Understanding the Process of Economic Change: Technology and Opportunity in Rural Tanzania

7/27/2019 Understanding the Process of Economic Change: Technology and Opportunity in Rural Tanzania

1/37


2/37

Understanding the Process ofEconomic Change: Technology andOpportunity in Rural Tanzania

By Maia Green

Special Paper 13/1


3/37

Published or: REPOA P.O. Box 33223, Dar es Salaam, Tanzania157 Mgombani Street, Regent EstateTel: +255 (0) 22 2700083 / 2772556Fax: +255 (0) 22 2775738Email: [email protected]: www.repoa.or.tz

Design: FGD Tanzania Ltd

Suggested Citation:Maia Green Understanding the Process of Economic Change: Technology and Opportunity in Rural Tanzania.

Special Paper 13/1, Dar es Salaam, REPOA

Suggested Keywords:

Economic change in rural setting, Agriculture & technology in rural setting, Kilimo Kwanza &technology.

REPOA, 2013

ISBN: 978-9987-483-20-4

All rights reserved. No part o this publication may be reproduced or transmitted in any orm or byany means without the written permission o the copyright holder or the publisher.


4/37

iii

Table of Contents

Table of Contents ............................................................................................................... iii

Acknowledgements ........................................................................................................... iv

Abstract .............................................................................................................................. v

Acronyms ............................................................................................................................ vi

1.0 Introduction .............................................................................................................. 1

2.0 Economic Change in Tanzania ................................................................................ 2

3.0 The Limits of Models ............................................................................................... 4

4.0 Economic Change in Ulanga District .................................................................... 6

5.0 Accounting for the ASCAs Appeal ........................................................................ 16

6.0 Conclusion: Smart Policy for Supporting Change ................................................ 19

References Cited ............................................................................................................... 20

Publications by REPOA ..................................................................................................... 24


5/37

iv

Acknowledgements

The title o this paper re ers to Norths (2005) text. This paper is based on research carried out inUlanga District between February and August 2012 while I was a senior research ellow at REPOA.I am grate ul to REPOA or providing the unds to carry out the research and or the opportunity towork with colleagues dealing with related issues o economic and social development in Tanzania.Help ul comments on the paper were provided by Lucas Katera, Blandina Kilama, Emmanuel Maliti,Jane Guyer, Paula Tibandebage, Amandus Muhariwa, and Brian Cooksey.

I am extremely grate ul to the residents o Ulanga and the sta o the District Council or their supportand assistance, and the Tanzania Commission or Science and Technology or their permission tocarry out the research. Thanks also to colleagues at the universities o Sokoine and Dar es Salaam

or their thoughts that have in ormed my analysis.

Amandus Muhariwa took time out rom a eld trip to help me understand important issues in livestock management. Abdul Mtumwa o Hei er International shared insights rom his experience. JosephMkude and Gepson Absalom o Ulanga District Council explained some o the constraints a ectinglivestock projects in the district. Additional insights were provided by retired agricultural o cer PeterKichelele, who had worked at the Council during the Irish Aid district development programme.

A long Skype conversation with Hugh Allen o Village Saving and Loan Associates helped me tounderstand some o the unique and generic attributes o Village Savings and Loans Associationsin Tanzania and elsewhere. The views expressed in this paper are those o the author and do notrefect the institutional position o REPOA.


6/37

v

Acronyms

ASCA Accumulating Savings and Credit Associations

NGO Non Governmental Organization

ROSCA Rotating Credit and Savings Associations

TASAF Tanzania Socio Action Fund

TAZARA Tanzania Zambia Railway Authority

URT United Republic o Tanzania

VSLA Village Savings and Loan Associations


7/37


8/37

1

1 Introduction This paper explores processes o economic change in rural Tanzania. It investigates some o the

actors that have contributed to the adoption o innovations in economic practice in Ulanga District,Morogoro Region, since the 1990s. 1 Understanding the actors which contribute to innovation isimportant or understanding how actual change occurs in rural areas. It also highlights importantlimitations in current policy models o change that are based on speci c assumptions about economicbehaviour and the impact o agricultural technologies. By ocusing only on individual behaviour orthe impact o technology, policy models o change ail to capture the multiple actors that generatechange in particular settings.

This paper is based on three months o eld research carried out between February and August2012. The research used a combination o ethnographic methods involving observation andin ormal discussions and semi-structured interviews with small armers and district sta to gainan understanding o changes in economic practices in the district. The research involved visits toseveral locations, including the villages o Nawenge, Msogezi, Mbagula, Mdindo, Idunda, Mavimba,Minepa, Mwaya, and Chilombola. Findings rom eldwork were compared with the results o asimilar study undertaken by the author in 1996 on economic and social change in the district.

The research revealed signi cant changes in everyday economic practices and a willingnessamong small armers to adopt new technologies. There were also important continuities in wayso managing livelihoods and household economies. Analysis o the actors leading to the adoptiono improved dairy cattle, small-scale pig keeping, and participation in the savings and loan groups,

which are relatively new to the area, demonstrates that productivity is not an inherent attribute o technologies or investments. It is generated by the economic and social relations in which theseare embedded. Success ul innovations are those that become integrated into daily practices andlivelihood strategies. As these strategies change in relation to emerging economic opportunities, theviability o certain activities is trans ormed. Smart policy or rural economic growth must ocus onincreasing the opportunities available or rural populations. This cannot be achieved by promotingan economic monoculture o agrarian uni ormity. Instead, it demands the care ul nurturing o actualand emerging opportunities.

1 The research on which the paper is based was conducted in Ulanga between February and August 2012 using aqualitative research methods derived rom anthropology. These include semi-structured interviews and group discussionsand observations, in addition to the use o historical materials and other sources. The research builds on the authorsprevious research in Ulanga on a variety o themes at various times since 1989.


9/37

2

2 Economic Change in TanzaniaIdeologies concerning the necessity o economic change have been central to the constitutiono the Tanzanian state since its inception as a territory appropriated by Germany, with a view topro table exploitation (Koponen 1995). As a League o Nations Protectorate a ter the First WorldWar, Tanganyika became the object o interventions intended to accelerate economic and socialchange. These included agricultural modernisation, community development, and large-scaleinvestments in commercial arming (Ili e 1979; Chachage 1988). Idioms o change were central tothe policy evolution o the independent government that pursued its strategy o social trans ormationthrough what were claimed to be reconstituted orms o local organisation (Maghimbi 1995; Coulson1982). Economic policy under Nyerere combined socialist theories o economic development witha commitment to the normative sel -reliance, not only o the supposedly traditional kin-basedcommunities that provided the rationale or Ujamaa policies, but the nation state itsel . Sel -reliance,the production o surplus, and the distribution o appropriate technologies would, in theory, driveeconomic trans ormation (Fatton 1985; Saul 2012). But this development never un olded. As aconsequence, Tanzanias economic crisis in the 1980s opened the national policy and scal spaceinitially to economic liberalisation and ultimately to the ideas linking democratic transition with market

reedoms promoted by northern powers at the end o the Cold War. 2

Although Tanzania as a non-aligned state had been a longstanding recipient o nancial assistancerom a range o governments, the move to aid harmonisation in the late 1990s consolidated

these policy directions with new types o programming in support o poverty reduction strategies(Wangwe 2010). Since the late 1990s, with the support o bilateral and multilateral donors rom the

global north, Tanzania has implemented a port olio o policies designed to oster economic andpolitical liberalisation. Despite the resurgence o relationships with alternative sources o nance andtechnology, o which the most signi cant is with China, Tanzanias present package o economicpolicies remains strongly infuenced by the neoliberal paradigms that have been advocated by itsdevelopment partners since the 1990s. Institutionalised through capacity-building initiatives (Phillips& Ilcan 2004) and substantial nancial trans ers provided by donor agencies, including the WorldBank, these policies are consistent with the orientation o much o the international developmentprogramming (Craig & Porter 2006). In addition to the standard neoliberal package o governancere orms, including decentralisation, Tanzania has implemented policies intended to ormaliseproperty relations and improve the unctionality o market relations through agricultural liberalisationand developing the vertical integration o commodities within transnational value chains (Tripp 1997;Ponte 2002; Lysons & Msoka 2010).

