Agriculture for development: new paradigmand options for success

Alain de Janvry∗

Abstract

The role of agriculture as an instrument for industrialization had been rigorously conceptualized in the 1960s and 1970s underthe classical paradigm of development economics. After many implementation failures under import substitution industrializationpolicies and protracted neglect of agriculture under the policies of the Washington Consensus that followed the debt crisis,agriculture has gradually returned in the development agenda, especially with the food crisis. We argue in this article that a newparadigm has started to emerge as to how to use agriculture for development, pursuing a broadened development agenda. Weexplore the specifications of this paradigm and discuss conditions for successful implementation.

JEL classifications: B20, O13, Q18

Keywords: Agriculture; Industrialization; Development; Governance

1. Introduction

In the classical paradigm of development economicsthat prevailed in the 1960s, agricultural growth washeld to be the key pillar for industrial growth, itself seento be synonymous with economic development. Theparadigm was anchored in compelling success stories,from the long history of the “Western Experience” tothe then-recent “Asian miracles.” And it was supportedby rigorous modeling exercises. But, in spite of severalsuccessful cases, the implementation of this paradigmwas running into increasing difficulty in the 1970s andearly 1980s as policy favored import substitution indus-trialization with strong antiagriculture price policy bi-ases. Integrated rural development strategies, designed

∗Department of Agricultural and Resource Economics, Universityof California, Berkeley, CA, USA.

Originally delivered as the Leonard Elmhirst Lecture at the27th IAAE Triennial Conference “THE NEW LANDSCAPE OFGLOBAL AGRICULTURE” International Convention Center, Bei-jing, CHINA, 16–22 August, 2009.

to meet the broadened development objectives intro-duced in the 1970s that included poverty and inequal-ity reduction, were also proving difficult to implementsuccessfully, in part because of the low profitabilityof agriculture and in part because of excessively com-plex state-led approaches. With the debt crisis of 1982,and the subsequent implementation of stabilization andadjustment policies under the Washington Consensus,use of agriculture as an instrument for developmentwas disregarded in favor of other approaches to de-velopment such as open economy industrialization toaccelerate growth and cash transfers or workfare pro-grams to reduce poverty. With a few notable exceptionssuch as China and Vietnam where smallholder-basedagricultural growth was pursued vigorously, the eco-nomic, social, and environmental costs of this neglectof agriculture have been huge.

In recent years, a number of economic, social, andenvironmental crises have attracted renewed attentionto agriculture as both a contributor to these problemsand a potential instrument for solutions. This return

c© 2010 International Association of Agricultural Economists

18 A. de Janvry

to agriculture is happening in a context where devel-opment objectives have been significantly broadenedand under contextual conditions to achieve agricul-tural growth that have changed markedly, implyingthat renewed attention to agriculture cannot be im-plemented by returning to the classical paradigm ofdevelopment economics. A new paradigm has startedto emerge where agriculture is seen as having the ca-pacity to help achieve several of the major dimen-sions of development, most particularly acceleratingGDP growth at early stages of development, reduc-ing poverty and vulnerability, narrowing rural-urbanincome disparities, releasing scarce resources such aswater and land for use by other sectors, and deliveringa multiplicity of environmental services. Yet, reneweduse of agriculture for development remains highly in-complete, falling short of political pronouncements.The underperformance of agriculture for developmentis associated in particular with continued under- andmisinvestment in agriculture by most governments andinternational donors.

We propose in this article the interpretation that thisgap in performance is due to two main causes: one is aninadequate reconceptualization of the role of agricul-ture for development to correspond to the new objec-tives and the new contexts; the other is an inadequateredesign of the approaches to ensure that agricultureeffectively contributes to development. Reconceptu-alization would involve: (1) Formalizing the comple-mentarities and trade-offs in the multiple functions thatagriculture has for development in the emerging con-text, (2) designing the process whereby agriculturalgrowth is achieved in order to attain the desired di-mensions of development beyond those derived fromthe market-driven economic competitiveness of privateagents, and (3) redefining the role of the state in set-ting social priorities among conflicting functions and inovercoming market failures for agriculture. Redesign-ing approaches for effective implementation would in-clude: (1) Experimenting with new approaches to us-ing agriculture for development, identifying the causesof success, and internalizing lessons for scaling up,(2) fixing the governance structure for agriculture toenable the state to achieve its new functions in us-ing agriculture for development, and (3) committing,through institutional designs, the state and the inter-national community to support the long-term role of

agriculture for development above price and politicalcycles.

We argue that this can and must be done, but thatsuccess requires recognizing and confronting the sub-stantial difficulties and exigencies of working with thenew paradigm of agriculture for development, includ-ing the needs for greatly enhanced levels of resources,expertise, coordination, and sustained commitments.Given the current and predicted world conditions forfood and agriculture, there is little choice but to meetthe urgent imperative of succeeding with this effort.

2. From classical paradigm of agriculturefor industrialization to contradictions and neglect

There has been a long and distinguished tradition indevelopment economics and in the practice of devel-opment in focusing on the role of agriculture in devel-opment (Johnston and Mellor, 1961; Mellor, 1976).1

In this classical development paradigm, strongly an-chored in both history and theory, agricultural growthwas seen as the engine of industrialization, structuraltransformation of the economy, and aggregate growth.The classical paradigm found its origins in the his-torical regularity linking industrial revolutions to prioragricultural revolutions in countries ranging from Eng-land in 1750 to Japan in 1880, the so-called “WesternExperience” (Bairoch, 1973). It also found strong sup-port in the then recent success stories of the “Asianmiracles” (Taiwan, South Korea) anchored in redis-tributive land reforms and rapid productivity growthin smallholder farming. This role was conceptualizedby the development profession in the 1960s in termsof the various contributions that agriculture can maketo industrial development through the generation andtransfer of an investible surplus and through the cre-ation of effective demand for industry on the domesticmarket (Morrisson and Thorbecke, 1990). The surplusof agriculture could take various forms: labor releasedfrom agriculture through productivity growth allowingindustrial employment (Ranis and Fei, 1961); lowerfood prices allowing lower nominal wages for indus-trial employers also as a consequence of technological

1 The evolution of thought in economic development is very use-fully presented in Thorbecke (2006).

Agriculture for development: new paradigm and options for success 19

change in agriculture (Lele and Mellor, 1981); foreignexchange earned by agricultural exports and availableto the central bank to import capital goods and inter-mediate products for industry; and taxes levied on agri-culture (in particular a land tax as in Japan) or forcedsavings on agriculture (through low-priced forced de-liveries as in the Soviet Union and India), or “invisibletransfers” (through managed low prices as in LatinAmerica via exchange rate and trade distortions) thatwere invested in urban public goods and industry. Agri-culture could also contribute through effective demandfor nontradable industrial goods created by rising agri-cultural incomes, inducing an Agricultural Demand-Led Industrialization process (Adelman, 1984).Productivity growth in agriculture was thus seen tohave strong growth multiplier effects on the rest of theeconomy through intersectoral and final demand link-ages. The empirical regularity supporting this visionof “agriculture on the road to industrialization” (Mel-lor, 1998) was the structural transformation (Johnstonand Kilby, 1975; Timmer, 1988). Along this transfor-mation, derived from Engel’s Law, which states thatthe share of food in consumer expenditures declinesas income rises, the share of agriculture in both to-tal employment and GDP declines as GDP per capitarises. The surplus, linkage, and market contributions ofagriculture, and the associated multiplier effects, couldaccelerate growth in the rest of the economy abovethe rate of growth in agriculture, leading to a relativedecline of the agricultural sector in employment andGDP. The role of agriculture in development was thusmeasured in its support to the acceleration of growth inthe rest of the economy, principally industry, with therelative decline of agriculture a symptom of missionsuccessfully accomplished.

