Wheels within Wheels: Poverty, Power and Patronage in the Cambodian Migration System

This article was downloaded by: [King's College London]On: 01 February 2015, At: 08:22Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

The Journal of Development StudiesPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/fjds20

Wheels within Wheels: Poverty, Powerand Patronage in the CambodianMigration SystemLaurie Parsonsa, Sabina Lawreniuka & John Pilgrimb

a Department of Geography, King’s College London, UnitedKingdomb Royal University of Phnom Penh, CambodiaPublished online: 05 Sep 2014.

To cite this article: Laurie Parsons, Sabina Lawreniuk & John Pilgrim (2014) Wheels within Wheels:Poverty, Power and Patronage in the Cambodian Migration System, The Journal of DevelopmentStudies, 50:10, 1362-1379, DOI: 10.1080/00220388.2014.940915

To link to this article: http://dx.doi.org/10.1080/00220388.2014.940915

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Wheels within Wheels: Poverty, Power andPatronage in the Cambodian Migration System

LAURIE PARSONS*, SABINA LAWRENIUK* & JOHN PILGRIM***Department of Geography, King’s College London, United Kingdom, **Royal University of Phnom Penh, Cambodia

(Final version received May 2014)

ABSTRACT This article presents evidence for an adjusted and refocused systems theory of labour migration inCambodia. Specifically, it seeks to highlight first, how migration in Cambodia may be understood as a multi-scalar phenomenon characterised by pragmatism and flexibility; secondly, it emphasises the undergirding role oftraditional rural norms in shaping and mediating the systematic process of labour movement; and finally, itpresents evidence concerning how these structures constitute a vessel of social change, not only from urban torural, but also from the rural to urban. In this way, a picture is presented of Cambodian migration as anadaptable, but nevertheless highly patterned process which is rapidly reordering the Kingdom’s cities and villagesalike.

1. Introduction

This article presents findings derived from a large-scale, mixed methodology research project con-ducted in Phnom Penh and a selection of villages in neighbouring provinces between 2008 and 2011.The results gleaned suggest that contemporary migratory systems thinking is unsuited to the complex-ity of migration in twenty–first-century Cambodia, a country in which explosive economic growth(World Bank, 2012) and worsening climate pressures (Davies, 2010; Oeur, Sopha, & McAndrew,2012) have prompted human migration on a scale not seen since the fall of the Khmer Rouge. Therapidity and volume of this change has produced an integrated rural–urban migration system shapedby culture, catalysed by livelihoods pressure and fuelled by the economy.

As such, the observed reality of rural–urban migration in Cambodia is one of simultaneousstructural rigidity and fluidity, with clear demographic patterns and structured behaviours underscoredand facilitated by a remarkable flexibility to circumstance. Migrants do not merely assess positive andnegative factors across a given time frame, as posited by Todaro (Harris and Todaro, 1970; Todaro,1969), but continually, as rural pressures compete or align with urban opportunities according to adynamic calculus of need incorporating both individuals and households.

Recently, this interdependency of traditional and non-traditional sectors, as mediated primarily byinformal, socio-structural forces, was thrown into sharp relief by the remarkable role they played inameliorating the effects of the global economic downturn (CDRI, 2012). Despite the garment sectoralone being forced to absorb a loss of over 53, 000 jobs between 2008 and 2009 (Arnold, 2013: 3), theanticipated livelihoods crisis did not materialise (Cambodia Development Resource Institute [CDRI],2012). Employment returned to pre-crash levels by the beginning of 2012 (Arnold, 2013, p. 3).

Nevertheless, the specific processes which facilitate such resilience and the manner in which theyinteract with both the modern sector labour market and traditional social structures have thus far

Correspondence Address: Laurie Parsons, Department of Geography, King’s College London, Strand, London, WC2R 2LS.Email: [email protected]

The Journal of Development Studies, 2014Vol. 50, No. 10, 1362–1379, http://dx.doi.org/10.1080/00220388.2014.940915

© 2014 Taylor & Francis

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

received little attention. The dynamic nature of the migrant labour market, for instance, or how villagelevel structures of patriarchy determine and are determined by the urban economy or the severity ofclimate shocks, have rarely been tackled. The research presented here constitutes an initial effort torefocus attention upon these smaller processes and to relate them to the Cambodian labour migration atlarge: to examine, as it were, the wheels within the wheels.

2. Theoretical Approach

The heavy debt owed by classic models of human mobility, such as Zelinski (1971) and Lee (1966) toparadigmatic economic models such as Thompson’s Demographic Transition (1929), Rostow’s Stagesof Growth (1960) and Lewis’s Dual Sector (1954), have seen migration thinking closely mirror theshifting economic paradigms of development, with the modernisation (for example, Rostow, 1960),dependency (for example, Frank, 1969) and neo-classical models all having left significant legacieswithin the migration literature. Such models have tended to share a conception of migration as a one-way process, strongly correlated with positive growth, but far more weakly associated with economicslowdown or recession in the sense that exit or re-entry to the modern sector is rarely considered.

In seeking to address this unilinearity, Mabogunje’s Migratory Systems Theory (1970) offers a morestructurally oriented viewpoint to the analysis of economic forces at the micro and meso scales. AsMabogunje explains:

within the systems framework, attention is focussed not only on the migrant but also on thevarious institutions (sub systems) and the social, economic and other relationships (adjustmentmechanisms) which are an integral part of the migrant’s transformation. (Mabogunje, 1970, p. 5)

Otherwise put, systems theory views labour movement as a process related in a persistent and dynamicmanner to the society and culture within which it takes place.

Since Mabogunje’s landmark publication, however, systems thinking has largely failed to progress‘beyond the usual focus on networks [in favour of] emphasising the importance of flows of informa-tion and ideas’ (De Haas, 2010, p. 1593). Furthermore, it has a tendency to ignore ‘various othercontextual feedback mechanisms’ (De Haas, 2010, p. 1593), particularly with respect to the analysis ofsystems as structures capable of fulfilling more than one role, such information sharing and hierarchyreinforcement. With some exceptions (for example, Kuhn, 2003), these remain under-investigatedwithin the literature, which has for decades tended to rely ‘more on the traditional referents ofmigration studies than on the formal propositions of systems analysis’ (Fawcett et al., 1989, p. 672).

Thus, although the notion of the migration system has become a staple within migration literaturereviews, the inherent complexities of its investigation have seen it ‘cast adrift from its conceptualmoorings’ (Bakewell, 2013, p. 1) in practice. Attempts at empiricism have therefore resulted in‘truncated versions’ of the systematic models delineated in theory, generally ‘either conflated withmigrant networks or elevated to the heights of macro-level abstraction, which divorces them from anyempirical basis’ (Bakewell, 2013, p. 1).