The extent to which these programmes have achieved their objectives is uncertain. Agriculturalliberalisation in particular has been viewed as problematic (Skarstein 2010; Bryceson 2010; 2012).State intervention continues to hinder Tanzanias competitiveness in the production o export crops,including cashews and co ee (Kilama 2013; Cooksey 2003). High transport costs and di cultiesin accessing markets are among the constraints that make returns or Tanzanian armers amongstthe lowest in sub-Saharan A rica (Binswanger Mkhize & Gautam 2010: 37). Productivity gains havebeen concentrated in emerging sectors o the economy, such as industrial mining undertakenby oreign mining companies, where a minority o the population are employed (Mkenda et al.2010). Large numbers o people earn their living in low-productivity sectors such as artisanal mining(Bryceson & Jonsson 2010) and agriculture, the primary source o income or over seventy percent o Tanzanians. Investments in agriculture are prioritised in the countrys new ve-year NationalDevelopment Plan as a means to unleash Tanzanias latent growth potential (URT 2011) and toincrease the inclusivity o growth.

2 Hence market re orm and liberalisation preceded political liberalisation in Tanzania. I am grate ul to Emmanuel Maliti orthis clari cation. For an account o liberalisation in Tanzania rom the bottom up, see Tripp (1997).


10/37

3

The plan builds on pre-existing policies in various sectors, including Kilimo Kwanza , a nationalinitiative launched in 2009 that aims to bring about a Green Revolution in Tanzania.

Tanzanias agricultural ambition has much in common with earlier incarnations o Green Revolutionthinking initiated in the 1960s in much o Asia. Agricultural modernisation through new technologies,notably improved seeds and inputs, is assumed to drive agricultural and wider economic development(Diao et al. 2008; Sanchez et al. 2009). In contrast to the Asian models o Green Revolution, in whichimproved seeds made available through public action led to increased productivity and hence tomarket growth, this new A rican instantiation o the Green Revolution emphasises the role o theprivate sector in driving agricultural trans ormation. Inputs to enhance agricultural productivity areto be provided through the market, via a private sector supported through occasional subsidies,as in the voucher programmes or input supply and ertiliser implemented in Malawi and Tanzania. 3 Facilitating this role or the private sector rests, in turn, on establishing an in rastructure or theproduction o seeds and input supply, which will make agribusinesses at di erent scales the ulcrumo national economic policy and o local social relations (Thompson 2012: 345). The private sectorenvisaged as lead actor in these policy scenarios is no longer imagined as a set o Tanzanian small-and medium-sized entrepreneurs, but as a diverse coalition o agents ranging rom small-scale

armers to multinational corporations.

Kilimo Kwanza is not a policy or a programme with detailed sector plans. It is a call to action, astrategic rationale or the prioritisation o agriculture. Its ten pillars build on the oundations established

through previous agricultural interventions with the intention o assimilating other programmes,including those beyond the agricultural sector, into supporting the agricultural strategy. 4 Suchprogrammes, and the policies which in orm them, are based on assumptions about economic andagricultural trans ormation that centre on the conceptualisation o individual armers as economicagents who can simply choose to change their practices in order to become more productive.In the case o armers, this entails managing their agricultural production as an enterprise orientedtowards expansion, something made possible through the ormalisation o tenure and access tocredit on the one hand and, on the other hand, access to improved, hence more productive,technologies (Sanchez et al. 2009). According to this kind o logic, articulated in the NationalDevelopment Plan, those armers who cannot attain the levels o productivity required by these

programmes should enhance their incomes through wage employment, hence the renewed statuso employment as a key policy area. These agricultural development policies set out in the NationalDevelopment Plan, like policy conceptualisations o change more generally, are premised on anunderstanding o how change occurs that is both strongly normative (Gregory 1997) and premisedon behaviourist and institutionalist oundations (Shove 2010). Capacity to e ect change is thoughtto lie with individual actors i they have access to the resources that would enable them to realisetheir economic potential. These resources may be material, in the orm o investments, or cognitive,in the orm o knowledge, attitudes, and practices. In this understanding o economic change basedlargely on normative theory the role o economic policy is to enable individuals to take the kind o action necessary to be con gured as economic agents by ensuring that the conditions are in placethrough which they can access those resources which will be catalysts or development. Moreover,in ensuring that the structural conditions or the e ective unctioning o markets and incentives arein place, economic policy also plays a role in building the proper institutional environments to enableproductive action (North 2005).

3 Public subsidies in such programmes are legitimate so long as they are smart, usually a shorthand or short term andtargeted.

4 Hence TASAF, the Social Action Fund, in its second phase provided nance or agricultural investments, including dairycattle.


11/37

4

3 The Limits of Models These policy interventions in Tanzania, as elsewhere, are not ounded on the examination o existinginstitutional arrangements through which practical economies are enacted. They are derived roma normative institutional architecture that is thought to yield the optimal organisational orm oran e ective market economy (Guyer 2004; Fairhead & Leach 2005: 889; Mitchell 2005). Feederroads, the construction o market buildings, the provision o credit, and the availability o improvedtechnologies, along with a strengthened private sector, are sectoral priorities which are broughttogether in the Tanzanian input supply programme that has sought to use the private sector as avehicle or ertiliser distribution. This component o agricultural programming highlights the practicallimitations o the institutional assumptions on which it is based. Neither rural agro supply dealersnor an established private sector o input supply exists in su cient quantity in rural areas. They haveto be generated through the subsidy and incentive schemes comprising programme interventions.

The extent to which this results in the establishment o a seasonal private sector that comes intobeing solely in response to programme subsidies rather than a trans ormed agricultural economyis open to question. This certainly happened in Malawi, where the experience o implementation o that countrys input supply programme clearly demonstrates that changing rural economies is morecomplicated than either the imposition o certain institutional orms (Chisinga 2011) or the promotiono technologies. Interventions aimed at increasing irrigated rice production or the use o power tillersin a bid to raise productivity have so ar met with limited success in Tanzania (Therkilsden 2011).

Similar scenarios have characterised agricultural development interventions internationally, whereincreases in the productivity o the small armer are sought as a means o accelerating the mass

uptake o modern arming. Such interventions are not new. They are characteristics o modernisationstrategies pursued by numerous states, irrespective o political philosophy, which assume thatthe productivity o traditional agriculture is a barrier to growth (Fairhead & Leach 2005; Richards1958; Berry 1993). Predicated on the assumption that modernisation, technology, and markets willestablish the systems in which productive agriculture can occur, programmes oriented towards themodernisation o traditional arming generally seek to replace existing traditional technologies,introduce new market relations, ormalise tenure, and make credit available (Rao 1986; Schultz1964; Fergsuon 1990). Small armers in Tanzania, a World Bank report con dently states, tendto invest in land improvement, livestock, trees, ertilisers and arm machinery..leading to a supplyresponse that can be very high in the short term (BinswangerMkhize & Gautam 2010: 8).

As Ellis and Mdoes work on the investment decisions o small armers in Morogoro region showed,such assumptions may be un ounded. Small armers tended to invest in o - arm enterprises toreduce the risks inherent in arming (2003), not only in Tanzania. This nding is echoed by research

ndings rom Ulanga today. In one village, a armer group that succeeded in raising output throughuse o a power tiller was choosing to invest in the motorcycle taxi ( boda boda) business in pre erenceto arming because returns were more certain and potentially available throughout the year, withmore limited seasonal variation than agriculture. In India, most investments in agriculture are madeby traders and wage workers who have the cash to invest in larger arms (Gregory 1997: 134). Thatprogrammatic e orts to accelerate change in rural economies are o ten unsuccess ul is not due tothe inherent conservatism o small armers, nor to the ailure o technologies which were thought tobe capable o generating modernisation, but to the theoretical model o change which underpinsthem. Linear models o change, which assume that small armers trade up asset levels to becomelarger armers or seek to adopt what economic theory may posit as more rational, entrepreneurial

orms o organisation, are not supported by the evidence (Platteau 1989: 644). Indeed, rationalorms o economic behaviour in ragile economic settings where people depend on diverse income

sources, including arming, wage work, and remittances rom distant relatives, are likely to di errom the kinds o strategies which would be predicted by economic theory (Bryceson 2002a & b).


12/37

5

Small armers prioritise keeping costs down and maximising certain over uncertain returns. Non-market orms o economic organisation persist in peasant economies not because they are notyet integrated into market economies but because o the terms o this integration (Bernal 1994;Platteau 1989). This does not imply that small armers are resistant to change. On the contrary,small armers seek to grasp emerging opportunities created by economic change to enhance thediversity and hence resilience o their own livelihood strategies. Understanding actual processes o change in rural economies reveals the actors that contribute to the uptake o success ul innovation.It also reveals gaps in the theoretical models that are driving the policies aiming to trans orm ruraleconomies.