The fact that rapid agricultural growth was possi-ble was itself derived from the work of Schultz (1964)who had shown that, in spite of poverty because of lackof access to assets, smallholder agriculture is respon-sive to price incentives and prone to adopt remunera-tive technological change opportunities. Hence, “get-ting the prices right” matters for agriculture (Streeten,1987), while a prolonged urban bias in the terms oftrade for agriculture can have devastating economiccosts (Lipton, 1977). It was also derived from the workof Hayami and Ruttan (1971) who showed the impor-tance of total factor productivity gains as a source ofgrowth for agriculture once land expansion had been

exhausted, and how the resource-saving bias of tech-nological innovations is responsive to the market sig-nals about relative factor scarcity and to public-sectorinvestment in agricultural research and extension. Fol-lowing the lead of Griliches (1958), rates of return toinvestment in agricultural technology were measuredto be typically far in excess of the opportunity cost ofcapital in public investment.

In the 1960s and 1970s, this classical paradigmhad been successfully implemented in Asia wherethe Green Revolution both averted major famines andhelped kickoff industrialization (Lipton, 1989). Whilethese successes focused attention in academic circleson the potential merits of the classical paradigm, it wasin fact not being put into practice in a way that could de-liver strong results across the developing world, and es-pecially to Sub-Saharan Africa (Dumont, 1962). Therewas in particular an underemphasis on the role of theprivate sector in the rapidly emerging agribusiness andfood-integrated value chains, an overestimation of thecapacity of the public sector to deliver quality pub-lic goods and services to farmers, and a suppressionof cooperative producer organizations that could givefarmers support in achieving market competitivenessand voice in policymaking. With import substitutionindustrialization and urban consumer interests domi-nating the political agenda, the terms of trade werestrongly biased against the agricultural sector, with ap-preciated real exchange rates and trade policies taxingtradable agricultural goods (Krueger et al., 1991). Theresult was underperformance of the agricultural sector.Public agencies serving agriculture such as parastatalagencies for access to inputs and markets, irrigationdistricts for free water, and agricultural developmentbanks for subsidized loans—that were in part actingto compensate agriculture for adverse macroeconomicand trade policies—were notably ineffective and fre-quently outright corrupt. And integrated rural devel-opment programs were assessed as largely unsustain-able, principally because they relied excessively on therole of the state as a coordinating agency and a sourceof subsidies (World Bank, 1997). The outcome wasbroad dissatisfaction with the performance of agricul-ture, inviting policy shifts toward alternative growthand poverty-reduction strategies.

With implementation of structural adjustment pro-grams and the policies of the Washington Consen-sus that followed the international spread of the 1982

20 A. de Janvry

Mexican debt crisis, sectoral policies were discardedto focus on the macro fundamentals and promote therole of market forces. The 1960s emphasis on the roleof agriculture in the classical development paradigm,and the subsequent contradictions in its implemen-tation in the 1970s and early 1980s, were followedfor 20 years (basically from 1985 to 2005) by theneglect of agriculture, which was seen as a sunsetindustry not competitive for public investment. Thisresulted in a sharp decline in public investment andoverseas development assistance going to agriculture,and dismantlement of many of the support agencies toagriculture including agricultural development banksand parastatal marketing agencies. In that perspective,agriculture was no longer seen as the most effectiveapproach to achieve industrialization that could bet-ter be obtained through Open Economy Industrializa-tion strategies based on international financial capi-tal movements, foreign direct investment, internationaltransfers of technology, and economies of scale in ur-banization (Rodrik, 2006). Combating rural povertywas seen to be more cost effectively achieved throughtransfer programs and urban migration. Cash trans-fers, often conditional on child health and educationas under the multibillion dollar “Oportunidades” pro-gram in Mexico and “Bolsa Familia” program in Brazil,and workfare programs such as the Maharashtra Em-ployment Guarantee Scheme, were put into place, be-coming the prime instruments for poverty reductionas opposed to gains in autonomous income earned bythe poor, as had been expected under the integratedrural development approach. Food security became apolicy concern of secondary importance as it presum-ably could be achieved via trade with declining inter-national food prices and by helping the poor accessfood through targeted assistance programs. Low inter-national commodity prices deterred public investmentin agricultural research and development. Agriculturewas also seen as a suspect for environmental dam-age, in particular the initial Green Revolution that hadmade extensive use of agrochemicals and damaged theenvironment. Finally, agriculture-based rural develop-ment projects had met with limited success as theywere complex to organize compared to industrializa-tion strategies such as duty-free zones and to welfareprograms for the rural poor based on transfers or food-for-work, while technical skills and experimentationwere lacking in designing agriculture-based projects.

While structural adjustment programs were successfulin removing biases in the terms of trade for agriculturethat had been caused by real exchange rate apprecia-tion and distortive trade policies, they were ineffectivein inducing supply response due to failure to under-stand the importance of the state for agriculture giventhe existence of extensive market failures and the spe-cial needs of smallholders to achieve competitiveness(Commander, 1989). This long lull in recognizing theimportance of agriculture as an instrument for devel-opment, and in recognizing the role of the state insupporting supply response in agriculture via publicgoods and institutional reforms, came to an end as aset of crises exposed in the mid 2000s the high costsof neglecting agriculture.

3. Agriculture in crisis and new demands onagriculture for development

Several major crises have brought agriculture backin the news headlines, restoring agriculture as a sec-tor of concern for policy makers and elevating it inpolicy agendas as an important instrument for devel-opment. These crises are in part the outcome of theflawed interpretations of the importance of agriculturefor development in the previous 20 years and of thepolitical economy of urban-industrial biases in policy-making. Crises include the following five aspects thatcharacterize the world food and agriculture situationtoday.

3.1. Rising food insecurity and hunger

The global food crisis, characterized by sharp in-creases in international commodity prices over the2005–2008 period, is presaging continued high pricesand high price volatility. Following a century of de-clining food prices and falling incidence of hunger,the world food situation is now more uncertain. Aftertripling between 2005 and 2008, the international mar-ket prices of rice, maize, and wheat are still 75% abovetheir levels of four years ago (May 2005 to May 2009,see World Bank, 2009). While some countries have hadhigh price transmission (as, for example, rice in Sene-gal in Fig. 1), transmission has been more limited inmany others (as, for example, rice in India and Mada-gascar in Fig. 1), following patterns that are difficult

Agriculture for development: new paradigm and options for success 21

Fig. 1. Transmission between international and domestic prices: High in Senegal, low in India and Madagascar (January 2005 = 100).

to predict and highly country specific (Abbott, 2009).Many countries have also experienced sharp price in-flation in nontraded foods (e.g., Mali and Senegal formillet in Fig. 2) due to related causes such as popu-lation growth, stagnant yields, rising domestic energyprices, and climate change (Daviron et al., 2008). Overthe last three years, world hunger in developing coun-tries has increased from 870 million affected peoplein 2005 to more than 1 billion in 2009 (FAO, 2009),taking the world sharply away from the MillenniumDevelopment Goal of reducing undernourishment to420 million people by 2010. Malnutrition is knownto be a major—yet neglected—cause of infant mortal-ity, with more than 5 million children dying each yearof malnutrition and related causes (Gross and Webb,2006). The recent rise in hunger is also associated withthe global financial crisis where loss of employment,

declining levels of economic activity in the informalsector, and falls in remittances are creating large num-bers of “new poor,” namely people previously nonpoorbut vulnerable to these shocks.