Furthermore, this investigative foreclosure is a problem not only of scale, but also of the categories ‘rural’and ‘urban’, which retain a conceptual grip upon migration research even in the face of a wideningappreciation (see, for example, Rigg, 2001, 2012) of their definitional porosity. In particular, this manifestsin a disjuncture between conceptions which emphasise the persistence of financial and social networksspanning urban and rural areas (Kuhn, 2003; Mabogunje, 1970; Potts, 2010; Watts, 1983), and thosefocussing upon the emergence of new, hybrid societies emerging in the urban space (for example, Erel, 2010).

This article begins to address this shortcoming by demonstrating, following Cook, Hall, and Larson(2012), the value afforded by multi-scalarity to a systems framework which, in accordance withMabogunje’s (1970) concept should seek to incorporate both ‘micro and macro elements, allowingsubsystems to nest within larger systems’ (King, 2012, p. 140). In turn, the ability to transcend socialand institutional scales allows the framework to accord more closely with Bylander’s (2013) view of

Poverty, power and patronage in the Cambodian migration system 1363

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Cambodian labour migration as a holistically social process, firmly embedded within the Cambodiancontext. In doing so, it serves also to heed De Haas’s (2010) and Bakewell’s (2013) calls for a modelof migration which bridges not only the scalar, but also the rural–urban divide.

As such, this article presents a view of migration systems characterised by three theoretical tenets. First,that population movement in Cambodia is driven by macro-scale conditions, but mediated via household-scale norms and networks in a pragmatic and flexible manner. Second, that thinking in terms of rural–urbanlinkages is insufficient, where a preferable conception would be of a fully integrated system spread acrosstwo or more geographical locations. Finally, this integrated model leads to the third proposition: that socialfeedbacks are a two-way process. Ergo, just as the rural is increasingly subject to urban influence, so too isurban migrant society continually recrafting itself in the model of the rural.

The aim herein is, therefore, to provide evidence directed towards fleshing out these areas ofenquiry Fawcett et al., 1989 which have been highlighted (by, for example, Bakewell, 2013; DeHaas, 2010; Fawcett et al., 1989) as lacunae within the migration systems literature. Beyond this,though, it is to demonstrate that it is not only networks which link these various subsystems, scales andflows, but the norms which underpin them. Otherwise put, the intention here is to demonstrate first, therole played by pragmatism and flexibility within the Cambodian migration system; second, the mannerin which this system is mediated and constructed according to a framework of traditionalism, duty andhierarchy; and third, the bi-directional nature of social feedbacks resulting from this relationship.

3. Methodology

The study was funded by the International Development Research Centre of Canada (IDRC) and conductedunder the auspices of the Graduate Programme in Development Studies of the Royal University of PhnomPenh. Fieldwork was conducted throughout between 2008 and 2011 at a primary site within Teuk Thla, anadministrative sangkat (commune) located in the western area of Cambodia’s capital, Phnom Penh. This sitewas selected from an initial three candidates put forward by a preliminary key informant, a former garmentworker now working for a charitable organisation related to the garment industry.

Following a preliminary transect walk and conversations with the inhabitants of each, it was agreedupon that Teuk Thla was the more representative of Phnom Penh’s migrant enclaves in general.Thereafter, in order to facilitate as holistic and detailed a perspective as possible, the study focused ona sub-area of the sangkat centred around Teuk Thla market and bounded by three main roads: Phlauv(streets) 110, 1019 and 2004 (see Figure 1), indicated by Teuk Thla’s inhabitants to be the generallyunderstood extremities of the migrant enclave.

The research programme was divided into four distinct but related phases of investigation. Theinitial phase of data collection comprised a large-scale quantitative survey to determine the basicdemography of the site. All available residences within the site boundaries as defined above weresurveyed, on a door-to-door basis, by a team of four researchers, working in pairs, who asked simplequestions in Khmer and recorded responses in English where possible. In total, this micro-censusreached 461 physical households and provided basic data on 1,642 inhabitants, who were each askedto give their age, gender, occupation, province of origin and date of arrival to Teuk Thla.

A second phase of data collection followed, in which a longer and more detailed quantitative ques-tionnaire survey was delivered to 195 of these initial respondents. These 195 respondents were randomlyselected from four occupational categories that were of prominent and principle interest at the site: 60garment workers, 60motodups (motorcycle-taxi drivers), 60 constructionworkers and 15 garbage workers.These four sets of respondents were selected because they constituted the largest occupational groups inTeuk Thla, as suggested by preliminary interviews and confirmed by the macro-census.

This second phase, conducted in Khmer and orally translated to a native English-speaking researcherin order to facilitate follow-up questions where necessary, captured a greater breadth and depth ofinformation. Its four sections focussed upon the respondent’s migratory history, current urban householdcomposition and livelihoods breakdown, reciprocal relations of communication and exchange with thesending area, and finally his/her rural household’s composition and livelihood strategies.

1364 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

One-third of the respondents taking part in this second smaller survey phase were then randomlyselected for invitation to take part in a series of extended semi-structured interviews as part of a thirdphase of investigation. In these interviews migrants were asked to talk at greater length about theirrural home, migration, the nature of urban life and work, and their future plans and aspirations. Duringthe same period, four focus group discussions comprising six to eight members of each of the majormigrant occupations (garment workers, motodup, construction workers and garbage workers) wereundertaken based on their willingness and availability to participate.

Subsequently, interviews were conducted with a number of key urban informants at this stage of theproject, including landlords, trade union officials, local community elders and elected administrativeoffice holders such as urban village chiefs and their deputies. These interviews were arranged either onan ad hoc basis, as in the case of landlords, many of whom reside close to their rented rooms, or inadvance, in the case of government and trade union officials. As in the rest of phases two and three,these interviews were conducted in Khmer and translated periodically into English to facilitate follow-up questions in relation to the wider data.

The fourth and final phase of the field research programme was a rural follow-up study that visitedthe sending villages and households of 20 migrants involved in the prior round of interviews. Thecollection of ‘linked in-depth interviews in origin and destination’ was intended to provide insight,following Kuhn (2003, p. 11), on ‘the forms of economic and social exchange that facilitate migration,as well as migration’s impact on economic development, risk diversification and power formation’.The team sought permission from urban migrant interviewees to visit the village on an agreed day,whereupon the migrant’s family and neighbours were interviewed, in addition to village officials suchas the village head, where available.

The left-hand image in Figure 1 shows the primary field site, framed by three main roads, inNovember 2000: a few factories (prominent buildings with blue metal roofs) are visible, surroundedby small-scale low-quality residential settlements and some areas of agricultural land. The right imageshows the area 10 years later, in August 2010.