13/37

6

4 Economic Change in Ulanga DistrictUlanga District is a rural district in Morogoro Region. It is the largest district in the region in termso area, 24,000 km2, but its population density is relatively low. District estimates, based onthe previous census in 2002, give the population as 234,000. This is likely to be a considerableunderestimate, 5 due to extensive in-migration over the past decade, particularly in the valley parts o the district where large-scale rice production is possible. The southern part o the district consists o rocky highlands and narrow valleys where armers grow a variety o crops, including hill rice, maize,beans, groundnuts, cassava, bananas, sweet potatoes, and increasingly sesame and sunfowers.

The northern part o the district is situated in the ertile Kilombero valley, an area subject toagricultural intensi cation accelerated by the infux o agro-pastoralists and entrepreneurial armerswho are taking advantage o the lucrative uture market in valley land. The capital o the district isMahenge, an expanding administrative centre that is, like other smaller towns adjacent to areas o economic intensi cation, bene ting rom increased private and public spending (Bryceson 2010).Improvements to the main road, better public transport, and the extension o mineral extraction inthe south o the district and paddies in the north are trans orming the capital rom a remote ruralcentre into a small-scale outpost o the new, rapidly urbanizing Tanzania.

Originally comprising what are now the districts o Mahenge and Kilombero, the present-day UlangaDistrict was ormed in 1974 (Green 2003: 28). Relatively inaccessible due to poor road transport,until recently at least, and bordered by the Selous game reserve to the south, the district has apatchy history o incorporation into national development policies. An economic cul de sac romwhich natural resources and agricultural products were extracted, the colonial economy under the

German administration was based on trade in wild rubber and hippo teeth (Ili e 1979; Larson 1976). As a market-oriented agricultural economy was created in the early twentieth century, accompaniedby orced resettlement into permanent villages under successive colonial regimes, rice, then cotton,became the main exports. 6 In the 70s and 80s, while inputs were subsidised and buyers guaranteed,cotton was important or valley armers. Rice has fourished in the past twenty years in the ertilevalleys o the north and west o the district an intensi cation uelled by population movement

rom other districts, in part encouraged by the TAZARA railway (Monson 2009), by the cultivationstrategies o Sukuma armers who use cattle ploughs to arm large areas, and by buoyant regionaldemand. Rice land in the valley areas is increasingly contested. Maize, beans, and groundnutscontinue to dominate production in the highlands areas. Sesame is undergoing a revival as a cropproduced primarily or sale in many parts o the district.

Despite these more recent changes, the organisation o the agricultural economy exhibitsundamental continuities with what existed in the 1990s. The majority o armers rely on household

labour and occasional kibarua , use hand hoes to cultivate small plots, and incorporate new crops,such as sesame, into their existing crop cycles. As the anthropologist o technology FrancescaBray has shown with re erence to the persistence o simple technology in Asian rice productioncontexts, reliance on hoe agriculture is not necessarily an indication o a arming system waitingto evolve into something more technologically advanced and hence productive. The relationbetween technology and productivity is not direct. Low levels o technology, such as hand hoes,may in act be optimal or certain conditions, including wet rice arming where irrigated plots are

small and where transplanting is per ormed by hand (1978: 15; 1983). Technological change o the sort associated with modernisation is not an inevitable consequence o the availability o technologies. The widespread adoption o modern mechanised arming in northern Europe, Brayargues, depended on the pre-existence o certain social relations, including developed land andlabour markets and the concentration o holdings into large, centrally managed units o production(1983:9). Such concentration and the organisation o labour are unsuited or the demands o wet

5 Results o the 2012 census were not yet available at the time o writing in March 2013.6 See Lorne E. Larsons (1976) A History of the Mahenge (Ulanga) District , PhD thesis, University o Dar es Salaam.


14/37

7

rice production, which requires care ul on-going supervision within small areas. In such situations,the costs o reorganising production exceed the likely pro t rom adopting mechanised technologies.Consequently, even in Japan and China, which have massively increased the productivity o rice

arming, production continues to be organised through amily arms using basic implements (1983:9). Technology is only adopted, and can only become productive, when it is enmeshed within the socialrelations that are amenable to its e ective utilisation.

Brays insight that local economic priorities and relationships determine how technology is usedand what new technologies are adopted enhances our understanding o the process o economicchange in Ulanga District. In lowland areas, tractors are used where people can a ord to hire them,but generally only or preparing the land and or harrowing o the same small shamba areas asthey would have armed using hired labour. The tractor substitutes or labour organised throughpiecework. It has not prompted a move to large arms nor the consolidation o holdings. Farmers inrice-growing areas make use o weed killer to reduce the labour inputs required, i they can a ord it.In both maize- and rice-producing parts o the district, ertiliser is not widely used. Where it is usedby individuals as a result o the National Agricultural Input Voucher Scheme, its use is likely to berestricted to a small area. This is partly to do with cost and also because many armers claim that

ertiliser is not only unnecessary in the areas where they are arming, but they ear that its use coulddamage the land or create an on-going need to use it in subsequent years, thus raising the costso production and trapping armers into dependency on the supply o inputs. Because armersproduction strategies prioritise fexibility, ertiliser is something to be avoided despite the potential or

higher yields. Small armers in Ulanga, like their counterparts elsewhere in A rica, aim to grow severalvarieties o maize or sesame in a small area, pre er crops which can be staggered throughout thegrowing season, and prioritise not getting locked in to costly or infexible production systems. 7

A New Service Economy The economy o the district today is mixed, comprising agricultural products, notably rice andsesame, mineral extraction, rubies and gold, illegal ivory and meat rom the Selous, and anexpanding trade and service sector. The production o local maize beer ( tekawima) remains animportant source o income or women. Mens incomes are supplemented by the trade in bamboo

alcohol, ulanzi , during the wet season (Green 1999). Local economic expansion is acceleratedby the increase in unding to districts through projects such as TASAF and the road und, whichuse some local contractors and procurement, and an expansion in the number o public sectorsta and school students. The number o public servants almost doubled between 2003 and2012, rom 1,256 to 2,162, largely through an increase in the number o teachers and healthworkers as a result o donor- nanced policies. The number o registered villages, and hence o the executive o cers who administer them, increased by around one third, to ninety-one. 8 Thenumber o shops large enough to be registered at the district level, that is excluding most smallvillage shops, increased rom 204 to 479 9 during the same period. This increase refects twoimportant legacies o liberalisation, the legalisation o private sector pharmacies and the morerecent creation o agricultural input supply shops, 10 as well as its unexpected cultural legacy, the

7 Berrys point is that pre erence is or timing o multiple varieties and crops, not a single variety, even where high yieldingvarieties were made available to armers.

8 Figures supplied by the Planning O ce, Ulanga District Council, June 2012.9 Data supplied by the Trade O cer, Ulanga District Council, July 2012.10 These two legacies were part o the NAIVI programmes. Because input supply is contracted to the private sector, award-

basis private sector suppliers have come into existence around the programme, which provided incentives in the orm o capacity building and certi cations.


15/37

8

proli eration o male and emale hair salons. 11 The number o secondary schools and the numbero students and teachers has increased hugely, rom a single government secondary school anda Catholic seminary in the 1990s to thirty-two in 2012, o which ve are private boarding schools.

The presence o teachers and public sector sta contributes to the local demand or goods andservices, including the sale o ood crops or home use and trade, which a proportion o localgovernment sta are engaged in. This activity has been made possible by another institutionalchange: the availability o credit or public sector workers through private sector providers suchas Faidika and Bayport. Once perceived by public sector workers as an unattractive, largelymarginal and inconvenient posting, Ulanga is now seen as a place where those with a combinationo capital and initiative can make money. The expansion o secondary schooling has economic

e ects, creating demand or services such as hair cutting and cooked ood in the vicinity o schools,and increasing demand among non-waged people or cash income in order to meet the costsassociated with schooling. In villages near Mahenge town, where there are three ward schools andthree private secondary schools, 12 the indigenous categorisation o cassava as a ood crop o lastresort is changing. Nowadays, mihogo ni biashara (cassava is business). Because cassava can beharvested fexibly, it can either be sold in bulk or in small quantities once it is matured, providing ause ul everyday source o cash or small armers.

