intervention through AD. To do this, I part from the
assumption that officially, the state operates separately from extra-legal groups. The common
perception among the interlocutors in my fieldwork, was that the state, the FARC, and the
paramilitaries were different types of authorities. In addition, the boundaries between the extra-legal
and legal types of authority and the extent of their collusions are difficult to measure. To my analysis,
I demarcate boundaries between the state and extra-legal groups, based on the form of land use and
farmers’ general description of the situation “before” and “after” illegal crops were grown in the
community. I take the form of land use i.e., legal vs illegal crops, to signal which authority was
officially in control. Similarly, I interpret farmers’ descriptions of which authority was sanctioning
theft and resolving tensions over land ownership in the community (i.e., an extra-legal group or the
police), to mark whether it was the state or an extra-legal group who held a localized monopoly of
force, and therefore the acting stationary bandit. The advantage of this interpretation is that it offers an
alternative explanation for why AD can succeed without official state intervention, which parts from
the same assumptions made by policymakers i.e., that state and extra-legal groups are mutually
exclusive providers of governance.
The Olsonian model of stationary – roving banditry is therefore a useful caricature that parts from the
general assumption among policymakers that the state and extra-legal groups are separate entities.
This coincides with how they are perceived in the farming communities I studied, prior to their
voluntary substitution of illegal crops. It is important however, to note that there are disadvantages to
using the Olsonian model as a proxy. Firstly, it helps to understand only part of the picture. And
secondly, it can emphasize certain aspects of a phenomenon according to human interpretation, which
ultimately is limited and subjective.
215
Working hypothesis
To check if this theoretical mechanism holds true in the communities I visited, it is important to
establish that farmers had fared well in the illegal trade. Favourable geographical conditions for coca
leaf production and access to effective governance institutions, provided by an extra-legal group, are
indicators of a healthy illegal trade. This is because the illegal trade would have been bountiful and
well governed. It is also vital to be sure that when these farmers eventually switched to legal
agriculture, and embraced state governance in their community, it happened because of a
malfunctioning governance of the illegal trade and not because of an inherent disadvantage in coca
leaf production in their location. It is crucial to establish this point, to discard the possibility that AD
succeeded because the state improved the conditions for legal trade or forced farmers to comply with
their terms.
In sum, if the deterioration of extra-legal governance explains AD’s success in some communities
with bountiful coca leaf crops, then the experiences of former coca leaf farmers in these communities
ought to be consistent with the following two hypotheses:
A. The illegal trade had previously flourished with extra-legal governance
B. The stationary bandit changed its behaviour, weakening its governance and making the illegal
trade less efficient, prior to official state intervention
In addition to observing whether these hypotheses are supported by the data, I will analyse the data to
offer a reasonable explanation for why the extra-legal groups, in each community, allowed this to
happen. I would like to find out why the stationary bandit, if there was one, ceased to provide the type
of governance that had suited its own long-term business interests in coca leaf production. Did it
demand too high a protection fee from farmers, stifling their productivity? Did it fail to protect them
from disruption? Coca leaf is profitable so long as farmers retain an incentive to produce, so why did
the stationary bandit lose sight of this? In short, by what mechanism did the stationary bandit, with a
vested interest in coca leaf farming, lose farmers’ trust and neglect its own long-term business?
216
2.4 Findings
My fieldwork took place just after the Peace Agreement in 2016, but the chronology of conflict that is
analysed in this chapter pre-dates the Peace Agreement and concerns the situation prior and during the
paramilitary- FARC war in the 2000s and 2010s, and what followed up until the Peace Agreement.
The following analysis thus refers to the breakdown of FARC’s armed control before it officially
demobilized in 2016 and does not describe the more recent happenings of new extra-legal groups,
composed of FARC dissident combatants, resuming the armed struggle.20
Hypothesis A: The illegal trade had previously flourished with extra-legal governance
My findings in all three communities confirmed hypothesis A. At the time when the coca leaf trade
had flourished in each community, FARC had offered coca leaf farmers an environment conducive to
productivity and well-functioning extra-legal governance. For a while, it had sanctioned anti-social
behaviour among farmers, including theft, fighting, and kidnapping. In the veredas studied in Valle
del Guamuez and in Santiago de la Selva, FARC had allocated land and employment to newcomers,
and had facilitated trade by buying coca leaf directly at farmers’ doorsteps. This improved the
efficiency of the illegal transactions as it reduced farmers’ risk of being mugged in transit, and it
saved them time that they had previously spent transporting their produce to sell it elsewhere. FARC
had also organized collective action among community members in La Carmelita to improve public
goods such as schools, pathways, and water sources. From this, I deduce that FARC had operated like
a stationary bandit during the coca leaf bonanza that each of these communities experienced.
Santiago de la Selva
Santiago de la Selva is a small vereda located in the Caquetá department, now home to about 142
families. The closest road had been paved only four years before this study, and it was a 17-hour drive
away from Bogota, a four-hour drive away on a dirt road from the closest city, Florencia, and a one-
hour drive away (on the same road) from the closest urban centre, Valparaiso. Its geographical
20 For a detailed description of my unit of analysis in this study, see the “Units of analysis” section in the Method and Data chapter.
217
remoteness, tropical climate, and very fertile soil ensured that coca leaf could grow plentifully in the
1990s. The community’s remoteness meant that the crops were less exposed to surveillance and
forceful eradication by law enforcement.
The land had been a virgin jungle area until farmers, who had been displaced by violence in other
parts of the country, began to settle there in the 1960s. In the 1970s and 1980s they began to plant
coca leaf. Coca leaf turned out to be such good business that the farmers who were able to invest their
savings, would buy more land, or cut down trees, to further expand their coca leaf plantations. Many
brought their relatives, mainly from Nariño, Huila, and Tolima, to come and live with them and help
them farm and process coca leaf or help with the domestic work on their coca plantations.
However, as the wealth and population increased in their community, the village became disorderly.
Farmers had had to carry sacks full of coca leaf by foot, out of Santiago de la Selva, on two-hour
walks to sell it. They then had to do the same on their return journeys but with cash on their backs.
With the value of all goods increasing, there was a local price bubble and thus much higher levels of
money supply in the local economy. It was common for thieves to hide in the bushes, wait for farmers
to pass and mug them. They often killed them if farmers resisted. Sometimes these sacksful of goods
were very heavy to carry across the jungle. Some farmers recalled these challenges:
In those days, coca would be sold to buyers who would come with big sacks of
money to pay a week in advance. At first, they could take sacks full of coca leaf
without any problem. Then thieves started to emerge, mugging people carrying
coca into town. They would kill the buyers to steal the money they were carrying
[Maria Lucia, 62]
I had to carry the coca leaf out of Santiago on a two-hour walk among the bushes
and hide from thieves who would be waiting on the main pathways. More than one
got killed doing that carrying job. After doing it twice I gave up as it was too
stressful [Jose Pablo, 70]
218
As coca leaf money flowed into the community, new bars and taverns opened. Some farmers reported
that it became more common for people to drink alcohol excessively, as they had more money and
time available for leisure. Farmers associated this with increased frequency of gun, machete, and knife
fights in the community at that time. At that point, farmers generally felt there was increasing need for
law and order in their community.
Around the 1990s, FARC arrived in Caquetá and established itself as the governing authority. It
incentivised more people to come and work in the region by offering to give them work as
raspachines on arrival, and then after three months allocating land for them to grow their own coca
crops. When a FARC battalion arrived in Santiago de la Selva, it became the protector of the illegal
trade. Farmers described how the group forbade unruly behaviour, such as gun fights and anti-social
drinking and drug taking. It enforced property rights, making it easier for farmers to leave their goods
unattended. It bought coca leaf directly from farmers on their doorsteps, which meant they no longer
had to risk losing it all by personally transporting it into town. And FARC defended farmers from the
rare visits made by law enforcement. These farmers’ accounts describe these events:
When FARC entered the village, they issued warnings that they would set up their
fronts here, and that people should not fear them. But people were scared because
FARC were known to kill thieves… However, FARC turned out to be helpful, and
bought coca locally, which was easier for us. They would also charge others a tax
for buying coca [Alirio, 72]
When people produced coca in the early days, they also started to open bars,
getting drunk and drugged. It wasn’t until the guerrilla [FARC] arrived that they
began to forbid cocaine consumption … When busloads of manual eradicators
came, the guerrilla placed landmines around the coca plantations and covered the
crops in Furadan [a deadly poison]. So, when the police dogs entered the
plantations, they got very ill and had to be sent out in helicopters. The eradicators
also left in a hurry about a week later [Maria Lucia, 62]
219
Coca leaf farmers felt less exposed to risk with FARC’s protection. They would not be deterred by the
occasional visit by law enforcement, or seduced by the occasional offer of state subsidies, as one
farmer recalled:
In the early 1990s the army sometimes visited and told people to stop growing
coca, but there was a lot of coca money floating around so they would be easily
bribed. In the early days when there were AD efforts [which tended to follow
aerial eradication campaigns] people would take the fertilizer they were given by
the state and sell it to those still growing coca. At that time the guerrilla would
organize meetings warning us not to be deceived by AD as nothing was as good
business as coca was. They said they would continue to protect those of us that
wished to stick with coca [Marta Elena, 42]
In return, FARC demanded regular and means-tested protection fees from farmers, sometimes known
as vacunas, as this farmer described:
The guerrilla would visit all coca farms around Galilea, Andalucía and Nueva
Granada and expect to be given a moderate cut, called a “vacuna” [Marta Elena,
42]
It was common for farmers to keep their savings from coca leaf as cattle assets, so FARC demanded a
fee based on the number of cattle they owned, as this farmer described:
Vacunas were introduced by FARC, so people started paying COL$10,000 [about
$3 USD] per cattle head [Alirio, 72]
During the period of 1980s and 1990s, the coca leaf trade flourished in Santiago de la Selva.
La Carmelita
La Carmelita is a vereda in the Putumayo department. The closest town to la Carmelita is Puerto Asis,
which was until recently a two-hour walk to the river, and then a four- to seven-hour boat ride away
(as there is only one precarious bridge). Today it is a two-hour drive away from Puerto Asis. Like in
220
Santiago de la Selva, coca leaf, could grow abundantly in the area like many other crops, thanks to the
remoteness and the fertile soil, as this farmer described:
This land is of good quality, it is just not well connected. It has a lot of water
sources (yacimientos), when it is sunny it never dries up [Alvaro Emigildo, 55]
Farmers recalled that when there were just a few residents, the community had been very cohesive.
Then, people began to grow coca leaf in the 1980s and sell it in Puerto Asis. At the time, Puerto Asis
was the commercial centre for the coca leaf economy of Lower Putumayo (Ramirez 2011, 84).
Raspachin work was a well-paid job because it entailed hand-picking coca leaf by coca leaf, one at a
time. So, more and more people started to arrive, to work as coca leaf pickers until they could
eventually buy big plots of land to grow their own coca crops. Landowners hired people who did not
own land to crop share with them, in exchange for lunch and coca seeds.
However, farmers recalled that as the population and wealth of the community grew, so did theft and
neglect of their common resources, such as the paths in and out of their community and water
supplies. Around this time, the most active and profitable FARC unit, called the Frente 48, arrived in
La Carmelita and asserted its authority. Farmers felt that FARC improved the efficiency of the illegal
trade because it helped farmers move in and out of Puerto Asis more easily, as this farmer described:
FARC arrived and controlled people’s access to Puerto Asis. They organized the
community to fix schools and the pathways that were overgrown with bushes so
that they could be less reliant on mules to carry their bundles and start replacing
them with motorbikes [Magula, 51]
Farmers recollected how during this period, FARC organized collective action in the community, to
fix public goods which improved their transit and saved them time. As the above farmer went on to
describe, they made sure farmers’ basic daily needs were met:
FARC would set deadlines and punish day laborers who did not comply with
fixing the roads and building septic tanks and generally cleaning and protecting
221
water sources from contamination (because they were very zealous about keeping
water sources clean) [Magula, 51]
FARC also enforced property rights effectively, by building trust in the community. As the farmer
described, people felt their assets were safe with FARC in charge:
FARC kept order in this way. They ensured there were no thieves, and people
could leave their doors always open. I had 20kg of coca worth COL$40 million
[about $11,700 USD], and no one would come to steal from me [Magula, 51]
In return for these governance services, farmers paid FARC a protection fee. This farmer recalled she
was taxed on each of her cattle head:
The guerrilla asked for a cattle head here and then but were never destructive
[Sandra Mireya, 35]
So, farmers recalled that while FARC provided this well-functioning extra-legal governance, coca leaf
production had thrived in the vereda from the 1990s until the 2010s.
Valle del Guamuez
Valle del Guamuez is a municipality in the Putumayo department, close to the border with Ecuador,
comprised of many veredas that together form what used to be one of the largest coca leaf hotspots in
Colombia and an important drug trafficking corridor for two decades.21 Most veredas are located about
a 30 to 45-minute drive from the urban centre, called ‘La Hormiga’, and in some cases a river must be
crossed by foot to get there. Like in the other communities, soil in this area is very fertile. Similarly,
the veredas are hard to access by road and have no electrical infrastructure.
In the 1980’s many people from different places of the country and from Ecuador, moved to this
municipality to farm coca leaf. However, as this farmer described, when they arrived, they found that
kidnapping was common:
21 Farmers interviewed in this area came from different veredas including ‘El Oasis’, ‘El Placer’, ‘Brisas del Palmar’, ‘Alto Palmira’, ‘Miravalles’, ‘Los Guaduales’, and ‘La Pradera’.
222
In 1980s we had bought cattle and our two sons were kidnapped, one at a time.
They asked for COL$1million [about $290 USD], then for COL$3 million [about
$880 USD] for our son. We managed to raise the money and they returned our
son. A year later he was returned but, in a wheelchair, and I was forced to move
to Pasto to take care of him [Laura Elisa, 77]
In the 1990s FARC’s Frente 48 unit also gained territorial control of the area and asserted their
authority. As this farmer looked back, people were extremely obedient of them:
When the guerrilla came to live in ‘El Placer’ they calmed things down for years
[Laura Elisa, 77]
Farmers recalled that in return for annual protection fees, FARC provided coca leaf farmers with
order and a peaceful environment for trade. There were no fights and kidnapping and theft became
crimes of the past. FARC also encouraged newcomers by offering them work on arrival as
raspachines. As this farmer described, after a three-month trial, FARC would allocate land to
newcomers so that they could grow their own coca crops:
The deal was that after passing a three-month trial, the guerrilla would help you
out [Mirta Marta, 42]
Coca leaf production in the area flourished during this period because farmers felt that with FARC’s
protection from law enforcement it was safe to cut down the forest area to expand their farming as far
as possible. Plantations of 10 to 15 hectares could provide employment for 20 to 30 people each.
Farmers believed this protection had transformed the local economy for a period. It made it possible
for them to pay for their children to go to school and to improve central roads, which reduced their
need to move around on horseback. Farmers felt that the coca leaf business did so well, that the
occasional aerial eradication campaign would not deter them from persevering with illegal crops. As
this farmer described, they would swiftly replace their coca leaf crops or prune and replant the ones
that survived:
223
There were various episodes of aerial eradication but people, including us, would
go back to farming coca leaf [Jose Felipe, 75 & Leopoldina, 75]
In sum, the arrival of FARC in the 1990s filled a necessary governance vacuum that led to the triumph
of coca leaf farming in the area for several years.
Hypothesis B: The stationary bandit changed its behaviour, weakening its governance
and making the illegal trade less efficient, prior to official state intervention
This hypothesis questions three things. First, it intends to ascertain whether it was the breakdown in
extra-legal governance which created a governance vacuum and made state governance more
attractive to farmers. Second, it checks that farmers did not cease to farm coca leaf because they were
seduced by the state’s promises to improve infrastructure, investment, and subsidies. And finally, it
verifies whether law enforcement was involved when this change happened, so that we cannot
conclude that farmers were forced in any way by the state to renounce the illegal trade.
In two of the three communities included in this study my findings confirmed hypothesis B: FARC
was a stationary bandit that turned into a roving bandit when confronted by a rival extra-legal group.
Previous studies have described the region in which these communities are located, to have undergone
‘paramilitarization’ between 1997 and 1999, and then militarization between 2000 and 2004 (Hough
2011). However, in both Santiago de la Selva and in the veredas studied in Valle del Guamuez
farmers’ experience was that FARC’s extra-legal governance had deteriorated when paramilitary
groups had entered their community. And, although the timing of the conflict differed, in both cases,
this happened prior to any official state intervention.
In Santiago de la Selva, the paramilitaries arrived in the very early 2000s. At this time the
paramilitaries were not locally mobilised but operated under the AUC authority. They came from
other regions to purposely disrupt the FARC controlled areas. This supports farmers’ description of
events. The violence caused by this confrontation displaced the entire vereda in 2002, and residents of
Santiago de la Selva fled mostly to the closest town, Valparaiso, until the situation was calm enough
to return home which was at least a year later.
224
In the veredas studied in Valle del Guamuez, the peak of the conflict occurred a bit later, in the mid
2000s, just after the AUC paramilitaries had officially been demobilized. It was therefore a new
generation of paramilitary groups, including “Los Rastrojos” and “Los Milicianos”, who the farmers
recalled had confronted FARC in these veredas. Farmers in both Santiago de la Selva and in the
veredas studied in Valle del Guamuez believed that FARC’s confrontation with the paramilitaries had
caused the levels of violence to rise. And this coincided with the time when FARC changed its
behaviour towards local people. Instead of reassuring farmers who were being preyed on by the
paramilitaries, FARC began to demand farmers’ loyalty by force.
While this happened, farmers felt coca leaf farming became too dangerous and stressful. In addition, it
was financially risky because farmers had less predictable profit, if any. Farmers in these two places
confirmed that by the time the state had entered the scene, they had already grown tired of FARC’s
behaviour and could no longer distinguish them from the paramilitaries. FARC’s ability to govern the
illegal trade was thus weakened prior to any official state intervention. By the time law enforcement
increased its presence, the illegal trade had been disrupted. Most farmers had already fled elsewhere
to escape the violence, which other farmers had been growing coca leaf under duress.
When the army arrived, promising respite from the conflict, farmers felt they had an option to return
or stay on their land and avoid the continued high levels of violence and displacement they had
recently experienced. Farmers that had opted to stay on or return to their land, chose to accept the
terms of the state because they no longer felt they could count on FARC’s protection to grow coca
leaf. Farmers stated that they accepted AD subsidies as compensation for the income shortfall they
had incurred anyway as coca leaf production had waned in their locality. Crucially, these AD benefits
had not been their main incentive for switching to the legal trade. Coca leaf was already too risky a
business.
Conversely, the experiences of farmers in La Carmelita, did not support hypothesis B because farmers
recalled that the army had arrived in their community almost simultaneously with the paramilitary
groups who had behaved like roving bandits over a prolonged period. In this case, despite the onset of
various waves of attacks by the paramilitaries and law enforcement throughout the 2000s, causing
225
violence to peak in the late 2000s, FARC had remained a stationary bandit. Unlike in the other two
communities studied here, FARC did not turn into a roving bandit and people had continued to feel
protected by this extra-legal group up until they had been ordered by FARC to stop growing coca leaf.
FARC did not change its behaviour toward the local people in La Carmelita but simply ceded its
authority to the state when it reached a Peace Agreement with the government in 2016. At this point,
FARC allowed a temporary campsite known as a Zona Veredal to be established in La Carmelita for
former FARC members to disarm and demobilize, in exchange for a salary and training in skills that
would make them more employable. This suggests that in this community there was no pre-existing
governance vacuum for the state to fill. Instead, farmers switched to the legal trade voluntarily
because the state had effectively absorbed FARC’s stationary bandit authority and forced them to
comply with the law. Farmers noted that as a result, the illegal trade had shifted elsewhere.
One farmer in La Carmelita mentioned an incident of a change in FARC’s behaviour prior to the
army and paramilitaries’ arrival. Their experience was of some rogue members of FARC who had
abused their power and demanded that farmers pay unprecedented fees. However, this experience was
not widely shared, and from the other interviews, it did not appear to significantly affect farmers’
overall profitability in the illegal trade.
Santiago de la Selva
Farmers in Santiago de la Selva confirmed hypothesis B. Their description of events was that the coca
leaf bonanza flourished when FARC had governed like a stationary bandit. This was the case until the
early 2000s when paramilitary groups began to arrive. The latter planted bombs, burned villagers,
terrorized, and shot many farmers dead, in their attempt to force them to pay them a rent and sell coca
leaf only to them. The paramilitaries seemed to behave like roving bandits to aggressively compete
with FARC for the local illegal trade. Everyone in the community, not just coca leaf farmers, was
affected by the raised levels of violence and extortion:
When the paramilitaries arrived in the 2000s, they were much more violent than
FARC, they charged extortion fees and killed many people who didn’t pay. Paying
COL$10,000 [about $3 USD] per cattle head to FARC was one thing but having
226
to pay extortion fees to both armed groups was too costly… sometimes I could
negotiate with them and offer to pay COL$1 million [about $290 USD] as a
maximum payment. Sometimes they would accept, sometimes they would not. If
someone refused to pay, they were first threatened with cattle theft [Alirio, 72]
Everyone, not just coca producers would be affected by the violence. The
paramilitaries would put bombs, and burn villagers…eventually my vereda was
displaced [Pedro Pablo, 55]
Farmers recalled that to start with, FARC had promised to protect them from the paramilitaries. But
FARC were increasingly unable to prevent their rivals from harassing and stealing from coca leaf
farmers and cattle ranchers. Farmers described how at this point FARC became more daring and
exploitative. It demanded more unreasonable protection fees from coca leaf farmers and even from
those who had stopped working in the illegal trade altogether and had become exclusively cattle
owners. Like the paramilitaries, FARC began to demand peoples’ loyalty by force, and they sought to
be the exclusive buyers of coca leaf. The violence soared as a result. As Santiago de la Selva became
an acute conflict zone, the entire vereda was displaced in 2002, mostly to Valparaiso, the closest
town. These farmers describe the events of this time:
At first the guerrilla [FARC] only charged a tax to those who produced cocaine
but did not charge anyone for their cattle. But from around 2002 they became
more daring, and they started charging “vacunas” and extorting every coca
producer…[eventually] my in-laws decided to stop their coca production and I
helped them pull out the crops [Rigoberto, 52]
Around 2002 the paramilitaries arrived, and there was conflict for 20 days as the
guerrilla confronted them. The village, along with 10 other veredas, was displaced
to Valparaiso. Some guerrilla recruits came to warn us to leave in advance
because they [FARC] were going to burn people’s houses in revenge for hosting
the paramilitaries. … this memory is still very hard to talk about… even for the
227
community… the population decreased significantly after the mass displacement
[Marta Elena, 42]
Farmers felt threatened by both extra-legal groups and put their lives at risk every time they tried to
sell their coca leaf to anyone. They would be chastised by one of the groups, either way. Sometimes
they would be intimidated even when they hadn’t sold to the other group. As these farmers recalled,
they began to make less profit from coca leaf and farming it no longer seemed worth the risk:
When the paramilitaries arrived all the armed groups began to ask for extortion
fees and to kidnap the young… I had to sell part of my land to stop someone from
being kidnapped. To terrorise people, they would destroy farmers’ subsistence
crops, and they cut off all the limbs off a man and left him there. No one could say
anything for fear of being attacked. The price of coca drastically dropped as a
result and it wasn’t worth the work as it didn’t cover my expenditures [Arcadio,
67]
My brothers started to switch to cattle ranching and sugar cane only to avoid
trouble when the armed groups came to visit [Marta Elena, 42]
I switched to only cattle ranching as I saw that coca production attracted violence
and trouble at the canteens... I preferred a bit more peace of mind even if I did not
earn as much, but armed groups would always take a cut of our profits anyway…
[Hipolito, 72]
Farmers described that the state had increased its presence in the vereda after the paramilitaries had
done so, which means it would have been after 2002. The arrival of the army and coca crop manual
eradicators had resulted in a full-blown conflict. The residents of Santiago de la Selva felt unprotected
by their former protectors and were displaced en masse from their homes.
Later there was a full-blown conflict between the different armed groups on my
land. 400 FARC fighters against 300 paramilitary fighters, so the FARC won. But
lots of people were killed and all my cattle was killed and stolen. Then, when they
228
began aerial spraying, I stopped growing it as they [FARC] had threatened me
too, so I left [Arcadio, 67]
People stopped growing coca when the violence got too heavy. With two armed
groups competing to buy coca, they would kill people and steal cattle from those
they found selling to the other group. Cattle was often stolen. People started
selling their land and migrating elsewhere. The population started decreasing.
Then the aerial spraying began and took place six or seven times, until 2008 or so.
Eventually that helped to end coca production here [Alirio, 72]
In the mid-1990s there was occasional aerial eradication, but we persevered with
our coca production. Then there was intense fighting among different armed
groups. We produced coca there until 2007 when the army entered… they blocked
the entry of petrol, and the exit of money, which made it hard for the coca
business…then there was another aerial spraying which finally ended with all
coca production in the area [Cornelio, 55]
After the community was displaced in 2002, the state
gained full territorial control of Santiago de la Selva and
people returned to their homes. It was not until then that
AD interventions began.
As can be seen in Figure A, official records from
UNODC show that coca production in the municipality
of Valparaiso, Caquetá, where Santiago de la Selva is
located, declined sharply in the early 2000s. This
coincides with farmer’s accounts of the difficulty they
faced in trying to farm coca leaf amidst the violence
created by the confrontation between FARC and the
paramilitaries. The lowest record of coca leaf is the
Figure A UNODC annual figures of hectares of coca leaf recorded in the municipality of Valparaiso from 1999 – 2014. Dashed line marks start of official AD interventions. Shaded area marks the national average number of hectares of coca leaf per municipality.