These drastic changes in the world food situationhave three major policy implications regarding the cur-rent role of agriculture for development. The first is theneed to give much greater attention to the supply sideof agriculture at both the global and local levels, byachieving sustainable productivity gains and greaterresilience to shocks. While the world was food securefrom the mid 1970s to the mid 2000s, food security isnow not only a matter of access to food for the poorbut also of food availability globally and locally. Be-sides population growth, threats to food availability arerelated to three phenomena: (1) new supply-side con-straints associated with rising water and land scarcity,

22 A. de Janvry

Fig. 2. Transmission between imported and local cereal prices in Mali and Senegal, 1993–2008 (January 1993 = 100).

soil depletion, rising energy prices, and climate change;(2) changes in diets toward animal products that de-crease the effective aggregate supply of calories forhuman consumption; and (3) diversion of land towardother uses such as biofuels production and urban con-struction. With population growth and changing di-ets, food production in Sub-Saharan Africa needs todouble over the next 20 years, posing a huge chal-lenge. The second is the need to raise anew the issueof food security as a major policy concern. Achievingfood security is complex as it requires the coordinatedmanagement of policy instruments that include freertrade (including completion of the Doha DevelopmentRound), restocking and management of national foodreserves (Wright, 2009), increased domestic produc-tion to achieve a higher level of food self-sufficiencyto respond to changes in relative prices, comprehensivesocial safety nets, and an increasing role for subsistencefarming in securing access to food for large segmentsof unprotected rural populations. The third is the needto focus food assistance not only on the chronic poor,but also on the nonpoor vulnerable to price and in-come shocks. This includes a surprisingly large shareof the rural poor who are net buyers of food in spiteof access to some land. We find, for example, that inIndia and Guatemala, respectively, 77% and 81% ofthe poor who were negatively affected by rising foodprices were in fact rural households or urban farmerhouseholds (de Janvry and Sadoulet, 2009). Existingsocial safety nets are largely designed to cater to thechronically poor. They are inadequately organized to

recognize the vulnerable nonpoor and reach them ontime to avoid irreversibilities in child health and edu-cation, and in asset decapitalization, when a price oran income shock occurs. Major adjustments need tobe made to the way these safety nets are designed ifthey are to help provide food security to the transitorypoor.

3.2. Continued stagnation in Sub-Saharan Africaagriculture

In spite of some improvement starting in 2004, agri-culture in Sub-Saharan Africa continues to be largelystagnant relative to population growth and falling be-hind progress made in the rest of the world. Per capitaagriculture value added has stopped declining, but re-mains stagnant (Fig. 3). Land expansion, which hadhistorically been the main source of growth of cerealproduction in Sub-Saharan Africa (Fig. 4, left panel),is now compromised by population growth, and landis becoming scarce faster in Sub-Saharan Africa coun-tries than in any other region (Fig. 4, right panel). Yet,yields of cereals remain stagnant (Fig. 5). Fertilizeruse is by far the lowest in the world (Sub-SaharanAfrica uses only 11 kg/ha compared to 130 kg/ha inSouth Asia and 271 kg/ha in East Asia), and only3% of the cropland is irrigated compared to 39% inSouth Asia and China (Fig. 6). The large exploitableyield gaps (measured as the difference between aver-age yield in farm demonstrations and national yield) inSub-Saharan Africa create, however, an opportunity for

Agriculture for development: new paradigm and options for success 23

Fig. 3. Agriculture value added per capita across regions (1970 =100).

rapid productivity growth compared to the rest of theworld where gaps are small and prospects weak in theshort run for major technological revolutions. Neededis an adaptation of technological opportunities to het-erogeneous local conditions, giving solutions to thelarge gaps in adoption of existing technological inno-vations, and supporting adoption through “smart” sub-sidies when there are market failures for the poor andlarge externalities in adoption derived from learningeffects and/or economies of scale in fertilizer markets.Insufficient growth in African agriculture as popula-tion continues to rise rapidly results in an increasingabsolute number of rural poor (Fig. 7), and an inci-

dence of rural poverty that will make the MillenniumDevelopment Goal of halving the poverty rate by 2015simply unreachable.

3.3. World poverty still overwhelmingly rural

Developing countries, poverty is still a stunning 75%rural, even though rural population is now only 58% ofthe total (Ravallion et al., 2007). While rural povertydecreased by 153 million people between 1993 and2002, as the world economy was growing rapidly, thissuccess was mainly due to only one event—China’ssuccess—that was sufficient to compensate for fail-ures to reduce poverty in the rest of the world. In EastAsia, rural poverty thus declined by 184 million peo-ple, while it increased by 33 million in South Asiaand Sub-Saharan Africa. Rural poverty remained un-changed in the rest of the world. Renewed attentionto agriculture as a source of income is thus needed ifthe Millennium Development Goal of halving povertyis to be met outside East Asia. This renewed focuson agriculture is justified by the rigorously substan-tiated observation that economic growth originatingin agriculture can be particularly effective for povertyreduction (Ligon and Sadoulet, 2007), which is notsurprising given where the poor live, what they do,and the labor intensity of most of world agriculture.With a high share of the rural poor having access toland and relying on their own production for at leastpart of their food consumption, increasing yields and

Fig. 4. Sources of growth in cereal production and rising land scarcity across regions, 1961–2007 (1961 = 100).

24 A. de Janvry

Fig. 5. Cereal yields across regions (1961 = 100).

diversifying production in subsistence farming shouldbe an important, yet largely ignored, part of the re-sponse to a food crisis. Subsistence and smallholderfarming also serve as safety nets for urban populationswhen jobs are lost in crisis conditions, providing whatOwen called “farm-financed social welfare” (Owen,1966), with the recent telling example of some 20 mil-lion unemployed Chinese workers returning to theirvillages. Linking subsistence farmers, and most par-ticularly women farmers, to markets and transformingthem into net sellers through one-time initial capitalendowments and sustained productivity gains in thefarming systems they can use, has to be one of the

most effective short-run instruments for world povertyreduction.

3.4. Increasing rural-urban income disparities

Rising gaps between rural and urban incomes havebeen a major source of political tensions in rapidlygrowing countries such as India and China (Hayami,2005). In India, rural populations have slipped down-wards relative to the distribution of urban incomes overthe last 22 years, concentrating at income levels thatcorrespond to the poor (but not the extreme poor) in theurban sector (Fig. 8) (Crost et al., 2009). In China, asindustrial growth accelerated, the difference betweenurban and rural mean incomes increased by 150% be-tween 1980 and 2002 after cost of living adjustments(Ravallion and Chen, 2007). These disparities havedominated electoral platforms in India and been thesource of widespread grassroots demands for incomeconvergence in China. There is also increasing sup-portive evidence to the proposition that rising inequal-ities can contribute to reducing aggregate economicgrowth, making rising disparities not only a politicalbut also an economic efficiency problem (World Bank,2005). Raising rural incomes is part of the solution tothe sectoral disparity problem. This includes bringingthe Green Revolution to favorable rain-fed areas whereit has not yet arrived, diversifying production toward

Fig. 6. Fertilizer use and irrigated cropland across regions (1961 = 100).

Agriculture for development: new paradigm and options for success 25

Fig. 7. Rural poverty across regions (1993 = 100).

higher-value agriculture, and increasing employmentopportunities in agricultural value chains and the ru-ral nonfarm economy. While rural education and mi-gration are part of the solution, especially for regionswith limited productive resources, improved incomesin agriculture and the rural nonfarm economy are themain ingredients to reduced disparities in the short andmedium run (Lanjouw and Murgai, 2008).

3.5. Increasing resource scarcity due to excessive useand misuse in agriculture, and underprovision ofenvironmental services

There is a growing resource crisis in agriculture,and it imposes heavy costs on the other sectors of theeconomy and the world. Agricultural growth is increas-ingly constrained by water scarcity, soil exhaustion,destabilization of climatic patterns, and rising temper-atures. Overuse and misuse of resources by agriculturehas led to environmental degradation both in areas ofagricultural intensification (through agrochemical pol-lution, animal waste, and depletion of undergroundaquifers) and in areas of extensive agriculture (throughdeforestation, desertification, and loss of biodiversity)(World Bank, 2007). Both make important contribu-tions to climate change, with agriculture accountingfor between 25% and 30% of global greenhouse gasemissions.

These pressures of agriculture on resources are fre-quently reaching crisis levels, compromising the sus-tainability of yields in some of the world’s main bread-baskets such as the Indian Punjab where water tablesare falling due to overdraft and water is polluted by

Notes: The horizontal straight line is the density function of urban expenditures acrossquantiles for 1983, 1993/1994, and 2004/2005. The other lines are the densities of ruralexpenditures corresponding to the urban quantiles for the same year.

Fig. 8. Increasing rural-urban income disparities in India, 1983 to 2005.