Greater need or cash or schooling along with an expanded range o goods and services in themarket, including transport and telecommunications, as well as subtle changes in aspiration andtaste, make getting cash more important, and, with the crippling infation in the country, hardto keep hold o . Although many aspects o quality o li e are much improved compared to theearly 1990s, popular perceptions among the poor are that li e has become more di cult. Thecontemporary economy in Tanzania is highly unequal, with rural rates or casual labour throughpiecework averaging between three and ve thousand shillings a day; a kilo o sembe costs 1,200shillings and a kilo o meat 5,000 shillings. 13 Opportunities or wage labour are limited and poorlypaid. Guest-house workers earn around orty thousand shillings a month. Rates o pay or unskilledlabour in the district, and indeed in Tanzania generally, even where they exceed minimum wage, donot provide a plat orm rom which rural people can trans orm their livelihoods. For highlands armersin particular, who do not have large rice shambas and who lack the capital or labour to open up newareas or cultivation, opportunities to get more cash income rom agriculture are limited. Against this

background it is not surprising that they are seeking other ways o making money, and respondingenthusiastically to the changed structure o opportunity created by wider changes in the district andnational economy, as well as those brought about through government and NGO projects aimedat rendering local economic practices more productive. The savings and livestock practices o men and women small armers are oriented towards relatively short-term cycles o return to meetother demands, such as schooling, rather than invested in scaling up production. The time scale o investment strategies, generally less than one year, is rein orced by the parallel institutional logics o school calendars and savings groups and the growth cycle o domestic pigs.

Modern Cattle

Although both dairy cattle and pigs have been kept by a small number o people or some time,having been rst introduced by Catholic missionaries rom Germany and Switzerland in the early

11 For accounts o the growth o hair salons a ter liberalisation, see the work o Amy Stambach (1999). In addition to thecultural arguments these authors make about ashion and identity, salon businesses, especially male barbershops, are alow-cost business to establish. Owners o salons usually charge barbers a ee to use their premises.

12 All are operated by the Catholic Church. These are Kasita Seminary, St Josephs Boys Secondary School, and St AgnesGirls Secondary School at Kwiro. There is another girls boarding school, operated by the church at Ruaha, about 10 km

rom Mahenge town.13 Prices are rom June 2012. The meat price quoted is or bee .


16/37

9

twentieth century, the past ten years has witnessed a striking proli eration o households keepingpigs and a steady increase in those keeping one or two dairy cattle. Dairy cattle, popularly re erredto as ngombe wa kisasa , or modern cattle, are di erentiated rom indigenous breeds kept bypastoralist groups and subjected to radically di erent management regimes. 14 So-called moderncattle are improved or grade cattle, bred rom crossing exotic strains with East A rican Zebu. Theyare kept primarily by smallholders or milk production under strict conditions o zero grazing. 15 Designated as improved breeds, these animals have potential daily milk yields in excess o teenlitres, ar higher than that o indigenous cattle, but are less resilient and have lower resistance todisease.

In 1990 the main source o resh milk or the highlands part o the district, including Mahenge town,was the arm run by the Catholic seminary, with perhaps three other households keeping two orthree cattle. By August 2012 the seminary herd was greatly reduced and milk was supplied rom atotal herd o around one hundred animals kept by some ty smallholders in the villages surroundingMahenge. 16 A urther twenty-two dairy cattle were kept by groups and individuals in the village o Mwaya some orty kilometres to the south. 17 Whereas pigs are usually bought by people themselves

rom stock bred locally by other smallholder armers, the majority o dairy cattle in the district havebeen distributed to individuals or groups o armers through donor- unded and government projects

or improving agricultural technology. The rst dairy cattle made available through such schemeswere allocated to a number o women armers through an Irish Aid district development programmein the late 1990s. Since then a number o dairy cattle smallholder schemes have been implementedin di erent parts o the district by organisations, including Caritas, working in partnership withHei er International, and the district agricultural department through district agriculture developmentprogrammes and the second phase o the Tanzania Social Action Fund (TASAF). 18

The introduction o crossbred dairy cattle in this district is challenging. Long-term residents o Ulanga, living close to an area historically subject to in estation by tsetse, do not have a cultureo livestock keeping. 19 Not only do ew people have an a nity with domesticated animals, butlivestock such as cattle or goats do not play a culturally elaborated role in economy and socialli e (Akumuluka & Madulu 2006). Among Pogoro people who live in the Mahenge area, marriageexchanges were historically negotiated through the institution o brideservice, in which a son-in-law

provides labour to his parents-in-law or an agreed period o time, rather than the trans er o cattle.20

Marriage trans ers today take the orm o cash payments, as they have since the 1950s

14 What are categorised as dairy cattle, as opposed to indigenous cattle, in Tanzania are Friesian, Jersey, and Ayrshirecattle, which may have been crossbred with indigenous East A rican Zebu. An estimated 70 per cent o milk productionis derived rom indigenous cattle (Njombe & Msanga n.d.: 7).

15 In peri-urban areas where there is lower risk o disease they are o ten kept under a semi-intensive regimen o grazing andsupplementary eeding. .

16 According to gures supplied by the district agricultural o cer or Ulanga, in December 2012 there were a total o 102dairy cattle in the villages o Makanga, Nawenge, and Isongo, close to Mahenge town. In 2002 there were 38. Interest-ingly, the number o animals in Isongo declined during this period rom eight to six.

17

In May 2012 there were a total o 22 dairy cattle in Mwaya. Fi teen belonged to a TASAF project that had provided twoarmer groups with animals in 2008. The remaining seven animals were descended rom the original six cows provided bythe Irish project in 2002 or 2003. TASAF had originally provided each group with six animals. All the cattle are now lookeda ter by a single herdsman, paid by the cattle owners, who supplied this in ormation.

18 TASAF is a multi-donor supported social action und, now entering its third phase. Phase one ocused on building localin rastructure, phase two supported in rastructure and investments which would promote economic growth, and phasethree is oriented towards supporting vulnerable groups through social cash trans ers.

19 Tsetse was the rationale or resettlement under the British colonial administration.20 Some material items were provided by the grooms amily, including hoes and cloth. For an account o marriage ex-

changes in the Ulanga valley in the 1930s, see Culwick & Culwick (1934).


17/37

10

(Green 2003). The reproduction o a herd is not oundational to social reproduction as it is amongagro-pastoral and pastoral communities. Moreover, although milk is becoming a prestige ood amongthe East A rican middle class, associated with urban consumption practices (Urassa & Raphael n.d.:2; Nkya et al. 2007; Tebug et al. 2012), milk products are not a routine part o local diets. Finally,household labour and gender relations are not structured around responding to livestock needs.

The needs o stall- ed animals are substantial. These animals are looked a ter as i they werepeople, one armer told me. We have to bring them ood and medicine. They are treated like kings!

Modern cattle, or those descended rom stock bred or the cold pastures o northern Europe,are not well adapted to tropical conditions. Susceptible to overheating, to skin cancer rom harshsunlight, and to tick borne diseases, they are also vulnerable to ingesting poisonous plants. Giventhe risks to these animals when grazing, the projects which promote these animals insist that theanimals have what an extension o cer called room service, that is that they are stall- ed. Stall-

eeding cattle is labour intensive, especially where armers rely on grasses cut rom common pasture. A single animal requires between ty and seventy kilos o cut grass each day. Grasses alone do notprovide the level o nutrients needed or milk production. Cows diets must be supplemented withseveral kilos o maize bran. In addition to regular worming, dairy cattle must be sprayed weekly toprevent tick borne disease. Sprays and worming have associated costs, but these are not signi cantcompared to the labour time involved in the daily task o accumulating su cient odder.

Originally directed by the Irish development programme at poorer emale household heads, on the

assumption that access to milk would impact on child nutrition and on womens incomes, thosehouseholds managing to maintain animals over the longer term are generally those with enoughcash to hire labour to eed them (Akumuluka & Madulu 2006). 21 In Ulanga these may be male or

emale headed. For households without access to labour or the cash to hire it, securing su cienteed is di cult, as is nding the time to sell milk beyond the villages. Most milk is not sold within

villages, where its price puts it beyond the reach o the majority o households. Rather, it is soldto restaurants (mgahawa) and to wage workers in the district centre, many o whom come romcommunities where consumption o dairy products is an accepted part o local diets. Becauselivestock have minimal social value except as sources o milk or sale, the distribution o rightsin stock is not a means o distributing the labour inputs needed or their upkeep, as it is amongpastoral communities. Few armers manage to increase their stock o animals. In act, the oppositeis not unusual, with armers o ten losing animals to disease or seeking to slaughter stock to meet thecosts o maintaining their remaining animals. In Mwaya, where cattle belonging to individuals wereherded by a paid stockman who also looked a ter the animals that were the property o two TASAFsupported armer groups, animals had been slaughtered to meet his wage costs which could notbe recovered via milk production.