0
500
1000
1500
2000
2500
1999 2001 2003 2005 2007 2009 2011 2013Year
Coc
a (h
a)
Valparaíso
229
same year the mass displacement of Santiago de la Selva happened. Presumably a similar pattern
occurred in other close by veredas in the municipality.
The records for AD interventions are at the municipal level, and not the vereda level, so there is no
exact date for when AD began in Santiago de la Selva. However, we know that there was AD in the
municipality of Valparaiso in 2003 with Familias en Accion, Programa Familias Guardabosques
(which ran until 2013) and Proyectos Productivos (which ran until 2016). Then there was another
wave of AD interventions which occurred in 2012 and 2015 with the Poserradicacion y Contencion
initiative. These interventions are marked by a dashed line in Figure A. From this we can infer that
there was no AD in Santiago de la Selva prior to their mass displacement and drop in coca leaf
production, as the earliest AD could have been introduced was in 2003. By this time, those that
returned had already ceased to grow coca leaf as requested by the state. Subsequent increases in coca
leaf production illustrated in the graph may reflect farmers who moved elsewhere to continue coca
leaf production away from veredas like Santiago de la Selva.
The main reason farmers gave to explain their decision to stop farming coca leaf at home was that
they no longer felt protected by FARC. Without FARC’s governance they felt more exposed to law
enforcement. Farmers said they remained unconvinced that trade in legal agriculture would be more
profitable than coca leaf had once been because the land was still too remote to facilitate trade. So,
their acceptance of state governance was simply because there was no extra-legal governance to
protect the coca leaf trade:
Then the army came and manually eradicated my coca crops… And there was
aerial eradication also - I felt like I was in an airport with all the aerial
eradication airplanes that would fly over us. My crops were all dried up and
without protection I didn’t want to run the risk of losing it all again by starting
from scratch... Others moved their coca production to Cauca and Nariño [Oscar,
54]
230
Most farmers had also come to associate the coca leaf trade with violence and extortion. They felt that
in switching to legal farming they would be able to live in a more peaceful environment. The prospect
of non-monetary benefits, like peace, may explain why even those farmers who did not meet the
eligibility criteria for AD subsidies, abandoned illegal drug production. As one farmer summarized:
Coca leaf is “the cancer” … instead of working to live people worked for money.
And war, death and theft followed from that. So, what’s the point of money? The
army and armed groups came…and living here was unbearable… [Jose Pablo,
70]
Farmers felt that the peaceful environment had resulted in improvements since the times of violence.
They now had better local education services including free meals for children, which helped families
that could only afford sugar water for breakfast. This had motivated many farmers to send their
children to school. They also had some access to electricity in the last two years as well as access to
regular health brigade visits.
And, significantly, there had been large-scale private investment in their land. Nestle established four
dairy storage facilities in Santiago de la Selva a few years after the state gained territorial control (four
years before this study). This had prompted many farmers to grow pasture for cattle and milk farming
instead of persevering with coca leaf crops. As one farmer described:
After the conflict and coca production dried up in the area, the land was
destroyed. But cattleranching does well in the land because the “Medicula” grass
grows well there. Nestle established 4 dairy storage facilities in Santiago. And
there is a cheesemaker. Unlike crops, where one must wait for a harvest every six
months to a year, people are paid fortnightly [Aurelio, 67]
In sum, FARC’s change in behaviour in Santiago de la Selva had begun before the arrival of law
enforcement. Unable to protect the coca leaf trade from theft and extortion by the paramilitaries, it
saw its time horizon in the community contract. With less foreseeable time to profit from farmers, it
extracted higher rents by charging them higher protection fees and demanding their loyalty by force.
231
The abuse that followed, decreased the incentives farmers had to grow coca leaf and increased the risk
of doing so. When the army arrived, the levels of violence were heightened which provoked the mass
displacement of the community.
As the state gained territorial control in the early to mid 2000s, returning farmers sensed it was too
risky to continue growing coca leaf without any extra-legal protection. FARC, as their stationary
bandit, had turned on farmers, which had deterred them from continuing to work in the illegal trade.
They associated state governance with a peaceful environment. In retrospect, they felt this shift had
eventually restored their quality of life and had attracted new opportunities for legal agriculture in the
community which had staved off coca leaf farming in the community.
Valle del Guamuez
Farmers in the veredas studied in Valle del Guamuez recalled a similar series of events as those in
Santiago de la Selva. The paramilitaries arrived in the mid 2000s and began to harass farmers working
in both legal and illegal farming. They demanded protection fees and farmers who refused, or could
not afford, to pay them were kidnapped for days at a time. Sometimes they would tie people up in the
mountains, until they could find a way to pay. FARC was unable to shelter farmers from the
behaviour of the paramilitaries. So, if farmers did not pay the extortion fees they were asked for, they
would be killed by the paramilitaries. This farmer described how she felt about the behaviour of the
incoming paramilitaries before she was forced to flee:
I remember that from a very young age, the guerrillas [FARC] had always lived
here… I say to my children… I would have preferred ten thousand times that the
guerrilla lived there, and that the paracos [paramilitaries] had never come…
when the paracos began to enter, first we would hear bombs and gun shots at
midnight in a neighbouring village…then when they came to where we lived, they
started to blame innocent people… they would say, “yes, you have something to
do with the guerrilla, so take that!”. Thank God we were lucky to escape with no
harm, but we left the day after there had been vicious gunfighting all day, and all
232
night… We left our house the following day at three in the morning, I remember
that Mommy woke all four of us… [Sandra Lorena, 24]
Despite this, FARC began to demand higher protection fees from the same people. Farmers struggled
to pay multiple protection fees and lived in constant fear of being attacked by both extra-legal groups.
Some farmers felt it was pointless to persevere with coca leaf when they were not be able to make the
profit they had once been able to. So, the excessive extortion and theft seem to have made farmers
grow disenchanted with the coca leaf trade, as these farmers described:
With higher “vacunas” growing coca did not make business sense anymore. Once
you add up the cost of the petrol pumps and supermarket food and wages for day
labourers, and clothes, our earnings did not go far [Jose Felipe, 75 & Leopoldina,
75]
The guerrilla demanded an annual “vacuna” of COL$140 to COL$150 million,
which reduced our profits. I wouldn’t have had such a problem with paying just
one group, but six months later the paramilitaries charged me more than I had,
and I was married and had four children… Many of my school friends were killed
for not paying all the “vacunas” … In 2007 they kidnapped me for four days and
tied me up in the mountain until I could pay. At that point I realized I was working
for other people and decided to give up working with coca. When I was freed, I
sold everything and became a day labourer... [Pedro, 42]
We had 10 hectares and every two to three months depending on the weather and
the crops, our coca harvest would earn us COL$5 million [about $1,400 USD].
And we would save it. But the situation started getting tricky when the
paramilitaries came and people’s property were robbed... gangs of rapists formed,
there were rumours that they were raping women, and I felt threatened… The
guerrilla started charging “vacunas” and because we were doing well, they
233
kidnapped my husband for a week. At this point we packed up and moved to where
my parents were working [Monica, 30]
In the mid 2000s the paramilitaries came to ‘El Oasis’ [a vereda] and harassed
me for a “vacuna” which I tried not to pay. I considered giving them fake money
to trap them but felt at high risk if they found out. I got death threats from the
paramilitaries and the guerrilla [FARC] alike… Sometimes these groups would
protect us and warn us when strangers at night were going to intrude…but we
didn’t really know who to trust. This went on until law enforcement cracked down
around 2010 [Orlando, 56]
In addition, the paramilitaries competed with FARC as the main buyers of coca leaf. FARC began to
use violence indiscriminately to discourage farmers from selling their coca leaf to the paramilitaries.
Farmers could no longer choose who to sell their produce to. Sometimes FARC would come to
farmers’ doorstep to buy coca leaf from them, preventing them from going into town to sell it to the
paramilitaries who paid higher prices for it. Sometimes women took on the role of going into town as
people felt men were at a higher risk of being killed. Without protection, it became impossible for
farmers to continue selling coca. As some farmers described, many fled from their land and these
hostile living conditions:
During the coca period there came a point when people were too afraid to leave
their homes. They [armed groups] would kidnap neighbours and tie them…I
would have to go to El Placer to claim them back… After that passed, not many
people are left… only the poor were left behind [Marianeli, 46]
For a while we persevered with coca production through some of the violence, but
the violence got so bad that at one point we could not even leave our homes. Men
could not sell coca as they would get killed regardless so women had to be the
ones transporting merchandise in and out of our homes. We were forced to sell to
the guerrilla who would come and buy it at our doorstep. If we sold any to the
234
paramilitaries in town, even though they paid more, we would have been killed by
the guerrilla afterwards. With our savings we’d bought a small car in which we
would bring food back from town. But the guerrilla would stop us on our way in
and once threatened to kill us because we refused to hand over our car… it was
our life savings. Then we were accused of cooperating with paramilitaries. So, we
had to escape in the car to my mother-in-law’s in Samaniego. We were displaced
for six years [Rubi, 46]
Farmers described that abuse was rife and people were killed in cruel ways by both FARC and the
paramilitaries, each accusing them of cooperating with their rivals. This meant that farmers could no
longer tell the different groups apart. As some farmers described, people felt caught in between
competing buyers, and were confused as to who they could trust. All this added to their feeling of
being unprotected:
We had no way of telling FARC apart from the paras. The paramilitaries settled
for six years on my farm. We would just hide when we heard gunfighting. Armed
groups would tie farmers up to frighten them... People were taken in vans and
killed by the river… I left for a while but then eventually came back [Laura Elisa,
77]
I had saved up a total of about COL$100 million [about $29,000 USD] from all
my coca leaf farming. With that I built my house and bought some small land plots
in La Hormiga... the rest was all stolen by the armed groups. They stole
everything, even the jewels I had bought my wife…worth COL$800,000 [about
$232 USD]. They threatened to kill me for cooperating with one group or the
other... They would say they were the guerrilla, but people would never know…
The only way that armed groups could be distinguished from each other was when
the guerrilla [FARC] would wear an arm band with Colombia’s flag when they
were in fighting mode. Otherwise, they could not be told apart from each other. I
had COL$48 million [about $13,900 USD] I had saved to buy a house in Ipiales
235
but in 2001 the guerrilla conducted an armed robbery and stole COL$35,000
[about $10,000 USD] from me. Then the paramilitaries wanted the same amount
and kidnapped me for a day claiming I was sided with the guerrilla. They
threatened to kill me, then my wife and then my daughter… but I had no money
left. I was taken to a place called Puerto Amor further above from El Placer,
where they were supposed to chop people up and drop the bodies into five holes.
They were ready to chop me up with machetes, but the community came to rescue
me. Between friends and family, they managed to raise COL$25 million [about
$7,200 USD] to rescue me. The paramilitaries still broke my arm and stole my
new Mitsubishi car and my motorbike [Jose Gerardo, 60]
Farmers recalled it was after the confrontation between FARC and the paramilitaries had already
begun, that the state intensified its law enforcement efforts. This raised the levels of violence further
and caused further displacement in the late 2000s. Some farmers who had wished to remain on their
land, but who no longer felt protected by FARC, were tempted to switch their loyalty to the state,
when they were offered AD subsidies in return for substituting their coca leaf. But at this point FARC
had issued death threats to anyone considering siding with the state. This meant that for a short time,
many continued with coca leaf production because they were afraid that if they did not, there would
be reprisals from FARC. This was unusual as farmers had never felt forced to farm coca leaf by
FARC, it had just made better business sense to them. In the absence of credible extra-legal
protection, farmers began to seek protection from the state, as one of them described:
As village leaders, we signed up to a state subsidy programme pledging to eradicate the coca
in the vereda. However, women leaders had to do it at night, to protect men who were more
likely to be killed if they were found doing it. But then when the women eradicators started
receiving death threats, we had to call the army in to protect us. The guerrilla opposed
voluntary eradication and threatened me. I had to leave my children for six months [Lucila,
51]
236
Eventually the extra-legal group scenario shifted. The paramilitaries left and FARC demobilized.
Farmers who still wanted to profit from coca leaf, moved to nearby river areas such as ‘El Venado’
and ‘Siberia’, about an hour away from ‘La Hormiga’. In these places FARC retained territorial
control and paramilitaries had not yet disrupted the coca leaf trade. As this farmer described:
People have stopped producing coca here, but most have continued elsewhere.
Those that wanted to continue growing coca (about 40% of the vereda – so a
significant proportion) agreed to do it in other veredas to not obstruct the state
subsidies for their home community. They purchased two or three hectares of land
informally and sell their coca to the highest bidder [Lucila, 51]
The farmers who stayed behind accepted the decline of the coca leaf trade in the late 2000s along with
the new AD subsidies that were on offer a few years
later.
Figure B shows that coca leaf production in the
municipality of Valle del Guamuez declined in 2002,
which again coincides with farmers’ accounts that
fighting between extra-legal groups hit the drug trade
economy. The people in the veredas I visited recalled
that the paramilitaries had arrived in the mid 2000s, that
coca leaf farming had continued in their veredas until
the late 2000s, and that AD subsidies had been offered
as early as 2004 in some veredas, and as late as the mid
2010s in others.
Records provided by the UNODC Colombia country office, show that AD interventions in the
municipality of Valle del Guamuez began in 2003 with Programa Familias Guardabosques (which
ran until 2013), and Proyectos Productivos (which ran until 2016). Then in 2012, 2013 and 2014 new
benefits and AD subsidies were offered in Valle del Guamuez through the Poserridacion y
Figure B UNODC annual figures of hectares of coca leaf recorded in the municipality of Valle del Guamuez from 1999 – 2014. Dashed lines mark the start of official AD interventions. Shaded area marks the national average number of hectares of coca leaf per municipality.
0
5000
10000
15000
1999 2001 2003 2005 2007 2009 2011 2013Year
Coc
a (h
a)
Guamuez
237
Contencion initiative. And in 2015 the AD programme offering land titles called Formalizar Para
Sustituir was rolled out, followed by the Programa Nacional Integral de Sustitucion de Cultivos
Ilicitos (PNIS) in 2017. These are marked by dashed lines in Figure B.
However, it should be noted that these are the official dates of the programmes, and the
implementation of these programmes did not take place in every vereda in the municipality, nor did
everyone qualify for these subsidies, nor did they run continuously for their official period of
duration. This brings us back to one of the reasons it is difficult to measure the effectiveness of AD in
general terms and at the municipal level.
But even if it is feasible that AD incentives had been offered prior to the arrival of the paramilitaries,
according to my interviews, people in their veredas had only stopped farming coca leaf once they had
become caught up in the violence caused by confrontations between FARC and the paramilitaries.
Without their usual protection from FARC, the situation became unbearable and forced many to flee
in the mid and late 2000s. After this, the state became increasingly present through repeated law
enforcement visits. During this period farming coca leaf, or even just living, in their vereda was very
risky, and only a few persevered with coca leaf farming amidst such violence. It was after the violence
had subsided and FARC was no longer in command, that the state rolled out the bulk of AD subsidies
in their veredas in the early 2010s. By this time, most people in their veredas had already given up on
coca leaf.
When asked why they had decided to substitute their illegal crops in their specific veredas, the overall
answer from the farmers I interviewed was that without effective protection for the illegal trade, there
was no money in it. Without a stationary bandit to deter law enforcement, farmers felt they would be
more exposed to future aerial spraying which damaged everyone’s agricultural produce.
Moving to places where it was still safe to grow coca leaf and possible to make a profit from it, did
not guarantee farmers that things would not turn sour like they had in their home community. They
were desperate to avoid the violence and extortion they had recently experienced and perceived to
come hand in hand with the illegal trade.
238
Farmers were aware that the cash, machetes, spades, grass seeds, corrugated iron sheets for roofing,
water tanks and wire for fencing offered to them by the state would not make them nearly as much
profit as they had from coca production. Nevertheless, they accepted them as some form of
compensation for their drastic income shortfall they had already incurred. Previously, six hectares of
coca leaf would earn a farmer COL$1 million [about $316 USD] to COL$2 million [about $632 USD]
every three months, that is about $52 USD per week. At the time of this study farmers’ average
weekly outgoings were about COL$60,000 (about $19 USD), if they only ate meat once a week.
Many still struggled to earn this much with legal farming in their small land plots.
Nevertheless, compared to other areas in the country Valle del Guamuez has continued to be a hub of
illegal crop farming. Therefore, the accounts from the few veredas that I visited where coca leaf
farming had already subsided cannot be generalized to the experience of everyone living in Valle del
Guamuez. The question that follows then, is why did these veredas stop farming coca leaf despite the
municipality of Valle del Guamuez still being an important coca leaf hub? Or framed differently, why
did FARC’s behaviour change in these veredas by becoming a roving bandit, whereas it continued to
behave as a stationary bandit in other areas of the Valle del Guamuez municipality? One explanation
may be related to their location.
239
Figure C Map of Valle del Guamuez, Putumayo. The yellow stars indicate the location of the seven veredas included in this study: Alto Palmira, Brisas del Palmar, El Cairo, El Placer, Los Guaduales, Miravalles and El Oasis
The map in Figure C shows the location of the seven veredas I included in this study. In relation to the
rest of the municipality of Valle del Guamuez (borders highlighted in red) it is evident that they are all
in very close proximity to the main highway roads compared to much of the municipality where there
is hardly any road infrastructure. Some are very close to highway 45 and others to Via La Hormiga
(highlighted in white). Some are closer to the road than others but they are all situated between a 30 –
45-minute drive away from La Hormiga, the main town in Valle del Guamuez. One explanation for
this is that alternative livelihoods to coca leaf farming are more realistic in places where it is easier,
and therefore cheaper, to transport agricultural products to an urban centre, to sell them in the market.
This brings us back to the overall discussion covered in the introduction of this thesis about what
makes AD effective. Easy market access increases the likelihood that AD investments will have more
lasting effects on reducing farmers’ financial dependence on illegal crops (Dion and Russler 2008;
UNODC 2015).
La Hormiga
240
However, this explanation assumes that legal alternatives to coca leaf can become, if not as profitable
as coca leaf, then at least a good enough business for people to substitute their coca leaf crops for
good. The fieldwork interviews did not confirm this was the case even for farmers who lived in these
veredas. Although relatively closer to the main roads and to La Hormiga than most places in
Putumayo, farmers still felt highly marginalized from the agricultural markets. And compared to other
places that were closer to the town markets and where there had been greater private investment in
legal agricultural alternatives to coca leaf, these veredas were still at an economic disadvantage to
compete in the legal agricultural market.
I would argue that the more plausible explanation is deterrence. The location of these veredas makes
them more accessible for more frequent visits by law enforcement. It is easier for the state to monitor
places that are located relatively close to the main road. In addition, there is a regional UNODC office
located about a thirty-minute drive away from these veredas. So, it is probable that without the
protection they once felt from FARC to grow coca leaf, the people living in these areas feel under
greater surveillance by the state if they are caught using their land for illegal crops again.
La Carmelita
Farmers’ accounts in La Carmelita failed to confirm hypothesis B. According to farmers’ accounts,
there was a slightly different evolution of events from what happened in Santiago de la Selva and the
veredas studied in Valle del Guamuez. Farmers in La Carmelita recalled the army and law
enforcement had arrived intermittently during the 2000s and sometimes at the same time as
paramilitary groups. The paramilitaries had started to compete with FARC for the local coca leaf trade
by offering farmers a better buying price for their produce. At the same time as the aerial eradication
campaigns had intensified.
When the paramilitaries and law enforcement had first arrived, FARC had assembled the community
and ordered them to continue selling their coca leaf only to them. At the same time, the paramilitaries
had threatened people who would not sell it to them. Farmers described how they were confused and
scared about who they should sell their coca leaf to:
241
The guerrilla assembled the community and ordered them to only sell to the
FARC. Then the paramilitaries sent out flyers also ordering people to only sell to
them and offering to pay COL$200 [about $59 USD] more per kg for people
bringing their coca all the way to Puerto Asis. The paramilitaries then controlled
Puerto Asis, which is where they would hold their meetings whereas FARC would
hold their meetings in rural areas, they had long controlled. So, coca farmers
were confused and scared about who they sold their produce to. Innocent people
were constantly killed for suspicion of siding with the paramilitaries or with
FARC on their visits to Puerto Asis [Magula, 51]
Nevertheless, farmers had continued to trust FARC as their protector and had not felt terrorized by
them, as had been the case in Santiago de la Selva and the veredas studied in Valle del Guamuez. On
the other hand, they did not trust the paramilitaries or the army:
At one time the whole community was caught in between the guerrilla, the
paramilitary groups, and the army. The paramilitaries killed many people
wrongly, whom they accused of being guerrillas [Alvaro Emigildo, 55]
People would be given 24 or 36 hours to vacate their homes by the paramilitaries.
My father was attacked at night for information about the FARC, and my sister’s
home was raided… my young daughter got into trouble for falling in love with a
FARC militant... There seemed to be an alliance between the army and the
paramilitaries, because it seemed strange that the paramilitaries would be able to
pass through where the army was meant to be keeping order. Some soldiers would
come at night in uniforms and identify themselves as paramilitaries. But people
were too scared to denounce them to the police [Magula, 51]
As they continued to feel protected by FARC, farmers had initially persevered with coca leaf
production, despite increased law enforcement. When they had glyphosate sprayed on their crops,
they would prune the crops to restore them to life:
242
During this period there was a lot of aerial eradication, but we persevered with
coca if we could [Magula, 51]
Figure D shows that coca leaf production in Puerto
Asis, the municipality in which La Carmelita is
located, decreased intermittently throughout the 2000s
and early 2010s. And marked by the dashed lines can
be seen the AD interventions that took place in Puerto
Asis. These officially began as early as 2003, with
Programa Familias Guardabosques and Proyectos
Productivos. These were followed by the
Poserradicacion y Contencion initiative n 2012 and
2014. Again, these initiatives were not implemented in
every vereda in the municipality, nor did everyone
qualify for these subsidies, nor did they run continuously for their official period of duration. So, from
this data, no deduction can be made about the effect of AD on coca leaf farming in La Carmelita, at
least in the long term.
However, what we can infer from my fieldwork interviews, is that farmers in La Carmelita persevered
with coca leaf farming up until the mid 2010s This was despite previous visits by the paramilitaries
and by law enforcement and AD programmes being rolled out ad hoc in other parts of the
municipality. It was only in the mid 2010s when FARC advised farmers to accept AD in their vereda
that they finally did. Their explanation was that as FARC was repeatedly confronted by the
paramilitaries and the army, sometimes at the same time, there came a point when they could no
longer protect farmers from the heightened violence:
The army confronted the guerrilla [FARC] and there was a lot of violence. Once,
the army came into our school and there was a shooting there by the guerrilla.
Another time a bomb was dropped from an army helicopter that landed on a
chicken hut roof [Sandra Mireya, 35]
Figure D UNODC annual figures of hectares of coca leaf recorded in the municipality of Puerto Asis from 1999 – 2014. The dashed lines mark the start of official AD interventions. Shaded area marks the national average number of hectares of coca leaf per municipality.
0
5000
10000
1999 2001 2003 2005 2007 2009 2011 2013Year
Coc
a (h
a)
Puerto Asís
243
So, in 2016, the local FARC battalion relinquished their power in that vereda, as part of the wider
Peace Agreement the government reached with FARC. La Carmelita became a zona veredal, a special
area hosting 500 FARC combatants from other areas in Caquetá who wanted to demobilize. At this
point, under the instruction of FARC, the community finally agreed to stop farming coca leaf, and
accept the new AD subsidies on offer by the state. The farmers described that the community had
been loyal to FARC until they were advised by FARC to switch to legal agriculture or move to grow
coca elsewhere:
FARC held meetings with 62 communities and announced their ceasefire and
encouraged people to live peacefully after the Peace Process… some people chose
to continue producing coca leaf, and moved to ‘San Miguel’, a vereda located an
hour away [Sandra Mireya, 35]
Those who opted to stay, had no alternative but to accept state governance, and go along with the
Programa Nacional Integral de Sustitucion de Cultivos Ilicitos (PNIS) launched in 2017. This meant
that they received COL$1 million (about $298 USD) per month in AD subsidies in return for getting
rid of their coca leaf crops. This programme would only run a year and was meant to tide them over
until they got their legal alternative use of land up and running. Farmers’ acceptance did not mean that
they trusted the state as their protector, but they saw no harm in getting some compensation for
foregoing their established livelihood. And they had been prompted by FARC to accept. Some
farmers continued with the illegal trade but in other places away from La Carmelita where they felt
they could rely on FARC’s protection:
We had coca crops until this year but were forced to eradicate by the army…
people could choose whether to eradicate their crops on good or bad terms… So,
we were one of the 111 families that chose to sign up for the AD subsidies. We
signed contracts and they came to measure our land with GPS to help them to
manually eradicate our coca crops…but we’d camouflaged coca so well among
our other crops that there was still some coca hidden that the GPS couldn’t find
… we told the manual eradicators about it and they didn’t believe us at first but
244
we showed them where it was located…we feared that if we hadn’t alerted them
about it, sooner or later they would have found out [Sandra Mireya, 35]
Theft has increased a lot and there have even been stabbings for stealing
motorbikes. This is because the guerrilla is no longer around… they used to make
sure that there was no theft. Before, people could leave their front doors open.