Inevitably for a project of this scope, the various methodologies employed herein contain a number oflimitations and compromises. On the quantitative side, for instance, the collection of arrival data in themicro-census is not ideal, given the skew generated towards more recent arrivals by those who havealready left. If this is borne in mind, however, it is believed that these data nevertheless retain significantvalue. In addition, it may be viewed as a disadvantage that, in order to facilitate the necessary volume ofresponses, the micro-census excluded a number of key variables, ultimately focussing on those fivewhich would provide a broad picture of movement, occupation and demography.

Similarly, although from the point of view of sampling, every effort has been made to provide arepresentative section of Teuk Thla’s inhabitants, the decision to focus on four key occupations rather

Figure 1. A decade of change in Teuk Thla.Source: Google Earth, 2013.

Poverty, power and patronage in the Cambodian migration system 1365

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

than randomly selecting informants from the entire micro-census is an exception to this. However, as onewhich was made in pursuit of depth as a foil to the micro-census’s breadth, it is one which was deemedpreferable to the collection of surface level data on a number of Teuk Thla’s smaller occupations.

4. Pragmatism and Flexibility in Teuk Thla

4.1 Arrival

Throughout the course of this research is has become clear that the household plays a majorcontrolling role in almost all migrations. Whilst curiosity, a desire to follow peers, or to gainindependence are, as detailed by Derks (2008), undoubtedly important factors, they are not the drivingforces behind migration decisions. Migrants do not, in general, insist upon moves to the city. Rather,an increasing willingness on the part of young people to migrate (as outlined also in Lim [2007])should be viewed as a factor which reduces the utility cost of labour movement, not as one whichrenders the decision itself an individual one. Regardless of the desires of migrants themselves, theirneeds are taken account of only in the context of the household as a whole.

As such, to draw a clear distinction between the roughly 60 per cent of migrants motivated by a lackof work in the village or household and the roughly one-quarter who ‘could no longer farm due toflooding’ (Focus group, 20 June 2010) or another natural disaster, is unsatisfactory. Instead, apreferable assessment might read that shock factors proved the catalyst in the pursuance of migratorylivelihoods strategies, but that these tended to be underscored by chronic poverty. As one garmentworker commented, regarding her migration some 18 years previously:

I faced the most difficult situation during the flood of ’93. Being the oldest sister, I had to beresponsible for my family. When there was no rice to eat, I had to collect water greens so that Icould get money in order to buy rice to feed my siblings. I did not care about myself as long asmy siblings had enough rice [for themselves]. (Focus group, 20 June 2010)

Given the ecological and economic interlinkages which characterise many rural areas in Cambodia,this emphasis on climate shocks within the qualitative data would anticipate a degree of covariance inmigrant arrivals. Nevertheless, some heterogeneity might be expected in response to differing condi-tions at the receiving end, especially for migrants entering different (or no) occupations. As shown inFigure 2, however, any such difference is in fact very limited, with migrants across all surveyedoccupations exhibiting roughly covariant patterns of arrival to Teuk Thla.

Figure 2. Arrivals to Teuk Thla disaggregated by Occupation.

1366 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Instead, the clear peaks shown in Figure 2 demonstrate an increased volume of arrivals associatedwith the years 2000, 2005 and 2010, across all occupations. Controlling for the increased volume ofarrivals over time, both a consequence of a real increase in numbers and the data’s inherent skewresulting from the methodology employed, these peaks appear to be correlated with both economicgrowth and the scale of natural disasters experienced by Cambodia (calculated via the proxy ofeconomic damage per year from climate events). Indeed, even using national level data alone(CDRI, 2012; EM-DAT, 2013), a regression model of arrivals incorporating both these factors givesan r-square statistic of 0.27 (N years=11, p < 0.05).

Although the use of national data in the composition of this regression carries inherent short-comings, these are somewhat mitigated by the breadth of migration to Teuk Thla, which containsmigrants from every province in Cambodia. Furthermore, disaggregating the data by province revealsmarkedly similar patterns of arrival to Teuk Thla from all points of departure, and no significantdifference in rural visitation patterns, other than a slightly longer average duration of stay undertakenby migrants from more distant provinces. The lack of correlation demonstrated with other indicatorssuggests that the significant role indicated by the regression is at least to some extent a valid one, evenif the r-square statistic should not necessarily be taken at face value.

As such, the bivariate correlation alluded to by these data not only makes a case for the balance ofpush and pull factors as determinants of migration in Cambodia, but also serves to highlight the extentto which pull factors are subject to filtration by rural conditions at a variety of scales. The notablesimilarity in arrival patterns for motodup, construction and garment work, for instance, suggest thatmigration decisions across occupational groups are linked together in ways beyond that which can beexplained by employment signals.

Rather, a perspective which highlights migrants’ inter-reliance provides a more satisfactory basis foranalysis. Migration in Cambodia almost always (that is, in roughly 99% of cases) involves residencewith at least one family member, and as such would not necessarily be expected to correlate solelywith urban opportunity, but also with rural labour availability at the household and local levels. Indeed,this is borne out by the practice of sending household members, who are neither needed on the familyfarm nor otherwise employed in the village, to live with relatives in the city in order both to ease theconsumption burden on the household and expedite the acquisition of a job.

Such arrangements represent one aspect of the densely knitted familial ties linking migrants fromvarious occupations, as well as the connectedness of rural and urban livelihoods. Persistent thoughthey may be, however, these ties, linkages and networks are not static, nor do they harbour flows ofinformation or resources whose directionality is unchanging. Rather, the linkages binding rural andurban are in constant flux, changing their function according to circumstantial requirements at bothends of the rural–urban continuum.

4.2 Rural–Urban and Inter-occupational Networks

It is a common pronouncement amongst commentators on migration, whether national or internationalin scope, that rural–urban migrants are dislocated from their adopted environments (Derks, 2008;Kuhn, 2003; Nishigaya, 2002; Pahl, 1966). In Cambodia, the ‘physical and social distance’(Nishigaya, 2002, p. 28) between migrants and their families has been argued to generate ‘changesin their behaviour patterns’, which may rapidly, or over time, result in rifts developing between parentsand (especially female) children as the latter are seen to have violated cultural norms in their newenvironment (ibid.).

Nevertheless, this perspective, though valuable in its elucidation of the dangers faced by migrantwomen in particular, services only an atypical sub-set of migrations in Cambodia. In addition to thefact that some 98.7 per cent of Teuk Thla migrants surveyed live with at least one family member(with 73% living only with family), in the age of telephony, the distances involved in rural–urbanmigration are rarely a barrier to continued personal interaction between migrants and their ruralhouseholds, as shown in Table 1.