26 A. de Janvry

agrochemicals. Overuse of water in agriculture, whichaccounts for 85% of developing countries’ fresh wa-ter withdrawal, is also running in conflict with urbanwater demands. It is well known that the rural poorare the main victims of climate change that affects dif-ferentially more tropical and high-altitude areas wheremany are located. This is because most of the worldpoor depend on agriculture for their livelihoods, andbecause they have limited access to instruments forrisk management (e.g., diversification of sources ofincome) and risk coping (e.g., credit and insurance).Clearly, environmental damage is among the currentcrises of agriculture that brought it up in the newsheadlines (e.g., expanding livestock and soybean oper-ations in the Brazilian Cerrados and Amazon at the costof large-scale deforestation). Perverse price and sub-sidy incentives, incomplete property rights, weak stateregulatory capacity, and inability to act collectively forthe management of common property resources arehuge problems that need attention. Missing marketsfor environmental services such as avoided deforesta-tion, biodiversity conservation, soil conservation withzero tillage, adoption of water-saving technologies, andterrace maintenance for watershed management con-tribute to underprovision of socially beneficial exter-nalities. These markets urgently need to be put intoplace, from the certification of environmentally soundproducts for consumers, to payments for watershedmaintenance by local water users, and to global mar-kets for carbon capture and storage for the internationalcommunity. Investments in more sustainable technolo-gies have met with success (such as zero tillage inconservation agriculture and sylvo-pastoral systems inagroforestry), but have also been lagging due to freerider problems in technology generation and lack ofincentives in adoption when externalities are not inter-nalized.

These major crises are pushing agriculture in thenews headlines and up on policy agendas for govern-ments and international development agencies. This isreflected in the explicit support to agriculture in thepolicy guidelines of the New Partnership for Africa’sDevelopment (NEPAD) and its Comprehensive Agri-culture Development Program (CAADP). The WorldBank made commitments to double its assistance toAfrican agriculture by 2010 and channeled $1.2 billiontoward agriculture and food security under its GlobalFood Crisis Response Program. Crises, with their huge

social costs on the poorest, have at least been effec-tive in inducing policy responses and resource com-mitments toward improved food security.

Crises are also met by opportunities to achieve bet-ter results than in the past in using agriculture fordevelopment. Most important among those are: (1)An improved macroeconomic context for all sectorsin the economy, a positive outcome of two decadesof structural adjustment policies. While these policiesmisunderstood the special needs of agriculture for aproactive state and a sectoral policy in supporting sup-ply response, they at least removed many of the pricedistortions that were due to exchange rate apprecia-tions and trade policies taxing agriculture (Andersonet al., 2008). (2) New approaches to rural developmentand the design of agricultural projects with emphasison decentralization, community participation, the roleof civil society organizations, and public-private part-nerships (Binswanger, 2006). These approaches, oftencouched under the rubrics of community-driven de-velopment (CDD) and territorial development, showpromise in achieving a degree of success that the inte-grated rural development approach failed to deliver. (3)New agents in development, particularly private actorsin integrated value chains that can serve for smallhold-ers as sources of innovation, inputs, contracts provid-ing credit and insurance, and access to new markets(Reardon et al., 2009); and producer organizations thathave in countries such as Senegal, Mali, and Ecuadorbecome major proactive actors in assisting their mem-bers overcome market and government failures, givingthem voice in domestic policy making and interna-tional negotiations, and striving to hold governmentsand public providers accountable for their actions. Alsoimportant is the presence of new philanthropic organi-zations such as the Bill and Melinda Gates Foundationand the FARM Foundation that are bringing efficientbusiness models into the development approaches thathistorically have been the privy of civil servants, devel-opment agency personnel, and nonprofit organizationsthat rarely had any personal business experience. Thisis frequently complemented by private business ini-tiatives motivated by the pursuit of corporate socialresponsibility, as for example under the leadership ofthe World Economic Forum.

Crises and opportunities thus combine in puttingagriculture back on the development agenda, as botha need and a possibility. This second chance in using

Agriculture for development: new paradigm and options for success 27

agriculture for development has introduced a broadlyendorsed new paradigm, even if still short of consis-tently formulated and massively implemented (Byerleeet al., 2009).

4. Putting agriculture back on the developmentagenda: toward a new paradigm of agriculture fordevelopment

Putting agriculture back on the development agendais happening in the context of two major changes thatrequire a new approach. The first is a change in theobjectives of development, from growth via industri-alization to a multidimensional agenda including GDPgrowth, poverty and disparity reduction, food security,and environmental sustainability. This broadening ofdevelopment objectives occurred first in the 1970s asgrowth had happened vigorously during the previoustwo decades while poverty had failed to decline cor-respondingly and inequality had increased sharply. Inresponse to this, development had been redefined fromindustrialization for growth in the 1950s and 1960sto growth for poverty and inequality reduction in the1970s. This was embedded in new growth strategies,particularly Chenery et al. (1975) “redistribution-with-growth” and Adelman’s (1978) “redistribution-before-growth.” Development had thus become a multidi-mensional agenda, a new purpose strongly endorsedby the World Bank that shifted its main objectivefrom accelerating growth to achieving “a world with-out poverty” (McNamara, 1973). This implies that thedevelopment objective that had motivated the 1960sclassical development paradigm—agricultural growthfor industrialization—is now set too narrowly to matchexpectations regarding what agriculture can do for de-velopment. The second is a drastic change in the struc-tural context where agricultural growth is to occur,including globalization of the food system, emergenceof integrated food value chains, major institutional andtechnological changes for agriculture, increasing re-source scarcity and climate change, and the demandson agriculture to serve as a provider of environmentalservices.

These two changes imply that the way agriculture isused for development has to be adjusted to the new ob-jectives and context, quite different from the ones thatprevailed when the classical paradigm was in force.

This calls for a new paradigm where agriculture is in-strumental for development in its multidimensionalityof functions and in the context of the opportunities andconstraints that now exist.

5. Why the continued underuse of agriculturefor development?

Putting agriculture back on the development agendahas been urged by the set of crises that pushed agricul-ture up in the news headlines. The call to overcome 20years of contradictions and neglect and to reprioritizeagriculture in policy agendas has been widely heededin political discourse. But rhetoric has been followedby very incomplete implementation. Antiagriculturebiases in the terms of trade have been sharply reduced.In Africa, the nominal rate of assistance to agriculturefell from a negative 13% in 1975–1979 to a negative 7%in 2000–2004, with much of the improvement comingfrom better fundamentals in macroeconomic policies(Anderson et al., 2008). But a Green Revolution forSub-Saharan Africa is still hardly in the making. Whilesome countries such as South Africa and Ghana areshowing relatively better performance, growth in agri-cultural value added is overall unchanged (3.2%/year inthe 1990s and 3.0%/year in the 2000s) and cereal yieldsthat have shown recent improvement remain by far thelowest in the world (1.3 tons/ha in Sub-Saharan Africacompared to 2.6 tons/ha in South Asia, 3.5 tons/ha inLatin America, and 4.7 tons/ha in East Asia). The la-bor force continues to leave agriculture in accordanceto the regularities of the structural transformation, butwithout the expected associated GDP per capita growth(Fig. 9). Agriculture in India still has a lackluster per-formance in spite of priority status in the 11th FiveYear Plan 2007–2012 (a growth in agricultural valueadded of 3.2%/year in both the 1990s and 2000s), andthe sectoral income disparities continue to deepen. InLatin America, population growth and land concentra-tion have pushed the rural labor force to the cities, butwithout sufficient employment creation in urban-basedindustry and services. In fact, comparing the structuraltransformation across countries in Asia, Sub-SaharanAfrica, and Latin America and the Caribbean in Fig. 10shows the surprising regularity that, while successful inAsia, it has been about equally unsatisfactory in LatinAmerica as it has been in Africa. In Latin America,

28 A. de Janvry

The structural transformation by region

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

4.5 5.5 6.5 7.5 8.5

Log of GDP per capita (constant 2000 US$)

China

East Asia w/o China

South Asia

SS-Africa

Normal pattern

Latin America

Fig. 9. The structural transformation across regions.

the result has been rising illegal international migra-tion and informal employment. Where poverty has de-clined in middle-income countries, this has been dueto massive social programs as opposed to rising earnedincomes by the poor. The result is a gap between dis-course and achievement, and reproduction of a set ofconditions related to the inadequate performance ofagriculture and conducive to the emergence of recur-rent crises.