Now people are afraid of having their motorbike stolen. The police are sometimes
able to rescue stolen property, but the criminals don’t get caught. People don’t
trust the police as much as they did FARC. FARC also used to say that people
should only worry about their own property, and everyone knew who owned what,
they did not even need to interfere [Sandra Mireya, 35]
Only one farmer experienced a change in FARC’s behaviour before the arrival of rival extra-legal
groups or the army. She described that there had been rogue FARC members pretending to be
operating on behalf of their officials and extorting farmers on their own initiative. She was told that if
she did not cooperate, they would spread false rumours about her family among their superiors so, she
had obeyed:
FARC would also ask for a cut and a cow here and there. Some of them took
advantage and falsely in the name of Frente 48 [the FARC unit that was in
charge] would take people’s motorbikes and cars and go out with women and ask
for COL$1 million (about $298 USD) in two days’ time. They were random
demands, with no reason. But people cooperated because they had been
threatened that if they didn’t, these guys would spread false rumours and tell their
superior FARC officials, that they [farmers] were allied with the army and that
they had passed on information to them. This meant automatic shooting. People
just wanted to avoid trouble [Magula, 51]
It is difficult to infer from the small sample of interviews if this example of FARC’s abuse of power,
which was only mentioned by one individual, was a side incident or perceived as usual behaviour by
245
most farmers. The behaviour resembles that of self-seeking individuals that seek to increase their
share of existing wealth within groups without creating new wealth (Kruger 1974; Tullock 1967).
Smith & Varese (2001) argue that impostors periodically free ride on the Mafia’s reputation to seek
payments from entrepreneurs, and perhaps this is the phenomenon reflected by this experience.
However, this type of incident was not widely reported and from farmers’ accounts, did not seem to
impact the overall efficiency of the illegal trade in that community. The overall picture that the
interviewees provided for the events in La Carmelita, was that FARC had retreated when the state
arrived with the paramilitary forces. They had encouraged farmers to give up on coca leaf and accept
the terms of the state. Farmers did not feel let down by FARC, like they had in other places. This
community was one of the places that FARC agreed to withdraw from in the Peace Agreement with
the government.
246
2.5 Discussion
All three communities in this study had favourable soil and climate conditions for growing coca leaf.
This meant that the illegal crops had optimal conditions. In addition, all these communities reported
that for a significant period, a single, extra-legal group, namely FARC, had governed coca leaf
production effectively. It had behaved like a stationary bandit and profited from the drug farming
business from the 1980s and at least until the 2000s.
Farmers recalled that FARC had at one point provided a stable environment, invested in local
infrastructure, and organized collective action for community goods. It should not be a surprise then,
that when FARC had been in charge, local coca leaf trade had flourished in each of these places. This
was the case until the early 2000s in Santiago de la Selva, until the late 2000s in the veredas studied in
Valle del Guamuez, and until the mid 2010s in La Carmelita. This supports previous research that
found FARC to have strong political legitimacy at the local level because it protected coca-leaf
farmers from other extra-legal groups and from the army, and because it implemented state-like
taxation of many coca leaf producing regions (Hough 2011). Similarly, increased aerial fumigation of
coca leaf crops helped to legitimize the role of the guerrillas as coca leaf farming protectors (Ramirez
2011, 61).
Sometimes the state can take the credit for AD success
Increased state investment and increased military presence are often the two reasons used to explain
why farming communities in Colombia voluntarily decide to substitute their illegal crops with legal
alternatives. The experiences of farmers in La Carmelita, support this traditional state-centric
explanation of AD. Farmers in this community had not detected changes in the behaviour of FARC to
the point that they ceased to regard them as their protectors (except for one incident of FARC
members abusing their power and extorting farmers beyond the agreed protection fee). It was not until
the army and the paramilitaries had simultaneously arrived and repeatedly attempted to sabotage drug
farming, that FARC’s authority, and ability to govern the trade, was disrupted. Farmers in this
community recalled that the army had ultimately forced FARC to relinquish its territorial control and
247
had outperformed FARC in its ability to protect farmers. As a result, the state could claim victory for
AD success in that community.
Sometimes stationary bandits become roving bandits
Nevertheless, in Santiago de la Selva and the veredas studied in Valle del Guamuez, farmers
experienced a different mechanism by which AD had succeeded. The downfall of the illegal trade in
these communities occurred by the stationary bandit turning into a roving bandit. This section
describes how this may have happened.
Thriving illegal trade attracts rival extra-legal group
It should not come as a surprise that communities where coca leaf grew abundantly attracted
competition from FARC’s rivals. A study shows that the onset of paramilitary attacks increased with
coca crops (Cortes and Montolio 2013, 11). The lucrative illegal drug trade is bound to draw
likeminded entrepreneurs who seek financial gain in some way. In these communities FARC’s rivals
were primarily paramilitary groups. In the case of Santiago de la Selva, the paramilitaries came from
other regions and operated under the AUC authority. In the case of the veredas studied in Valle del
Guamuez, the AUC paramilitaries had officially been demobilized so the paramilitary groups that
came to disrupt FARC’s territorial control were those that were locally mobilized, including “Los
Rastrojos” and “Los Milicianos”.
As discussed above, extra-legal groups can behave like stationary bandits or like roving bandits. In the
case of the latter, they use violence to deliberately create demand for their protection services
Figure 10 Box plot showing numbers of victims (including kidnapping, death threats, sexual violence, displacement, homicide) by municipality where with different extra-legal group presence from 1999 – 2014. It draws on panel data provided by CEDE Research Centre at the University of Los Andes in Bogota, and demographic data on population from the Colombian National Administrative Department of Statistics (DANE 2018). ‘AUC’ denotes all paramilitary groups.
248
(Gambetta 1993; Shortland and Varese 2016; Skarbek 2012). Farmers’ experiences in Santiago de la
Selva and in the veredas studied in Valle del Guamuez confirmed that the paramilitaries behaved in
such a way, as they had been the ones to originally raise the levels of violence. Figure 10 shows that
the number of victims between 1999 to 2014 in Colombia, was highest in municipalities where
paramilitaries were officially present (shown in grey and labelled AUC for convenience). These
include victims of kidnapping, death threats, sexual violence, as well as displacement and homicide.
As seen previously in Figures 7 and 8, the highest rate of homicides and displacement occurred in
areas that were contested between various extra-legal groups, (shown in green). This suggests that the
paramilitaries’ tendency to behave particularly abusively from the start, exhibiting the highest levels
of kidnapping, death threats and sexual violence, is supported by Figure 10. They did not become
protectors because they had no ambition to develop political legitimacy in the community. As was
discussed in the theory section, their strategy for extracting long-term rents was not by offering an
ideological alternative to the state, but by infiltrating the state institutions instead.
At this point it should be noted that if one parts from the Olsonian stationary – roving bandit model
and takes a more comprehensive understanding of the state that defines it as a network that can
operate through its affiliation with extra-legal groups e.g., the paramilitaries, then a more complex
explanation emerges. It can be argued then that agents of the state did play a role in effecting change
by covertly manipulating their ties with extra-legal groups and causing the disruption to FARC’s
stationary banditry. In this sense, agents of the state may have been responsible for the arrival of the
paramilitaries in these places, and it was not simply because of the paramilitaries’ greed to compete
with FARC. Testing this theory, however, is beyond the scope of my research so the analysis of my
findings is based on farmers’ perceptions and on record of official state interventions only.
Stationary bandit changes its behaviour
Some argue that when a stationary bandit of an illegal trade is faced with competition, it can reach a
market sharing arrangement with their rivals so that the trade can continue to operate in a relatively
stable environment, and provide maximum output which is in their long term interests (Varese 2010).
Indeed, this was the experience of some coca leaf producing communities in the 1990s and 2000s,
249
when the paramilitaries and the guerrilla groups divided up the areas where the state was absent, so
that the guerrilla groups controlled the majority of the production side of the coca and poppy crops
while the paramilitaries were predominant in their refinement and trafficking (Thoumi 2012).
However, this was not the experience of farmers in Santiago de la Selva and the veredas studied in
Valle del Guamuez. Instead of adapting to the pressure of intruding paramilitaries and sharing the
local drug market with them, FARC had attempted to compete with them and raised additional
resources to increase its protection of the territory. Idler (2019) argues that “where multiple groups
compete to maximize profit, general distrust persists, which makes group interactions more volatile”
(Idler 2019, 38). Previous fieldwork in Putumayo has shown that different extra-legal groups recruited
youngsters to inform on others, which meant that people didn’t trust anyone as they could not tell
which side people were sided with (Ramirez 2011, 75). This distrust is unfavourable to business.
The increase in violence in these communities was not only the effect of increased confrontation
between extra-legal groups, but also due to changes in FARC’s behaviour towards farmers. FARC
added to the disruption of the existing peaceful environment when they began demanding higher
protection fees from farmers. Given farmers in these communities were unable to pay increased and
multiple protection fees to both extra-legal groups, FARC began to terrorize them into paying and to
demand their loyalty by force. Farmers recalled that FARC had started to use intimidating forms of
violence to enforce unacceptable terms of trade on them. This change in FARC’s behaviour had a
destabilizing effect in the communities, who were unable to farm coca leaf profitably without the set
of institutions they had become reliant on.
This echoes what Hough (2011) has found in other communities, where FARC had to diversify their
coercive extractive activities when faced with intensified confrontation by paramilitaries and the state
“to extract more resources, labour power, and commitment to recruit than the local population was
willing to actively consent to” (Hough 2011, 401). FARC began to replace targeted acts of violence to
more indiscriminate forms in their efforts to collect their taxes and to provide governance services.
This included massacres, forced displacement of entire villages, collective death threats, taking
hostages, public bombings and, according to Hough, public torture (Hough 2011, 402). Hough’s study
250
is based on a longitudinal data set22 and on primary data collected in multiple fieldwork visits in
Bogota, Florencia, and nearby villages in Caquetá between 2002 and 2009. It should be noted,
however, that the paramilitaries’ use of violence is widely believed to have been more horrific,
indiscriminate, and terrorizing, and on a greater scale, than what FARC inflicted on their victims. In
the words of Gutierrez Sanin, “FARC is much more intent on fighting than the AUC, kidnaps much
more and massacres less” (Gutiérrez Sanín 2008, 6). Putting aside the difference in the scale and
extent of the murderous violence inflicted by each of the two types of armed groups, FARC’s shift
towards more indiscriminate forms of violence still meant their legitimacy was undermined at the
local level in many places (Hough 2011, 403).
The accounts of farmers in these two communities suggest that when two or more extra-legal groups
compete to control an illegal trade, it can make it difficult for the original one to continue providing
governance effectively as a stationary bandit. Related to this is a study that argues that a stationary
bandit can only maintain property secure if its coercive power is unquestioned (Murtazashvili and
Murtazashvili 2016). This may be because prolonged high levels of violence between rivals reduces
the rents available to the extra-legal groups, and eventually runs their business dry (Snyder and
Duran-Martinez 2009, 258).
Put differently, the arrival of rivals shrank the time horizon in which FARC could expect to continue
profiting from the illegal drug trade in the community. This transformed their behaviour to that of
roving bandits because the shorter the period an extra-legal group can expect to control an area, the
lower the value of its assets, and the more likely it will loot tom maximize its short-term gains (Cheng
2018, 62). In other words, short term horizons mean that extra-legal groups don’t care about their next
year’s income from a given area, and so will loot opportunistically instead of continuing to invest
resources to govern it and protect its long-term business.
Moreover, previous research has shown that extra-legal groups adopt more terrorising forms of
violence to communicate the fiercest reputation when faced with increased competition (Gambetta
22 The data set is comprised of the “Banco de Datos de Derechos Humanos y Violencia Politica”, compiled by a Bogota-based peace institute (the Investigation and Popular Education Centre, CINEP) that documents incidents of FARC-perpetrated violence between 1987 and 2001, in addition to data from newspaper report from El Tiempo between 1975 and 1987.
251
1993; A. Smith and Varese 2001). The expectation is that by conveying a fierce reputation they can
frighten their subjects into paying protection fees, attract members to their group and deter
competitors from entering their market, ultimately economizing on the need for violence to enforce
their commands (A. Smith and Varese 2001, 365). The indiscriminate and more brutal use of violence
only subsides when one of the groups gains greater control of a given area (Snyder and Duran-
Martinez 2009; Kalyvas 2006).
Although this theory is for entrepreneurs that have fixed location and legal enterprises, I suggest it can
also be applied to shifting illegal enterprises, such as coca leaf production.
The demise of the illegal trade
The influx of the paramilitaries in these two communities, combined with the deterioration of FARC’s
governance, altered the transaction costs involved in the illegal trade. As FARC demanded higher
taxation from them, and the paramilitaries demanded additional fees, farmers felt extorted by both
extra-legal groups. Extortive taxation results in predation or coerced trade, instead of increased
productivity, and becomes burdensome to an illegal trade (Skaperdas 2001, 182–89; Konrad and
Skaperdas 1998).
This happens in two ways. Firstly, as people are forced to pay increasing protection fees for nothing in
return, any investment they could make in the trade is diverted to further protection, which is a purely
unproductive activity. This means the operation costs of an illegal trade increase, while the profit
available to reinvest in the business decreases or even disappears. Secondly, excessive extortion is
detrimental to economic productivity in that property rights become weak and are not enforced.
Protection fees will no longer be perceived as fair by entrepreneurs, but as extortionist, and will
eventually become socially unacceptable (Skaperdas 2001, 187; Schelling 1984). This removes
entrepreneurs’ incentive to persevere with the illegal trade.
Another way in which the coca leaf trade was compromised was that both extra-legal groups used
targeted, terrorising, indiscriminate and frequent use of violence, which had a high social cost for
farmers and reduced market competition. Violence began to play a big part in determining the value
252
of coca leaf, so farmers had to sell to the extra-legal group that would threaten to kill them instead of
the one that offered the best price for it. Farmers felt exploited and lost the incentive to produce.
Farmers also felt the illegal trade became less efficient without the set of informal constraints that
guaranteed buyers for coca leaf and coca paste, set market prices, and enforced codes of behaviour
that facilitate collective action, enforced contracts, and protected the from opportunistic behaviour. As
the transaction costs increased, their profits from coca leaf farming declined and it became a riskier
business.
As a result, farmers recalled they could no longer distinguish FARC from the paramilitaries. Previous
studies have shown that extra-legal groups may attempt to punish non-paying entrepreneurs and re-
establish their reputation in the face of competition, which leaves entrepreneurs uncertain as to who is
truly able to protect them (Smith and Varese 2001). Similarly, in these two communities, FARC lost
the trust of those they had once sought to protect. Like other businesses that are caught in ongoing
violence, it became difficult for coca leaf farming to operate well in the midst of the uncertainty,
incomplete information, and property destruction (Skaperdas 2001, 188).
So, when coca leaf farmers in the veredas studied in Valle del Guamuez and Santiago de la Selva no
longer knew who they could rely on for protection amidst the intolerable levels of extortion and
violence, they diverted their resources away from coca leaf farming and towards survival strategies.
Some members of the community responded to this dual roving bandit behaviour by moving
elsewhere, where they felt more protected. Others by ceasing production and simply trying to survive
on whatever they could grow in their small plots of land. Subsistence farming became more attractive
than during the coca leaf bonanza, as it felt safer, even if it meant a huge income loss. This echoes
land use in the Democratic Republic of Congo, where farmers are forced to grow bitter cassava when
there is no protection to grow sweet cassava, to cut their losses (Banea et al. 2015).
A governance vacuum is created for someone else to fill
It was only after FARC changed its behaviour in Santiago de la Selva and the veredas studied in Valle
del Guamuez, that farmers recalled the state had intervened. The struggle had been exclusively
between FARC and the paramilitaries until the state sent the army in. This should not be surprising
253
given that violence attracts increased policing (Gambetta 1993). In addition, optimum conditions for
the state to gain the monopoly of violence arise when there are multiple extra-legal groups and are
thus unable to pose a threat to the state by acting alone (Snyder and Duran-Martinez 2009, 257).
A struggling illegal trade made it less costly for the farmers who stayed behind, to embrace the state’s
authority and agree to substitute their remaining coca leaf crops with legal alternatives. They did not
consider subsistence farming preferable to coca leaf, but at least they felt sheltered from the violence
and had some form of governance for a different kind of business. Thus, farmers accepted the
subsidies the state had to offer even though they were not optimistic about being able to earn as much
from legal alternatives to coca leaf. Some farmers who had fled during the violent period, returned to
their home community, to make a peaceful living with the next best thing to coca leaf. Those who
prioritised profit, moved elsewhere to another coca leaf hotspot where FARC, or another extra-legal
group, could offer them governance and unrivalled protection to continue working in the illegal trade.
As they left the veredas, they also left behind land plots, that they either sold or rented out to others
(which presumably slightly improved the conditions for subsistence farming).
To summarize, in these two cases, the stationary bandit became a roving one when it resisted a market
sharing arrangement with its rivals. This damaged the illegal trade of coca leaf which it used to rely
on. A governance vacuum was then created in these communities, which was then filled by another
stationary bandit: the state.
254
2.6 Conclusion
Much of the literature on extra-legal groups focuses on their formation, and their effectiveness for
trade, when they behave like stationary bandits. This research builds on this, to present a hypothetical
mechanism for how extra-legal groups can also become disorganized. It argues that when they stop
providing effective extra-legal governance, they can neglect the illegal trade on which they relied and
facilitate a state takeover. Olson’s stationary and roving bandit framework best illustrates some
situations where AD has succeeded in Colombia. My study suggests that it is not always the strength
of state governance that stamps out illegal drug production, but that the breakdown of extra-legal
governance can pave the way for the state to gain authority.
Unlike state incentive and deterrence efforts, the dynamics of extra-legal groups is a concept that has
rarely received attention in explanations for successful AD. This chapter illustrates that according to
my interviews, in some communities, extra-legal groups’ behaviour was more relevant to farmers’
local decision-making than what the state offered or threatened. Two of the three cases presented in
this study, highlighted that the extra-legal group, that had acted as a stationary bandit, became unable
to set informal constraints that ensured the coca leaf trade could work smoothly, and ceased to enforce
codes of behaviour for farmers to cooperate, before the state intervened.
This is because the extra-legal group that had behaved like a stationary bandit in these communities,
opted to compete with incoming rival extra-legal groups, instead of adapting to them. By FARC
demanding higher protection fees and increasing the threat of terror to secure farmers’ loyalty, the
efficiency of illegal drug farming was reduced, becoming not less profitable than legal farming, but
less costly for farmers to switch to legal farming. This pushed farmers away from drug production and
increased the attractiveness of the state’s authority.
As discussed in the theory section of this part of the thesis, the Olsonian nature of statehood is a black
and white simplification of what is in fact the grey reality of the Colombian state and its violent
interventions. In real life there are nebulous boundaries between agents of the state and the extra-legal
groups in Colombia. So, my analysis provides only a limited account of how agents of the state and
255
extra-legal groups really operate in Colombia, and therefore of how AD can succeed in some
communities. When an extra-legal stationary bandit breaks down, and loses control over a territory,
such as in some coca leaf farming communities, some studies have shown that often there is more
going on than meets the eye (Arias 2010b; Arias and Goldstein 2010; Arias 2010a; Pearce 2010). The
breakdown of armed control in a locale is partly the conflict between armed groups but is also the
result of agents of the state intervening in a complex web of agreements with extra-legal groups. The
tacit realignment of the relationships between armed groups was not visible to the farmers I
interviewed, but it was probably as real as the violent confrontations they did witness.
However, I did not attempt to measure how these relationships reaccommodate behind the scenes. My
objective was to illustrate, with a simple Olsonian stationary-roving bandit model, how the success of
AD is perceived by members of some communities to stem from fighting between extra-legal groups
and not from the comparative advantage that legal farming poses over coca leaf farming. In this sense,
my analysis is only part of the picture but still a helpful depiction of how different bandits lose and
assume official power in coca leaf farming communities irrespective of any long-term economic
advantage of legal farming over what coca leaf farming offers.
My analysis leads to three main conclusions. The first conclusion is that sometimes, extra-legal
groups inadvertently fail to look out for their own long-term interests. This is because they stop
enforcing the set of informal institutions that ensure a stable environment for an illegal trade to
prosper. The second is that AD success can also be explained by the breakdown of extra-legal
governance, rather than the mere strengthening of state governance.
Once farmers lose faith in their local stationary bandit, they may become more responsive to state
governance. If the state strategically capitalizes on the hardship of coca leaf farmers who wrestle with
their low profits, it may lead to shifts in their hearts and minds. In such situations, increased law
enforcement, accompanied by the promise of state investment, results in a better chance that farmers
will desist from the illegal trade altogether. Farmers in these places will be seeking an immediate
hiatus from the high social costs imposed by extra-legal groups who do not offer any governance for
their trade. Thus, they will move their loyalty away from their former stationary bandit and towards
256
the state more easily. A similar pattern has been documented by insurgents in Iraq (Berman, Shapiro,
It is not possible to determine how typical this phenomenon is as these veredas do not represent the
whole of Colombia’s coca leaf growing area, or the places where there have been AD interventions.
However, the plot presented in Figure 11 shows that most AD interventions have been deployed in
places where there were no extra-legal groups at the start of the intervention (shown in the yellow
bar). This Figure is based on data provided by the UNODC office in Colombia on AD interventions,
and the panel data provided by CEDE Research Centre at the University of Los Andes in Bogota on
presence of armed groups. To calculate it, I examined all the AD interventions that were rolled out in
Colombia between 2003 and 2017 at the municipality level, against presence of extra-legal groups.
Figure 11 AD interventions between 2003 – 2017 in municipalities with different extra-legal group presence [data source: UNODC Colombia and CEDE Research Centre, University of Los Andes, Bogota]
The phenomenon highlighted by Figure 11, raises two important questions. Firstly, why there are so
many interventions in places where there are no extra-legal groups. Based on my research, it would
seem these may be areas where extra-legal groups had competed for the coca leaf trade to a point
where the stationary bandit turned into a roving bandit. This negatively affected the profitability of
257
coca leaf farming and most coca leaf farming had thus shifted elsewhere by the time the state
intervened. It also meant that these areas became “safe” for the state to enter and introduce AD
programmes, even though the levels of coca leaf had massively dwindled anyway. This is part of the
argument made by Daniel Rico, a Colombian pundit who has worked closely with the state in AD
efforts in his past role as a policymaker. He found that the bulk of the state’s AD efforts have tended
to occur in the places where there is very little, or no coca leaf farming (D. Rico 2017; 2020).
In addition, Figure 11 raises the question of how the state has been able to roll out AD programmes in
places where there were still multiple extra-legal groups present, and where one would expect it is
more difficult to succeed and overthrow extra-legal groups that profit from the illegal drug trade. As
this research proposes, AD does not have to make legal agriculture as profitable as illegal drug
production to succeed. AD efforts may be more effective in these places precisely because there is
heated confrontation between extra-legal groups, and illegal drug production has been affected as a
result. Rivalry between extra-legal groups creates governance vacuums, as they fail to govern because
of fighting with each other. This in turn creates strategic opportunities for the state to offer alternate
protection and governance to coca leaf farmers. I would argue that AD efforts stand a better chance of
succeeding in places where this is the situation.
In view of the research that documents ties between the state and the paramilitary groups that I refer
to in the theory part of this study, my empirical findings also help to raise questions about whether
paramilitary groups have served as the advance guard of the state. This would reflect a deliberate
strategy used by the state to gain territorial control in these communities. The government has in the
past blamed the high levels of violence in coca leaf farming communities on farmers’ ties to the
guerrilla, but this glosses over the impact of the arrival of the paramilitaries in these communities in
Putumayo, for instance, and the intense political violence that ensued (Ramirez 2011, 81).
I do not pretend the dynamics of extra-legal groups can single-handedly explain the success of AD.
For instance, this study does not explain why paramilitary groups’ behaved in this way in some
contexts but not in other places where FARC retained territorial control. Although the informal
institutional dynamics are surely a prime mover in the success of AD in these communities, it cannot
258
be the only factor. We cannot understand AD success only by changes in the behaviour of extra-legal
groups. But neither can we attribute AD success exclusively to the state’s intervention.
Clearly, there are various factors that might cause a decrease in coca leaf production, sometimes
unrelated to the actions of the state or to the lack of effective extra-legal governance. For example, the
price of gold and the US dollar exchange rate, influence the illegal activity extra-legal groups choose
to prioritize, causing them to allocate their resources to illegal mining or elsewhere (Rettberg and
Ortiz-Riomalo 2016; Garzon Vergara 2019). However, the usual explanations for AD failure are
poverty and law enforcement issues. Less discussed is where extra-legal governance of the illegal
trade has diminished. Analysing this may present further insights into the barriers and opportunities
that may help determine the success of AD.
The third conclusion is that the state must ensure there is long-term trade profitability in legal
alternatives, otherwise there is little that can be done to prevent farmers from returning to coca leaf
production either at home or elsewhere. As I illustrated in the first part of this thesis, the displacement
of illegal drug production elsewhere may occur despite state intervention. This suggests that the AD
success referred to here, does not mean that illegal drug farming did not also shift elsewhere as a
result. Wherever extra-legal groups can reorganize to provide efficient governance, the illegal coca
leaf trade will very likely be restored. Therefore, the state should take note of the importance of
making their AD efforts ongoing for long-term solutions.
259
Part III: Gender Norms & Alternative Development
3.1 Abstract
Why may women find coca leaf farming particularly attractive? In this chapter I review how the work
situation of women compared before and after their communities decided to stop farming illegal drug
crops. I present the case that equal access to property rights and education, have so far been the focal
points of policymakers’ concern about gender equality in contexts of Alternative Development (AD).
However, to understand the full picture of what limits women’s economic opportunities in AD
contexts, it is also essential to study the influence of gender norms. This is because gender norms can
shape other institutions in unquantifiable ways and therefore continue to make women more
economically disadvantaged than men when they stop working in illegal drug farming. Awareness of
this may be vital for AD policymakers who want women to support the substitution of illegal drug
crops in their community.