Poverty, power and patronage in the Cambodian migration system 1367

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Moreover, rural–urban communication is not limited to telephony. A notable proportion ofmigrants receive inverse remittances, comprising rice (in particular), fish or vegetables, usuallygleaned from the rural household’s farming activities (Table 2). These are often transported via theknown taxi drivers which form a key artery between more established migrant enclaves and ruralsending communities of which the driver is usually a member. In addition to goods, these crucialtransport links also contribute to a much higher frequency of rural–urban visitations by familymembers than might be expected, with migrants receiving an average of 2.7 visits per year fromfamily members during their urban residence.

Furthermore, the relative proximity of migrant and household facilitates a somewhat flexible (re)allocation of labour throughout the year. Rather than the handful of return days per year whichrepresent the average in many high-migration societies such as Bangladesh (Kuhn, 2003),Cambodian migrants in general retain a strong physical presence in their sender villages. Around60.8 per cent of migrants (Table 3) contribute labour to their rural households upon returning,amounting to a mean 47 days per year per migrant.

This commitment to providing rural labour is a significant factor, also, in urban inter-occupationalmovement. Rather than remaining for extended periods in a job at odds with their household’s needs,migrants display a propensity to shift between employment sectors in search of more suitable work.In particular, this manifests in a flow away from garment work which, despite offering a stableincome conducive to high remittances, is viewed by many as intolerably inflexible. As one suchworker noted:

Table 1. Frequency of migrant contact with rural household of origin

Percentage of migrants

Migrants who make telephone calls to rural householda

– at least one call each week 48.8– at least one call each month 90.6– at least once call each year 93.5Migrants who make return visits to rural householdb

– at least one visit each week 6.5– at least one visit each month 29.7– at least one visit each year 99.5Migrants who receive visits from rural household members in Phnom Penhc

– at least one visit each week 5.9– at least one visit each month 16.2– at least one visit each year 65.4

Notes: aN=172; bN=185; cN=185.

Table 2. Bi-directional remittance flows

Percentage of migrants sending torural householda

Percentage of migrants receivingfrom rural householdb

Any form of remittance 94.8 72.0– money 88.6 1.0– rice 3.6 64.2– other food 55.2 35.2– clothes 27.1 0.5– medicine 6.1 0– household goods 14.6 0

Notes: aN=192; bN=193.

1368 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

If we are absent from work for even one day, our premium [the bonus paid to workers for a fullmonth’s attendance] will be taken away from us, despite the fact that other factories considerthree days [acceptable]. If you are late, your card will be guarded…They monitor our work and ifwe make a mistake, we will be scolded. (Focus group, 20 June 2010)

Furthermore, as a construction worker commented:

I used to work as a garment worker, motorcycle taxi driver, and a construction worker … I wouldsay being a construction worker is much better … Working in factory gets a low salary … Alsowhen a parent gets sick, it is really difficult for us to ask for permission [to visit them]. (Focusgroup, 20 June 2010)

Far from being merely an inconvenience, or as suggested by Vickery (1984) a mismatch with a culture atodds with working to heavily prescriptive schedules, such inflexibility frequently proves incompatiblewith what many migrants view as their responsibilities to their household. Indeed, above and beyond thecommitment to remit, and to participate in major festivals such as Pchum Ben (the annual festival forancestors) and Khmer New Year (for which even garment workers are given universal leave), theprovision of labour during key agricultural periods generally remains a significant duty for migrants.

As shown in Table 3, although the proportion contributing rural labour varies markedly betweenoccupations, it is significant even for garment workers. For all other surveyed occupations, rural labourprovision represents at least one-sixth of an average year, rising to around three-quarters in the case ofgarbage workers. For these more flexibly employed migrants, urban work is viewed as merely oneelement in a portfolio of income strategies, often centred on rural agriculture.

Indeed, migrants frequently delineate seasonal migration plans, such as that ‘I work on my [family]farm during the rainy season, but during the dry season I come back to work as a construction worker’(04 August 2010) or, as stated by a motodup (16 July 2010) that ‘I first left my home due to a drought,[but] I go back to work on my farm every rainy season. When I have done my work on the farm, Ihave to come back to work in Phnom Penh.’

Moreover, the ability to retain a life centred around the rural is often expressed as a mitigating factorin the uptake of otherwise difficult or unpleasant occupations such as garbage or construction work. Injobs like these, health problems are endemic, prompting complaints that ‘when we work too much, weget sick. We vomit and sometimes we even cough up blood. We also suffer skin irritation’(Construction Worker, 9 June 2010). Similarly, despite the fact that ‘everybody looks down on garbageworkers’ (Depot Boss, 7 April 2010), the lack of constraints it offers are viewed as partial compensa-tion. As one garbage worker (05 May 2010) commented, ‘[migrants] are all here to earn money. WhatI receive doing this is similar to the other migrant jobs but it allows me the freedom to go home andsee my children when I choose.’

The advantage of flexibility is not, on the basis of this study, enough to encourage anybody into garbagework from another occupation. However, in times of livelihoods pressure it appears sufficient both tofacilitate the initial rural–urban move – mediated via informal networks from the depot owner’s homevillage and those near it – and to retain the majority of those already involved. Indeed, whilst some people

Table 3. Migrant contributions to rural household agricultural production, disaggregated by occupation

Percentage of migrants

Total migrants contributing to some form of agricultural 60.8production at rural household (N=194)– garment workers (N=54) 44.4– motodup (N=61) 70.5– construction workers (N=61) 60.7

Poverty, power and patronage in the Cambodian migration system 1369

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

were found to have moved out of garbage work, the majority of current garbage workers, driven by arurally focussed pragmatism, had no intention of doing so themselves. As the depot owner explained:

Nobody disrespects garbage workers in the home village. People respect them a lot for going toPhnom Penh and working hard for their family. People see the houses that their families buildfrom their remittances and they respect them. (Owner interview, 28 September 2010)

Neither are the depot owner’s words empty in this regard. The lower cost of urban living afforded byliving rent-free amongst the piles of sorted recyclables in the depot mean that garbage workers are ableto remit the highest remittances per urban day of any entry-level migrant. This efficiency of earningmay be additionally advantageous, given that the number of days spent in the rural household is theonly statistically significant factor in the (self-reported) improvement of overall rural householdwealth. Gross migrant earnings, by contrast, exhibit no relationship, suggesting that the ability toprovide free labour during the planting and harvesting periods – the main reason for extended returnaccording to the qualitative data – is of even greater importance to the household. Figure 3 highlightsthe strength of this relationship.

Nevertheless, despite the persistence and foundational importance of rural–urban linkages, it wouldnot be correct to suggest, as Lim (2007) for instance have, that migrant enclaves are essentiallynegligible social entities. Whilst Derks’s (2008) ethnography of Cambodian migrant women goessome way in refuting this assumption on an individual and small-scale basis, Lim’s (2007, p. 42)conclusion that migrant interaction with the urban environment is ‘sorely shallow’ requires furtherchallenge. In doing so, what follows shall endeavour to afford migrant networks a degree of socialcontext.