This gap is most visible in the continued low and of-ten misguided public expenditure going to agriculture.While successful former agriculture-based countries2

typically allocated no less than 10% of their publicexpenditures to agriculture, and NEPAD chose thisbenchmark as a guideline for the Sub-Saharan Africacountries, few current agriculture-based countries arereaching this level (Fig. 11). Data in recent public ex-penditure assessments for African countries show 3%for Uganda and 4% for Nigeria. The gap is also visiblein the share of agriculture in Organization for Eco-nomic Cooperation and Development official develop-ment assistance (ODA) to Africa that declined from20% in 1990 to still less than 5% today, with no ma-jor aggregate reversal (Fig. 12). Rising support hashappened in some countries, most notably in BurkinaFaso (7.9% of ODA going to agriculture), Malawi and

2 Agriculture-based countries are defined in the WDR2008 ascountries where agriculture contributes approximately one-third ofoverall growth and more than 70% of the poor are in the rural sector.

Mali (5.3%), and Uganda (5.2%). However, these per-centages are disproportionately low compared to thebenchmarks provided by the share of agriculture inGDP, the share of the rural population in total poverty,and the budget shares allocated in historically success-ful countries. In that perspective, the G8 announcementon July 9, 2009, that foreign aid would give greater em-phasis on investment in agriculture, with a $20 billioncommitment over the next three years, is a welcomeand long-awaited turning point (G8, 2009). The hope isthat this may be signaling a return to greater attentiongiven by ODA to investing in agriculture in support offood security, nutrition, and sustainable development,particularly in the countries least able to respond to afood crisis.

More research is needed in estimating the rate ofreturn to public investment in agriculture, the GDPgrowth value of a dollar of investment in agricultureversus other sectors of the economy, and the condi-tions under which these returns can be increased. Wehave extensive evidence about the rate of return frominvestment in agricultural research showing, in spite ofselection biases in the cases analyzed, that returns canrepeatedly be above opportunity cost for public funds(Alston et al., 2002). But we also know that many verypoor investments have been made in agriculture, withpublic funds diverted to private subsidies in response torent seeking accounting for 75% of total public expen-diture in India and 40% in Latin America (World Bank,2007). Agricultural investment projects have also beenfaring poorly on a comparative basis, following mis-guided approaches such as the training-and-visit ex-tension system, subsidized credit, and integrated ruraldevelopment that have since been discontinued.

What are the main causes of the continued underuti-lization of agriculture’s potential for development, inspite of statements to the contrary and some remark-able comebacks to investing in agriculture? There areof course many reasons, and they are idiosyncratic tospecific contexts. Yet, two fundamental causes can besingled out for lack of implementation of the agri-culture for development paradigm. One is insufficientreconceptualization of the role of agriculture for de-velopment to correspond to the new development ob-jectives and to the new contexts and conditions foragriculture to perform. The other is insufficient re-designing of approaches for effective implementationof the agriculture for development paradigm.

Agriculture for development: new paradigm and options for success 29

Structural transformation in Asia

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

4.5 5.5 6.5 7.5 8.5Log of GDP per capita (constant 2000 US$)

Cross-sectionBangladeshCambodiaHong Kong, ChinaIndiaIndonesiaKorea, Rep.MalaysiaMyanmarPakistanPapua New GuineaPhilippinesThailandChinaVietnamTaiwan, China

Structural transformation in Sub-

Saharan

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

4.5 5.5 6.5 7.5 8.5

Log of GDP per capita (constant 2000 US$)

Cross-sectionAngolaBeninCameroonChadCongo, Dem. Rep.Cote d'IvoireKenyaMadagascarNigerNigeriaRwandaSenegalSouth AfricaTanzaniaTogoZambiaZimbabwe

Structural transformation in Latin America and

Caribbean

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

4.5 5.5 6.5 7.5 8.5

Log of GDP per capita (constant 2000 US$)

Cross-sectionArgentinaBoliviaBrazilChileColombiaDominican RepublicEcuadorGuatemalaHondurasHaitiMexicoPeruParaguayEl SalvadorVenezuela, RBx

Notes: The smooth line is the cross-country normal pattern of structural transformation.

Fig. 10. The structural transformation across countries in Asia, Sub-Saharan Africa, Latin America, and the Caribbean.

6. Moving forward in using agriculture fordevelopment: two options for success

Agriculture has thus come full circle from beingthe main pillar for industrialization, seen then as syn-onymous to development, in the classical develop-ment paradigm, to implementation failures and pol-icy neglect, and back to being seen as a pillar ina new paradigm of agriculture for an enlarged per-spective on development under the push of cumulativecrises and the pull of new opportunities. The emerg-

ing new paradigm of agriculture for development isquite different from the classical paradigm as bothdevelopment objectives and contexts and conditionsunder which agriculture performs are quite differentfrom the way they were in the 1960s. Visibility in thenews headlines, pronouncements of political support,attention by some international and philanthropic de-velopment agencies, and G8 resource commitmentshave been achieved. Yet, successful implementation isstill elusive, with continued lack of support in publicbudgets and in overseas development assistance. The

30 A. de Janvry

Fig. 11. Public expenditures on agriculture.

Fig. 12. Overseas development assistance to agriculture.

economic, human, and environmental costs of delayedimplementation have been enormous, and will keepon mounting. For Africa in particular, there is no op-tion but success in implementing the agriculture-for-development paradigm if major humanitarian crises areto be avoided. What will it take for success to happen?We argue that there are two major lines of action thatrequire attention for success to follow: reconceptual-ization of the role of agriculture for development andredesign of approaches for effective implementation.

6.1. Reconceptualization of the role of agriculturefor development

The role of agriculture in the classical developmentparadigm was well conceptualized and informed, withhistorical support, formal modeling, quantitative sim-

ulation exercises, and a set of policy prescriptions. Thenew paradigm currently falls short of these forms ofsupport. The reconceptualization task is particularlydaunting as the functions of agriculture for develop-ment are now multiple, the very process through whichagricultural growth is achieved matters if it is to yielddevelopment in addition to higher piles of cereals, andthe role of the state in relation to the market and civilsociety needs to be reimagined. We discuss each ofthese issues in turn.

6.1.1. Formalizing the complementarities andtrade-offs in the multiple functions of agriculture fordevelopment in the emerging contexts

The context where the role of agriculture for de-velopment is to be modeled is one characterized byglobalization (of trade, capital movements, migration,transfers of technology, emissions, and even land trans-actions), integrated value chains with large economiesof scale and market power, technological innovationsthat include major advances in biotechnology and inInformation and Communication Technologies, insti-tutional innovations with progress toward more com-plete financial services and more effective producer or-ganizations, and environmental pressures with greaterresource scarcities and increasing risks associated withclimate change. Models are needed to capture thecomplementarities and trade-offs in the multiple de-velopment outcomes as agriculture performs in thesecontexts. The models we have today are, however,generally still quite removed from meeting theseneeds, and attention given to this research theme by

Agriculture for development: new paradigm and options for success 31

the economics profession has declined and is vastlyinsufficient. Major efforts must be made to bringagriculture into estimated (and hence beyond cali-brated computable general equilibrium) models thatcan help us understand what determines the perfor-mance of agriculture and how it translates into differ-ent development outcomes according to the contextsthat prevail. This is a tall order, and work is in progress(e.g., Fan, 2008). Success in this work is a conditionfor effective policy design in using agriculture for de-velopment that needs to be of much greater concern tothe economic profession.