To assess this argument, I conduct desk research and analyse the experiences of 28 women living in
communities where coca leaf was substituted for legal alternatives in South West Colombia. I also
draw on 48 interviews with men and 5 interviews with heterosexual couples that I conducted during
my fieldwork. I find no signs of formal or informal gender discrimination in property rights to land
and access to education, however I find that gender norms nevertheless played a part in favouring
men’s employment in AD contexts. The illegal trade had presented women with economic
opportunities that were socially transformative as they helped them challenge local gender-role
attitudes. Paid employment was more locally available to women which helped many of them to work
as more equal business partners with men, to branch out into better paid and non-domestic work, and
to be less financially dependent on men. Conversely, the ceasing of coca leaf production in the
borough, rolled back their opportunities for paid work. In addition, women encountered a relapse of
residual gender norms that specialized them in unpaid work again and favoured men in paid
agriculture.
260
My findings highlight the importance that AD programmes address the factors that affect women’s
economic opportunities. Failing to do so can not only disadvantage women once their community
stops farming coca leaf but can give them added incentives to resist crop substitution programmes, or
to return to illegal drug production, later, or elsewhere.
3.2 Introduction
Many gender policies are concerned with the rules that create bias. Indeed, in AD policymaking
circles, efforts to prevent gender discrimination have focused on the rules that may obstruct women’s
access to land and to education. Important as these factors are, this approach fails to address the more
implicit ways in which women’s economic behaviour may be constrained after their communities
substitute illegal crop farming. Individuals’ payoffs may still be shaped by residual norms that
adversely affect women, even if there is no gender bias in formal institutions.
In this study, I will present specific opportunities that illegal crop production offered to women and
the issues that women faced when they stopped growing illegal drug crops in their community.
Researching the informal gender institutions in communities that began and stopped farming coca leaf
is important on two accounts. It may illustrate an important factor that may be keeping the illegal
economy up and running. It may also further our insight into why it is so difficult for AD to succeed,
and what issues need to be addressed for women to find it less costly to stop working in coca leaf
production. A more comprehensive understanding of women’s experiences before and after their
community stopped growing coca leaf, is key to understanding what their incentives are to work in
illegal drug production. This may help explain why some communities persevere with coca leaf
farming despite the violence and risks associated with it.
I do not propose to have an all-encompassing explanation for gender inequality in AD contexts, but I
seek to raise important questions about policymakers’ traditional approach to making AD gender
sensitive. I will begin with an overview of how norms differ from rules, and the importance of taking
them into account when considering the institutions that create gender bias. I will then provide some
context for how gender issues have so far been covered in the analysis of AD and make a case for
261
what has been missing in research until now. I will then lay out my research objectives and finally
present my findings and discussion.
3.3 Literature Review
About rules and norms
New institutionalism stresses that shared institutions, namely rules and norms, shape actors’
behaviour. Ostrom defines rules in the regulation sense, that is, something laid down by an authority
as required of certain persons. They are “shared understandings by participants about enforced
prescriptions concerning what actions (or outcomes) are required, prohibited or permitted…all rules
are the result of implicit or explicit efforts to achieve order and predictability among humans by
creating classes of persons (positions) who are then required, permitted or forbidden to take classes of
actions in relation to…outcomes or face the likelihood of being monitored and sanctioned in a
predictable fashion” (Ostrom 2005:18).
Norms, on the other hand, she defines as the shared concepts of what must, must not, or may be
appropriate actions of outcome types of situations. They are patterns of behaviour shared by most
people. Norm-following individuals consider other people’s interests as well as their own, but they
vary in the extent to which they do so. In other words, norms represent the intrinsic benefits and costs
of doing something, and as such have more control over behaviour than rules do.
Ostrom’s grammar of institutions, or the ADICO syntax, is a logical tool which helps us differentiate
rules from norms. According to it, rules are composed of the following five attributes. First, they
distinguish to whom the institutional statement applies and assumes participants have been assigned to
positions. Second, they permit, oblige, or forbid and usually use words such as may, must and must
not, to represent the perceived costs and rewards of obeying or breaking a prescription. Third, they
define the amount of action or outcome or the process of the action, which must be physically
possible and its negation physically possible. Fourth, they define when and where to act, and if
262
nothing is specified then they assume that it is always and in all places. And finally, they assign a
consequence for rule disobedience albeit a physical sanction or implicit shame or guilt.
This final attribute is the defining factor that demarcates rules from norms. While norms have first
four attributes in common with rules, the former do not assign a physical sanction if disobeyed. In
general terms, rules can be boiled down to obligations and prohibitions whilst norms are not enacted
as written law and are more informal. Instead, they are self-enforcing through social sanctions such as
ostracism.
While much interest has been given to the rules that may affect women adversely in the context of
AD, there has been little research into the norms that do so. This is an important gap in our knowledge
because men and women are not undifferentiated individuals (de Clercy 2010). Gender norms can
define people’s behaviour differently depending on who they are. Van Staveren & Odebode’s
understanding of gender is as “a complex, multi-dimensional institution shared unequally for men and
women, with men generally benefiting more than women in terms of access to and control over
resources” (Van Staveren and Odebode 2007, 908). Institutions are gendered when they express this
power dynamic (Mackay, Kenny, and Chappell 2010, 580).
Van Staveren & Odebode (2007; 2014) identify two types of institutions: those that have a symmetric
effect for everyone, such as traffic rules and universal primary education, and those that have an
asymmetric effect on different groups, like gender norms (Van Staveren and Odebode 2007, 41: 903).
Identifying the influence of gender norms on other institutions is important because they can shape
other rules as well as other norms, that are seemingly gender neutral.
In India, legal reform to favour gender equality in property rights was still circumvented by norms
that continued favouring sons’ inheritance of the household land (Roy 2011). Similarly, in Nigeria
and Ethiopia, women are legally allowed to buy property, but the law also requires them to have male
approval to do so. Given the laws that allow for early marriage, the big age differences within the
marriage means that men effectively continue to control women’s access to resources (Odebode and
van Staveren 2014, 8). These are some example of gender norms that have shaped rules.
263
In the context of the Yoruba women in Nigeria, norms that express values of independence and equal
opportunity, such as men and women earning their own living, do not result in gender equity because
gender norms in the household are at play. Norms that prevent women from inheriting land and from
working the same hours as men in paid work as well as encourage in women a greater sense of
responsibility for spending their money on their children altogether create an asymmetric effect on
women, despite the supposed equal opportunity to work (Odebode and van Staveren 2014, 8). These
are examples of gender norms that also shape other norms.
Although it would be expected that economic norms might offset the influence of shared beliefs about
gender roles, this does not happen. The reason is that norms do not compete but shape one another.
Symmetric institutions that are applied equally to everyone, may be imbued by asymmetric, or
discriminatory, institutions that have a different effect on different social groups (Van Staveren and
Odebode 2007; Odebode and van Staveren 2014; Martin 2004, 1256–58). In other words, norms do
not operate in a zero-sum game, whereby one norm must yield in influence for another norm to
dominate. Strengthening a norm that expresses values of independence and shared contribution to
household expenditures does not necessarily weaken another norm operating in the household which
expresses male advantage (Van Staveren and Odebode 2007, 41:920). Instead, norms are entangled
and interrelated, so a new norm, such as women’s independence, does not automatically cancel out
older norms, but enters a medley of norms.
These are some ways in which gender norms can determine the distribution of wealth and the
household division of labour and thus influence the economic process. Likewise, the economic
process can influence gender norms, in a two-way relationship (Veblen 1931; Odebode and van
Staveren 2014). And yet, like the concept of power, gender norms are often unmentioned in the
literature of institutional analysis (Thelen 2004; Braunstein 2008; Mackay, Kenny, and Chappell
2010; Knight 1992; Moe 2006).
Policymakers too have shown a bias in their approach to gender inequality. Specifically, those
concerned about making AD more gender sensitive, have limited their appeals to the rules that
264
allocate property rights to land and that promote universal education. In the next section, I will
describe their rationale for this approach, before presenting my research objectives.
Gender and property rights to land
The German development aid agency (GIZ) and the Global Drug Policy Programme of the Open
Society Foundation published a joint report in 2019 that summarized the experiences shared by a
handful of female drug farmers from Bolivia, Colombia, Mexico and experts from Afghanistan and
Thailand who were invited to a workshop in Mexico. One of the conclusions was that all these women
faced wider structural barriers and discriminatory social norms that are present in most rural
communities. They pointed to “a clear relationship between the limited access to land or alternative
economic activities, and the involvement in the illicit drug crop cultivation… [and so] without any
[land] security, women are further excluded from potential economic development opportunities”
(GIZ 2019, 3). The suggestion is that strengthening women’s legal land tenure, either with joint titles,
if they have partners, or individual ones, if they head their household, will enable them to access land
in a way they have not been able to, and thereby become more productive. In other words, these
organizations hoped to correct the drawback of gender norms that prevented women from owning
land, with rules that ensured universal access to land.
Gender inequality in property rights to land has previously been identified as a barrier for women to
generate an income. Many studies support the argument that land ownership and land security
increases people’s investment in land (Eggertsson 1990a, 35; Galiani and Schargrodsky 2010). From
this it follows that women with weaker property rights to land will necessarily be less productive than
men because their land will be less valuable. Again, the Yoruba women in Nigeria are a case in point
as they have no individual property rights and cannot inherit land. This means that despite having the
same norms of economic independence as men and of contributing equally to household income, they
are unable to accumulate as much wealth as men and so have less access to credit and fewer
opportunities to save and invest in their assets. And gender norms further give women an additional
265
sense of financial responsibility for children. These differences thus make them unable to move to
better paid work and make them less productive overall (Van Staveren and Odebode 2007, 41:919).
Inequality in property rights stem not only from rules but also from norms. Rules focus on addressing
unequal distribution in who has actual land titles, but norms can continue to establish qualitative
differences in the type of land they own. For example, in Peru women are increasingly obtaining land
titles but their land plots tend to be smaller than men’s and their land is less likely to have access to
irrigation (Fuentes and Wiig 2009). So, even with legal access to land, gender norms still influence
the respect given to women’s formal and informal ownership and the quality of land available to
women. And what land women end up getting gives men a higher income potential than women.
Similar findings have emerged from studies in Africa, as gender norms skew women’s ability to
maximise the productivity of their land assets in ways that run contrary to the rules. For example, in
Akwapim, Ghana, women have weak property rights so are at greater risk of being expropriated from
their land. This makes them leave their land fallow for shorter periods than is technically optimal thus
reducing their soil fertility and agricultural output (Goldstein, Bank, and Udry 2008, 116:983).
Likewise, in Burkina Faso, because men can afford to buy fertilizer, they use it on their own land plot
but not on their wife’s, and because women fear losing their land, they do not consider renting it to
their husband, which makes them less productive (Duflo 2012). The point is that gender norms
influence how property rights are used and enforced, which ultimately results in men’s land being
more productive.
These examples illustrate how rules that seek to redress gender inequality in property rights can clash
with existing gender norms, defeating the objectives of the law. As we have seen, besides being
unable to address the power imbalance, these laws can perpetuate economic inefficiencies.
Nevertheless, policymakers continue to place gender inequality in legal property rights at the centre of
their efforts to address gender issues in former drug farming communities.
266
Gender and education
Similarly, low education attainment has been reported by policymakers as a driver for women to
become involved in the illegal drug trade (UNODC & Colombia Ministry of Justice 2019; UNODC
2018b; OAS & CIM 2014). Although these reports accept that higher education levels do not
necessarily make it impossible for women to participate in the illegal drug economy, they believe that
women who have recently stopped working in the illegal industry are more economically and socially
vulnerable than men because of lower education levels (UNODC & Colombia Ministry of Justice
2019).
Gender inequality in education is an obvious barrier for women’s earning opportunities. As
Braunstein (2008) elegantly summarizes, various econometric studies, mainly from the 1990s, have
shown a strong positive correlation between women’s education and economic growth. In Venezuela
an additional year of schooling resulted in 9.1% increased earnings for men and 11.1% increase
earnings for women (Psacharopoulos, G. and Tzannatos 1991). Hill and King argue that women’s
education results not only in better employment opportunities and higher earnings for them, but also
in healthier children.
Better educated women are not only better informed about hygiene and nutrition but also have fewer
children which means there are more resources available for each child they raise (Hill and King
1995, 24–25). A comparison of 127 countries from the 1970s to the 1990s found that gender
inequality in secondary schooling has a negative effect on economic growth (Dollar and Gatti 1999).
The World Bank has linked the gender pay gap to barriers in women’s access to and participation in
education (Blackden 1999). And cross-country evidence has shown that gender inequality in
education lowers the average level of human capital as it appears to reduce investment and increase
population growth, and this has a long-term effect on economic growth (Klasen 1999). Increasing
girls’ schooling leads to higher levels of labour productivity across countries, and gender inequality in
education is an obstacle for economic development (Knowles, Lorgelly, and Owen 2002). Given the
above research, it has not been controversial for AD policymakers to promote universal rights to
267
education to increase gender equality in former drug farming communities. It is clearly seen as one of
the most obvious ways to nourish local economic development. However, this approach fails to
recognize some less tangible ways in which women may still be adversely affected when they can no
longer benefit from the economic opportunities that the illegal drug trade had once offered them.
Gender norms and economic behaviour
The above literature review suggests that gender inequality in property rights, including in access to
land titling, the respect given to ownership and inheritance laws, and differences in the quality of land
men and women own, are ways in which women may experience discrimination. Similarly, it is
feasible that women face greater barriers when trying to access education. Therefore, we should not
discard the possibility that inequality may exist in the rules that determine women’s land ownership
and their access to education, in the former coca leaf farming communities included in this study.
These are rules that may be obstructing equal economic opportunities for women, and if so,
policymakers should continue to review them in their quest for reducing gender bias.
However, as we have seen, there are other possible ways in which gender norms can skew economic
playing fields that seemingly uphold gender equality. Therefore, it is essential to observe with equal
attention the less tangible ways in which gender norms may shape women’s economic behaviour in
communities that have stopped farming illegal drug crops. This will allow us to capture a more
accurate picture of why women may, or may not, have a vested interest in illegal drug farming and
why some women persevere with it despite the violence and risks associated with it. Greater
awareness of this, will make it easier for AD policymakers to address the subtle obstacles that women
face when trying to make a living in former drug producing communities in Colombia. So, in the
interest of better-informed AD policies, my analysis shall observe the following three themes in the
fieldwork interviews and desk research I conducted:
• Theme A: property rights to land (both formal and informal) – whether there were gender
differences in how they were respected
268
• Theme B: education - whether there were measurable gender differences in school attainment
and attitudes to education
• Theme C: economic opportunities – whether there were differences in women’s economic
behaviour before and after their community substituted coca leaf and the role that gender
norms played in shaping these
3.4 Findings & Discussion
Theme A: Gender differences in respect for property rights to land
Carruthers & Ariovich’s five-dimensional framework is a useful tool for the analysis of property
rights in relation to gender in these farming communities before and after substituting coca leaf crops.
The first dimension is the object of property rights. In other words, what can be owned, which in this
case was rural land plots, in various sizes. The second dimension is the subject of property rights, that
is, who can own it. This varies across societies. Some grant ownership rights to selected individuals,
while others recognize collective groups or corporations as owners (Carruthers and Ariovich 2004,
27). Here, the subject of property rights I observed was men and women as individuals, and as joint
owners, such as siblings or partners. The third dimension of property rights is the articulation of use
of property, that is, what can be done with it. Contracts can shape how owners use their property (J. L.
Campbell and Lindberg 1990). Other times, restraints are imposed through informal channels such as
social and cultural influence (Carruthers and Ariovich 2004, 28; Grassby 1995). I studied this aspect
of property rights in these communities, by observing whether women were able to use their land as
they pleased and whether there was general agreement among men and women, who owned land
jointly, about how to use their land. A fourth dimension of property rights is the transfer of rights, in
other words how property can move between owners. This may be through market exchanges,
through nominal ownership by the state, through conquest or colonization. Alternatively, rights can be
transferred as a gift such as an inheritance, noting that inheritance rules are usually shaped by social
conventions that favour specific types of heirs differently across different traditions (Carruthers and
Ariovich 2004, 30–31). In this case, I analysed the transfer of rights with respect to gender restrictions
269
both formally and informally on who was able to acquire land, how and when. And the final
dimension of property rights is the enforcement of rights and restrictions. In other words, how
property rules are maintained, either through formal or informal conduits. In this case I looked at
whether there were gender differences in who had land titles and who did not. My overall analysis in
this chapter is mainly focused on the second, third, fourth and fifth dimensions described above.
Women in rural Colombia have been legally entitled to land ownership in the same way as men,
gradually since the 1990s, and definitively since 2002. The campaign for gender equality in property
rights to land in Colombia began over thirty years ago. At that time, although women had been able to
own land, women from rural farming communities had been excluded from owning land in other
ways. For example, they could not own land jointly with their male partners and when they had no
partner, they were not given priority in land distribution programmes (Deere and Leon 2004, 206;
Donny Meertens 2000). In the 1980’s international development agencies, such as the Food and
Agriculture Organization of the United Nations, the Economic Commission for Latin America and the
Caribbean and the United Nations Development Fund for Women, promoted a gender-focus in
Colombia’s rural development planning. Following this, women’s associations were formed, such as
the Asociación de Mujeres Campesinas, Negras e Indígenas de Colombia (ANMUCIC) which
advanced women’s interests and increased their participation in the land reform negotiations.
Eventually this led to the creation of Law 30 in 1988 which allowed for joint land titles for couples
over the age of 16, which then became mandatory in 1995. In addition, Law 160 in 1994 gave female
heads of households who found themselves in a socially and financially “unprotected state” as victims
of violence, priority as land title recipients (Díaz Suasa 2002; Deere and Leon 2004, 181–209).
Between 1961 and 1991, only 11% of the beneficiaries of Colombia’s agrarian reforms had been
women, whereas between 1995 to 1998 after these legal reforms were made, 45% had been women
(Deere and Leon 2004, 209; 2001). In 2002 it was further codified in Law 731 that rural women who
had been abandoned by their permanent partners were entitled to their joint land title (Díaz Suasa
2002, 45).
270
Historically, these legal reforms were met with some social resistance in the 1980s and 1990s from
women’s partners and their own rural communities, who regarded the legal recognition of women’s
property rights to land to be mutually exclusive with the interests of the family household and the
“campesino” identity per se (Sanudo Pazos 2015, 16–17). Nevertheless, in the farming communities
included in this study, there were many signs of social acceptance of women’s property rights to land,
suggesting the above legal reforms had long been assimilated by the community. This woman’s
account of how her mother’s informal land acquisition was accepted by the community illustrates this
point:
Yes, people could own land, because people would take it [referring to baldios
(state vacant land)], but that is not to say that we had it writing, or that we had
paid to register it in the cadastre… no, nothing like that…
INTERVIEWER: But it was as if it were yours.
INTERVIEWED: Yes.
INTERVIEWER: Nobody could take it from you?
INTERVIEWED: No. they would even help you take care of it, to fence the
boundaries, and they would say this property belongs to Fredesminda, that was
my mother's name (Ligia, 33)
There were no indications in any of the interviews of there being a conflict within heterosexual
couples in how they used their land both formally and informally. When women had acquired land
prior to marriage, and their partners wanted to grow something different, they did so on separate land
they owned or rented, and women’s ownership was not transferred to their husbands.
When couples owned or rented land together, the choice of land use was generally dictated by what
was most profitable, or what they had experience in farming, and this criteria for land use were shared
equally by men and women. When women and men shared their land, couples generally decided how
to use their land and when to start and stop growing different crops together, including coca leaf.
271
Women usually had influence over where to buy their land with their joint savings. This woman’s
account was echoed by many others:
My husband didn’t want to buy land here, so I said to him, “no, we need to leave
[Putumayo] as it’s getting violent, so let’s buy land here [in Nariño]”, so we
bought our land here…. I told him that I would come here in advance, and he
should try to sell our land there, and he followed me six months later (Lilian, 34)
I only came across occasional disagreements in couples about when they should to stop growing coca
leaf. From my observation, they were usually because women wanted to avoid the violent
environment that coca production was attracting in the community. In these cases, the women had
remained free to migrate elsewhere and retain their property assets equally, as this woman described:
In the 1980s my husband suggested we get married and come here [to Putumayo]
to work. So, we did. We arrived at a coca farm, where I earned COL$7,000 pesos
[$2 USD] for cooking for 25 workers and he earned COL$20,000 pesos [$5.7
USD] for keeping the workers' accounts … after eight months with the money
we’d saved we bought two hectares of land and a coca crop of 80 arrobas of the
so-called caucana [a variety of coca leaf] … and we became independent and
began to earn a living… after that my first daughter was born in 86 so we built a
little ranch on the land that the same employer had lent us to live on, because the
little land we had bought ourselves did not have a ranch…so there we lived until
we had another daughter in 98 and by then we had another little farm … then I
decided to separate from my husband and came to live on this land with my girls
(Magula, 51)
When siblings owned land jointly or shared the land use rights of their parents’ property, it was usual
practice to devise boundaries for everyone to choose what crops to farm in their own sub-plot. There
was one example of a woman transferring her brother’s land rights to herself because he was mentally
impaired, and she was looking after him.
272
With regards to transfer of property rights, social norms respected women’s rights to inherit land and
to leave land as inheritance, even when they lacked land titles. As this man’s account suggests,
individual property rights were respected within couples too:
I’m trying to get a land title…my wife already has one for some land she inherited
(Huber, 63)
Adult children respected their parents’ ownership and choice of land use until their death, including
when their elderly mothers were old and fragile. But the land would not always be distributed equally
by siblings on the assumption that women did not need as much land, as this man described:
When my mother died, we split the land between the five of us… well the two
sisters took the house and us three brothers split land equally (Sabalon, 60)
This may be linked to the gender norms that “protect” women from working in agriculture which will
be discussed later in this chapter. Having said this, the man believes each siblings’ inheritance was of
equal value. And siblings often respected each other’s inheritance even when they did not have a legal
land title. When there was dissatisfaction and disagreement it was common for women to assert their
rights, as this woman’s account reflects:
…yes, I did [inherit land from my parents] but there were problems with my six
brothers because my father had left them half his land and the rest to my mother.
My father had asked my mother to leave the rest to us six remaining daughters, so
she did. My brothers were angry and said that my father had given it to them in as
a gift whilst he lived…so they are still mad at us. They still don't talk to me, my
brothers (Marta Elena, 42)
Siblings would often help each other out regardless of their gender. Occasionally, brothers who had
‘colonized’ vacant land or bought land informally, would give their sisters a piece of land to grow
their own crops. However, it was more common that they would offer them employment in cooking
and cleaning on their farm.
273
There was only one example of a woman who had had a serious disagreement with her husband
during the coca leaf bonanza about who could keep the land they owned after separating. The woman,
who had been expecting a second baby, wanted to split the land equally with her husband but he
threatened to kill her to keep the entire land plot. In this case, an armed group intervened and
protected the woman’s claim to land ownership because she had also invested in it and had a child.
They warned the man that if she was killed, he would be too. Other than this example, separated
couples, who were not the most common scenario, tended to divide land equally.
Overall, from the interviews I conducted, there was no identifiable difference in the subject,
articulation of use and transfer dimensions of property rights to land among men and women.
Women’s property rights to land, both formal and informal, were equally respected as those of men.
Theme B: Gender differences in school attainment and attitudes to education
It was not possible to gather systematic data on how long the women and men in these communities
had stayed in school. In any case, from this sample this data would not have been in any way
statistically representative. For the interviewees that gave me information about their education, I
worked out that there were more men than women who had had no education at all, and there were
more women than men that had finished their secondary school. In terms of school attendance, it had
been about equal for men and women. See Table 3 for a general idea of how long the people I
interviewed recalled being in school:
Table 3 School attainment among people interviewed
School attainment Women Men
None 6% 11%
Up to second grade 53% 48%
Up to fifth grade 12% 26%
Finished secondary school 29% 9%
Further educated 0 6%
274
This means that the conclusions I make in terms of school attainment are devised exclusively from
interviewees’ experiences and their versions of life in their community, before and after substituting
coca leaf. I compared this information with desk research about education in general in rural
communities in Colombia.
Education has been obligatory for all children aged 5 to 15 years in Colombia since the 1994 General
Law on Education (Aguilar- Barreto et al. 2017, 87). Nevertheless, school attainment in rural areas in
Colombia has historically been very low. Only 48% of children in rural areas remain in school
compared to 82% of children in urban areas, and in rural areas 12.5% of people over 15 are illiterate
compared to 3.3% of the national average. Moreover, academic performance is systematically lower
in rural areas than in urban areas in Colombia (Martínez-Restrepo, Pertuz, and Ramírez 2016;
Delgado Barrera 2014). Reasons for these rural-urban discrepancies include children working during
harvest times, and the long distances that they must travel to access schools (Tieken 2014).
On the other hand, there are no significant gender discrepancies in who goes to school. The school
attendance rate for girls aged between 5 and 16 living in rural areas in Colombia in 2015 was 91.3%
(DANE 2015). Hill and King (1995) found that countries in Latin America and the Caribbean were
exceptional, compared to other developing countries in that school enrolment of girls was not
significantly behind that of boys (Hill and King 1995). Based on Colombia’s 2013 National Survey on
Quality of Life, school attainment in rural areas was found to be similar for women (36%) and men
(32%) between the age of 18 and 24, whereas in urban areas 72% of women in the same age group
had finished school compared to 63% of men (Martínez-Restrepo, Pertuz, and Ramírez 2016). Some
explain school desertion among teenagers by the availability of jobs in farming for boys and teenage
pregnancy for girls (Martínez-Restrepo, Pertuz, and Ramírez 2016). Either way, legislation and
national statistics do not reflect gender inequality in women’s access to education in Colombia.