5. Norms and Hierarchies

5.1 Urban Migrant Hierarchies

Although it encompasses a great variety of occupations, living conditions and livelihoods, Teuk Thla,taken as a whole is a community characterised by youth and femininity. The mean age of inhabitantswithin the survey area is just over 24, with 69 per cent being female. To an even greater extent,though, it is characterised by migrancy. Teuk Thla’s inhabitants, 94 per cent were born outside PhnomPenh, having arrived, on average, four years prior to the study’s undertaking. It is an area churningwith movement.

Nevertheless, although social linkages between migrants and non-migrants1 in Phnom Penh are notcommon, neither are they insignificant. Moreover, despite the very strong linkages to the ruralhousehold detailed above, migrants should not be conceived merely as social annexes of their sendervillages. Teuk Thla migrant society is far from amorphous, being instead characterised by variousoccupationally determined hierarchies associated with financial and social differentiation. Figure 4

Table 4. Proportional remittances disaggregated by occupation

Yearlyearnings(mean)

Yearlyremittances(mean)

Remittances asa proportion ofearnings (mean) Mean rural household

days per year$ $ %

All migrants– garment workers (N=48) 1075 365 34.6 15– motodups (N=33) 1295 375 32.2 35– construction workers (N=48) 1350 380 30.1 33– garbage collectors (informal)(N=13)

735 230 39.0 139

1370 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

illustrates the various social hierarchies in evidence in the Teuk Thla migrant community andconstitutes a heuristic sketch compiled from various signifiers of status, merit and opinion. Inrecognition of the invalidity of traditional indicators of rural status in the urban setting, the socialstructure it depicts was formulated using the following criteria:

● co-habitation: do the various occupations and levels therein cohabit?● qualitative responses: responses to questioning on perceptions of the various occupations;● income and ladders of promotion: where a specific ‘career ladder’ exists (in the case of garment

and construction workers), it has been assumed that the increased money and power (for example,control over hiring and firing) connotes an increase in urban social status;

● movement between occupations: where migrants frequently and bi-directionally move betweentwo occupations, is has been assumed that neither can be considered significantly inferior to theother.

As demonstrated in Figure 4, not only are the basic (entry) levels of each of the three major migrantoccupations regarded as more or less equally desirable by migrants, but they are associated also withfrequent occupational movement. Given that migrants engaged in garment, motodup or constructionwork frequently co-habit and intermarry, it has been concluded that there is no significant difference inthe status of these three occupations or their employees.

In the case of garbage work, whilst there does appear to be something of a disparity in ‘status’compared with other occupations, it should be noted that this is not a clear hierarchical distinction.Rather, those garbage workers who consider themselves to be inferior do so as a result of largelypractical reasons, claiming that ‘walking around all day in the hot sun or rain is very tiring and affects[their] health’ (Kumpta, 13 April 2010). Friendship, or even marriage, between garbage workers andmigrants working in other occupations are not rare, despite the fact that garbage workers are perceivedas ‘poor’ (Chantou, 21 November 2009) by the other inhabitants of Teuk Thla.

Indeed, despite this nuanced distain for garbage work, the basic levels of migrant labour in TeukThla are not hierarchically distinct. Above these, however, powerful individuals are viewed as thecommunity’s hierarchical points of reference. Some of these positions are accessible, over time, tomigrants and are viewed by those with plans to remain longer in the city as realistic aspirations.Construction foremen, specialist workers, depot chiefs and garment factory administrators, for

Figure 3. Mean time spent in home village versus rural livelihoods improvement.

Poverty, power and patronage in the Cambodian migration system 1371

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

instance, lead notably different social lives to the people working below them in the hierarchy, tendingto live either alone or with their nuclear family rather than fellow workers, and often counting non-migrants amongst their friends. However, whilst these individuals play a central role in migrants’ lives,the real power lies above them, with rented room owners, government officials and (far above both ofthese) the Chen (Chinese factory owners).

5.2 Power and Authority in Teuk Thla

Although the structure of formal urban authority in Cambodia is designed to mimic that found in ruralareas – with village, commune and district chiefs comprising the lower levels of government control inthe cities – in practice migrants to Teuk Thla reside within a unique power structure. The roughly 250households for which the urban village chief has responsibility appears comparable to a rural village,but in an area such as Teuk Thla, heavily populated by migrants, he or she may in fact be responsiblefor many times more. Indeed, because only permanent residents are required to formally registerthemselves, the village chief often does not even possess an estimate of the population within hisadministrative zone:

The owners of [rented rooms] have responsibility for migrants from my point of view. If theywant to register migrants they can. If they don’t, they don’t have to. (Deputy Village Chief, 31July 2010)

As a result, migrants are essentially invisible to formal, urban authority, relying instead upon theowners of their rented rooms, who tend to be the only authority figures available. Over time, thissituation has become institutionalised to the extent that rented room owners have total responsibilityfor migrants and formal authorities are disbarred from intervening in their affairs unless requested todo so by the former:

I am not allowed to enter rented rooms without permission. I am not allowed to intervene indisputes unless the owner asks me to. (Deputy Village Chief, 31 July 2010)

garment workers contruction workers

motodop

[motobike taxi

drivers]

line leaderschieung

[skilled workers]

foremenunion

officials

office/

admin

village leaders

landlords

remorque drivers

garbage

collectors (municipal

and informal)

village

garbage truck drivers

recyclables depot

owner

Figure 4. Migrant occupational trajectories and social hierarchies in Teuk Thla.

1372 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Though informal, this arrangement is credited as playing a major role in the ongoing peace andsecurity of the community:

Teuk Thla has a strong community and this is based on [two things:] the relationship between therented room owners and the migrants, and the relationship between the rented room owners andthe other permanent residents. (Deputy Village Chief, 31 July 2010)

The owners of rented rooms are the big people in the community. If migrants have problems theygo to the owners, who go to [other] high ranking people in the community such as policemen.(Deputy Village Chief, 31 July 2010)

Migrants themselves are not only expected keep a distance from formal authority, then, but even fromordinary non-migrants, a perception supported by the data’s indication that only very few long-termmigrants possess social connections to non-migrants of any kind. Whilst this leaves migrants almostentirely dependent on the benevolence of rented room owners, no migrant interviewed during thecourse of the study reported a failure to live up to this responsibility, with many considering theirlandlord the first port of call for money lending as well as dispute resolution.