6.1.2. Designing the process of agricultural growthto achieve development

Some of the development objectives in using agricul-ture for development, growth in particular, can largelybe driven by market incentives in the context of asupportive investment climate. Reduction of resourceoveruse in agriculture can also be achieved throughmarket incentives with the proper allocation of prop-erty rights and pricing of resources. Putting into placeexchange platforms to trade water can be a powerfulincentive to reduce water use in agriculture once ithas an opportunity cost in other applications to whichit can be sold by the rightful owners. The provisionof environmental services, as one of the developmentfunctions of agriculture, can be transformed into mar-ket responses by offering payments for environmentalservices, eventually institutionalized into marketplacesthat allow to trade externalities (as for C02 and waterservices). However, other functions of agriculture fordevelopment such as poverty reduction, the narrowingof sectoral disparities, gender equity, and intergenera-tional sustainability in resource use must be obtainedthrough social choices made about the very processthrough which growth is achieved (short of pure ex-post income redistribution). In the new paradigm, pro-cess thus matters along with product if the multipledimensions of development are to be achieved. Thisimplies that social choices must be made in weight-ing the various development outcomes of agriculture.Small farms may be preferred over (or in associationwith) large farms not only on efficiency grounds (if aninverse relation exists between total factor productiv-ity and farm size), but also because they allow incomegeneration by the rural poor, even if sometimes at the

cost of a growth trade-off.3 Technologies that createemployment in agriculture and in agribusiness maybe preferred over labor-saving technological changesas the labor market becomes an increasingly impor-tant instrument through which productivity gains inagriculture are translated into welfare gains for ruralpopulations. Agroecological approaches may be pre-ferred when negative externalities in chemical-basedfarming cannot be internalized, even if at the cost ofyield and risk trade-offs, and even if they are more de-manding in public assistance to achieve a productivityrevolution for Sub-Saharan Africa. Particularly in theworld of international development agencies, we havetoo long wished for a world where win-wins dominatethe outcome set, in part to avoid protracted politicaldebates (as for example in the growth-equity debate,see Banerjee et al., 2006). But win-wins have to be rec-ognized as the exception more than the rule. Trade-offsin the process of agricultural growth thus require mak-ing difficult social choices, expectedly based on solidinformation and democratic participation to establishpriorities, requiring a set of institutions and practicesthat typically need to be reinforced in developing coun-tries.

6.1.3. Redefining the role of the state in setting socialpriorities among conflicting functions and inovercoming market failures for agriculture

Rightly or wrongly done in practice, the state wascentral to the classical development paradigm in the1960s and 1970s. Differentially across countries, itplayed a key role in implementing land reforms, in-vesting in research and extension, stabilizing prices,providing access to inputs and credit, managing irri-gation schemes, and coordinating rural development.Under the Washington Consensus in the 1985–2005period, stabilization, and adjustment policies severelyreduced the size and functions of the state in agricul-ture, minimizing in particular the role of sectoral poli-cies. Expectation was that the private sector would pickup the functions left open by the state. As we by nowknow well, this happened very partially at best, leaving

3 Dercon (2009) for instance suggests that stimulating agriculturalproduction in smallholder farming in Africa may be needed for socialinclusion even if other sectors can me more effective for growthbecause market failures prevent the rural poor from moving out ofpoverty via the labor market.

32 A. de Janvry

huge gaps in institutional support for agriculture andespecially for the smallholder sector. Largely ignoredby the agents of the Washington Consensus was thatagriculture suffers from major market failures and isunusually dependent on state support to achieve growthand development, a painful lesson in irreversibility de-rived from the OECD farm policy experience wherestate support was extensive. The result has been under-performance relative to expectations, opening the doorto a series of crises that we analyzed above.

Redefining the role of the state in using agriculturefor development is thus essential, and also a major in-complete task. Lack of clarity has been a source of ma-jor inefficiencies, as ill-defined and ill-performed statefunctions stand in the way of private sector investmentand performance, and undermine the effectiveness ofproducer organizations. Along with the emerging con-text for agriculture to perform its development func-tions, a new state needs to emerge. This state must havethe capacity to work with the private sector and pro-ducer organizations, regulate externalities and marketcompetitiveness, provide public goods, be the guardianof processes in the way agricultural growth is achievedso that it will deliver development, set priorities amongcompeting functions of agriculture for development,provide safety nets, and insure sustainable outcomes.For the state, agriculture is no longer the road to indus-trialization, but an instrument in a multidimensionaldevelopment process. Lack of attention to this redefi-nition will be a major cause of delays in the successfuluse of agriculture for development.

6.2. Redesigning approaches for effectiveimplementation

Development requires that reconceptualization beaccompanied by effective redesign of approaches foreffective implementation. We discuss three aspects ofredesign: experimentation for learning, fixing the gov-ernance structure for agriculture, and committing pub-lic support to agriculture over price and political cy-cles.

6.2.1. Experimenting with new approaches andinternalizing lessons for scaling up successes

As opposed to what is often said in activist donorcircles, it is a serious mistake to believe that we know

what should be done, and all that is left to do is do-ing it. Because objectives and contexts are novel, weare entering unchartered territory that needs to be re-searched and experimented with. Extraordinary newopportunities exist to successfully invest in agriculturefor development, but they must be carefully identified.What we know is that a “Green Revolution” for Africawill have to be different from what it was in Asia,and hence that it is left to be designed. As opposed toAsia, Sub-Saharan Africa has a mainly rain-fed agri-culture with many different agroecological conditionsand it also has highly fragmented social systems. Het-erogeneity and local adaptation must consequently bepart of the solution. In addition, conditions for suc-cess are demanding. Soils are generally exhausted,infrastructure (roads, water control) is weak and of-ten non-existent, levels of education and health amongfarmers are low, the private sector in value chains isincipient, the investment climate is frequently uninvit-ing, and many countries are small requiring regionalcollaboration to achieve economies of scale in pro-viding many public goods. A Green Revolution forAfrica must also address challenges that did not existat the time of the Asian Green Revolution. It must dealwith sustainability and environmental friendliness as aprocess, and hence go beyond a seed-fertilizer-waterpackage to extend toward agroecology, agroforestry,and conservation agriculture. It must go beyond ce-reals to encompass locally specific food crops suchas roots and tubers and banana plantains, as well ashigh-value activities such as fruits and vegetables, live-stock, and fish. It must address brand new challengessuch as climate change (particularly vulnerability toclimate shocks and the resilience of farming systems)and the forces of globalization (particularly competi-tion with producers dispersed across the world and ben-efiting from uneven state support, increasingly volatileprices, and rapidly changing requirements for gradesand standards in value chains and supermarkets). Itmust redefine the roles of the state and civil societyin support of agriculture for development, recraftinginstitutions beyond structural adjustment as part of“second-generation reforms” (Stiglitz, 2006). And itis under pressure to succeed urgently given the rapidchanges in the world food situation and the recurrenceof crises.

Innovation, experimentation, evaluation, and learn-ing must thus be central to devising new approaches to

Agriculture for development: new paradigm and options for success 33

use agriculture for development. This requires puttinginto place strategies to identify causal impacts as weproceed with new options. Too much of our econo-metrics still reports unidentified “determinants” thatcannot be used for policy advice because they mea-sure correlates instead of causalities. To date, rigorousidentification in agricultural economics remains morean exception than the rule, perhaps more so than inother fields of application of economics because ofgreater difficulty in doing so compared to health andeducation where most of the impact analysis has beenconfined. This is a serious and insufficiently recog-nized handicap in introducing new options as to howto use agriculture for development.

If additional resources are effectively going to flowtoward agriculture following the G8 call for a scaled-up use of agriculture for development, it is funda-mental that lessons be derived from past mistakes inimplementing the classical paradigm, and that newapproaches be devised, experimented with, and con-firmed before scaling up is attempted (Easterly, 2006).More investment will only yield more development ifnew ways of using agriculture for the broadened de-velopment objectives and in the new context where itmust perform have been established. There is otherwisea huge risk of resource wastage and serious disappoint-ments if we return to agriculture without knowledge ofbetter ways of how to use it for development.