The interviews in the rural communities included in this study reflected very similar trends to the
above. There was no sense that the men had higher levels of school attainment than the women
interviewed. Most people interviewed had either not had a local school to attend or had had to stop
attending at an early age to work and help their parents out, cooking, cleaning, picking up the manure,
275
milking cows, harvesting crops, moving the cattle, or scaring pests off. Often people had started to
work at home as early as seven or eight years of age. These statements were echoed by many of the
people I interviewed, both men and women:
I studied up to third grade in primary school and when I was twelve, I began
working in agriculture with my parents (Horacio, 65)
When my father was killed, I was 11 so I stopped going to school and became the
breadwinner with my mother…I never learnt to write (Luz Mila, 55)
Both men and women, felt strongly disappointed about not having had a good education, which they
saw as the cause of their poverty, as this man describes:
When I started to work, I farmed what my grandmother told me to. “Millet has to
be sown because that's what we are going to eat!” she would say. Then I farmed
coffee, banana, sugar cane, and corn or whatever because that is all I could do.
When I left school, I screamed because I couldn't study anymore… at that time we
didn’t have the facilities there are now, it was more difficult to go to school (Jose
Mauricio, 42)
And this sentiment seemed to influence their concern that their children be able to finish their studies.
Many would share with pride when they had been able to finish their studies. My interviews suggest
these communities mirrored the broad patterns in rural Colombia in terms of the levels of school
attainment and did not present a tangible gender bias affecting girls’ access to education more than
boys.
Therefore, my main research interest in this study was not to compare men and women’s literacy
levels but whether there were attitudinal differences in the communities about girls’ and boys’ access
to education which may be creating gender bias in any way. After all, it is not uncommon in some
countries for parents to have lower aspirations for their daughters than for their sons, and this
translating into teenage girls having lower aspirations than teenage boys (Duflo 2012, 1056).
276
However, gender norms did not seem to permeate peoples’ aspirations for their own or their
children’s education. Mothers expressed concern about ensuring their children had the resources to
continue their education, more frequently than fathers. This was reflected in what they talked about
and how they intended to use their loan if they had taken one out or were going to apply for one.
However, couples generally paid for their children’s education together, as this woman described:
I had two farms, and I had to sell them, one... because we were forced by the
paramilitaries to sell it. And the other farm we sold ... to help pay for the
children’s’ education… we haven’t finished paying for it… it’s been hard…. right
now, I work as a school guard and my husband works as a bus driver, but this is
still not enough to support them. (Lucila, 51)
The profits from coca leaf farming had made this possible for most, as this woman described:
So, we started growing coca leaf… and thank God that gave us enough money to
send both our daughters to school… without coca it would have been hard (Rubi,
46)
There were no gender preferences in who of their children should go to school, although there were
two instances of parents prioritising their daughters’ education over their sons’ because they showed
more interest in further studies. Overall, there was no identifiable difference in the school attainment
and attitude to education between men and women.
The findings discussed so far do not support policymakers’ claims that gender inequity in AD
contexts stems from gender disparities in access to land and education (UNODC & Colombia
rights were codified in the law and the interviews revealed that women’s property rights to land were
respected by informal norms. Background research on education levels in rural Colombia also
discarded the premise that there were formal barriers to girls’ education compared to that of boys. The
interviews reflected these general patterns of low school attainment but gender parity in school
277
attendance in rural Colombia. They further suggested that people’s attitudes towards education in
these rural communities did not informally favour the exclusion of girls from education.
Theme C: Gender differences in economic opportunities before and after coca leaf
farming
Many women had irregular incomes from individual jobs, so systematic data on women’s earnings
from before and after the contraction of the coca leaf economy was unobtainable. Without accurate
estimates of their monthly earnings, it was not possible to ascertain changes in their incomes.
Therefore, my analysis focused on how women perceived their economic opportunities had changed
with the arrival and passing of the coca leaf bonanza.
Most people I interviewed experienced an income shortage after coca leaf ceased to be grown in their
community. In addition to this, many women felt increased financial dependence on male
breadwinners. This should not come as a surprise given the gender pay gap is a phenomenon
experienced in all societies (Jayachandran 2015). However, more than inequality in rates of pay, the
general perception was that the opportunities for any paid employment had drastically decreased for
women when the communities stopped farming illegal crops.
Next, I will present my findings relating to this theme and discuss the role that gender norms played in
changing women’s economic opportunities over time, and vice versa. I will argue that the deep-seated
gender norms that specialized women in unpaid work were briefly challenged by the local coca leaf
economy as it increased women’s paid economic opportunities. This, in turn, helped to change some
people’s behaviour and gender-role attitudes. However, when the communities stopped producing
coca leaf, paid economic opportunities were greatly reduced for women. I will show how this
perception was associated with the reinforcement of gender norms.
Gendered specialization of work
To understand why women in these former coca leaf farming communities experienced increased
financial dependence on men when their community stopped producing coca leaf, it is crucial to
observe how gender norms interplayed with women’s economic opportunities. One such norm, is the
278
segregation of work by gender, whereby women carry the burden of unpaid housework and specialize
in domestic services.
The gender norm that assigns women a greater share of the unpaid housework has long influenced
men and women’s behaviour differently in many parts of the world, as well as Colombian rural
communities. In 2013 women carried out 79.1% of all unpaid domestic work in Colombia (DANE
2013). National Administrative Department of Statistics in Colombia (DANE) used replacement cost
method to work out the value of women’s unpaid housework, and what it contributed to the national
GDP. By estimating how many people and how many hours they each spend on unpaid childcare, care
work, household management, cooking, cleaning and administrative work, and then estimating what
people in the paid workforce are remunerated per hour for similar tasks, they worked out that while
women’s unpaid housework represented 16.3% of the country’s GDP in 2012, men’s only accounted
for 4.1% (DANE 2012). DANE estimates that the situation is even more unequal in rural areas.
Today, women in rural areas spend more time on domestic and unpaid care work than people in urban
areas, approximately 36 hours a week (DANE OXFAM 2020).
The former coca leaf farming communities I visited were not exceptional to other rural communities
in Colombia, and across the globe, in that women were expected to do most of the housework. The
women I interviewed revealed they had always assumed a high burden of unpaid work in the
household, which men did not share much. This was principally cooking, cleaning, food sourcing,
water collection, laundry, and childcare. As this woman describes, women were shaped for these roles
from a young age:
As a girl I preferred to do agricultural work, my family always said that I was a
tomboy because I liked to work the land and I liked to fish and so they would
punish me because I was seldom in the kitchen [laughs] (Sandra Lorena, 24)
Women usually assumed these roles in full when they married, as this woman described:
My husband was older than me and was jealous, so I could not work like he did to
help raise our nine children, but then thank God I became a young widow! Then I
279
was able to work [in farming] and I raised my children… they are all workers
now and they all own their homes (Maria Isabel, 75)
It was not unheard of for housework to be shared out between men and women, but as this man
described, it was the exception to the norm:
We come from a very poor family and there were no women, just seven boys and I
had to help my mother with both women’s and men’s work, because I had to look
after my younger brothers and help to cook and clean at home, now I always pull
my weight at home, I mean if my wife is doing something, I will keep busy and do
something else, if she is cooking then I am sweeping or mopping the floor or
washing clothes! I mean we help each other out with whatever needs doing
(Misael, 55 – speaking next to his wife)
In addition, most women in these communities had always found it easier to work in paid domestic
services than in paid agriculture. Before the coca leaf bonanza, women who were able to take up paid
employment were limited to domestic services in town, as this woman described:
I left home when I was 15 to work as a domestic servant for a family in Cali… it
was the only thing, really the only thing you can find work in as a woman if you
haven’t got an education… I worked there for about 9 or 10 years before I could
buy a land plot here and start farming (Derly, 40)
Although to a large extent most women in rural Colombia acquired farming skills as men did, when
they were young children, gender norms excluded grown women from paid jobs in agriculture. Before
the 1990s, it was common practice for children to help their parents with farming jobs. Rural women
were therefore not unskilled in agriculture. In 2016, it was estimated that 36.6% of rural women with
paid work in Colombia worked in agriculture, while 29.5% worked in services (DANE 2016).
Nevertheless, it was common for women in these communities to do subsistence farming but feel
discouraged from taking up paid farming jobs. The reasons they gave for this was because agriculture
was hard-physical work and women should avoid exposure to chemicals in the fertilizers and
280
pesticides during pregnancy. Therefore, before and after the coca leaf bonanza, agricultural work was
mainly considered men’s work, and women were protected from it as far as possible, as this man
described:
For a woman, agriculture would not be a good career... [I hope my daughters]
will study for a technical career and stay out of working in agriculture… working
in the country means living with your back to the sun, and it is hard. I’ve been
doing this since I was nine years old, putting my back to the sun and one already
feels tired, like the body no longer can handle it… so wishing that on a daughter
would not be good. It would be better if they studied and made a living without
having to turn their backs on the sun, as they say (Saul, 43)
This man only had daughters, perhaps he would not have wished agriculture as a career on a son
either. Nevertheless, there was a general belief that men were tougher and more able to withstand the
physical nature of farming work. When women did undertake paid farming work, it was in addition to
their unpaid domestic work. Like women often do, they put in more working hours than men, often
double shifting between paid work and housework (Hakim 2001).
Illegal trade and more economic opportunities for women
The gender norms of work specialization prevailed during the coca leaf bonanza. However, as the
economy began to grow, women in these communities encountered more paid work that was locally
available to them. Something that came up with some women, was that thanks to this increase in paid
employment, they were able to break with their gender roles. There was greater teamwork among
couples, which sometimes increased women’s household bargaining power. Some women were able
to diversify into jobs that broke with gender role stereotypes. And others felt their increased financial
independence gave them freedom to live a life they valued.
Most of the women who were interviewed had come to work in the coca leaf economy after they
married or began to cohabit with their partners. They then migrated with their partners or relatives to
281
another community. They had either “colonized” land to eventually grow coca leaf on it or they had
moved to work in an already established coca leaf plantation.
The bulk of coca leaf farming, including picking, processing, and selling crops, was left to the men.
Farming coca leaf included heavy-duty chores such as digging the earth, pruning crops with heavy
electric scythe mowers, and spraying them with pesticide and fertilizer. These jobs required people to
carry weighty pumps on their backs and expose their face to the chemicals. Most women described
these conditions to explain why they had continued to specialize in cooking and cleaning for men
daily. As this woman confirmed:
INTERVIEWED: at first when there was plenty of coca, there were many people
who came to pick and process coca leaf. My family employed them and my mum
and us [other sister] had to make food for them.
INTERVIEWER: And was it mainly men who worked on the plantation?
INTERVIEWED: Men.
INTERVIEWER: Were there women?
INTERVIEWED: No.
INTERVIEWER: Why?
INTERVIEWED: Because it was most of all men who did that job.
INTERVIEWER: Why?
INTERVIEWED: I don't know, on our farm there were just men, my brothers hired
men… the only women were us three [sisters] and my mother. Between us we took
turns to make lunch and breakfast or dinner for the men (Carmen, 43)
282
Often women would do this kind of work for up to 25 or 30 raspachines who worked on the coca
plantation.23 The gender demarcation of employment continued during the coca leaf bonanza. The
difference was that now this work was mostly paid for and there were more jobs for women.
Women were still not paid as much as men. In 2018 in Cauca, a woman could earn COL$16,666
[$4.35 USD] per day for cooking and cleaning in a coca leaf plantation24, while a raspachin could earn
around COL$65,000 [$19 USD] a day for picking coca leaf. Coca leaf farm accountants and
managers, also traditionally men’s job, were paid approximately three times as much as cooking and
cleaning. Although cooking and cleaning jobs still had less value than jobs in coca leaf farming, there
were many more paid jobs locally available for women in cooking and cleaning. They were often
more convenient because they were more local and flexible, and thus easier to combine with childcare
than paid domestic service in town.
Overall, local economic growth brought about by coca leaf production in the community created more
economic opportunities for women. Women were still largely expected to carry out the domestic
work, but at least they were getting paid for doing it.
Adaptation of gender roles
For some, this new income helped to erode gender-roles in three ways. Firstly, it increased some
women’s household and business bargaining power as it offered more employment for women who
also had caring responsibilities. Secondly, it enabled some women to pursue jobs that broke with the
stereotypes of unpaid housework or lower paid domestic work and increased the returns of their work.
Thirdly, it made it easier for some women to live independently from men if that is what they wished.
Better paid employment opportunities for women strengthened many women’s bargaining power in
the household and in business. Women who could be paid to cook and clean on a farm where their
husbands worked, were an asset as men were offered farming jobs more easily if they came with their
wife. This was the case for farms with legal and illegal crops. However, coca leaf farming expanded
23 While the UNODC and Colombian Ministry of Justice 2019 report did reflect these women’s experience in that they had a subordinate role to men in the drug trade, it did not reflect the principal motivation of women working in illegal drug farms which was working for the family business. Instead, it stated very generally that women’s main motivation for participating in the illegal drug trade was their inability to meet their needs or because they had not option and were pressured into it (UNODC 2019: 124). 24 Paid as $500,000 Colombia pesos [about $144 USD] per month.
283
this type of opportunities and because there was more demand for this work on coca leaf farms, than
say for coffee farms, the job came with fewer restrictions for women who had their own domestic
responsibilities, as this woman described:
I used to cook for labourers while my husband harvested coffee, and because I
had a small daughter I stayed in the kitchen and had to cook to be able to look
after her... it was hard because the landowners were difficult, and I had to keep
my toddler locked in a room because in bigger towns it is different to the
countryside you know? Here were are all equals whereas in town we are not,
there are big social differences and it gets more complicated for people like me
with a daughter….and it was hard work, I had to wake up very early, at 3am, to
cook for 100 men and would work until 10pm and every day it was the same, so it
was very tiring… but then we found work in a coca leaf farm…my husband said
“they’re looking for a young woman to cook” so we went and they hired me as a
cook for fewer labourers, and they hired him too as a day labourer … they paid
me a bit more and I could look after my children at the same time (Rubi, 46)
Couples started to work in what they often referred to as joint business partnerships. Regardless of
whether they had both farmed coca leaf, or the woman had cooked and cleaned for other farmers,
increased opportunities for paid employment enabled women to contribute a larger share to their
household savings. This increased their husbands or male partners’ co-dependence on them, as this
woman described:
At 15 I married and went to work on my parents’ in law’s coca leaf farm. I
sometimes worked as a day labourer, or when my husband hired day workers, I
would cook for them…but I didn’t get used to that place so I came back home with
my five-year-old daughter and some savings, and my husband soon followed
because he said he could not work alone (Leudania, 35)
284
And this working partnership sometimes increased women’s bargaining power in the household, as is
illustrated in this woman’s story:
Before he met me, my husband played and played billiards all day, he would leave
all the farming work to his day labourers, but when we got together, I said no, my
life was not going to be like that because when I was a girl, we were very poor,
and we would sometimes go hungry… so my mum taught us that you have to work
hard in life… I began to work at the age of 8 in a family home and then as a
cleaner in a hotel. So, he [husband] said to me to leave that work behind, but he
would still wear new clothes every day. So, the deal was that I would leave my job
but that we would work together farming [coca leaf]. At first, he said no, that I
should leave the work to the day labourers, so I said that I wanted to help my
mother out and I wanted to work, and I didn’t want to live my life depending on
someone else, that just wasn’t what I wanted. So yes, we started farming ourselves
and he stopped partying and buying clothes…I was about 19 years old when we
got together and … we worked, and we saved our money… and after we bought
those 10 hectares of land we kept working [in coca leaf farming] and paid for
more land with our savings.
INTERVIEWER: Were those joint savings?
INTERVIEWED: It was like a family job, most of the money was for the farm
which was for both of us… and that’s how we came to live here (Monica, 30)
These partnerships also increased women’s bargaining power in the family business. Some women
were able to work in coca leaf picking and processing, as long as it was alongside their partners, as
this woman described:
I went with my female cousin to work [in Caquetá] in coca leaf picking but I first
had to work in a shop…my brother did not let me work [as a raspachin] he said it
was very hard work and it was men’s work and given there were other jobs in
285
town he said I should work in a shop, so I worked there for two years and he
would drop me off and pick me up…until I married… and then with my husband
we worked as raspachines (Mirta Marta, 42)
Gender-roles also began to shift because of women’s increased participation in the paid workforce in
that they were able to assume jobs outside the “domestic” remit. With more money in the community,
women were able to be more entrepreneurial. They could sell home-made snacks on their doorstep, or
do laundry service, or sell cheese and bread at the weekends. Better paid service jobs were also more
available to women including as shop, bar and restaurant keepers or school gatekeepers. Some women
became able to save up their own capital and diversify into even better paid jobs as well as other kinds
of work outside their conventional gender roles. Some bought land and moved onto more lucrative
farming work including livestock and cattle ranching or their own coca leaf production. These jobs
often only required a couple of days’ work every three weeks to tend the crops or to check on the
cattle but would still earn them the equivalent of a week’s work in something else. Sometimes they
could use the rest of their time to do other jobs such as coca leaf picking, cooking, or washing clothes.
Some women were able to open businesses and farms and become bosses and employers themselves.
In addition, parents had the means to help their children finish secondary school and pursue further
studies to become teachers, beauticians, nurses, police officers and pharmacists. Besides helping them
find better paid jobs, investment in their children’s education helped some women to break
conventions of what counted as men or women’s work, as these farmers’ experiences illustrate this:
By the time I had 80 arrobas of coca leaf… I left my parents in charge of all the
crops and moved to Popayan to get a health diploma and train as a nurse… my
parents helped me out with savings while I worked as a rural nurse linked to a
hospital in Balboa (Daniel, 46)
I continue to work as a raspachin to support my wife and 14-year-old daughter.
She is still at school but would like to continue her studies after graduation to
become a policewoman (Jesus Efren, 37)
286
I would like to be a pharmacist but right now I cannot afford my education...
perhaps when my five-year-old daughter is older I can then work as a raspachin
to pay for it (Milena, 23)
Finally, the increased economic opportunities brought about by the coca leaf bonanza helped more
women increase their financial independence from men. If they no longer wanted to cohabit, they
were freer to live independently from their partners. More women had the option to leave as they had
the means to earn the minimum level of income, they needed to cover their rent and their own
expenses. Single mothers were not wholly dependent on income support from their parents because
there were more jobs locally available to them. This would not have been an option if they lacked the
financial resources to leave.
Contraction of the coca leaf economy
After the local economy contracted, following the eradication of coca leaf crops in their community,
the employment situation changed for everyone in these communities. Women who had not been able
to break out of poverty, however, seemed to be more adversely affected by the income shortfall than
their male counterparts. They encountered additional challenges when trying to find paid work that
men did not.
There were no rules that forbade women from competing on par with men for jobs. However, the
barriers women faced to earn as much as men were implicit barriers created by the expected codes of
behaviour. Women felt they had less time for paid work, fewer locally available job vacancies
requiring their skills and that the agricultural job market made demands on them that they could not
meet. And women who were not dependants and who had dependants themselves, were the people
who seemed to be most negatively affected by unemployment. The combination of fewer local work
opportunities with no supplementary income to cover their children’s costs, made some women more
reliant on earnings from their relatives or their children if they could work.
Gender norms explain in great part why women perceived there were fewer economic opportunities
available to them. On the one hand, home production became more labour-intensive. Gender norms
that created the expectation that women would carry the burden of unpaid housework and subsistence
287
farming, meant that women were more affected by this than men. With more housework, women had
less time available for paid work. In addition, there was a shift in local employment opportunities
away from services and towards agriculture. Gender norms that specialized women in local and
domestic work helped to give men a comparative advantage as employees in these roles.
I. More labour-intensive home production and subsistence farming
After coca leaf was eradicated and substituted for legal alternatives in their communities, there were
lower household incomes. This meant there was less money available to buy food from shops and to
pay for time-saving appliances, and sometimes even to pay for electricity and running water (when it
was available). Wealthier households too had less spare money for people to eat out and pay for
services like laundry, packaged food, and utensils they had been used to during the coca leaf bonanza.
This meant everyone was increasingly reliant on subsistence farming to feed themselves.
The perceived effect was that without coca leaf money, home production became more labour-
intensive, and the share of this unpaid workload increased disproportionately for women. Cars and
motorbikes had been sold, which meant it was more expensive and took people longer to source
groceries such as soap, oil, and salt, from urban areas. Without money to buy better quality clothes
and shoes, like they had during the coca bonanza, women now had to wash and mend clothes more
regularly. From these changes, it follows that women had fewer hours available for paid employment
outside the household.
In addition, people were more reliant on subsistence farming than when there was coca leaf. As men
were more likely to migrate for work, women assumed a greater share of this unpaid subsistence
farming. It meant that they had to assign more time for growing more food from seed than they had
before. There was a general sentiment that subsistence farming was more time consuming and harder
work than before. Many complained that the soil had become less fertile after it had been used to
grow coca leaf. And day labourers had become costlier. This resulted in women having fewer hours
for off-farm work, as this woman described:
When we stopped growing coca leaf, our basic family necessities changed because
when we used to harvest coca leaf here, I could work two, three days for our
288
neighbour and be paid at the weekend, then I would go into town and buy
groceries. Now that there is no coca leaf, there is a drastic change in our daily
food… now we must work to produce our food because we don’t have enough
money to buy the things, we were used to consuming… (Seneida, 42)
These experiences echo the findings of a study on non-migrant women in rural China who assumed
more unpaid farming work when their male partners migrated for work to urban areas (Mu and van de
Walle 2011). It also relates to findings from studies on AD communities in Peru where, after
substituting their illegal drug crops, women typically took charge of the first plots of alternative
agricultural production while many of the men sought work elsewhere to maintain the family income
(USAID 2012, 16).
A lower gender wage gap has been found to decrease women’s unpaid workload (MacPhail and Dong
2007). My findings suggest the same phenomenon but in reverse: women’s reduced economic
opportunities appeared to increase their unpaid workload.
II. Shift in local paid labour from services to agriculture
After these communities stopped producing coca leaf, people felt the local paid employment
opportunities shifted away from services and towards agriculture. No longer were there cooking and
cleaning jobs available on coca leaf farms. As local butchers, shops and restaurants went broke after
the local economy contracted, there were fewer local paid jobs in services and domestic work. And
jobs that were available in services, were paid less than they had been during the coca bonanza, as this
woman described:
It is difficult to be here, because here there is no work anymore. I worked in a
restaurant, but after coca leaf farming, I lost my job so now I am trying to make
ends meet by selling yogurt and baked goods, but it is not enough to get by
(Hortencia, 45)
The gender effect of this shift away from paid jobs in services was magnified by two factors. One was
that the women in these communities were less flexible to work away from home. The other was that
jobs in agriculture favoured men in different ways. I will expand on these next.
289
Less flexibility to work away from home
Women who had the time to seek paid work, still had to go to urban areas to take up these jobs, as this
woman described:
There are always more ways to earn a living in town than in rural areas, at least
as a woman you can find work in restaurants or in a family home (Sandra Lorena,
24)
Women who were able to migrate to take up better paid service jobs in town did so. However, the
women interviewed in this study were those who were not able to do this.25 Some found it was hard
for them to take up jobs in town because it was still expensive to travel into town and to continue
living on their land. As discussed in the land ownership chapter, there was often a strong attachment
to land in these rural communities.26
Others said that paying for rent and services in town was costlier than living in their rural homes.
Some found they could not combine urban jobs with agricultural work and were uncertain if they
would earn more in town than from farming, as this woman described:
The few of us who have stayed on our land have had to work with what God has
given us, our land…but our land has stopped producing plantain and yuca and the
yuca grows with stains. The same with the plantain, it becomes yellow, so the
merchants won’t buy it in town…so we’ve tried farming chickens and things
because you think this is going to move things along but what happens? You go to
the market and you say “I have some chickens to sell” so they say “great, bring
them in” but then a day later they call you and say they no longer need any, bring
them another day, and on and on…so there’s nothing that takes off…I’ve had to
work in other places because chicken farming does not pay enough but even this is
25 I cannot give the proportion of people in these communities who moved to the cities or to coca-growing regions. As I explain in my methods chapter, my sample is not meant to be representative, see page 43. However, the point here is to focus on the situation of the people who continue living in the community where AD efforts have taken place. 26 This phenomenon may arise as rural farmers often value land as a bundle asset which offers them food security and a pension. Ghatak et al say this explains some Indian farmers’ hesitation to sell it even for a high price (Ghatak and Banerji 2009; Ghatak and Ghosh 2011a, 67).
290
hard because if you go into town to work in a kitchen then you have to leave your
land behind and if the merchants come here to buy our produce then you’re not
able to sell much because you’d miss each other, you have to stay at home to be
available to sell … it’s been hard for me because my husband has stomach cancer
and cannot work (Marianeli, 46)
Added to this, gender norms played a role in preventing women from taking up jobs outside their
locality. Firstly, many of the jobs in town such as motorbike-taxi driving, bus driving and construction
work were still largely socially out of bounds for women, especially older ones.