Regular calving is necessary not only or expanding herds but also or maintaining levels o milk production. Although calving is the route to pro table dairying, it is perceived by small armers ascostly. These costs derive rom the ees bull owners charge or insemination, 22 and the short-termloss o the animals milk-producing capability, as what is valued is not the potential increase in stock through calving but the daily income rom milk sales. Although purebred exotic cattle are claimed tohave high milking potential with yields o up to thirty litres a day, and crossbred ones around teenlitres, the reality o inadequate eeding means that average yields are closer to six to eight litres. 23

21 For a detailed account o labour issues in smallholder dairy arming in Tanga, see Swai et al. (2005).22 As o 2012 these ees were between teen- and twenty thousand shillings.23 These yields are not unusual in Tanzanian smallholder dairying due to the problem o eed. See, or example, Urassa &

Raphael (n.d.: 5); Swai et al. (2005), Nkya et al. (2007: 629), and Tebog et al. (2012) or a comparable account o thesmallholder dairy sector in Malawi.


18/37

11

This kind o management regime rarely breaks even unless an owner has other income sourcesto subsidise his or her dairy business to the point o pro tability (c . Nkya et al. 2007: 631). Similardi culties have been reported or smallholder dairy armers in other parts o Tanzania (Swai etal. 2005; Urassa & Raphael n.d.). 24 Where labour is the limiting actor and indeed is likely to havecontributed to household poverty as a criterion o initial bene ciary selection, the viability o stall-

ed dairying is open to question. Despite these constraints, dairy cattle remain popular on theparticipatory wish lists generated though national district planning processes as icons o modern

arming (c . Haugerud 1993). 25 The district as a whole was home to a total o around 44 dairycattle in 2002. At the close o 2012 there were 124. This increase in number is not explained bynatural reproduction. Rather, it is project driven. 26 Extending dairying has political and ideologicalsupport in the categorisation o dairy cattle as an improved agricultural technology, in assumedlinkages between livestock and higher income, and access to milk products and improved nutrition(URT 2011). 27 These correlations between income and nutrition do not exist because armers makedairying a source o income and either buy better ood or use milk to improve the diets o householdmembers. They exist because better o armers are more likely than their lower income neighboursto be able to sustain dairy arming. 28 Livestock keeping, or some households, is an e ect ratherthan a determinant o household socioeconomic status.

Pig KeepingDomestic pigs, like dairy cattle, were also introduced to the Mahenge highlands by Catholicmissionaries, whose involvement in the district continues to exert an infuence on patterns o economic development. Current patterns o religious a liation in Ulanga refect ormer missionarycontrol over the delivery o basic education and health services. In the highlands area surroundingMahenge, close to the centre o missionary infuence, the majority o the population are a liatedto the Roman Catholic Church. In the valley areas, on the other hand, the population is morelikely to be a liated to Islam. Religious identity structures the new economy o pig production inthe highlands, where producers and consumers o pork meat are Christian. Until the mid-1990schurch-related enterprises, such as the seminary arm and piggeries owned by individual priests,supplied the bulk o locally consumed pork meat. The small number o individuals who kept pigswere public sector or mission employees who kept ve or six animals or occasional slaughtering.

Pork meat was not routinely available or sale. It was sold at Christian estivals and on holidays. Pigswere slaughtered at the owners home, and buyers came to take away the meat. Pork and bee soldat the same price per kilo, but pork was valued or its atty content and association with sikukuu .Changed tastes and practices, along with greater numbers o wage workers and boarding students,have created daily demands or pork meat that are no longer reserved or sikukuu but sold as kiti

moto , ried pork. Today probably around three quarters o households in villages in the highlandsarea, where the majority o people are Christian, keep on average one or two pigs, usually in raisedwooden enclosures situated adjacent to their homes. Demand or pigs in the orm o meat and

24 Nkya et al. state that, or Tanga Region, [ ]amers spent on average...between 39% and 77% o income rom milk oneed costs yearly (2007: 627). They continue, in some arms . eed costs were 100% o milk income (op. cit.: 632).

25

For an overview o participatory planning in Tanzanian rural development, see Green (2009).26 This would seem to be the case or elsewhere in Tanzania. For example, Swai et al. state that 61.5 per cent o armssurveyed in their Tanga study had ewer than ve years o dairy arming experience, and over hal o study householdshad attended some orm o training, These gures suggest that entry to dairy arming or this category was acilitated byexternal projects (2005: 7). A study o the adoption o crossbred dairy cattle in Mbeya and Iringa between 1978 and 2001

ound that price was not a signi cant determinant o armers decisions to adopt dairy arming. This was due to the actthat the crossbred cows were mainly introduced by oreign development agencies (Abdulai & Hu man 2005: 653).

27 The Livestock Policy o 2006 aims to double per capita consumption o milk, rom 40 to 80 litres (Njombe & Msanga n.d.:8).

28 Hence, it was not the primary source o income or armers in the studies cited on Tanzanian dairying.


19/37

12

piglets or rearing is consistent. Piglets are o ten sold long be ore they are weaned, a ter which theirnew owners take them rom their mothers to rear at home. The price o weaned piglets rose romthirty- ve to orty thousand shillings between February and June 2012. While some people continueto slaughter and sell meat rom their homes a ter putting up notices advertising when and wheremeat will be available, it is now more usual or kiti moto entrepreneurs to seek out pig keepers andtake the live pig away or slaughter or to buy all the meat, except or the head and intestines. Pork meat is sold at our thousand shillings a kilo. Fried pork sells or ve thousand. Pork meat is slightlycheaper than bee and much cheaper than goat meat. Butcheries in the town sell only bee . Wildmeat, including elephant, is widely traded in villages. Pig meat is sold either rom peoples housesor rom kiti moto outlets.

The daily sale o pork meat and the evolution o a kiti moto meat sector is a very recent phenomenon,originating in the urban ashion or ried pork sold in larger cities, but made possible by a combinationo technology and changes in the disposable income o a proportion o consumers in what is lessa rural centre than a new urban periphery. The availability o electricity and re rigerators or salein the district centre, along with access to credit with which to purchase them, makes innovationslike the kiti moto trade viable. This combination o credit, Christianity, paid workers, electricity, andtechnology means that a kiti moto seller can sell daily and deal with unsold volumes o pork. Changesin the wider economy that create a daily market or pork meat have undoubtedly contributed to theincreased uptake o pig keeping in Ulanga since the 1990s.

This kind o increase is paralleled in other parts o Tanzania. A study on smallholder pig productionin Iringa ound that 97 per cent o households kept an average o 2.9 animals. The majority o surveyrespondents, some 58 per cent, had acquired their rst animal recently, that is, between 2001 and2006 (Karimuribo et al. 2011). Other actors are also relevant. Unlike modern cattle, mixed breedpigs o the kind which are locally available are relatively cheap both to procure and to rear (Lekule& Kyvsgaard 2003). Although they gain more weight with higher quality inputs, including nutritionand quality o housing, they can be kept adequately in basic wooden enclosures. Pigs do notrequire costly sprays and injections. With regular worming, vitamin A supplements, and occasionalwashing, they can be expected to thrive until they are ready or slaughter. As with cattle, eedingpigs is the big constraint or highlands armers, who do not actively integrate pigs ood needsinto their shamba management strategies. Average household waste in such production systemsusually provides most, but not all, o the ood needs o a single animal (Lekule & Kyvsgaard 2003).Fed on scraps, bananas, cassava, and greens, pigs also require a daily portion o bran, pre erably

rom maize. For the majority o households that depend on the market or access to maize or mucho the year, this has to be bought. Local armers said that a pig requires at least one kilo o maizebran daily. One debe , actually a twenty-litre bucket, nominally equivalent to twenty kilos, sells ortwo thousand shillings. 29

The perceived costs o eeding pigs impact the ways that pig production is managed by smallholderswho keep costs to a minimum by having only one or two animals and prioritising maturing orslaughter over breeding. Farmers may slaughter sows rather than breed rom them, using the money

rom the sale o meat to buy a weaned pig to raise. Similar strategies o pig management can bediscerned rom a study o smallholder pig production in Iringa, where the majority o households

29 Weights and measures are an issue o contestation in Tanzania, as the government attempts to promote and en orcethe use standard measures by weight. Rural conventions are less exact, with volume pre erred over weight and standardconventions about weight equivalents involved or each measure by volume rom small kilo equivalent measures o pusu/ kimbo (1 litre) to debe/ndoo and gunia (sacks). Traders pre er volume measures because these can be adjusted upwardsor downwards depending on how the measure is lled. For a good description o volume measures used by markettraders in Kilimanjaro, see Pietila (2007).


20/37

13

kept growers, or maturing pigs, in pre erence to either piglets or adult animals (Karimburo et al.2011). Such approaches are characteristic o what Lekule & Kyvsgaard have called the paradox o traditional pig keeping, in which this mode o husbandry, despite its high losses and low o take, isnevertheless more sustainable than intensive pig keeping. Traditional pig keeping, they maintain,cannot be understood simply in the pro t and loss calculation o the small agricultural enterpriseenvisaged in development policy: The pig has unctions that are not refected in a simple economicbalance. The pig is a source o capital income, which can be realized at times o the year whenmajor expenses are oreseen, and it can also be used as a way to put aside small amounts o moneywhich alternatively might evaporate (2003: 114).