6.2.2. Fixing the governance structure for the state tofulfill its new functions for agriculture

The new functions to be assumed by the state insupporting effective use of agriculture for develop-ment are daunting. They include addressing currentforms of market failure (e.g., complementarities be-tween public and private research, intellectual prop-erty rights for biotechnology innovations, security ofaccess to land in Sub-Saharan Africa, excessive pricevolatility in guiding investments and insuring food se-curity, and lack of risk-sharing options to invest in agri-culture), regulating competition in the private sector,engaging in public-private partnerships and the man-agement of “smart” subsidies, and supporting effectivedecentralization for agriculture. Correspondingly, re-designing the structure of Ministries of Agriculture andthe public sector in support of agriculture remains anunresolved and unsatisfactory dilemma in using agri-

culture for development (Wiggins and Cabral, 2007).The “new ministries” must be able to not only man-age agricultural policy but also integrate it with therest of policymaking and coordinate multisectoral ap-proaches. In the perspective of agriculture for devel-opment, ministries should largely have a multisectoralterritorial function, with agriculture only one (if themost important to them) sector of economic activity. Aterritorial approach would allow us to take very differ-ent perspectives on such fundamental issues as pricevolatility (with better compromises between the inter-ests of producers and consumers, including the manynet buyer smallholders), biofuels (with better balancebetween delivering energy, protecting consumer inter-ests, and managing environmental impacts), and foodquality based on proximity (with better trade-offs be-tween consumer health and cost concerns). To serve fordevelopment, the “new ministries” must consequentlybe concerned not only with agriculture, but also withthe welfare of rural populations, consumer food safety,and the provision of environmental amenities. The cor-responding administrative structures are still largely tobe imagined.

Governance cannot make agriculture work for de-velopment without giving a fundamental role to civilsociety organizations. Progress with governance mustconsequently be accompanied by parallel efforts to pro-mote effective producer organizations that can havevoice in public affairs. Yet, little emphasis is gener-ally given by governments and development agenciesto the role of rural organizations in development thatthey often see as challengers and lobbies rather thanpartners. Learning how to work with organizations isthus part and parcel of the institutional reconstructionof agriculture in support of development.

6.2.3. Committing the state and the internationalcommunity to support the long-term role ofagriculture for development above price and politicalcycles

Food crises have been recurrent in time and spaceover the last century, in spite of secularly decliningprices (Gardner, 1979), and conditions are in place formore and deeper crises to come. Like all crises, theyhave huge costs, particularly on the poor. Policy instru-ments to respond to a food crisis typically include trade(reducing tariffs if an importer and increasing taxes if

34 A. de Janvry

an exporter), destocking of public reserves, consumerprice subsidies, safety net programs for the poor, andshort-term boosts to supply response in agriculture(Delgado, 2009). Yet, poor countries with weak fiscaland administrative resources have little capacity to usethese policy instruments, with the result that most oftheir poor remain unprotected when crises come, par-ticularly their rural populations. For these countries,essential is consequently to invest in preventing crisesfrom occurring (Timmer, 2009). If preventing crises isa necessary option, how should this be done, given thepast failures? Public investments in R&D and in in-frastructure such as water, roads, and market facilitiesare the key to agricultural growth. When prices fall,due to effective public investments in agriculture at aworld scale, there is a long cycle whereby agriculturebecomes neglected by governments and donors. Thisneglect becomes the source of future crises. Becausethere is a long lag in how public investment in agri-culture translates into output response, avoidance offuture crises (as population and income effects con-tinue to increase demand, in addition to demands forbiofuels and climate change) requires continued pub-lic investment in sources of productivity growth andresilience of agriculture even when the prices are low.We must thus learn that low market prices are not theproper signal in allocating public resources to agricul-ture. In electoral cycles, short-term competition overpublic budgets dominates over long-term considera-tions. Hence, what countries need is a commitment de-vice that can elevate public investment in the sourcesof productivity growth in agriculture above price andpolitical cycles. Countries that have made long-termcommitments to public investment in agricultural re-search and education, such as the United States withits Agricultural Experiment Station system and Brazilwith its Agricultural Research Corporation, Embrapa,have reaped large benefits when food crises return.

Hence, we need to take the long view when investingin agriculture given the existence of extended lags inproductivity effects, and to resist the tendency to recur-rent neglect by governments and donors when pricesare low. With neglect, as observed over the 1985–2005period, high prices can return with a vengeance. Sus-tained investment in agriculture is a political choice,since market signals fail in guiding public investmentover the long run at prevailing discount rates. The po-litical economy of food crisis avoidance is thus one of

committing public expenditures to agricultural produc-tivity gains over and above price signals and over andabove short-term government budget priorities. Mak-ing public expenditures in sources of agricultural pro-ductivity gains into a state policy is thus a necessarycommitment device. This requires giving public agri-cultural R&D institutions budgetary guarantees and ad-ministrative autonomy to protect resources from short-term market signals and political cycles.

7. Conclusion: the way forward

The classical paradigm of agriculture on the roadto industrialization that prevailed in economic thoughtin the 1960s and 1970s no longer matches needs asdevelopment objectives have broadened and the con-texts and conditions for implementation have changeddramatically. The subsequent 20-year neglect of agri-culture seen as a sunset industry for development hascontributed to serious crises with huge economic, so-cial, and environmental costs. Crises and opportuni-ties have prompted the emergence of a new, yet stillincomplete, paradigm of agriculture for developmentthat has been shown to work in several countries. Con-ditions are currently favorable to its implementation asa consequence of high public consciousness about itsrelevance with agriculture repeatedly in the news head-lines, political pronouncements of concern with foodsecurity in response to electoral demands and popu-lar discontent, and attractive investment opportunitiesfor producers with high prices and new dynamic mar-kets. Yet, putting the paradigm of agriculture for de-velopment into practice has been lagging in too manycountries, especially in Sub-Saharan Africa where suc-cessful implementation is both critical and urgent.

We have advanced the interpretation that current lagsin implementation have been due to incomplete recon-ceptualizations and insufficient redesign of approachesin using agriculture for development. Reconceptual-ization will require: (1) Scholarship applied to formaleconomic modeling of complementarities and trade-offs in the development outcomes that can be achievedthrough agricultural growth; (2) priority setting andcoordination to use the process of agricultural growthas an instrument in achieving nonmarket developmentoutcomes; and (3) participation to redefine the rela-tive roles of the state, the market, and civil society

Agriculture for development: new paradigm and options for success 35

in using agriculture for development. Redesigning ap-proaches for effective implementation will require: (1)Experimentation, evaluation, and learning to identifycausalities in new approaches to achieve the desired de-velopment objectives; (2) institutional innovations toredefine the shape of governance needed for agricultureto achieve its development functions; and (3) crediblecommitments by governments and international orga-nizations to sustain investment in agriculture aboveprice and political cycles.

What we derive from this analysis is that recon-ceptualization and redesign of the approaches to suc-cessfully use agriculture for development are complexand demanding. The biggest mistake one could makeabout using agriculture for development is the beliefthat it is easy to do and that we already know all weneed to do it. It is not and we don’t. It requires signif-icant additional resources, new expertise (which is inserious deficit at all levels, including in internationalorganizations), coordination among many actors, andsustained commitments. Half-way measures and dis-continued efforts will not produce results. Recent com-mitments by several governments, international orga-nizations, and philanthropic donors show promise. But,like in the classical theory of the Big Push proposedby Leibenstein (1957) in the 1950s and recent applica-tions of concepts of multiple equilibria to developmentby Sachs (2005), payoffs will only happen if the ef-fort is sufficiently massive, concerted, and sustained.Lessons must be derived from past mistakes, and newapproaches devised and evaluated. In that sense, addi-tional resources are necessary, but far from sufficientin insuring success. While complex and demanding,there is little choice and much urgency to the need tosucceed in this effort given the current and predictedworld conditions for food and development.

References

Abbott, P., 2009. Development dimensions of high food prices. De-partment of Agricultural Economics, Purdue University.

Adelman, I., 1978. Redistribution before Growth: A Strategy forDeveloping Countries. Martinus Nijhof Editions, The Hague.

Adelman, I., 1984. Beyond export-led growth. World Dev. 12(9),937–949.

Alston, J., Chan-Kang, C., Marra, M., Pardey, P., Wyatt, T.J., 2002.A Meta-Analysis of the Rates of Return to Agricultural R&D:Ex Pede Herculem? International Food Policy Research Institute,Washington, DC.

Anderson, K., Masters, W., Kurzweil, M., Valenzuela, E., 2008. Howfar has Africa come in reducing its anti-agricultural policy bias?School of Economics, University of Adelaide.