Another common reason that women gave to explain their reluctance to move was that they had
dependants. As women were expected to carry most of the care responsibilities, those with young
children or with sick dependants, were less able to travel or relocate elsewhere to work, as these
women recounted:
Well, the lady called me again to go [to her coca leaf plantation] this weekend to
cook for her [30 – 35 workers] for one, two or three months, but my daughter has
an appointment in Popayán, so I cannot… Now I am looking for other ways to
earn money, like selling yogurt and food here … whenever I go there [to the
plantation] my little girl says to me: Mummy, don't go there anymore, please
Mummy, she says let’s find other ways to pay for our food, so I say darling I don't
have money, where am I going to get money to take you to your medical check-up
in Popayán for instance [Hortencia, 45 – single mother]
They killed my husband, so I had to go [and work as a raspachin], now I have no
support and the city is a difficult place to live in, it’s hard to live there with
children (Luz Mila, 55)
Previous research has found that women in developing countries are less likely to migrate for work
than men (Mu and van de Walle 2011; Amuedo-Dorantes and Pozo 2006; Mendola and Carletto 2009;
Lokshin and Glinskaya 2009; Rodriguez and Tiongson 2001). Part of the reason may be that they are
291
not able to combine childcare with paid work. As illustrated with the above accounts, coca leaf
farming was so useful to women with children precisely because of this reason. It had given them the
opportunity to find jobs on farms, which were easier to combine with childcare, than town work.
Studies based on US census data, show that geographical proximity to mothers or mother-in-law has a
positive effect on the labour supply of married women with young children due to the increased
availability of childcare services and unscheduled emergencies (Compton and Pollak 2014). As many
of the people in the former coca leaf farming communities I studied had migrated to their current
abode and had often left their parents behind elsewhere, it is possible that without their extended
family nearby, free childcare was less readily available to them.
Research has also shown that women with young children and without childcare available, are more
likely to be involved in informal work which they can combine with childcare, which means they
forgo opportunities to carry out more productive forms of employment (Duflo 2012, 1059). This
means that they may not take up formal employment in town because they are not able to take their
children with them to work. The implication was that women without a male breadwinner and with
dependent were the most affected. Not only did they have to carry the burden of childcare costs alone,
but the more profitable work opportunities in town were especially out of bounds for many of them. A
similar trend may have been occurring in the former coca leaf farming communities in this study.
Uneven playing field for jobs in agriculture
Turning now to the agricultural job market, there were no rules that restricted women from working in
paid agriculture. Women were clearly skilled in agriculture and did much of the subsistence farming
in these communities as men. However, men had an advantage in paid agricultural jobs in great part
due to gender norms that kept women at home, and specialized women in domestic work.
a) The physicality of agriculture
Men’s generally greater physical strength made it easier for them to carry heavy equipment and pick
more crops during harvest times. This was often given as a reason why they should take up the paid
work while the women stayed at home. When seasonal farming jobs were locally available, women
said it was heavy duty and harder work for them than for men. They could not use machetes and carry
292
heavy spray pumps on their back from 7am to 4pm as easily as men. This was a disincentive for some
of them to seek employment in local agricultural jobs.
b) Off-location of opportunities
Better paid seasonal farming jobs in large plantations also tacitly favoured men because they were less
local. People believed it was easier and less risky for men to travel to take up paid jobs. When coca
leaf had been farmed in the home community, these contract farming jobs did not require people to
travel. Besides finding it difficult to work away from home because of their dependants, women had
less time to travel in the aftermath of coca, as they were occupied with more unpaid housework.
As has been discussed above, women often lacked extended family to support them, which made them
less mobile. In addition, women in general felt vulnerable when travelling alone to rural places far
from home. This was especially the case for work in coca leaf plantations, as they feared being
mugged or taken advantage of, as this woman describes:
Since my husband died, I’ve had a very hard time trying to find whatever job is
available, [working as a raspachin] is the only thing that provides an income…
it’s harder for women who are the breadwinners, they must scramble around for
any job… you have to be the father and the mother at the same time… and these
are jobs for men, you know women are a bit more sensitive…and I feel afraid
when I go there…I normally have to travel there with some male colleagues… and
I need to go with references because they [coca plantation owners] won’t let just
anyone in… (Luz Mila, 55)
These fears for personal security when women travel is a global phenomenon. It is possible that these
fears may be heightened in certain rural contexts where men find it easier to assault women in transit
because their anonymity is better protected, as was found in Nepal (Neupane and Chesney-Lind
2014). As I will later reflect on the gendered dynamics of violence in these communities more
generally, there was a general lack of trust resulting from rivalrous extra-legal groups and the
governance vacuum they had created (see chapter on extra-legal groups). This heightened fear of
293
violence against women may have prevailed and helped to confine women to the domestic sphere and
reduce women’s spatial mobility.
c) Forms of payment
Many of the jobs in lucrative cash crops, such as coffee and coca leaf, favoured men because they
were paid by performance i.e., the load of cash crops collected instead of the hours worked. This
resulted in fewer equal pay opportunities for women in agriculture. While domestic jobs were paid by
the day, the rate at which the more lucrative daily farming jobs were paid was ultimately based on
physical performance rather than the number of working hours. People were paid about COL$15,000
to COL$20,000 [$4.2 to $5.7 USD] per day for washing clothes in Putumayo in 2018, while they were
paid around COL$6,000 [$1.7 USD] for the arroba of coca leaf collected. Coca farming was not easy
because of the weight people had to carry and the hand strength they required. Men estimated they
could pick between 15 to 20 arrobas of coca leaf per day, while women estimated they could pick
between 2 and 15. Agricultural employers thus had greater incentives to hire men than women, given
their greater efficiency for the jobs required.
Reinforcement of gender norms
I observed that with fewer economic opportunities available to women, following the local
substitution of illegal drug crops, women’s identity as dependants of men was reinforced. Women
mentioned that they had fewer resources to diversify into better paid work than housework or to pay
for their own education and training, as this former coca leaf producer and nursery teacher explained:
There is no alternative [out of poverty]. I have experience of working in a school
but now they ask for a degree for this kind of work and I'm missing four semesters
to finish my studies, worth COL$ 20,000,000 million [$5,700 USD]. So how can I
overcome my situation? Most people here can barely finish school, and that is as
much education as they can get (Carmen, 43)
As they were less able to become financially independent from men, the remoulding of gender-role
attitudes that had started during the coca leaf bonanza, stopped and gender norms were reinforced.
294
Not only did gender norms influence their economic opportunities after the coca leaf economy
contracted, but women’s lower participation in the paid workforce, reinforced the same gender norms,
in a kind of symbiotic relationship.
The barriers faced by women who try to participate in paid agriculture in other contexts, have also
been found to consolidate gender norms. For example, Carranza found that agricultural jobs that
required deep tillage in India created more demand for male labour which simultaneously decreased
women’s bargaining power (Carranza 2014). And Alesina et al. suggest that male bias in agriculture
can have a lasting effect on gender-role attitudes (Alesina, Giuliano, and Nunn 2013). They argue that
views among second-generation US migrants that men make better political leaders stem from
whether their ancestors came from places where the soil required the use of the plough, favouring
male labour, as opposed to hand tools which women were able to use as effectively. They say this
determined female participation in the workplace, which had a knock-on effect on gender-role
attitudes among their descendants (Jayachandran 2015, 6–7). Conversely, when there are fewer
barriers for women to earn as much as men from the same work, gender norms have been found to
lessen as a result. For example, in 1970s China when tea production flourished, women had an
advantage over men in picking tea leaves, so more women were employed (Qian 2008). This then
increased their share of household income as well as their bargaining power in their families
(Jayachandran 2015, 6).
My inference mirrors, albeit in reverse, Jayachandran’s (2015) theory of how economic development
reduces gender inequality. She describes how higher household incomes result in more efficient and
less labour-intensive home production which frees up women’s time. She echoes findings that
technologies such as central heating, electricity and running water, in the US reduced the time women
spent on fetching water, coal for heating and other chores (Greenwood, Seshadri, and Yorukoglu
2005). A similar effect has been found in other countries (Ramey 2009; Dinkelman 2011; Coen-
Pirani, León, and Lugauer 2010; Meeks 2014; Devoto et al. 2012). Jayachandran also explains that the
shift in labour from agriculture and manufacturing towards services, means that less physical strength,
or “brawn” is required so female labour productivity increases with development (Jayachandran 2015,
295
5). This builds on Galor & Weil’s theory that capital shifts the relative returns from physical intensive
tasks to mentally intensive tasks, so men cease to have a comparative advantage over women and so
the gender pay gap is reduced (Galor and Weil 1996).
The danger of gender norms being reinforced, is that it is difficult to redress their effect with
“seemingly neutral” rules and laws. However much universal property rights and universal access to
education is available in former illegal drug farming communities in Colombia, equal financial
opportunities for men and women may still be absent because of the gendered segregation of work.
When gender norms have been internalized by people in their identities and roles, the economy will
continue to privilege men over women (Kenny 2007).
Women’s dependency on the family unit: source of efficiency or hindrance?
Some may argue that women who lived in conventional family units were equally affected as men by
the economic recession that followed the suspension of coca leaf farming. After all, women could rely
on male breadwinners, there was a strong sense of cooperation within the family in these
communities, and household incomes were often pooled. While women generally carried out more
unpaid work, men worked equally hard and in more physically demanding jobs. The family household
was a unit that was entitled to AD subsidies and land titles could be joint, as well as belonging the
individual men or women. And women were not overlooked by state subsidies. Twice as many
women received AD benefits from the Familias Guardabosques programme, compared to men
(Cortes Yepes 2007, 272). In families where resources were shared equitably, any income drop or
decrease in economic opportunities that resulted from the abandonment of coca leaf farming in the
community, would therefore have affected all the family equally.
From a new institutional economics perspective, Coase (1937) and Becker (1991) held that
hierarchies in the firm and in the family can lead to economic efficiency and are a solution to
coordination problems (Coase 1937; Becker 1991; Braunstein 2008, 959). They would probably argue
that despite the existence of a degree of work specialization and a gendered household division of
labour, it was the whole family that suffered the income shortfall, not just the women. In addition, the
296
family unit structure may have ultimately protected women from being more financially
disadvantaged than men.
However, as Braunstein points out, Becker’s argument that the family behaves as an altruistic unit,
not only overestimates the gains to work specialization but presumes that the persistence of gender
hierarchies in the household result in the greatest social good, when arguably they simply further the
interests of men at the expense of women’s (Braunstein 2008; Braunstein and Folbre 2011). Becker’s
view also assumes that if men contribute a fair share to the household expenditures, they are not also
keeping a higher share for their personal expenditures compared to women (Van Staveren and
Odebode 2007, 41:920). Kanbur & Haddad point out that the unitary household does not always
distribute resources equitably within the household, so a decreased household income can
disadvantage one household member more than another (Kanbur and Haddad 1994, 457).
Although some women in these former coca leaf farming communities revealed a strong sense of
household bargaining power, in terms of choosing where to live and sometimes how to spend their
family’s money, it was not strong enough to bargain over gender norms such as unpaid housework.
There were some women, like this one, who complained about their husbands, sons and brothers
spending their money how they wanted, such as on alcohol, gambling and even prostitution, and not
saving it for household expenses:
INTERVIEWER: And with what you saved from coca production, were you able to
buy land? Or did you lose everything when you had to stop...?
INTERVIEWED: Ehhh, well look here, I think coca just gave us enough for our
daily bread, my husband at that time really liked his liquor, so the money just
covered his drinks and the food he brought us… now we try to make ends meet
from my husband’s pennies [savings] (Maria, 58)
In addition, there was an unequal sense of responsibility for children among many couples. Mothers
often assumed greater responsibility for their children’s education than fathers, especially those who
no longer lived with their children’s fathers, as this woman’s experience illustrates:
297
…the truth is I’m the one who is there for my four children
INTERVIEWER: What about your husband?
INTERVIEWED: My husband’s harvest only comes once a year… so I really must
cover the cost of our girl who is going to study in Popayán
INTERVIEWER: What will she study?
INTERVIEWED: She is going to study nursing and wants to go to the private
university… as she’s the only one [of our children] … I’d like to give her
something better, but the truth is I do not earn enough from the store… [so I will
take out a loan for my daughter’s study] … I see that she really shows an interest
and that she wants to study, so with the help of God somehow, I will have to pay
back the loan so that she can study…and one day I hope to buy some land because
that is the other dream of mine (Derly, 40)
Therefore, in the new institutional economics sense, the gendered segregation of work could be
regarded as a structure of voluntary cooperation that has emerged to overcome collective action
problems and reduce uncertainty but only up to a point. Mackay et al argue that institutions are more
complex than this and can sometimes prioritize power and domination over efficiency maximising
cooperation (Mackay, Kenny, and Chappell 2010, 574). Braunstein argues that institutions that create
male dominance and privileged access to resources is effectively patriarchal rent-seeking behaviour,
which is ultimately detrimental to the economy, as well as being intrinsically unfair. “Social norms
like the sexual division of labour, are not simply solutions to the problem of coordinating family
production, but rather a way to organize family labour in terms that benefit men” (Braunstein 2008,
967). Braunstein, as other feminists, does not refute that a degree of gender work specialization in
childcare is economically efficient, but argues that there are also incentives for men to perpetuate
norms that force women to overspecialize in this work. It gives men greater control over resources
and diminishes women’s bargaining power, and further strengthens hierarchies within the household
which eventually leads to distributional friction and economic inefficiency (Braunstein and Folbre
298
2011; McCrate 1988). A feminist’s response to Coase and Becker’s argument that norms of gendered
labour division maximise efficiency, is thus that they also shape behaviour in a way that captures
power asymmetries and benefits men at the expense of women’s freedom (Krook 2003; Kenny 2007,
92). Therefore, it should not be assumed that gendered work segregation and women’s dependency on
the family unit are institutions that have emerged solely for economic efficiency.
This suggests that even if one argued that most women were financially “safeguarded” by the family
unit when their community stopped farming illegal crops, they were nonetheless demoted back to
their conventional roles, as men’s support workers. The argument here is that it is irrelevant whether
women or men felt the income shortage more acutely, because the women who stayed behind, and
assumed the burden of unpaid work, were effectively relegated to a secondary role. This may be a
non-monetary reason why women, who develop greater aspirations whilst working in the illegal drug
trade, may find it difficult to revert to social expectations, and may instead prefer to follow the illegal
trade elsewhere.
Gendered dynamics of violence
During the 1980s and 1990s the drug trade brought the conflict to many rural areas in Colombia, like
the ones where I conducted my fieldwork. The conflict between armed groups exposed women in
places like these, where farmers had colonized the land relatively recently and lived a precarious life,
to particular forms of violence and presented them with additional new risks (D. Meertens 1997, 248).
Apart from noting women’s anxiety about travelling for work opportunities, my research has not
examined the gender violence that emerged from coca leaf farming in these regions. I did not set out
to assess the weight that violence puts on women from the start because the focus of the study was on
women’s income and economic opportunities after their communities stopped farming coca leaf.
Therefore, there is scarce data from which I could generalize about these women’s exposure to
violence before, and after, their communities abandoned coca leaf farming.
Nevertheless, it is important to appreciate that coca leaf farming, and the presence of armed strangers
in the communities I studied, brought risks upon women, as well as economic opportunities. There are
studies of the gendered dynamics of violence in some of the same communities where I conducted my
fieldwork, such as in Valle del Guamuez, as well as nearby and similar rural places when they farmed
coca leaf. They confirm that prior to giving up coca leaf farming, women had been subject to
specifically targeted, and sometimes less visible, forms of violence that only applied to them.
299
For example, a study of violence against women in Bajo Putumayo, found that between 1999 and
2006 more than 150 women disappeared and hundreds of children were orphaned (Guatame Garcia
and Delgado Machega 2016, 64). It further found that most of the women who disappeared between
2006 and 2014, were farmers in Orito, Valle del Guamuez and San Miguel, places where
paramilitaries and coca leaf crops were present and not far from where I conducted part of my
fieldwork (Guatame Garcia and Delgado Machega 2016, 51).
In 1993, violent death was the second highest cause of death for women in Colombia between the age
of 15 and 39 years (D. Meertens 1997, 227). And this is just what was publicly reported. Besides the
loss of life and suffering, a huge repercussion of women’s deaths and disappearances, were their
families’ wellbeing. The study in Bajo Putumayo found that the grandmothers and relatives who took
care of the numerous orphans left behind by the forcibly disappeared and killed women, had been
unable to meet the children’s basic needs because they lived in poverty. They said the high levels of
poverty were created not only by the conflict, but by the fumigation of coca leaf used by law
enforcement, which besides coca crops, destroyed other forms of subsistence farming on which they
relied for their daily food (Guatame Garcia and Delgado Machega 2016, 56).
The sexual nature of violence is another way in which women were disproportionately affected,
because women who work in the illegal drug trade are at a higher risk of sexual abuse (Navarrette
2019a). Some of the women I interviewed revealed they had been victims of sexual violence from the
time when they had worked in coca leaf production and still carried the trauma. Some have argued
that sexual violence was used by armed groups in Colombia’s conflict, as a metaphor of their
supremacy and to show their territorial control (Guatame Garcia and Delgado Machega 2016, 64).
The use of sexual violence, particularly by paramilitary groups, has been recorded by academics and
official historians alike.27 During the seven year period in which the paramilitaries ruled in El Placer
community, in Valle del Guamuez, and in the nearby veredas, women’s bodies became the object
through which they asserted their territorial control in various ways (CNMH 2012, 173). Firstly, they
created a sexual trade in places such as La Hormiga, in Valle del Guamuez, during the coca leaf
bonanza. Women, mostly underage and with little education, came to work where armed men paid for
sex. Despite the economic opportunity, many exposed their bodies to diseases, and a high proportion
of sexual workers were murdered if found to have a sexually transmitted disease (Guatame Garcia and
Delgado Machega 2016, 53). Others were punished for misbehaving with torturous sanctions such as
by being tied naked out in the sun in public view (CNMH 2012, 175, 193, 197, 199).
27 It is worth noting that sexual violence is not an inevitable side effect of people living in contexts of conflict. As some studies have noted, guerrilla groups in Colombia, such as FARC, constrained the use of sexual violence possibly because they had a high fraction of female combatants, higher military discipline than many armed groups, and or relied on local support (Wood 2006).27 Therefore, it should not be assumed that gender violence is a standard reality of life among armed groups.
300
Secondly, local women suffered in various ways from being forced to live in the presence of hundreds
of armed men in their community. It was common for them and their children to be raped, and
sometimes taken as sexual slaves, under the threat that if they did not accede, their relatives would be
killed. Sometimes they did not tell their fathers and husbands that it was rape for fear of reprisal, so
their communities didn’t interpret these acts as rape and ostracised them for having what looked like
consensual relations with paramilitaries (CNMH 2012, 204, 214–16, 218).
Even when there was no homicide or physical injury, a life in a state of constant insecurity, had an
impact on women’s wellbeing and their social relations. As Meertens put it, violence destroys not just
bodies and assets, but women’s identity, personal integrity and social relations because where
violence is suffered it can also be reproduced (D. Meertens 1997, 226). For example, mothers in El
Placer asked the brothel owners to please turn around the TVs showing pornographic videos in full
view, so that their children did not have to watch it from the street, to which the brothel owners
replied that their children needed to learn from a young age how to behave towards women (CNMH
2012, 187–89). Women had to adapt their way of life, their schedules, their routes, how they moved
and greeted men and how they dressed, in their fear of the armed men and their sexual behaviour
(CNMH 2012, 206).
In addition to being targeted with violence, women living in conflict situations similar to the ones I
studied, experienced fear, anguish and extreme rage when they faced the loss and injury of male
relatives (Zubillaga, Llorens, and Souto 2019). As is often the case when men are killed, women
suffer in the aftermath, as they become witnesses of violence, widows, heads of households, and
relatives of inmates and disappeared men. Their suffering when their husbands and children are taken
by the state to jail or by armed actors to fight, and when their freedoms are curtailed by fear of
violence, is but another form of violence (Wilding 2010, 739–40).
Besides the psychological traumas and emotional shock of these experiences, women are often forced
to take on additional financial responsibilities they are unprepared for. For example, a 1997 study
based on fieldwork interviews with campesino women from numerous places in Colombia where
there had been conflict between armed groups in Colombia (including in Guaviare, Monteria,
Barrancabermeja, Florencia, Villavicencio, and Cordoba) found that women suffered the indirect
consequences of homicide. Not only did they lose their previous means of subsistence, but they
became responsible for the financial as well as the emotional wellbeing of their dependants, and
sometimes had to resort to prostitution in their desperation (D. Meertens 1997, 233).
Moreover, the mere fear of their loved ones being harmed drove many women in these places to
displacement in the 1990s, resulting in about 30% of displaced households being headed by a woman
(D. Meertens 1997, 230). The consequence of this phenomenon was that women were forced to re-
301
build their lives in a new setting where their support network and primary relations were absent, at a
huge cost to their personal lives.
These are some examples of the gendered dynamics of violence in communities like the ones in my
study, during the coca leaf bonanza. It is worth reflecting on them because they are an aspect of
conflict and the drug trade that seldom receives adequate attention, and which generated explicit as
well as hidden risks for women. It has been noted that initial discussions of “new violence”,28
excluded important forms of violence, such as less visible and non -lethal violent practices in the
family and the private sphere, as well as omen’s coerced or intentional roles in violence. These forms
of violence are more likely to go unreported because there is a tendency to trivialize or regard them as
less important (Wood 2006; Hume 2008).
The importance of acknowledging these forms of violence in research is twofold. Firstly, they are
intrinsically important in the pursuit of the wellbeing of women. As Hume succinctly puts it,
“separating the domestic from the public [study of violence] is more than a statement of legality. It is
a statement of power” (Hume 2008, 68). And secondly, understanding that women are not simply the
supporters or opponent of the use of firearms, but are linked in a range of relations with armed actors
sheds light on women’s own role in the proliferation of violence. Women who live in constant of fear
of their male loved ones being potential victims, will devise ways to cope that transform them into
active participants in violence as they condone or even exacerbate it (Wilding 2010, 722). For
instance, a study of mothers’ survival strategies in Caracas, where the homicide rate is rife in poor
areas, found that women who opted not to isolate themselves from the violence, were able to broker a
local ceasefire but also to teach violence and use gossip to protect their loved ones at the cost of
perpetuating violence (Zubillaga, Llorens, and Souto 2019). Similarly, the use of violence by mothers
against children in El Salvador is considered good parenting and these values reproduce violence
(Hume 2008, 64). This suggests that women’s role in limiting as well as proliferating violence in
conflict situations should not be underestimated.
In sum, there were risks as well as the economic opportunities that coca leaf farming presented for
women. Whether the risk of gender violence increased or decreased after these communities stopped
farming coca leaf was beyond the scope of my study but clearly, there was a trade-off between the
dangers and the opportunities that women brought to women. There are two implications to this brief
discussion. One is that the extent and the various forms of violence, not just lethal violence, and the
role that gender plays in the reproduction of violence, are an area that merits further research (Wood
28 “New violence” refers to the proliferation of social and economic violence by a spectrum of actors across rural and urban contexts and across social classes in Latin America. It has emerged from the breakdown of the state’s monopoly on the legitimate use of violence and other factors including the attraction of “easy money” from working in the illegal drug trade, and the state has an interest in violently maintaining the status quo (Wilding 2010, 724; Pearce 2010). It is complex and multi-dimensional and not on a single spectrum of conflict (Hume 2008).
302
2010; 2006; Wilding 2010; Hume 2008). The number of homicides is a highly limited measure of
violence (Pearce 2010). And hidden and ignored forms of violence against women need to be
unpacked. The other implication is that AD policymakers should not underestimate the psychological
strain that gender violence has in limiting women’s potential to earn an equal income as men. Not
only can experiences of violence negatively influence women’s wellbeing but they can also hamper
their tenacity in the workforce.
3.5 Policy implications
So far, gender-sensitive AD policies have included supporting women in Bolivia with the production
and marketing of alternative crops, supporting women in Afghanistan to form associations and
enterprises, promoting women’s leadership in Peru and providing micro-credit for women in Vietnam
and Lao (UNODC 2015). These efforts to support women’s businesses make sense because the labour
market can use gender roles as entry barriers, which then promotes further gendered job segregation
(Elson 1999). In other words, the barriers women encounter when entering the business labour market
can reinforce traditional gender roles.
The insights from this chapter highlight the importance of these types of AD efforts to help reduce
gender bias in the economic opportunities of former drug farming communities. As my findings
illustrate, women have a surprisingly big stake in the coca leaf economy in Colombia. And drug
farming can increase women’s economic self-reliance and to redistribute power between men and
women. These are additional incentives that may keep communities contently farming coca leaf.
Therefore, besides supporting the creation of women’s cooperatives and associations, AD efforts
should also strive to address the informal gender-role attitudes and norms that divide labour by
gender. Not to do so, risks perpetuating the gender pay gap, leaving at least half of the working
population economically disincentivised and vulnerable to participating in the illegal drug trade again.
Tackling residual gender bias in legal property rights and universal education are other quantifiable
ways in which policymakers and aid organizations are already addressing gender inequality and
helping women build social capital. However, as this study has shown, the norms of gendered work
segregation in former coca leaf farming communities can persist despite there being overall gender
equality in access to property and education rights and the opportunity to join women’s associations.
303
In addition, behaviours that challenge gender-role attitudes can revert to the status quo when the
economic incentives are no longer present. In other words, gender norms can overrule previous
changes in behaviour when there are fewer economic incentives for these behaviours to continue.
Policymakers should therefore broaden the ways in which they tackle the issue of resilient gender
norms which create economic disincentives for women and undermine their motivation to support the
substitution of illegal drug crops in their community. Education campaigns that advocate change in
the social expectations of women should not be underestimated. Similarly, investment in ways to
lessen the burden of housework may contribute to reducing women’s comparative disadvantage in the
workforce. Childcare support could help reduce women’s domestic workloads and facilitate their
entry into the paid workforce. Time saving appliances such as labour-saving machines and food
processing machines have been recommended for improving women’s participation in agriculture in
Nigeria (Ogunlela and Mukhtar 2009). And recognising ways in which to support women who may
have experienced violence in many ways, some of which will be less visible, in the period when their
communities farmed coca leaf, is a key way to enable women to make the most of the economic
opportunities that are available to them.