Indeed, interviews with rented room owners suggest that many landlords assume a role similar tothat of a wealthy rural patron in the sense that, much as a Cambodian rural landlord tends to remain ‘afarmer amongst fellow farmers’ (Ledgerwood & Vijghen, 2002, p. 114), room owners frequently livein close proximity with their tenants, albeit in larger and superior accommodation. They are often aperpetual presence during the day, filling the cisterns from which residents obtain their water fordrinking, washing and cooking, and tending to the communal areas of their property. Furthermore,many rented room owners go to great lengths to emphasise the responsible and protective role theyplay in the lives of their tenants, claiming for instance that:

I visit every one of my rooms every five to seven days to make sure everybody who lives there iswell and to make sure everything is in order and there is no trouble. If somebody has a seriousillness, I will call a doctor or nurse for them, although they will have to pay for their treatmentthemselves.

I keep an eye on who comes in and out of my block. If somebody unknown appears I ask themwho they are, where they are from, whether they have any relatives here, what they do usually,what they are doing here etc. I don’t want bad people to come here and do illegal things.

There has never been a proper conflict in my rooms, but there have been minor problems such asTVs being too loud and keeping other residents awake. In those cases the affected residents willusually come to talk to me and I will intervene on their behalf so that the police or village chiefdon’t need to get involved. (Rented room owner 3, 29 July 2010)

This protective role incorporates a notably territorial aspect, with residents often locked into the rentedcompound at night (most rented rooms are gated, although the gated area usually includes all theblocks owned by the landlord, providing a useful proxy for their ‘territory’) in order to ensure that noharm comes to the inhabitants of their property. As two informants stated:

I would prevent anyone from entering the community who wished to do harm to the residents.(Rented room owner 1, 29 July 2010)

Nobody from my rented rooms has ever had a fight, but I have seen it happen elsewhere. Theowner usually stops the fight and dismisses them from their room. To avoid anything like that Ialways lock the gate at night to keep my residents safe. (Rented room owner 2, 30 July 2010)

Poverty, power and patronage in the Cambodian migration system 1373

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Although important, however, these are not the only type of patron in the city. From the point of viewof labour supply, a similarly patron-like role is played by construction foremen, who has control overrecruitment, as well as continued employment from site to site. Since foremen usually call workerspersonally to re-employ them once they have found a new job, this makes them a central authorityfigure in construction workers’ lives.

As such, the nature of horizontal relationships in migrant society depends on the case in question.Construction foremen tend to extract only minimal rents (on the order of $5 to $10) for access tolabour opportunities, whereas factory administrators and translators, who in theory have little to norole in recruitment, are often reported as charging anything from $20 to $200 for a job. Beyond these,higher ranking ‘patrons’, like policemen or local government officials, who generally provide notangible benefit to the migrant beyond the prestige of their acquaintance, are nevertheless jealouslysought as associates and frequently courted with gifts and invitations to social events (Motodup 8, 28March 2010). In this way, connections of various sorts are of vital importance to Khmer migrants,whose ‘need to be in ksae [networks] with knorng [patrons]’ (Ledgerwood & Vijghen, 2002, p. 51),survives dislocation from the rural sphere intact.

6. Bi-directional Feedbacks

6.1 Reshaping the Urban

In spite of the value of social connections in the city, migrants to Teuk Thla tend not to arrive intoorigin-based networks extending beyond a handful of family and/or friends at most. Indeed, manymigrants begin their urban residence knowing only their initial point of contact in the city – an aunt orsister already working in a similar job, for instance – and, partly as a result of a widespread perceptionof the danger of the city (Sol, 28 June 2010), do not spatially expand their social networks muchbeyond this.

However, migrants are well aware that connections to useful individuals are vital and their absenceis frequently bemoaned as a barrier to career change or advancement. It is a typical pronouncement, forexample, that ‘though I am a construction worker, I would prefer to be a [truck] driver. Unfortunately Idon’t have ksae though, so that way isn’t open to me’ (Motodup 4, 15 May 2010).

Similarly, another interviewee bemoaned that:

I used to be a construction worker before saving the money to buy a motorbike and become amotodup. After a year I realised that I couldn’t earn as much doing this so I want to go back toconstruction work. Unfortunately, I cannot because I no longer posses ksae. (Motodup 1, 28March 10)

The strong affiliations to members of family and village, in addition to work-based groups, means thatmigrants’ social networks tend to include a number of occupations. Although migrants themselvestend to be spatially limited in their habits, these multiple bases of connection mean that their webs ofassociation extends beyond one factory or even community and are additionally augmented by thefrequency of inter-occupational marriage. As a result, migrants are relatively efficient conveyers oremployment information, as reflected in the frequency of movement, especially by men, betweenemployments.

Nevertheless, although from the perspective of employment signals the migrant network may havethe appearance of an integrated construct, the individualised components which comprise it depend toa significant extent upon the particular circumstances of employment. Construction workers possessthe largest (work-related) networks because they need to keep in contact with as many colleagues andsuperiors as possible in order to ensure continued employment between sites. Cintri [Phnom Penh’smunicipal waste disposal company] garbage truck workers’ networks are based around their truckteams. Furthermore, although motodup employment is ensured by their ownership of a motorbike, theymaintain key ksae with the other workers at their spot, who are a crucial means of ensuring them

1374 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

access to regular and secure income by protecting their informal right to occupy the same locationeach day and limiting the competition they face.

Notably, motodup tend to talk down their own power in this regard, whilst emphasising that ofothers. For instance, they generally responded that, should an unknown newcomer seek to occupy theirspot, they ‘would not like it since it would reduce the amount of money we all earn, but this is publicproperty, so what could we do about it?’ (Motodup 6, 16 February 2010). When asked, however,whether they would consider setting up next to a group of unknown motodup, they were extremelyreticent, even when they know they could make more money by doing so. The same motodup thatstated he could do nothing to prevent new workers joining his group later stated that:

I was advised that the roundabout in front of the Japanese Bridge is an extremely lucrative placeto work. Unfortunately I am unable to work there because I don’t know any of the other motodupthere. (Motodup 6, 18 February 2010)

Similarly:

I wait [at this spot] with fellow villagers … I previously tried to wait at a different spot, but themotodup already working there got angry with me and said that if I had somewhere else to wait,then I should go there. (Motodup 18, 31 March 2010)

As such, although motodup have no obvious need to group together into work teams, they neverthelessdo so and are protective of their spot, usually permitting only known persons join them where theywork. For instance:

I used to be a garment worker but I became a motodup when I saved up the money to buy amotorbike … I was able to work at this spot with these people because I used to pass them everyday in the street and talk with them when I worked at the factory. (Motodup 7, 24 February 2010)

Furthermore, this pattern of connection building is repeated even in the formal sector, where garmentworkers also tend to acquire ksae over time, building linkages with trade union officials and officeadministrators who are a vital means of obtaining both time off and overtime, as well as attaining asecure long-term contract. As one informant stated:

I had to pay the factory $200 in order to get the status of steady worker which gives me moreprotection against being fired. The money was paid to an interpreter who made the transactionwith the Chinese bosses on my behalf … Later, my sister paid $70 for the same service. (Garmentworker 3, 20 March 2010)

Similarly:

I belong to the trade union. It is very useful because it provides am pee [security from paycuts].(Garment worker 2, 20 March 2010)

Via the judicious acquisition and utilisation of key connections in this way, many migrants are able tobuild relatively secure medium- to long-term income generation strategies in Phnom Penh according tothe needs of themselves and their households. However, in doing so, they are acquiring more than amonthly income. The social implications of migrant work extend beyond its financial rewards, to theextent that they may be playing an increasingly important role in shaping and defining rural powerstructures at a number of scales. What follows shall seek to delineate some of these effects.