Bairoch, P., 1973. Agriculture and the industrial revolution, 1700–1914. In: Cipolla, C. (Ed.), The Fontana Economic History of Eu-rope: The Industrial Revolution, Vol. 3. Collins/Fontana Books,London, pp. 455–466.

Banerjee, A., Deaton, A., Lustig, N., Rogoff, K., 2006. An Eval-uation of World Bank Research, 1998–2005. The World Bank,Washington, DC.

Binswanger, H., 2006. Empowering rural people for their own devel-opment. Elmhirst Lecture, IAAE Conference, Gold Coast, Aus-tralia.

Byerlee, D., de Janvry, A., Sadoulet, E., 2009. Agriculture for de-velopment: Toward a new paradigm. Annu. Rev. Resour. Econ.1, 15–32.

Chenery, H., Ahuluwalia, M., Bell, C., Duloy, J., Jolly, R.,1975. Redistribution with Growth. Oxford University Press,London.

Commander, S (Ed.), 1989. Structural Adjustment and Agriculture:Theory and Practice in Africa and Latin America. Overseas De-velopment Institute, London.

Crost, B., de Janvry, A., Sadoulet, E., 2009. The structure of rapidlyrising rural-urban expenditure disparities in India over the last twodecades. Department of Agricultural and Resource Economics,University of California, Berkeley.

Daviron, B., Aubert, M., Bricas, N., David-Benz, H., Dury, S., Egg,J., Lancon, F., and Meuriot, V. 2008. La transmission de lahausse des prix internationaux des produits agricoles dans lespays Africains. CIRAD-FARM, Montpellier.

de Janvry, A., Sadoulet, E., 2009. The global food crisis andguatemala: What crisis and for whom? Department of Agricul-tural and Resource Economics, University of California, Berke-ley.

Delgado, C., 2009. Global food crisis: From urgent short-term re-sponses to viable longer-term solutions. Agriculture and RuralDevelopment, The World Bank.

Dercon, S., 2009. Rural poverty: Old challenges in new contexts.World Bank Res. Observer 24(1), 1–28.

Dumont, R., 1962. L’Afrique Noire est mal Partie. Editions du Seuil,Paris.

Easterly, W., 2006. The White Man’s Burden: Why the West’s Effortsto Aid the Rest Have Done So Much Ill and So Little Good.Penguin Books, New York.

Fan, S., 2008. Public Expenditures, Growth and Poverty in Develop-ing Countries: Issues, Methods and Findings. The Johns HopkinsUniversity Press, Baltimore, MD.

FAO, 2009. 1.02 billion people hungry: One sixth of humanity un-dernourished —more than ever before. http://www.fao.org/news/story/en/item/20568/icode/, accessed June 2009.

G8, 2009. L’Aquila joint statement on global food security. 2009Summit, L’Aquila, Italy.

Gardner, B., 1979. Optimal Stockpiling of Grain. Lexington Books,Lexington, MA.

Griliches, Z., 1958. Research costs and social returns: Hybrid cornand related innovations. J. Pol. Economy 66(October), 419–431.

36 A. de Janvry

Gross, R., Webb, P., 2006. Wasting time for wasted children: Severechild undernutrition must be resolved in non-emergency settings.Lancet 367(April), 1209–1211.

Hayami, Y., 2005. An emerging agriculture problem in high-performing Asian economies. Presidential address at the FifthConference of the Asian Society of Agricultural Economics, Za-hedan, Iran.

Hayami, Y., Ruttan, V., 1971. Agricultural Development: An Inter-national Perspective. The Johns Hopkins University Press, Balti-more, MD.

Johnston, B., Mellor, J., 1961. The role of agriculture in economicdevelopment. Am. Econ. Rev 87(2), 566–593.

Johnston, B., Kilby, P., 1975. Agriculture and Structural Transforma-tion: Economic Strategies in Late-Developing Countries. OxfordUniversity Press, London.

Krueger, A., Schiff, M., Valdes, A., 1991. The Political Economyof Agricultural Pricing Policy. Johns Hopkins University Press,Baltimore, MD.

Lanjouw, P., Murgai, R., 2008. Poverty decline, agricultural wages,and non-farm employment in rural India, 1983–2004. The WorldBank Development Research Group Poverty Team WPS4858.

Leibenstein, H., 1957. Economic Backwardness and EconomicGrowth. Wiley Press, New York.

Lele, U., Mellor, J., 1981. Technological change, distributive bias,and labor transfers in a two sector economy. Oxford Econ. Pap.33(3), 426–441.

Ligon, E., Sadoulet, E., 2007. Estimating the effects of aggregateagricultural growth on the distribution of expenditures. Back-ground note for the World Development Report 2008, The WorldBank, Washington, DC.

Lipton, M, 1977. Why Poor People Stay Poor: Urban Bias in WorldDevelopment. Harvard University Press, Cambridge, MA.

Lipton, M., 1989. New Seeds and Poor People (with RichardLonghurst). Unwin Hyman, London.

McNamara, R., 1973. The Nairobi speech: Address to the board ofgovernors. World Bank Group Archives, Washington, DC.

Mellor, J., 1976. The New Economics of Growth: A Strategy forIndia and the Developing World. Cornell University Press, Ithaca,NY.

Mellor, J., 1998. Agriculture on the road to industrialization. In:Eicher, C., Staatz, J., (Eds.), International Agricultural Devel-opment. Johns Hopkins University Press, Baltimore, MD, pp.136–154.

Morrisson, C., Thorbecke, E., 1990. The concept of the agriculturalsurplus. World Dev. 18(8), 1081–1095.

Owen, W., 1966. The double developmental squeeze on agriculture.Am. Econ. Rev. 56(1), 43–70.

Ranis, G., Fei, J.C., 1961. A theory of economic development. Am.Econ. Rev. 51(4), 533–565.

Ravallion, M., Chen, S., 2007. China’s (uneven) progress againstpoverty. J. Dev. Econ. 82(1), 1–42.

Ravallion, M., Chen, S., Sangraula, P., 2007. New evidence on the ur-banization of global poverty. World Bank Policy Research Work-ing Paper No. 4199.

Reardon, T., Barrett, C., Berdegue, J., Swinnen, J., 2009. Agrifoodindustry transformation and small farmers in developing coun-tries. World Dev. 37(11), 1717–1727.

Rodrik, D., 2006. Industrial Development: Stylized Facts and Poli-cies. Harvard University, John F. Kennedy School of Government,Cambridge, MA.

Sachs, J., 2005. The End of Poverty: Economic Possibilities for OurTime. Penguin Press, New York.

Schultz, T.W., 1964. Transforming Traditional Agriculture. YaleUniversity Press, New Haven, CT.

Stiglitz, J., 2006. Making Globalization Work. W.W. Nor-ton/Penguin, London.

Streeten, P., 1987. What Price Food? St Martin’s Press, New York.Thorbecke, E., 2006. The evolution of the development doctrine,

1950–2005. UNU-WIDER Research Paper, Helsinki.Timmer, P., 1988. The agricultural transformation. In: Chenery, H.,

Srinivasan, T.N., (Eds.), Handbook of Development Economics.Vol. 1. North Holland, Amsterdam.

Timmer, P., 2009. Lessons from food crises past. Working Paper,Center for Global Development, Washington, DC.

Wiggins, S., Cabral, L., 2007. Politics and the future of ministriesof agriculture: Rethinking roles and transforming agendas. Fu-ture Agricultures Briefing Paper, Overseas Development Institute,London.

World Bank, 1997. From Vision to Action. The World Bank, Wash-ington, DC.

World Bank, 2005. World Development Report 2006: Equity andDevelopment. Washington, DC.

World Bank, 2007. World Development Report 2008: Agriculturefor Development. Washington, DC.

World Bank, 2009. World Bank Commodity Markets Review.DECPG, The World Bank, Washington, DC.

Wright, B., 2009. International grain reserves and other instru-ments to address volatility in grain markets. Department ofAgricultural and Resource Economics, University of California,Berkeley.