Some of the inherent biases in agriculture that favour men over women, may be challenging to
address, such as the safety of women who travel to work and the form in which some cash crops are
paid. Nevertheless, there are some things that could be done to help reduce the imbalance created by
the physical nature of some work. Encouraging women’s participation in farm management and
decision-making, may help women influence how resources are used to invest in agriculture, to favour
women’s needs and interests, such as the acquisition of labour-saving machinery. Similarly, gender-
mainstreaming in all the AD projects, instead of promoting separate agricultural projects for women,
may be a more effective way to ensure a gender-sensitive rural development (Bock 2015).
Gendered work segregation, like other social norms, may be difficult to change (Mahoney and Thelen
2010, 4; Pierson 2004, 10–11). As everyone knows, cultural gender norms can persist even when
economic conditions improve, and women in developed countries continue doing a lot of the
housework. And no single policy may be able to dissolve gender work segregation (Jayachandran
304
2015). Nevertheless, it is not impossible for gender norms to change in a way that favours gender
equality. For evidence of this one can look back on how women have gained the right to vote, to
divorce, and to own property.
3.6 Conclusion
The experiences of the women I interviewed reflected the interrelationship between economic
opportunities and gender norms. Illegal drug farming played a significant role in shaping gender
norms in the community. However, when the opportunities for local paid employment faded with the
demise of the coca leaf economy, the gender norms were ensconced in the community once again,
leaving women with fewer economic incentives. My findings suggest that the first step in
transforming institutions, such as gender norms, is to change people’s behaviour. People’s behaviour
changes when there are economic incentives to do so (Greif and Laitin 2004, 640; Mackay, Kenny,
and Chappell 2010, 577). This is what people experienced in the communities I studied. The coca leaf
bonanza created more paid employment opportunities for women. This, in turn, enabled many
working women to work as equals with their partners, pursue different kind of jobs, and feel less
financially dependent on men.
It would not be fair to say that the illegal trade increased gender equity in all its forms. Women still
faced unique challenges that men did not, such as queuing all night for water, putting up with moody
labourers, lower remuneration and double shifting. However, in this process, some gender-role
attitudes began to evolve. Some women gained greater bargaining power in their households and
businesses. Some also described their ability to diversify into new types and better paid jobs. And
others described how they had become less dependent on male breadwinners and parents.
The theoretical point highlighted in my analysis, is that institutions require ongoing maintenance and
support to survive (Mahoney and Thelen 2010). If there are no economic incentives for changed
behaviour to continue, then deep-seated gender norms will resurface. This again, was illustrated in the
communities once they substituted their illegal crops.
305
After the passing of the coca leaf bonanza, most women encountered not only fewer jobs, but also
implicit barriers that kept them from participating in the workforce on par with men. Lower household
incomes meant more labour-intensive home production, using up more of women’s time for unpaid
work, including subsistence farming. With less money available to buy food and pay for time-saving
appliances, women had less time available for paid work. In addition, as the local economy dropped,
local paid labour shifted back from services to agriculture. This meant there were fewer paid jobs in
domestic work. And jobs that were available were not remunerated as well as they had been during
the coca leaf bonanza. In addition, although there were no rules that restricted women from working
in paid agriculture, men had an advantage in agriculture firstly because their physical strength made it
easier for them to carry heavy equipment and pick more crops during harvest times. It was easier and
less risky for men to travel to these jobs, and the form of payment favoured men. Many of these jobs
were paid by the load of cash crops collected and not by the hours worked. The increased flexibility of
men and wage premiums for physical strength gave men an advantage in the workforce and resulted
in fewer paid opportunities for women.
All this meant that women who were unable to break out of poverty during the coca leaf bonanza,
perceived a greater income shortfall than their male counterparts. Women who were not dependants
on parents or partners and who had dependants themselves, seemed to be the people most negatively
affected by the fewer job opportunities that were available to them. In addition, they had to carry the
burden of childcare costs alone. But even those women who lived in family units were adversely
affected by the contraction of the illegal economy. This was because it reinforced the segregation of
work by gender and restored women’s identity as financial dependants of men. As I have argued, it is
not necessarily more economically efficient for women to overspecialize in unpaid housework and
domestic services. Besides, it limits women’s freedom of choice on how to live a life they value.
It should be noted that the sample of people on which these observations are based, is more likely to
favour profit over traditional customs, such as the segregation of work by gender. Conceivably,
people who migrate to work in coca leaf production are risk takers and more entrepreneurial than
average and therefore, they might be more open to women working. However, despite this possible
306
sample bias, my findings suggest the opposite effect when the local economy contracted in these
communities. When they stopped producing coca leaf, the norms that promoted gendered work
segregation were reinforced, even among a potentially more entrepreneurial sample of people.
This research is highly relevant to AD policymakers in three ways. Firstly, they highlight the
economic opportunities that coca leaf farming created especially for women. For AD efforts not only
to be gender sensitive but sustainable over time, policymakers must seek to recreate these incentives
for women in the legal economy they promote. Failure to do so, may result in unsuccessful AD,
because women may find greater returns and equality in the illegal economy either later, or elsewhere.
Secondly, I offer new insights into the issues that women face in former illegal crop farming
communities. The findings in this chapter contrast with what is already known about women working
in drug farming in other countries. In opium poppy farming communities in Afghanistan, girls lack
access to school and women have low or non-existent wages (UNODC 2018b, 23–25). Whereas in the
Colombian coca leaf farming communities included in this study, there was equal respect for
women’s rights to education and women were able to earn better money during the coca leaf bonanza.
And contrary to recent generalizations made by policymaking institutions, formal and informal
restrictions to property rights and to an education, were not perceived as significant barriers to
women’s progress in the communities I studied. Women’s articulation of use, transferability, and
enforcement of their property rights to land were equally respected as men’s both by the law and by
informal norms. Similarly, the men and women included in this study had very similar levels of
school attainment and their attitudes to their children’s education did not appear to favour sons over
daughters.
Thirdly, this study highlights the pervasive role of gender norms that specialize women in unpaid and
domestic work. These can influence women’s economic behaviour asymmetrically despite there being
gender symmetry in institutions of property rights to land and access to education. Therefore, even
with equal access to land and education as men, women living in communities undergoing AD, may
find it harder to diversify into other types of better paid jobs and to reduce their financial dependence
on men. AD policymakers in Colombia should thus expand their focus from campaigning for
307
women’s increased access to land ownership and education, toward other ways of increasing their
economic opportunities so that they are not unfairly burdened with subsistence farming and unpaid
housework or held back from profiting from agriculture equally as men.
Overall, this research shines a light on the positive effect that economic growth has on gender equity,
even when that growth arises from an illegal crop, and the negative effect an economic contraction
has on reinforcing gender norms. In Latin America, the participation of women in the workforce has
grown fast and the gender gap in wages has narrowed (Duflo 2012, 1054). This makes the situation of
the women left behind in former coca leaf producing communities in Colombia even more disquieting
and shows importance of tackling the constraining effect that gender norms can have on women’s
economic opportunities.
308
Thesis Conclusion
I came back to Bogota from my fieldwork visits feeling tired and overloaded with information. It was
not only the long journeys I had undertaken. The windswept motorbike rides in the fog, the bumpy
drives in 4X4 vehicles across scenic valleys, the slow bus and truck rides along endless bendy roads
squeezed in between other passengers, the taxing treks up muddy mountains, and the long delays at
local airports, had inevitably added to the exhaustion that follows any adventurous challenge. The
hardest part of my readjustment to life in the city was having to reconcile with the stark contrast
between what life had been like in the veredas I visited, and what life was like in Bogota. I bought a
cup of coffee at the airport with the same amount of money I had budgeted for a large plate of lunch
in many of these rural places. I paid the cab driver for my ride home, with the equivalent sum that
many farmers I had interviewed were earning for a day’s work. My trainers and wind jacket, which
had stood out as bright and fancy in the veredas, now looked dirty and shabby on the streets of
Bogota. How different the standards of life were in those isolated pockets of rural communities. How
trusting the people had been to share their stories with me, expecting nothing in return except perhaps
the vague possibility that I might be able to convey their experiences to decision-makers in Bogota
and beyond. I pondered how to give meaning to the dozens of stories that I now had etched in my
audio recordings, my notes, and my head.
I recalled a panel discussion on rural development and land issues that had been organized by a
leading news magazine around the time when I had been carrying out my pilot enquiry in Bogota.29
Top policymakers were there in shiny suits and blow-dried coiffures nodding in agreement on how
evident it was that land titling was the missing link in farmers’ ability to succeed in legal rural
agriculture. There had not even been an odd farmer present to act as the “representative” of all
campesinos and to echo what their hosting pundits proffered, as I had witnessed in similar events.
After my fieldwork, I found it surprising that in policymaking circles the association between land
titles and bank loans was perceived as a no-brainer in rural farmers’ path to financial security.
29 Foros Semana (“Tierras y Desarrollo: Una Mirada Vanguardista a La Propiedad En El Campo” 2017) Gimnasio Moderno, Bogota 25 July.
309
My various trips to Bogota, between the pilot enquiry and fieldwork visits, had happened in parallel to
several news reports of violent confrontations between coca leaf farmers and the army. These had
occurred in different locations including in Nariño, Tumaco and Cauca, when the army had
accompanied manual eradicators who had been sent by the state to uproot illegal drug crops.30 The
follow-on interviews and discussions of these reports centred on the tough job the state faces in its law
enforcement task. The sentiment in Bogota was that armed groups were manipulating farmers in these
places to oppose the state’s endeavours to convert them to the legal economy. After my fieldwork, it
was clear to me that this was not the situation for many coca leaf farmers. At least the ones I had
interviewed, had freely chosen to farm illegal drugs and then to stop doing so. One lesson the state
can learn from communities like these, where people substitute their illegal drug crops with legal
alternatives voluntarily, is that their success in stamping out illegal drug production for good lies in
their offer of superior governance to what armed groups could provide. Only then will legal
agriculture be more attractive than the illegal trade. And occasionally, in some communities where
extra-legal groups compete, there is a small window of opportunity for this to happen almost
spontaneously.
Later, again in between fieldwork visits, I came to Bogota to spend Christmas with some relatives. I
watched how the housemaid gift wrapped some presents for her children and grandchildren. She only
went home once a year, partly because that was what was in her contract and partly because she lived
so far away that she could not afford to visit more regularly. She was thrilled to spend her year’s
savings on the trip and on clothes for her family. Despite living so far away, she was her extended
family’s main breadwinner, and she believed this made her sacrifice as a round-the-clock live-in
domestic worker worthwhile. I couldn’t help wondering how other women in rural communities like
hers, had made ends meet. Did they all have to deny themselves family life or else submit to a life of
poverty? Clearly, easier money could be made in illegal crop farming, and I wondered how significant
30 For some examples of news coverage see 26 February (Noticias Caracol 2017); 11 September (El Tiempo 2017a); and 7 October (France24 2017)2017. For further analysis of this phenomenon see (Garzon Vergara, Gelvez, and Silva Aparicio 2019).
310
women’s search for a better future, with greater economic opportunities, contributed to the resilience
of the illegal drug trade.
Considerations for theory and for AD policy
There seems to be a dissonance with what I witnessed in these communities and how policymakers in
Bogota, and beyond, perceive some of the issues of Alternative Development (AD). Clearly
policymakers do see the need to support their assertions with surveys carried out by their own
officials. At the time of this study, the effectiveness of the current land titling programme for AD was
being monitored with a survey that measured beneficiary farmers’ new economic activities, land
tenure situation, credit history and reasons for having desisted with the production of illegal crops.31 It
also categorized their family units’ socio-economic level and quantified their finances. However, this
survey had very simple multiple-choice answers that failed to detect the role of informal institutions
as I perceived it. For example, when asking farmers why they had stopped producing illegal crops, it
gave them the option of ticking any of the following answers: “a) because it is a crime; b) because it
harms others; c) because it generates violence; d) because I lose my land; e) because it generates
distrust between neighbours; f) because of law enforcement’s aerial eradication efforts; g) because it
does not set a good example to my children; h) because the community disapproves of it; i) because of
religious reasons; j) other”. This information, however, fails to convey the full story of how different
reasons may be interlinked and overlooks the nuances in how and why farming coca leaf generates
violence, by whom, and in what sequence.
Another blind spot in this survey was that of gender issues. Firstly, it only included people who had
been illegal crop owners, pickers, sellers, or refiners as those were who it classified as being directly
linked to illegal drug production. This dismissed off hand the participation of many women who are
vital in the drug supply chain, and who operate as domestic workers on coca leaf plantations.
Secondly, when measuring people’s economic activities, it failed to include non-income generating
activities such as childcare and housework. This meant that this type of work was not even considered
31 Survey called “Encuesta linea base para formalizacion de tierras, Informacion preliminar familias beneficiarias, programa de desarrollo alternativo”, designed by UNODC and the National Land Agency in Colombia, and used in 2017.
311
to be the alternative to which many women had been obliged to assign more of their time to, since
illegal crop farming had ended in their community. And thirdly, it measured people’s finances as a
family unit, and not as individuals, reflecting a biased measurement of what each person was earning,
and thus failing to detect women’s asymmetrically reduced income, compared to that of men.
I believe that I have captured a bigger picture by analysing people’s stories, and by listening not only
to those who have received AD subsidies, or who were enrolled in the land titling programme (these
people were predictably convinced of the value of land titling), but to the invisibles in these
communities: the landless, the wives, the daughters, the people who still worked as raspachines… It
was their testimony that gave me a more nuanced idea of what is at stake when people opt to stop or
continue farming coca leaf. It brought up the limited value that land titling offered these communities,
at least in terms of delivering the agricultural success they had been led to expect. It revealed the
mechanism by which the behaviour of armed groups and the deterioration of their governance had
facilitated AD in some places. And it uncovered how women’s economic opportunities worsen in
contexts of AD.
From my findings, I inferred that the state’s attempt to “formalize” land ownership was not having the
intended effect of disincentivising farmers from growing illegal crops in the long term. This presents
an important empirical addition to the ongoing discussion on property rights theory, as it raises
questions about the accuracy of De Soto’s argument that land titles infallibly increase the productivity
of land assets (de Soto 2000). It was clear that farmers already had functional informal property rights
to land, which gave them most of the benefits that land tenure security brings. This lends evidence to
the Ostromian thesis that informal institutions can be more effective than formal ones (Ostrom 2009).
Many have argued in different ways that secure land tenure makes it easier to sell and rent land, and
therefore increases the profitability of land (Demsetz 1967; de Soto 2000; Mueller et al. 1994;
Brasselle, Gaspart, and Platteau 2002). And some have claimed that land titles make land tenure more
secure (Fenske 2011, 139; Timothy Besley 1995; de Soto 2000). However, in the communities I
visited, farmers could already buy and sell their land with their informal property rights. Some have
argued that land titles increase people’s commitment to their land because they are less afraid that
312
they will lose it and so invest more in it (Galiani and Schargrodsky 2010; Brasselle, Gaspart, and
Platteau 2002). However, even without land titles, the farmers I interviewed had felt it was safe to
invest resources and commit long term to it. And while Fenske (2011) and Besley (1995) suggest land
titles give people more freedom to innovate in how they use their land, thus rendering it more
profitable, the farmers I interviewed did not feel restrained by the lack of land titles to use their land
in innovative ways because their de facto property rights were already secure. So, although land titling
was a symbolic landmark in the country’s social and political context, and was being welcomed by
farmers who sought recognition by the state, it was not in itself increasing tenure security and assuring
the ongoing success of AD.
I thus question whether policymaking institutions in Bogota, like the National Land Agency,
UNODC, the Colombian Drugs Observatory and the National Programme for the Substitution of
Illegal Crops, are overplaying the significance of land titling in AD. For most farmers I interviewed,
the use of their land as collateral for a bigger bank loan was not the long-term solution out of illegal
agriculture. This was a surprise given what I had learnt from the elite interviews I carried out with
high level policymakers during my pilot enquiry. It was clear from my fieldwork that the temptation
to work as a raspachin in neighbouring municipalities still loomed like a solid option for many
farmers. Others were inclined to grow some coca leaf away from their home land, in an allotment-
style, tending to their crops intermittently. In most cases, the rate of return of coca leaf continued to
make it a highly attractive crop, compared to its alternatives. Many coca leaf producers are not land
owners, they rent land and split the cost of coca leaf harvests with the landowners, or work as day
labourers in coca leaf plantations. Many people like this had already followed coca leaf production
elsewhere. And those that had been left behind, were generally not endowed with bountiful fertile soil
or with easy connections to market towns, even if they owned their land. They believed it would take
more than a land title-based bank loan to make a living from their characteristically small land plots,
let alone to get their children through school or to be able to retire with less uncertainty of how to pay
for their meals.
313
Some farmers with land titles had confided that they were already resorting to coca leaf farming
elsewhere to compensate for their income shortage. Perhaps most surprising had been the cases of
farmers who had mentioned they had already paid back their title-based loan with money they had
earnt by farming coca leaf elsewhere. Their experience flashed amber lights that when land titling is
used as an AD policy, it can not only fail to prevent the displacement of coca leaf farming but may
create a need for farmers to return to illegal crop production. In some instances, land titles risked
increasing farmers' debt and reliance on coca leaf production to pay back their loans. By encouraging
farmers to take out loans to make their legal alternatives more productive, it is possible that this AD
policy can have the opposite effect of what it intends.
On the other hand, this is not to deny the significance of land titles nor to negate their value for
farmers. I recall the countless times people told me that they wanted a land title “to feel that their land
was theirs”, like a certificate of status and pride. Evidently the symbolic value of land titles should not
be dismissed. They are emblematic of a kind of agreement between the state and farmers who have
had to live under the authority of self, or extra-legal, governance provided by armed groups, for a long
time. They mark some sort of durable change in these communities where the state now forms part of
their future. Except that for many, this future will most likely not be accompanied by the prosperity,
that is the implied promise.
A second broad insight I derived from my research is that the state is not always the initiator of the
institutional change that occurs in AD. Of the farmers I met, those who still lived in places where coca
leaf had once thrived, were those who had witnessed the worse cases of violence. Through my
interviews I asked many of them to remember the riches and, sadly, some of the horrors they
experienced when they had produced coca leaf at home. They had been caught in the fighting between
armed groups and it was evident that for this reason they felt relieved they no longer farmed coca leaf
and that the state was now largely in command. According to the stories of most people in two
communities, they had tired of farming coca leaf when it became too costly to do so. This happened
when rival extra-legal groups entered the community and disrupted the supply chain of coca leaf
production. It became too risky for them to sell and farm the crop, as new buyers entered the scene
314
and used terrorizing tactics to force farmers to sell to them, while their usual buyers no longer trusted
them and adopted a similar behaviour to demand their loyalty. In this murderous atmosphere it
became impossible for farmers to profit, let alone save their assets. They no longer felt protected by
their former stationary bandit and instead felt exploited by different roving bandits. According to
them, it was only then that the state had entered the scene and had eventually rolled out an AD
programme.
Their joint experience suggests that the institution of extra-legal governance can sometimes evolve by
itself and facilitate a state takeover, and not because of it. This questions the universality of the
assumptions made by policymakers from UNODC, as was illustrated in my pilot enquiry. It also
offers a different point of view to studies that have found official state interventions to be what causes
communities to side with the state (Berman, Shapiro, and Felter 2011; Revelo Rebolledo and Garcia
Villegas 2010).
Moreover, it shows that not all armed groups are infallible in maintaining the illegal drug trade
efficient. This finding supports Cheng’s argument that extra-legal groups fail to protect their business
interests when they sense a shorter time frame within which they can extract rents (Cheng 2018). The
breakdown of governance that a stationary bandit provides can create an opportunity for the state’s
AD efforts to succeed. If the state searches for similar events in others coca leaf farming communities,
this could help select where its intervention is more likely to stick. I wonder if the government is
already privy to this. This reminds me of a conversation with a former AD state policymaker, now a
publicly-known intellectual, who is convinced that most AD interventions are being rolled out in
places where there is no longer any coca leaf (D. Rico 2017; 2020).
And yet again, it is worth pointing out that I only heard the stories of the individuals who had stayed
behind and were still residing in these communities. The rest had already fled, either following coca
leaf jobs or escaping the violence, or both. This brings us back to the question of what it means for the
state to “succeed” with AD. It seems that when there is more money to be made in coca leaf farming
elsewhere, AD will always be confined to communities populated by people who are resigned to a life
of rural poverty. However, this “second best” livelihood will not be attractive to the more
315
entrepreneurial individuals. It would therefore be complacent to assume that communities where
extra-legal governance has broken down will endorse state governance indefinitely simply because of
the immediate hiatus from violence the state can offer. Farmers are attracted to the freedom that a
peaceful environment brings, but the state must not lose sight of other types of personal freedom that
the illegal economy can provide them too when it is well governed by a stationary bandit.
This thought brings me to the third broad insight I make in this thesis, and that is that in these places
where AD has so far stuck, there are informal institutions that make legal agriculture less attractive to
women relative to illegal drug farming. During the coca leaf bonanza, women had been able to
challenge gender norms that had restricted them to unpaid and domestic work. The abundance of local
and paid jobs it brought with it made it easier for some women to break away from the confinement of
subsistence farming, of unpaid housework and of financial dependency on others.
So, when the illegal trade crumbled, women in these communities were left with fewer economic
opportunities because of the simple fact that it was customary for them carry out most, if not all, of
home production and subsistence farming. This was even though women’s rights to own property
continued to be well recognized and girls continued to be encouraged to attend school as much as
boys were, wherever possible. As the local economy contracted, the burden of unpaid work increased,
and it was women who had to assume it. Added to this, there were fewer chances for women to work
as cooks and cleaners on local coca leaf plantations. And to make matters more unequal, men were
better adjusted to the jobs that were available in agriculture because many of them were paid based on
physical strength and loads harvested, instead of hours worked, and required location flexibility. All
this meant that paid farming jobs were less compatible with many of women’s other unpaid
responsibilities.
In short, gender norms in these communities that had stopped growing coca leaf, operated to favour
men’s economic opportunities despite there being no significant gender inequalities in peoples’ legal
access to land property and to education. This not only introduces a new angle through which to
review women’s empowerment in communities that are undergoing AD, but challenges the premise
on which policymakers are currently assessing gender equality in these contexts (GIZ 2019; UNODC
316
& Colombia Ministry of Justice 2019; UNODC 2018b; OAS & CIM 2014). I believe this is a useful
contribution to AD policymakers’ understanding of illegal drug farming communities. It shows how
working in the illegal drug trade can remain the more attractive option for women who want to gain
financial independence, and it explains why women who are able to, may prefer to work in the illegal
economy than persevere with AD.
These inferences are based on the women and men that I was able to interview in all the places I
visited. No doubt the more entrepreneurial women, with fewer family commitments or less attachment
to land, had already left the rural communities to make a better living than what they could at home.
But the issue of how economic opportunities in contexts of AD are skewed to advantage men, remains
an important one for policymakers to resolve so that their AD efforts are not only gender-sensitive but
more effective and sustainable.
In very broad terms, my findings provide further empirical evidence of how informal institutions
such as de facto property rights, extra-legal governance and gender norms, shape economic outcomes.
And the key to yielding better economic outcomes in legal alternatives for coca leaf farmers is not
always down to formal policy, state actions and legislation, but determined by these farmers’ self-
governance institutions at the local community level.
Reflections on methods and ideas for further research
My method and data open questions that, as discussed above, large surveys cannot capture. I visited as
many former coca leaf farming communities as I was able to and spoke to as many people as I could
in each of these places, as well as in Bogota. I made the most of the time that was available to me, by
using an interview technique that gave me a degree of control over the topic and gave the interviewees
the freedom to take the discussion to their priority topics. It allowed me to discover existing issues in
the substitution of coca leaf, from different people’s perspectives, and gave me a notion of how
recurring they were. This open-minded, interdisciplinary approach paid off in that I garnered a wealth
of detailed data, that brought to life what other rigid surveys with pre-determined measures cannot.
317
As a result of using this methodology, I was able to pick up on two other important aspects of AD
which I was then able to pursue as additional research questions. During the pilot enquiry I noticed
the frequency with which armed groups are blamed for the resilience of the illegal drug trade. From
my institutional analysis standing point, I became curious about how true this assertion was, and
decided to find out how people at the grassroots level considered the behaviour of armed groups to
have influenced their choice to farm illegal drugs. My fieldwork confirmed the opposite i.e., that
farmers in these communities were not forced to grow coca leaf by armed groups in any way but had
opted to do so because it made better business sense. And in a couple of communities, when armed
groups had no longer been able to guarantee a safe environment for them to farm coca leaf profitably,
they had freely chosen to stop doing it.
Then, in the early stages of data collection, I came to the question of women’s situation in former
coca leaf farming communities almost accidentally. Once I was sensitized to this theme, it became
more evident as my research went on that it was an area that merited careful attention. So, I made use
of my research methodology to collect data on the changes women had experienced in their day to day
lives after their community stopped farming coca leaf and adapted my interviews to also collect data
on their situation. In addition, I picked up on how men perceived women’s economic roles to be
different from their own. This revealed the more subtle ways in which women face a shortage of
economic opportunities in ways that men do not, when their communities stop farming coca leaf.