Savings Groups The pre erence or predictable returns rather than maximising potential pro t drives the extension o pig keeping in Ulanga. It also infuences the rapid uptake o another socio-economic innovation, a

orm o accumulating savings and loan group popularised in Ulanga and in other parts o Tanzaniaby the NGO Care International. These are a globally dispersed orm o nancial organising in whicha limited number o members commit to saving regularly or a xed period o time, usually one year.Group members are able to borrow a xed proportion rom group savings or a set period o time,usually three months, depending on the amount they have contributed to the und. Loan chargesare returned to savers in proportion to their contribution at the end o the groups cycle, giving areturn on investments that can be as high as thirty per cent. As part o its Access A rica Programme,Care established a project in Ulanga District in 2009 with the aim o ostering the uptake o its VillageSavings and Loan methodology and promoting the ormation o local groups (Hendricks & Chidiac2011: 135).

The Village Savings and Loan Associations (VSLAs) promoted by Care are a variation on what arere erred to in the literature on credit groups as Accumulating Savings and Credit Associations, or

ASCAs (Bouman 1995: 371). ASCA groups have been interpreted by some analysts as an adaptationo the Rotating Credit and Savings Associations (ROSCA), which are common in low-income settingsinternationally. ASCAs are di erentiated rom ROSCAs by members control over the timing o borrowing and the accumulation o the und, which can then be divided proportionally among group

members. The act that members can deposit variable amounts means that memberships o ASCA groups are likely to be less homogenous than those o ROSCAs, where contributions and loans arexed (Bouman 1995). ROSCA groups have a long history in sub-Saharan A rica, certainly rom the

early twentieth century, where their rapid expansion is linked to emerging commercial sectors andwage employment (Delancey 1977; Bouman 1995). In West A rica the evolution o rotating savingsgroups is associated with migrant labour and with orms o commercial credit among ethnic traderassociations, including those o the Nigerian Hausa (Yusu 1975; Miracle et al. 1980). Rotatingsavings groups are reported or East and Central A rica rom the 1950s, when waged workers in

Tanzania were members (Ardener 1963: 207).

While some analysts regard the ASCA as a kind o evolved version o the ROSCA, which is thuslikely to have emerged in contexts where the ROSCA orm and its variants were already established(Bouman 1995: 377), it seems to be the case that the ASCA was introduced to East A rica throughNGO programming. The institutional designs o the savings groups distributed through theseprogrammes are the product o hybrid combinations o indigenous modalities or the rotation o capital and labour (Delancey 1977; 319) with the mechanisms o bureaucratic audit. ASCAs wereintroduced to Kenya via a US NGO programme in the 1970s. There they have evolved into a semi-commercial orm o small-scale savings clubs that are externally managed by private sector agents


21/37

14

(Johnson et al. 2002). The VSLA model promoted by Care is inspired by a orm o small-scalesavings group originating in Niger 30 (Anyango et al. 2007: 11). Modi ed through the requirements o programming replicability and the need to manage per ormance, the VSLA methodology combinesthe elements o group contributions and audit with the material orm o the lockable cash box,an arte act that, together with personalised savings books, enables the public per ormance o transparency. This institutional orm is replicated by some other organisations in Tanzania, includingSEDIT through its vicoba system.

The VSLA methodology promoted by Care is a variation on the standard ASCA architecture. VSLA groups consist o a maximum o thirty members. Each member buys between one and ve weeklyshares and can borrow up to a maximum amount based on a multiplier o what they have saved or amaximum three-month period. Borrowers pay a xed ee based on the amount loaned. At the end o the cycle, usually one year, members receive what they have saved plus interest proportional to theirshare (Mkoma 2011; Maliti 2012). Cares innovation in disseminating this orm o ASCA has beento ormalise and standardise the methodology or group management, to introduce documentationsuch as a constitution that structures the ways in which groups are managed, and to ensure thatall transactions take place in public through the per ormance o transparency. This is symbolisedby the savings box which has three locks, the keys or which are held by ordinary members o thesavings group rather than by elected o cials, namely the treasurer, chairperson, and secretary(Anyango et al. 2007: 13). Care has also encouraged the ritualisation o group savings practicesthrough standardised chants and actions around running the group meetings, including counting

and depositing money. Members o VSLA groups meet weekly, sit in a circle acing the o cials andthe savings box in strict order o their membership number (one to thirty), and participate in thesequence o saving and lending according to strict ormulaic routine.

Members buy weekly shares, up to a maximum o ve, and loans are made or a maximum o three months on the basis o money deposited by savers at a rate o interest usually around veper cent. At the end o the cycle, income rom interest on loans is distributed in proportion to thevalue o members shares, generating returns which depend on the number o shares an individualhas purchased and on the amount o loans made (and hence interest charged) by the group.Saving is a requirement o VSLA membership. Members have to save every week, but they do nothave to purchase the maximum number o shares. The end o cycle calculation and allocation iscalled an action audit in the Care terminology (Anyango et al. 2007: 11), an apt description o theper ormative practice o public auditing through which these groups make saving sa e.

Care Groups in Ulanga ASCA groups, also known as hisa groups in connection to the shares that members pay in, haveproli erated in Tanzania over the past decade (Mkoma 2011; Maliti 2012). Some o this expansion isattributable to the e orts o donor organisations that have enthusiastically invested in microcredit asan intervention combining the ideal o entrepreneurship with aspirations o community development(Roy 2010). External unding alone cannot account or the mass expansion o savings groups inmany areas. What is truly remarkable about the ASCA groups as a social technology is their rapidspread and uptake in A rica even where there are no external interventions (Hendricks & Chidiac2011). This was the case or VSLA groups established by Care in Zanzibar, which have continuedto increase by 37 per cent annually several years a ter the programme ended (Anyango et al. 2007:14). A similar process is underway in Ulanga, where what are popularly called care groups continueto be ormed even a ter Cares two-year intervention ormally ended in 2011. According to the data

30 Hugh Allen, personal communication.


22/37

15

gathered by the umbrella organisation or savings groups initiated as part o the Care programme,in mid-2012 there were a total o 344 groups in the district.

In Ulanga, as in Zanzibar, there are various problems with the sustainability o the institutionalarchitecture envisaged by Care or the establishment o VSLAs in the districts nancial landscape(Anyango et al. 2007). The umbrella organisation struggles to cover costs, to bring memberstogether, and to provide the cash boxes and savings books demanded by newly ormed groups.

This is not surprising. One o the striking eatures o both ASCAs and ROSCAs as an organisationalorm is that autonomy is designed into the system (Bouman 1995: 381). Because groups use their

own resources to save and to provide loans, a group can be ully unctional solely through its ownmembership. There ore, while the ASCA can potentially be integrated into wider systems, as inthe Kenya example where groups are managed by external agents, or be incorporated into apexstructures, as in the Care programme design, incentives or groups to seek integration depend onwhether members perceive value added rom it. In Kenya, where ASCAs have many members andwhere co-presence ceases to be a practical means o enacting accountability, compliance, andaudit unctions are contracted out to an external agent (Johnson et al. 2002). For the Tanzaniangroups that were ormerly part o the Care programme, ritualised internal management and aper ormative audit at weekly meetings enact trust and make integration into wider systems lessattractive. Furthermore, the autonomy o groups that are not part o wider structures provides themwith the fexibility to adapt to members interests.

The original design o the Care programme anticipates the engagement o members in settinggroup rules o operation, including share prices, responsibilities o elected o cials, and ne amountsand other contributions that are annually reviewed at the start o a new cycle. Share prices vary,depending on the membership o the group. Although Boumans insight that ASCAs were likely toattract a more di erentiated membership than ROSCAs, because members can vary the amounto weekly purchased shares up to the maximum amount, is borne out to some extent in Ulanga,where income di erentials between members would seem to be the norm, there was signi cantdi erentiation between urban groups, which had higher share values, and more rural ones, whereshare prices were as low as 500 shillings. One urban group, with a membership comprising a highproportion o traders, had share prices o 2,500 shillings. At the start o its second cycle o operationthis group insisted that each member should pay in a lump sum o ty thousand shillings so thatloans could begin earlier and returns would be potentially higher based on the size o loans. Thisgroup also gave short-term loans at higher rates o interest in the nal quarter, when groups tend tohave ceased lending in order to recoup assets prior to the payment o dividends and shares.