Poverty, power and patronage in the Cambodian migration system 1375

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

6.2 Reshaping the Rural

As Derks (2008) emphasises, the ability to work away from the fields is, especially for women, lacedwith positive cultural connotations beyond those pertaining directly to comfort or even wealth. Factorywork has, for instance, been equated by several of Derkss’ subjects to the traditional practice of ‘goinginto the shade’, wherein betrothed girls spend up to six months without leaving their home by day inorder to lighten their skin for their wedding. Indeed, as one garment worker noted regarding those whomigrated and returned prior to her departure, returning migrants tend to be seen as highly glamorous.As she put it:

they became more fashionable. They had money and their skin was so white. These [things]really took my interest. (Focus group, 20 June 2010)

Such associations, with both duty and beauty, make certain types of migrant work, especially garmentwork, a source of respect in the village. Nevertheless, this rapid attainment of rural status throughmigration is not necessarily gladly received by all remaining members of the sender community, manyof whom complain of returning urban workers behaving in a haughty and aloof manner towards thosewho had remained, or of ignoring them completely. As one garment worker commented:

It is not the same. Before we were really close to each other but after having worked in thefactory, they seem to be arrogant or [to act] as if they have forgotten us. (Focus group 1, 21 June2010)

Nevertheless, despite the impression of wealth and self-indulgence projected by many migrants upontheir return home, personal spending is in fact heavily restricted by financial responsibilities, withremittances constituting not only a practical, but a vital social aspect of migration. A failure to fullycomply with remittance duties is discouraged to the point of distain, with migrants who remit nothingor who remit less than similarly aged siblings liable to be branded by family members as ‘lazy’ or‘selfish’, and to be berated for wasting money on consumables such as clothes, or food and drink (Sol,28 June 2010).

Moreover, although for some migrants remittance payments are the difference between life anddeath for elderly or sick family members at home, this is not usually the case. Rather, that a similarenforcement of homeward remittances is in evidence across almost all wealth strata demonstratesthe role that they play in long-term acquisition and the building of status within the community. Inthis latter respect, the increased prominence of migrants post-migration is mirrored in that of theirhouseholds. Perhaps most tangibly, this manifests in an upturn both in the number of invitationsthey receive to weddings and other ceremonies and the amount that they are able to pay at suchevents.

As a young CINTRI truck driver, who remits between $25 and $50 each month, explained, suchadditional income can have a significant impact on the status of a household: whereas prior to hismigration the family were ‘quite poor’ compared to the rest of the village, they are now ‘average’.The family has already noted a large increase in the number of social event invitations they receive.‘Since I came to work in Phnom Penh, my family have more money in the village so other peopleinvite them to their ceremonies because they have money to give. Before, they had to ignore invitesthey received.’ Another CINTRI worker, Sonim’s household has noted a similar substantialincrease in event invitations since migration, after ‘the rest of the village found out I had a joband money and could afford to give donations’. As Yorng, his colleague also explains, ‘before Ididn’t have much money to participate or donate, but now I have a lot more so I get a lot moreinvites’.

In addition to the changes in the positionality of both migrant and household within the socialworlds of their sender village, a third significant feedback constitutes a shifting in the balance ofpower within the household. Almost all migrants noted status within the home household had

1376 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

improved since their migration, with an increased influence over household decision-making anduse of household financial resources. Sonim, for example, 33 at the time of interview, migrated toPhnom Penh in 2003, when aged 26. He explained that whilst he did have ‘some’ say in householddecision-making before he migrated, his role in such areas was now ‘strong’, compared with‘before, when I was without work and not earning money [and] was not judged to be able tomake good decisions’.

Although such role shifting may be partly attributable to demographic factors, interviews withmigrants suggest it has much to do with their provision of support. For instance, that the tendency toreport an upturn in household status was universal across age and gender groups indicates that it is nota question of coming-of-age alone. Furthermore, case studies of sibling migrants demonstrated thatthose who shoulder the responsibilities of being a contributor to the home household gain more respectfrom their families, with those migrants who fail to do so being, as noted above, heavily criticised byfamily members and risking ostracism in the long term.

Moreover, interviews in rural areas suggest that remittances may also have begun to promulgatedeeper cultural shifts. Notably, the perception that women are more likely or more reliable remittersappears to have formed in rural villages, offering scope to recognisably enhance the role and status ofwomen, particularly daughters, in rural communities. The village chief of one of the largest sendervillages to Teuk Thla, Pilea village in Prey Veng province, keenly emphasised that, ‘now it is better tohave daughters than sons because daughters help the family’ (04 April 2010).

These shifts in rural power dynamics, away from age and the male and towards youth and thefemale, are symptomatic both of the size of the role now played by migration in the Cambodian socio-economy, and of the systematic pragmatism with which migrants and their households maximise thebenefit they derive from it. As the sources of rural merit move towards the urban, the traditional valuesof duty, self-sacrifice and care take on a new complexion. In response, the culture and structure of ruralCambodia are being re-thought day by day.

7. Conclusion

As the preceding sections have sought to demonstrate, the great cogs of Cambodian labour migrationturn in response to a multitude of smaller cycles. As such, it is a highly structured, patterned process,but one which nevertheless facilitates dynamic adjustment on the part of participants according tochanging household or personal circumstances. The birth of babies, rather than necessitating an end tomigration, may be accommodated by one or both parents seeking employment with more flexiblehours; the need to provision rural labour may be met via a range of strategies; and income shocks,whether pertaining to natural or economic disasters, may be absorbed effectively by the networks,resilient in their dynamism, which undergird Cambodian migrancy.