As it stands, my study relies on what farmers expressed at face value in communities and I
complemented this with further background research. Analysis at the individual level sheds light on
whether the theoretical mechanism by which land titles persuade people to stop growing coca leaf
holds true. This unit of analysis also reveals how women experience their economic opportunities
differently after coca leaf is no longer farmed in the community. An analysis of clusters of
individuals, at the community level, is useful for explaining why communities as a whole stop
farming coca leaf. This data has been invaluable to open new questions about the reasons why coca
leaf farmers decide to stop or to continue working in illegal drug production. If I had to give greater
validity to my impressions from the standpoint of quantitative research, I would replicate my study
318
with a random sample of former coca leaf farming communities that includes those that have
continued or never stopped farming it. This would allow me to compare the situation of those living in
different contexts and would facilitate greater generalization from my sample about the situation of all
coca leaf farmers in Colombia.
My study fills a gap in what is currently known about how land titles influence illegal crop farmers’
decision-making but is limited to places where farmers had not been forced to start or continue to
grow coca leaf by extra-legal groups, and where land titling had been offered specifically as an AD
incentive. As such, it reflects the situation of coca leaf farmers living in relative proximity to roads
and where other AD benefits were available. The use of random sampling, as described above, may
help us to further understand the significance of land titling in farmers’ decisions to stop or continue
farming coca leaf, in isolation from other and no AD subsidies. It would facilitate a comparison
between veredas that were included in the land titling programme, and some of their neighbouring
veredas where communities had also stopped farming coca leaf around the same time, and which
were in receipt of other AD subsidies, but which were not selected for land titling. And this would
disentangle how significant land titling is from other changes that were being introduced in the
communities, including greater state presence, surveillance, investment in roads and subsidies for the
equipment needed for alternative legal land use. Future research in this area could also look at the
long-term effect of land titling on farmers’ decision to stay out of illegal drug production. Perhaps a
research team could return in a few years’ time to the places I visited and ask those who acquired their
land titles if they had to resort to farming coca leaf elsewhere, or to work as raspachines at some
point, to make ends meet. This would give an indication of the extent to which land titling helped AD
efforts by increasing land productivity or by displacing coca leaf farming elsewhere.
I welcome policymakers to question whether armed group presence always prevents AD from
succeeding, and whether the state is always what drives communities to convert to the legal economy.
My thesis highlights exceptions to these widely held beliefs, based on data from three communities
that had been former coca leaf havens, thanks to their climate and soil characteristics, and which had
been governed by a single armed group at that time. Those wanting to enrich our understanding of the
319
wider role extra-legal governance plays in facilitating AD, could increase the scale of my research and
visit more than three former veredas where coca leaf once grew abundantly but no longer did. They
could also ask people more directly about how the armed groups had behaved prior to their decision to
stop farming coca leaf. These research enhancements would not only present more data on the
mechanism by which extra-legal governance breaks down but could indicate how prevalent the
phenomenon is.
Quantitative researchers interested in building on these findings could use a random sample of all the
communities with similar characteristics i.e., where coca leaf farming once flourished or has
continued to flourish, and where AD has been introduced - including places where AD has succeeded
as well as those where it has failed. Although not without its logistical difficulties, this type of
research could help to compare whether farmers’ reasoning is consistent in different scenarios. It
would also help to observe more carefully how the stationary bandit, assuming there is one, behaves
alone and in the face of rival extra-legal groups. Lastly, this kind of study could help us verify
whether the possible correlation between rivalrous extra-legal groups and successful AD, that I point
out in this thesis, is strong enough to merit further and more rigorous analysis.
And finally, my research sheds light on the importance of coca leaf farming for women’s economic
opportunities can in these communities and signals the less measurable ways in which women can be
adversely affected when their communities decide to stop growing coca leaf. If I had set out to study
this phenomenon from the beginning, I would have used a broader purposive sampling method. But
given I came to this research question at a later stage, I used purposive sampling based on the criteria
I required for the first research question. That is, people who no longer had illegal crops, who had
access to land where illegal crops used to be grown, who had applied for or had already received AD
benefits such as a land title, and who had stopped working for illegal crop farmers (including
domestic workers and coca leaf pickers). My insights are therefore derived from the women who met
these criteria. Doubtlessly there are many more stories from women who were affected by the coca
leaf bonanza, and who never worked directly with coca leaf, that would give a fuller picture of how
the end of the illegal economy can redress advances in economic gender equality.
320
I therefore invite further research into this topic that selects a different sampling method and that
designs the interviews exclusively about this topic, asking women more boldly about their personal
lives and household bargaining power. I suggest considering the experiences of women in
communities that still farm coca leaf and those which have returned to coca leaf farming after
attempting to substitute their illegal crops. Interviews could enquire about how the prospect of
economic opportunities featured in their decision to continue or return to the illegal economy, and
how this influenced changes in gender norms and in their household bargaining power. These
measures could yield much more detailed data from which to draw conclusions, and point to some
ensuing issues for AD.
In the timeframe and with the funding that was available to me, however, a study that included these
extra steps would have been unfeasible. It would have been ribbed with logistical and ethical hurdles
to be able to access and interview people in these places, especially in those where armed groups were
still present. Furthermore, a study of this scale would have delivered a colossal amount of data, that
would have been unrealistic for one researcher to analyse in the time frame of a PhD. My data may
come from a limited number of communities that met the criteria for research, but my thesis is
nonetheless a valuable policy-relevant exploratory study. I had the rare opportunity to hear the stories
of around 80 former coca leaf farmers and analyse in depth how they experienced three forms
informal institutions that influenced their choice to work in the illegal drug trade, or not. AD
policymakers know little about these, and their accounts thus make a significant contribution to the
field of drug policy.
Summary
The main take-home message of this thesis is that there are informal institutions influencing farmers’
participation in illegal drug farming, and policymakers may not be aware of them perhaps because of
the in-built assumptions their surveys rely on. Thanks to my methods, I have identified three
institutions in this study, that help explain the resilience of coca leaf farming in most of the country.
As policymakers strive to make AD efforts more effective, it is vital they understand the complexity
321
of these institutions and the problems that come with simply trying to replace informal with formal
institutions in the hope that they will stick because they are legal and therefore superior. The status of
‘legal’ or ‘illegal’ is relatively insignificant for people whose priority is to maximise their resources to
break out of poverty (Dua 2019). Individuals will cooperate to change or maintain institutions
depending on their relative stake in the situation. Until then, they will hold onto the institutions that
render the best outcomes.
I am certainly not the first to highlight the need to create greater economic incentives for farmers to
choose to abide by the law (Gaviria, Mejía, and Weiskopf 2017; Thoumi 2012; Dube and Vargas
2013). I suppose this study helps to stress the crucial importance that state efforts to make legal
alternatives more profitable in communities like these, are continued and realistic. It is also critical to
measure the scale of the displacement of illegal drug production elsewhere when claiming AD has
been successful. Without ongoing economic incentives for farmers to stay in the legal realm, it will be
less costly for them to return to the illegal one when the opportunity arises. And overestimating the
changes that AD subsidies and state institutions, such as land titles, will bring to the community, can
create false expectations among farmers, which eventually undermines their trust in the state and in
their own ability to make a living in the legal economy.
This is particularly the case for women, who may have additional incentives to return to the illegal
trade either at home or elsewhere, if they are able to relocate. AD programmes have made some effort
to address women’s financial situation by underlining the importance of equal property rights to land
and access to education and supporting women’s associations and cooperatives. However, if AD
interventions are to be assessed not only in terms of the reduction of illegal crops, but also in terms of
women’s empowerment and gender equality, as has been the intention of many national AD strategies
(UNODC 2015, 88), then there must be greater discussion of the gender differences in people’s
incentives to work in the illegal drug trade. Failure to acknowledge this runs the risk that AD is short-
lived in many rural communities when women experience greater scarcity of economic opportunities
and a regression of gender norms.
322
Finally, my thesis poses the possibility that institutional change can happen spontaneously in favour
of legality in places where extra-legal governance breaks down. Indeed, there is need for greater
analysis of this phenomenon than what this study has been able to provide. Observing the dynamics
between extra-legal groups can help the state identify opportunities for where AD efforts are likely to
be more effective and lasting. If the state focuses on being ready to swiftly introduce its institutions,
including land titles, and making sure legal alternatives to illegal drug crops remain the more
attractive livelihood, to men as well as women, in these communities, AD is more likely to endure.
323
Appendices
Appendix I: Original interview structure and topics
Main
Topics
Sub-topics Key Questions
Introduction About me
About my research
Duration of interview
Happy to go on record?
If there is anything you would prefer to say off the record that is fine, just let me
know
What is a
property
right
Is there a piece of land you consider yours?
What do you mean when you say it’s yours?
Can you choose how to use it? Can you keep all the profits? Can you pass it on?
If you leave it, is it still yours?
Can you exclude others from using it? At what cost?
324
Land
productivity
Alternatives
to illegal
crops
In your view, why do communities farm illegal crops?
How much income does coca/ marihuana provide per month?
What proportion of profit is given to farmers?
What are the legal alternatives to coca in this area?
Is coca better than the alternatives?
How much in wages would someone need to raise to make coca inoperable?
How much does a bag of coca cost?
How many people per hectare does coca farming employ? How about the
alternative?
How much land would legal agricultural alternatives to coca require? (plot size)
What alternative project would you propose for your land?
How much income do you expect that to provide per month?
How much credit would you need to make your business as profitable as coca?
How much is the state offering?
How long have there been AD efforts here?
Access to
credit
Is there credit available to you?
Does land titling may present new opportunities to access credit? (land as
collateral)
Given the chance, would you want to apply for credit and use your land as
collateral?
Would the credit applied for be sufficient to overcome the barriers to
development?
Does membership of a cooperative help? How?
325
Allocation of
resources
Has extra-legal land ever been treated as a free-for-all resource?
Do land users spend resources (time and labour) on self-enforcement of their
rights? Why?
Does/ would a land title make a difference? E.g., increasing safety and
protection? in freeing up time for more labour-intensive farming? Or boosting
quantity and quality of farming produce? In changing levels of displacement?
And levels of violence?
Have you ever felt at risk that the benefits of your land could be appropriated by
others?
Have you made big investments in improving your land for the long term? E.g.
expensive perennial crops, infrastructure, irrigation, soil erosion and
reforestation? Why or why not?
Do you fear of that investment in your land asset may go to waste?
Investment Have any private firms invested in the community? Is it for land improvement,
market mechanisms for scaling up agricultural production? Mega agricultural
projects or mineral resource extraction?
Do they own land in the area? Do they want to own land in the area?
Do they favour large land holdings?
How much employment does this provide?
How frequent are land market transactions?
Have formal land titles made any difference?
Would/ have they made you sell or rent your land?
Does it make a difference if you own the land on how productive it is?
Does small-scale farming make coordinated farming harder?
Does small-scale farming attract less investment from foreign companies?
Enforcement
of land
rights
SYSTEM
Property
Rights
Who owns this land?
What is the formal status of this land?
Does it matter? In practice, do things work differently? How?
Can the community exclude others from appropriating their land?
326
Land USE Are there clear boundaries to a land plot?
Who decides what land is used for? The state? The farmer? The community?
What happens if you do otherwise? Can you give me examples?
Do you know of anyone who has ever been expropriated from their land?
What are the consequences for rule disobedience? E.g., physical sanction,
implicit shame, or guilt
Land
PROFITS
Can anyone appropriate land anywhere? How? Who decides?
Do land users pay anyone to use their land? E.g., rent, tax. How much? How
often?
Who profits from its returns?
How are the returns split? i.e., between the farmer, taxes, the community, or
extra-legal group dividends
What happens if you do otherwise? Can you give me examples?
What are the consequences for rule disobedience? E.g., physical sanction,
implicit shame, or guilt
Land
EXCHANGE
Is most land owned or rented?
When and how was this land last sold?
How much is it worth?
Has it been exchanged formally or informally?
What happens if it is sold in a different way? Can you give me examples?
What are the consequences for rule disobedience? E.g., physical sanction,
implicit shame, or guilt
How land
TITLING is
administered
and
experienced
What has been the process and impact of impact of land titling practices?
Why do people want the formal titles or not?
What aspects of the property right were you missing that will be given through a
title?
Is it about selling?
Why not go for a collective right?
Are there any free-riding problems?
Are land titles easy to enforce?
327
Social
acceptability
of the
distribution
of property
rights
Tensions in
the
community
What are the sources of authority in the community?
Are there any external sources of authority over the community? How is this
experienced?
Is there peer pressure? Are extra-legal actors influential?
Have there ever been any tensions among different community members?
Is there general satisfaction with the current way that land is distributed?
Are there tensions between the community and the state?
Are there any tensions with large-scale capital private investors?
Are there tensions between the community and extra-legal groups?
Are any of these tensions connected to land? How?
How are these tensions resolved?
How many of these problems does a land title resolve?
What is your perception of equitable rights to land?
Warm down About the
research
Is there anything you would like to add?
Thank you! - Would you like to receive a copy of the theses when it’s finished?
Do you know other people who might be able to share relevant experiences?
328
Appendix II: List of Interviewees
Below is the full list of the details of the 90 interviews that were carried out from November 2017 to
February 2018.
Department - Municipality
Vereda Name/ Code
Sex Interview Duration
Age at time
Joint Date
Caquetá - Valparaiso
Santiago de la Selva
0077 Male 38:46 72 30/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0075 Male 38:19 67 29/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0078 Female 34:23 52 30/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0079 Male 1:04:47 55 30/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0068 Male 1:01:25 72 28/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0072/3 Male 1:00:15 70 29/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0069 Female 56:13 62 28/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0071 Female 1:20:17 42 28/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0076 Male 29:59 54 30/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0070 Male 33:20 55 28/11/2017
Caquetá - Valparaiso
Santiago de la Selva
0074 Male 31:22 52 29/11/2017
Cauca - Balboa
La Palma 0108 Male 31:44 46 30/01/2018
Cauca - Balboa
La Palma 0110 Female 17:10 40 30/01/2018
Cauca - Balboa
La Palma 0103 Male 28:20 63 29/01/2018
Cauca - Balboa
La Palma 0109 Male 11:11 65 30/01/2018
Cauca - Balboa
La Palma 0111 Male 23:38 63 30/01/2018
329
Cauca - Balboa
La Palma 0112 Male 16:00 42 30/01/2018
Cauca - Balboa
La Villa 0106 Male 15:30 42 29/01/2018
Cauca - Balboa
La Palma 0113 Female 12:09 35 30/01/2018
Cauca - Balboa
La Villa 0105 Male 11:30 50 29/01/2018
Cauca - Balboa
La Villa 0104 Male 41:38 45 29/01/2018
Cauca - Balboa
La Villa 0107 Female 23:34 53 Joint 29/01/2018
Cauca - Balboa
La Villa 0107 Male 23:34 55 Joint 29/01/2018
Cauca - Balboa
La Villa 0102 Female 32:00 42 29/01/2018
Cauca - Balboa
La Villa 0101 Male 30:57 39 29/01/2018
Cauca - Mercaderes
Esperanzas del Mayo
0117 Male 26:14 66 31/03/2018
Cauca - Mercaderes
Carboneros 0125 Female 22:03 49 01/02/2018
Cauca - Mercaderes
Esperanzas del Mayo
0122 Male 18:36 54 31/03/2018
Cauca - Mercaderes
Esperanzas del Mayo
0123 Male 22:11 40 31/03/2018
Cauca - Mercaderes
Sombrerillo 0116 Male 31:14 65 30/01/2018
Cauca - Mercaderes
Sombrerillo 0114 Male 15:22 39 30/01/2018
Cauca - Mercaderes
Esperanzas del Mayo
0119 Male 14:03 76 31/03/2018
Cauca - Mercaderes
Esperanzas del Mayo
0121 Female 26:35 33 31/03/2018
Cauca - Mercaderes
Carboneros 0127/8 Male 26:43 64 01/02/2018
Cauca - Mercaderes
Carboneros 0126 Male 22:56 48 01/02/2018
Cauca - Mercaderes
Esperanzas del Mayo
0124 Female 26:28 42 31/03/2018
330
Cauca - Mercaderes
Sombrerillo 0115 Male 30:10 46 30/01/2018
Cauca - Rosas
El Porvenir 0138 Male 30:08 52 02/02/2018
Cauca - Rosas
El Porvenir 0133 Female 11:24 38 02/02/2018
Cauca - Rosas
El Porvenir 0135 Male 26:09 48 02/02/2018
Cauca - Rosas
El Porvenir 0134 Female 25:31 45 02/02/2018
Cauca - Rosas
La Soledad 0132 Male 25:34 62 01/02/2018
Cauca - Rosas
La Soledad 0130 Male 30:41 37 01/02/2018
Cauca - Rosas
El Porvenir 0137 Female 15:48 55 02/02/2018
Cauca - Rosas
La Soledad 0129 Male 39:00 38 01/02/2018
Cauca - Rosas
El Porvenir 0136 Female 06:16 23 02/02/2018
Cauca - Rosas
La Soledad 0131 Male 20:32 43 01/02/2018
Nariño - Linares
Linares * 0086 Female 23:43 N/A 23/01/2018
Nariño - Linares
La Tola 0083 Female 21:28 34 22/01/2018
Nariño - Linares
Llano Grande Alto
0085 Female 54:17 34 23/01/2018
Nariño - Linares
La Tola 0080 Female 22:07 55 22/01/2018
Nariño - Linares
La Tola 0081 Male 28:30 60 22/01/2018
Nariño - Linares
Linares 0087 Male 06:13 48 23/01/2018
Nariño - Linares
La Tola 0084 Female 35:42 32 22/01/2018
Nariño - Linares
La Tola 0082 Female 10:03 38 22/01/2018
Nariño - Los Andes
Sotomayor 0095/6 Male 52:36 62 25/01/2018
331
Nariño - Los Andes
Huilque 0092/3 Male 38:52 50 25/01/2018
Nariño - Los Andes
San Vicente 0099 Male 13:01 73 26/01/2018
Nariño - Los Andes
Huilque 0094 Male 49:30 48 25/01/2018
Nariño - Los Andes
Sotomayor 0097 Male 22:55 69 25/01/2018
Nariño - Los Andes
El Alto 0091 Female 06:56 75 24/01/2018
Nariño - Los Andes
El Alto 0089 Female 27:43 35 Joint 24/01/2018
Nariño - Los Andes
El Alto 0089 Male 27:43 33 Joint 24/01/2018
Nariño - Los Andes
El Alto 0090 Female 19:52 36 24/01/2018
Nariño - Los Andes
Sotomayor 0098 Male 46:52 37 26/01/2018
Nariño - Los Andes
El Alto 0088 Female 23:44 40 24/01/2018
Putumayo - Puerto Asis
La Carmelita
N/A Male Written Notes
55 22/11/2017
Putumayo - Puerto Asis
La Pradera 0064 Male 26:45 53 23/11/2017
Putumayo - Puerto Asis
La Pradera 0062 Male 39:27 39 23/11/2017
Putumayo - Puerto Asis
La Carmelita
N/A Male Written Notes
41 23/11/2017
Putumayo - Puerto Asis
Agua Negra 0067 Female 47:33 42 24/11/2017
Putumayo - Puerto Asis
La Carmelita
0061 Female 1:09:45 51 22/11/2017
Putumayo - Puerto Asis
La Carmelita
0058/9/60 Female 54:59 58 Joint 22/11/2017
Putumayo - Puerto Asis
La Carmelita
0058/9/60 Make 54:59 62 Joint 22/11/2017
Putumayo - Puerto Asis
Santa Lucia 0065 Male 1:05:45 53 24/11/2017
Putumayo - Puerto Asis
La Pradera 0063 Female 46:03 46 23/11/2017
332
Putumayo - Puerto Asis
Puerto Asis 0066 Female 57:16 24 24/11/2017
Putumayo - Puerto Asis
La Carmelita
0057 Female 1:07:00 35 22/11/2017
Putumayo - Valle del Guamuez
El Placer N/A Male Written Notes
55 20/11/2017
Putumayo - Valle del Guamuez
Brisas del Palmar
0046 Male 27:17 76 20/11/2017
Putumayo - Valle del Guamuez
Alto Palmira
0048/9 Female 57:48 43 21/11/2017
Putumayo - Valle del Guamuez
Brisas del Palmar
0045 Male 53:46 32 Joint 20/11/2017
Putumayo - Valle del Guamuez
Brisas del Palmar
0045 Female 53:46 79 Joint 20/11/2017
Putumayo - Valle del Guamuez
Los Guaduales
0052 Male 52:43 75 21/11/2017
Putumayo - Valle del Guamuez
Los Guaduales
0052 Female 52:43 75 21/11/2017
Putumayo - Valle del Guamuez
Miravalles 0050 Male 1:04:27 60 21/11/2017
Putumayo - Valle del Guamuez
Brisas del Palmar
N/A Female Written Notes
77 20/11/2017
Putumayo - Valle del Guamuez
El Cairo 0047 Female 56:21 51 20/11/2017
Putumayo - Valle del Guamuez
Miravalles 0051 Female 40:57 46 21/11/2017
Putumayo - Valle del Guamuez
Los Guaduales
0055 Female 49:11 30 21/11/2017
Putumayo - Valle del Guamuez
Oasis 0044 Male 41:14 56 20/11/2017
333
Putumayo - Valle del Guamuez
La Pradera 0056 Male 1:08:29 42 21/11/2017
334
Appendix III: Key informants in pilot enquiry
Interview Organization Date
Francisco Thoumi (skype) Author of various books on drug policy in Colombia, Alexandria, Virginia
14/07/2017
Jorge A Restrepo Director of Resource Center for Conflict Analysis (CERAC), Bogota
27/07/2017
Thomas Mortensen Country Director of Christian Aid Colombia, Bogota 27/07/2017
Fredy Orlando Montealegre Martinez
Coordinator of GIT Land Planning, Mapping Division, Instituto Geografico Agustin Codazzi (IGAC), Bogota
28/07/2017
Diana Marily Rodriguez Administrative Director, Espinal Mayorship, Projects Bank, Land Planning Division, Espinal, Tolima
01/08/2017
Cesar Leonardo Picon Arciniegas
Secretary General of Rural Development, Ibague Mayorship, Ibague
02/08/2017
Marietta Bucheli Gomez Director of Faculty of Environmental and Rural Studies, Javeriana University, Bogota
04/08/2017
Maria Fernanda Sañudo Pazos Research Fellow, Instituto Pensar, Javeriana University, Bogota
04/08/2017
Fabio Hector Santos Duarte Advisor Liaison Global Program on Drug Policy and Development “Protección del Bosque y Clima / REDD+” at German Aid Agency (GIZ), Bogota
Marisol Gomez Peace Process Editor, El Tiempo newspaper, Bogota 08/08/2017
Daniel Rico Drug Policy Researcher, Fundación Ideas para la Paz (FIP), Bogota
08/08/2017
Herbert Infante & Oscar Ricardo Santana López
Advisor to the Directorate of Drug Policy on Drug Supply Reduction Issues, Observatory of Drugs of Colombia (ODC), Ministry of Justice, Bogota Information Manager at the Observatory of Drugs of Colombia (ODC), Ministry of Justice, Bogota
09/08/2017
Pedro Arenas Director of the Observatory of Crops and Cultivators Declared Illicit (OCCDI GLOBAL), Bogota
09/08/2017
Santiago Gonzalez Advisor to the Directorate of Policy Against Drugs, Ministry of Justice, Bogota
09/08/2017
Javier Soto & Maria Fuentes Advisors of Social and Technical Directorates, Land Restitution Unit (URT), Bogota
10/08/2017
Johanna Herrera Arango Lecturer, Observatory of Ethnic and Campesino Territories, Javeriana University, Bogota
10/08/2017
Andres Herreno ART - Agencia de Renovacion del Territorio, Asesor de Direccion, Bogota
10/08/2017
335
Sergio Coronado Former Deputy Director of Jesuit database (CINEP), Bogota
11/08/2017
Gabriel Tobon Senior Lecturer, Observatory of Ethnic and Campesino Territories, Javeriana University, Bogota
14/08/2017
Rafael Navarro Operational Coordinator of Early Warnings, Public Ombudsman (Defensoria del Pueblo), Bogota
14/08/2017
Eduardo Diaz Director, Comprehensive National Program for the Replacement of Illicit Use Crops (PNIS), Bogota
14/08/2017
Ana Maria Ibanez Professor of Economics, Universidad de los Andes, Bogota
15/08/2017
Carol Torres Security Advisor, Norwegian Refugee Council, Bogota 15/08/2017
Cesar Ortiz Director, Department of Rural and Regional Development, Universidad Javeriana, Bogota
16/08/2017
Miguel Samper Director, National Land Agency (ANT), Bogota 16/08/2017
Bo Mathiasen Country Representative, United Nations Office on Drugs and Crime (UNODC), Bogota
17/08/2017
Andres Murcia Adviser to the Chief Planning Advisory Office, Special Assets Company (SAE), Bogota
17/08/2017
Eduardo Medina Country Director, International Office for Migration (OIM), Bogota
18/08/2017
Felipe Fonseca Fino Director General, Rural Land Planning and Adaptation of Land and Agricultural Land Uses Unit (UPRA), Bogota
18/08/2017
Rafael Gonzalez Gordillo National Director, Solidarity Organizations (UAEOS), Bogota
18/08/2017
Padre Fernan Gonzalez Senior Researcher, Jesuit database (CINEP), Bogota 22/08/2017
Jose Ilario Gomez & Diana Velasquez
Advisor to Juan Camilo Sanchez (director of Legal Land Management Land Titling Programme), National Land Agency (ANT), Bogota Advisor to Javier Florez (director of Formalize to Substitute Land Titling Programme) National Land Agency (ANT), Bogota
22/08/2017
Ana Maria Puyana Director, National Association of Campesino Reserve Areas (ANZORC), Bogota
25/08/2017
Juan Fernando Lucio (skype) Director, Fundación Paso (Paz Sostenible), Cali 29/08/2017