Ulanga care groups had not, as yet, di erentiated themselves to a signi cant extent in order torestrict membership to those with similar investment capabilities. People with more money toinvest did not make larger investments by starting new groups that would sell more shares or moreexpensive shares. As in some other countries where small savings groups are numerous, peoplepre er to join several groups concurrently rather than increasing their shares in a single group (Geertz1962: 245; Yusu 1975: 171). Joining multiple groups remains a better option than membership ina single group with higher share prices, because returns on savings are only generated throughthe interest paid by borrowers. Having more invested in a single group will only yield a return i borrowing is also proportionately higher. For the ew individuals who are members o ve or sixgroups, returns on shares are likely to be substantial, as the number o potential borrowers, andhence ee income, is large.


23/37

16

Despite the strong policy narratives in Tanzania and elsewhere concerning the importance o accessto microcredit as the missing link in trans orming the livelihoods o the poor, borrowing was notthe primary motivation or joining Care groups in Ulanga. What people valued, as research onthe motivations o ROSCA members in many countries has shown, was the regularity o saving,especially as this was compulsory (Ardener 1963: 218), the act that money put away in a savingsgroup was o limits to the demands o others (Ambec & Triech 2007; Bouman 1995: 375), andthe option o crisis insurance through the groups social ( jamii ) und to which members contributeweekly (Geertz 1962: 247; Ardener 1963: 218). As or the ASCAs, such as the care groups, theirgrowing popularity in Ulanga is based on member hopes o not only saving but also making a returnon their investment. People with cash to invest by joining multiple groups can make good returns, aslong as other people keep borrowing. Because saving is pre erred over borrowing, and saving withadded return most valued o all, some Ulanga groups have made borrowing compulsory. Groupsalso undertake additional activities in order to increase the value o their shares. The most usualo these is the purchase and resale o soap powder to members at prices that are the same as orhigher than market rates. As with borrowing, this is o ten compulsory. 31

The move towards selling soap and other goods to members can be interpreted in various ways. A ormer project o cer working with the Care programme in the district regards this activity as anoutcome o the original training that group members received, which aimed to promote nancialliteracy and entrepreneurial skills. The intention was, he said, to give people experience o incomegenerating activities that could then be expanded into group projects. Group members themselves

were more likely to regard the compulsory sale o goods as another way o saving and borrowing.Members in e ect receive goods on credit and repay in instalments. Another interpretation is to viewcompulsory borrowing and purchasing as an entrepreneurial practice that consolidates the groupas the enterprise that generates income or the individuals in the group. The compulsory nature o the purchase by all members means that returns, as with saving, are proportional to cash invested.Some members cannot simply make pro t out o the cash needs o others. Investing in care groups,as long as they are success ul, guarantees a return to members in addition to the amount theyhave saved. Where returns can be as high or higher than the thirty per cent average or Tanzania(Hendricks & Chidiac 2011: 140), care groups are attractive as an income-generating option ormen and women, and one that works well or those, especially women, who have limited start-upcapital. 32 Ulangas care groups demonstrate some o the core characteristics that account or theinstitutional durability and rapid proli eration o the ASCA/ROSCA orm in many countries. The actthat integration into the apex is either partial or non- unctional allows groups to experiment withdi erent ways o organising according to member interests. Importantly, despite the promotion o groups as an organising principle o community development interventions, a large part o the appealo the savings groups is that both investments and returns are individualised. The group providesthe structure through which capital is amassed in order or loans to be disbursed, but the group isnever an end in itsel . It is the means through which individual members realise their objectives.

31 Compulsory borrowing is common in other countries as well. See Bouman (1995) or a discussion.32 Returns on pig raising are potentially higher, but the start-up costs are also much higher, as are the labour inputs required.

Beer brewing continues to be a viable income-generating option or women. However, start-up costs are quite high, andproducing beer is high risk. In 2012 in Mahenge, brewing sixty litres o maize beer rom twenty kilos o maize could cost14,500 shillings (excluding rewood), and i sold at ve hundred shillings a litre, brewing would generate a pro t o 15,500shillings. Such returns are not guaranteed. Beer can spoil or may not sell, and rotas in villages limit the amount a womancan brew even i she has access to the necessary capital and equipment (oil drum). For an account o the economics o brewing in the 1990s, see Green (1999).

5 Accounting for the ASCAs Appeal


24/37

17

The Process of Economic ChangeCurrent policy interventions in Tanzania are based on speci c theories o economic change andhuman behaviour. Interventions, particularly in the agricultural sector, have been designed on theassumption that improved technologies will increase productivity and that armers will engage withthe market in particular ways that prioritise returns on investments. As one would expect, giventhe ndings rom numerous studies o economic change in A rica, such assumptions are unlikelyto be realised in practice (Richards 2010). Analysis o the process o economic change in Ulangasupports the claims o Richards and others that the actors impacting on changes in economicpractices are complex. Economic practices in Ulanga exhibit a remarkable dynamism, as armersadapt their livelihoods to changing circumstances and to new technologies and ways o li e. Thisresponsiveness is not new. The twentieth century was characterised by economic opportunism, aslivelihood strategies in the district evolved to accommodate the move to settled villages, cash cropproduction, and di erent regimes o incorporation into the global economy.

This process is on-going. Ulangas residents today must adapt to their place in a transitioningperiphery whose terms o integration into the world and regional economic orders are uncertain.Part o this adaptation involves the kinds o investments people seek to make and the ways inwhich men and women struggle to get cash income where returns rom agriculture remain low andwhere households with their shi ting composition continue to be the unit through which agriculturalproduction is organised. These actors infuence the kinds o innovations that have been adopted inthe highlands part o Ulanga District since the 1990s. Success ul innovations depend on the socialcontext in which their uptake becomes viable in relation to values, attitudes and the organisation o labour, and on the wider economic context in which they are embedded. This context is continuallychanging. The viability o particular economic activities is not xed. It is constituted throughtrans ormations in economic relations at di erent scales. These trans ormations are variable andcontingent. The pig economy in the highlands becomes viable in its current orm as a result o changed tastes and pre erences and availability o money that sustains the daily market in kiti moto enabled by new technologies o electricity and re rigeration. Small-scale pig keeping has low barriersto entry and can be managed by households with a single adult. Mahenges peri-urban status, largeChristian population, and growing proportion o wage workers all contribute to the growth o apork market that depends on the local sale o pig meat. Small-scale dairying, in contrast, founders

because low-income armers cannot manage the investment necessary to make it productive in asituation where demand or dairy and other cattle products is limited.

Modern cattle, symbolic o desired agricultural trans ormation in East A rica, have ailed to trans ormlivelihoods or agricultural cultures. Dairy cattle, as improved breeds, are not simply an agriculturaltechnology. The productivity o dairy cattle depends on the state o the market or dairy productsand on the quality o the relations between people and the cattle. Productivity is not an inherentquality o an asset but an e ect o the relations in which it is embedded. As in other East A ricancountries, the expansion o the mini-dairy sector in Tanzania is largely politically driven (Haugerud1993). What one o cial re erred to as ngombe wa siasa (political cattle) have a tenuous placein Ulangas economy. Although more cattle are being brought into the district by well-intentioneddevelopment programmes, it is ar rom certain that more than a minority o Ulanga armers canadapt their arm and household management strategies to make dairying productive. Where cattlehave value only or a single product, namely milk, and where markets or milk are limited and di cultto access, low milk yields become a problem that exacerbates and is exacerbated by poor eedingpractices. Finally, care groups become popular or several reasons, including their transparencyand ormalisation, but also or what they enable: getting a return on money invested through the eeslevied on lending to other members o the group.


25/37

18

The most success ul innovations are those which not only t with the demands o household labourand capital constraints but which can be adapted to changing circumstances. Pig keeping is fexibleto a degree not ound in dairying. Pigs can be made to t with the ways that people organise theirhouseholds as units o production. Animals can be kept without the need or complex interventions.Pig keeping can be scaled up or down depending on resources. Farmers keep pig numbers to aminimum, slaughter early, and let the market bear the costs o reproducing the next generationo piglets or attening and sale. The market or piglets and growers, as well as adult animals ormeat, means that they can be sold at any time. Replacement stock can be obtained without theexpense o raising a litter. A large proportion o a pigs eed requirements can be met rom domesticresources.

Likewise, care groups also o er a degree o fexibility. Although members join or a year-long cycle,they can leave the group at any time, taking their loan as a return on shares. They can contributedi erent amounts by adjusting the number o shares. Should circumstances demand, they can missa couple o weeks saving. Moreover, people with more to save or seeking larger loans can investin several groups at the same time.