Nevertheless, the aim herein has not been merely to demonstrate that rural–urban migrants inCambodia respond efficiently to employment signals and climate shocks, but rather to shed somelight upon how they do so. Migrants continue to be influenced by the norms and hierarchies whichunderpin rural social life, and in many ways their persistent ‘ruralness’ is one of their defining features.What underpins their behaviour, though, is not conservatism but pragmatism. Trusted socio-familialnetworks of assistance and information constitute an effective insurance against risk, as well as aneffective channel of information. Hierarchies and patronage also have a role to play, as universallyintelligible loci of protection and favour in an otherwise alien environment.

This is not to argue, either, for the hegemony or intransigence of rural culture. Indeed, the samepragmatism which governs the systematic behaviour of migrants applies equally to their sendervillages. As a result, the increased respect and authority afforded to the women and the youngconstitute the symptoms of a slow sea-change in the Cambodian socio-economy. Similarly, theneed to create and maintain the horizontal social linkages which expedite and enhance migration,as opposed to the horizontal chains which have traditionally characterised Khmer society(Ledgerwood & Vijghen, 2002) may represent the beginnings of a shift in rural power structures,

Poverty, power and patronage in the Cambodian migration system 1377

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

as wealth and status are increasingly determined by connections beyond the village. Whether thissignifies a better life for all, or the beginnings of a schism between the connected and the poor,remains to be seen.

Acknowledgements

This research was made possible by a grant from the International Development Research Centre ofCanada. The authors would like to express their thanks for the assistance of both the IDRC and ourmany former colleagues at the Development Studies Faculty at the Royal University of Phnom Penh.

Note

1. For the purposes of this study, non-migrants are defined as people who were either born in Phnom Penh (or what is nowPhnom Penh given the expansion of the city), or who have lived there since the 1980s, during which time the city wasgradually repopulated.

References

Arnold, D. (2013). Workers’ agency and re-working power relations in Cambodia’s Garment Industry (No. ctg-2013-24). BWPI,Manchester: The University of Manchester.

Bakewell, O. (2013). Relaunching migration systems. Migration Studies, 23, 1–19.Bylander, M. (2013). Depending on the sky: Environmental distress, migration, and coping in rural Cambodia. International

Migration. doi:10.1111/imig.12087Cambodia Development Resource Institute (CDRI). (2012). Annual report 2012. Phnom Penh: CDRI.Cook, E. M., Hall, S. J., & Larson, K. L. (2012). Residential landscapes as social-ecological systems: A synthesis of multi-scalar

interactions between people and their home environment. Urban Ecosystems, 15(1), 19–52.Davies. (2010). “Cambodia: Country Case Study” in Lueng, S., Bingham, B. and Davies, M. (eds.) Globalization and

Development in the Mekong Economies (pp. 152–168) Cheltenham: Edward Elgar Publications.De Haas, H. (2010). The internal dynamics of migration processes: A theoretical inquiry. Journal of Ethnic and Migration

Studies, 36(10), 1587–1617.Derks, A. (2008). Khmer women on the move: Exploring work and life in urban Cambodia. Hawaii: University of Hawaii Press.EM-DAT. The international disaster database. Retrieved April 21, 2014, from http://www.em-dat.be (accessed December 9,

2013).Erel, U. (2010). Migrating cultural capital: Bourdieu in migration studies. Sociology, 44(4), 642–660.Fawcett, J. T., Hammar, T., Widgren, J., Gallagher, D., Martin, D. A., Goodwin-Gill, G. S., & Schmidt, I. (1989). Networks

linkages and migration systems. International Migration Review, 23(3), 671–680.Frank, A. G. (1967). Capitalism and underdevelopment in Latin America: Historical studies of Chile and Brazil. New York:

Monthly Review Press.Harris, J. R., & Todaro, M. P. (1970). Migration, unemployment and development: A two-sector analysis. The American

Economic Review, 60(1), 126–142.King, R. (2012). Geography and migration studies: Retrospect and prospect. Population, Space and Place, 18(2), 134–153.Kuhn, R. (2003). Identities in motion: Social exchange networks and rural-urban migration in Bangladesh. Contributions to

Indian Sociology, 37(1–2), 311–337.Ledgerwood, J., & Vijghen, J. (2002). Decision-making in rural Khmer villages in Ledgerwood (ed.) Cambodia emerges from

the past: Eight essays, 109–150.Lee, E. S. (1966). A theory of migration. Demography, 3(1), 47–57.Leung, S., Bingham, B., & Davies, M. (2010). Cambodia: A country case study. In S. Leung, B. Bingham, & M. Davies (Eds.),

Globalization and development in the Mekong economies. Edward Elgar Publishing.Lewis, W. A. (1954). Economic Development with Unlimited Supply of Labour. The Manchester School, 22 (2),139–191.Lim, S. (2007). Youth Migration and Urbanisation in Cambodia. Cambodia Development Resource Institute (CDRI).Mabogunje, A. L. (1970). Systems approach to a theory of rural–urban migration. Geographical analysis, 2(1), 1–18.Nishigaya, K. (2002). Female garment factory workers in Cambodia: Migration, sex work and HIV/AIDS. Women & Health, 35

(4), 27–42.Oeur, I. L., Sopha, A., & McAndrew, J. (2012). Understanding social capital in response to flood and drought: A study of five

villages in two ecological zones in Kampong Thom Province. Engaging for the Environment, 60, 60–83.Pahl, R. E. (1966). The rural urban continuum. Sociologia ruralis, 6(3), 299–329.Potts, D. (2010). Circular migration in Zimbabwe and contemporary sub-Saharan Africa. Rochester, NY: Boydell & Brewer.

1378 L. Parsons et al.

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5

Rigg, J. (2001). More than the soil: rural change in Southeast Asia. Great Britain: Pearson Education.Rigg, J. (2012). Unplanned Development: Tracking Change in Southeast Asia. London: Zed Books.Rostow, W. W. (1960). The stages of economic growth. A Non-Communist Manifesto. Cambridge.Thompson, W. S. (1929). Population, The American Journal of Sociology, 34 (6), 959–975.Todaro, M. P. (1969). A model of labor migration and urban unemployment in less developed countries. The American Economic

Review, 59(1), 138–148.Vickery, M. (1984) Cambodia, 1975–1982. Boston, MA: South End Press.Watts, M. (1983). Silent violence: Food, famine, and peasantry in northern Nigeria. Berkeley: University of California Press.World Bank. (2012). Doing business 2013: Cambodia – smarter regulations for small and medium-size enterprises: Comparing

business regulations for domestic firms in 185 economies (Doing Business 2013). Washington, DC: World Bank.Zelinsky, W. (1971). The hypothesis of the mobility transition. Geographical Review, 61, 219–249.

Poverty, power and patronage in the Cambodian migration system 1379

Dow

nloa

ded

by [

Kin

g's

Col

lege

Lon

don]

at 0

8:22

01

Febr

uary

201

5