The adaptability o care groups and pig keeping extends beyond the level o individual practice. The structure o care groups and their weak integration into the apex organisation means that theycan evolve fexibly, experimenting with new practices such as compulsory loans. Pig keeping hasevolved rom something carried out by a minority o people producing meat or occasional sale

into a micro-economy that ensures the requent availability o pork meat or local sale and the dailyavailability o kiti moto in the town. Dairy cattle, especially i they are stall- ed, are less amenable tofexible management or to producer autonomy. Crossbred dairy cattle require costly spraying andinoculations to protect them against disease. Becoming a dairy armer involves integration into awider system o extension services and agricultural inputs, as well as daily engagement with thesearch or a market or milk. Attitudes towards up ront costs and risk taking have an e ect on theways in which people manage small-scale dairying. The pre erence or low risk, fexible incomestrategies is not con ned to livestock. It in orms attitudes to ertiliser use as well as to the currentpopularity o cassava. The viability o cassava as a crop or sale is determined by its fexibility withina changed context o demand that creates a locally accessible market.


26/37

19

Cultures o economic practice are not static. They are constantly being remade. Practices becomeviable in relation to the contexts in which they are embedded. As demonstrated by the sociologistElizabeth Shove in relation to changes in everyday domestic practices internationally, studyingchange by ocusing on practices as clusters o behaviours and the institutional contexts whichmake them possible, as opposed to ocusing on either individual behaviour or technologies, revealsmore about the drivers o and constraints to change (2004; 2010). Contextual changes are theoutcome o multiple infuences that are not simply economic. Changed values create the contextin which certain changes are sought; changed relations may make them possible. Policy directedat changing economic behaviour must there ore pay attention to supporting the emergence o thecontexts that enable its occurrence.

For armers in Tanzania, who manage production as a household enterprise and who are subjectto uncertainty in external and internal environments and to the vagaries o terms o trade, fexibilityand adaptability are oundational to economic innovation and the sustainability o rural livelihoods(Berry 1993). This emphasis on practical adaptability through dispersed risk and the option to exit

rom economic endeavours may not t well with the imaginaries o the new armer enmeshed inagribusiness value chains and high technology. It has, however, contributed to the resilience o rurallivelihoods or actual small armers in Tanzania and elsewhere. As Jane Guyers long-term study o changing rural economies in Nigeria shows, adaptation takes place primarily through the adoptiono new crops and specialisations rather than intensi cation or increasing arm size. Rural economiesare never stagnant. They are continually evolving, albeit in ways that diverge rom conventional

models o growth derived rom radically di erent conditions (1998).

Examining the actual process o economic change and the relations through which this is structured,rather than through normative theory based on deductive generalisation, permits an alternativeevaluation that renders economic institutions in rural A rica not as ailures but as adaptively e cient(North 2005: 163) or the environments in which they work. Failure is not then an attribute o armerslack o capacity to adopt improved agricultural technologies and economic practices, but a producto inappropriate development interventions i.e. ones not designed to work with establishedpre erences or fexibility and keeping up ront risk to a minimum. Nor are they su ciently sensitiveto the changing contexts in which implementation occurs. Smart policy or rural economic growthmust ocus on growing opportunities or rural populations. This cannot be achieved by promotingan economic monoculture o agrarian uni ormity. Rather, it has to be built on the care ul nurturingo actual and emerging opportunity. Achieving this goal demands a new set o theoretical tools

or understanding economic change in rural A rica. Empirical studies by sociologists, historians,and anthropologists suggest the directions in which change occurs. This research has shownhow context, fexibility, and risk are the three actors that contribute most to the success o small

armers in adopting economic innovations. Interdisciplinary e orts should prioritise building bettermodels or underwriting more e ective policy and interventions. Smart policies aimed at promotinginclusive growth in rural areas must begin rom an analysis o the contextual actors that impacton the potential productivity o particular activities and investments. Those policies must result ininterventions designed to work with armers own pre erences or fexibility and risk reduction.

Conclusion: Smart Policy forSupporting Change6


27/37

20

Abdulai, A. & Hu man, W. (2005). The di usion o new agricultural technologies: The case o crossbredcow technology in Tanzania. American Journal of

Agricultural Economics 87(3), 645659.

Akulumuka, V. & Madulu, R. (2006). Assessment of income and socio cultural factors affecting technology adoption in Ulanga District . Ulanga District Council.

Ambec, S. & Treich, N. (2007). Roscas as nancialagreements to cope with sel -control problems.

Journal of Development Economics 82, 120137.

Anyango, E., Esipisu, E., Opoku, L., Johnson, S.,Markuu, M. & Musoke, C. (2007). Village savings andloan associations: Experience rom Zanzibar. Small Enterprise Development 18(1), 1124.

Ardener, S. (1963). The comparative study o rotating credit associations. Journal of the Royal

Anthropological Institute 94(2), 201229.

Bernal, B. (1994). Peasants, capitalism and (ir)rationality. American Ethnologist 21(4), 792810.

Berry, S. (1993). No condition is permanent. Thesocial dynamics of agrarian change in sub-Saharan

Africa . Madison: University o Wisconsin Press.

Binswanger-Mkhize, H. & Gautam, M. (2010).Towards an internationally competitive Tanzanian

agriculture . World Bank Dra t Report, paperpresented at the 15 th Annual Research Con erence o REPOA, Dar es Salaam.

Bouman, F. (1995a). Rotating and accumulatingsavings and credit associations: A developmentperspective. World Development 23(3), 371384.

Bouman, F (1995b). ROSCA: On the origin o thespecies. Savings and Development, 29(2), 117148.

Bray, F. (1983). Patterns o evolution in rice growingsocieties. Journal of Peasant Studies, 11(1), 333.Bray, F. (1998). Technics and civilization in lateimperial China: An essay in the cultural history o technology. Osiris 2(13), 1133.

Bryceson, D. (2002a). Multiplex livelihoods in rural A rica: Recasting the terms and conditions o gain ulemployment. Journal of Modern African Studies 40(1), 128.

------. (2002b). The scramble in A rica: Reorientingrural livelihoods. World Development 30(5), 725739.

------. (2010a). Agrarian undamentalism or oresight?Revisiting Nyereres vision or rural Tanzania. InHavnevik, K. & Isinika, A. (eds). Tanzania in Transition:From Nyerere to Mkapa . Dar es Salaam: Mkuki naNyota Press, 7191.

------. (2010b). The birth o a market town in Tanzania: Towards narrative studies o urban A rica. Journal of

Eastern African Studies 5(2), 274293.

Bryceson, D. & Jnsson, J. (2010). Gold diggingcareers in rural East A rica: Small-scale minerslivelihood choices. World Development 38(3), 379392.

Chachage, C. L. S. (1988). British rule and A ricancivilisation in Tanganyika. Journal of Historical Sociology 1(2), 199223.

Chisinga, B. (2011). The Farm Input Subsidy Programme (FISP) and the seed industry inMalawi: The making and unmaking of an effectivedevelopmental state? Paper presented at A ricanStudies Association Annual Meeting, Washington,DC.

Cooksey, B. (2003). Marketing re orms? The riseand all o agricultural liberalisation in Tanzania.Development Policy Review 21(1), 6791.

------. ( 2012). Politics, patronage and projects: The

political economy of agricultural policy in Tanzania .Working paper 040, Political Economy o AgriculturalPolicy in A rica (PEAPA), Future AgriculturesConsortium.

Coulson, A. (1982). Tanzania. A political economy .Ox ord: Clarendon Press.

Craig, D. & Porter, D. (2006). Development beyond neoliberalism?: Governance, poverty reduction and political economy . London: Routledge.

Culwick, A. T. & Culwick, G. M. (1934). The unctionso bridewealth in Ubena o the Rivers, A rica . Journal of the Royal African Institute 7(2), 140159.

Delancey, M. (1977). Credit or the common man inCameroon. Journal of Modern African Studies 15(2),316322.

References Cited


28/37

21

Diao, X., Headey, D. & Johnson, M. (2008). Towarda green revolution in A rica: What would it achieve,and what would it require? Agricultural Economics 39(s1), 539550.

Ellis, F. & Mdoe, N. (2003). Livelihoods and ruralpoverty reduction in Tanzania. World Development 31(8), 13671384.

Fairhead, J. & Leach, M. (2005). The centrality o thesocial in A rican arming. IDS Bulletin 36(2), 8690.

Fatton, R. (1985). The political ideology o JuliusNyerere: The structural limitations o A ricanSocialism. Studies in Comparative International Development 20(2), 324.

Geertz, C. (1962). The rotating credit association: A middle rung in development. EconomicDevelopment and Cultural Change 10(3), 241263.

Green, M. (1999). Trading on i

Understanding the Process of Economic Change: Technology and Opportunity in Rural Tanzania

Documents