The “Cotton Problem” in West and Central Africa: The Case for Domestic Reforms

  • Downloads

Cotton subsidies have received considerable attention duringthe past four years, primarily triggered by the excessivegovernment support received by the cotton sectors in theUnited States and the European Union. In response to thatsupport, four cotton-producing countries in West and CentralAfrica—Benin, Burkina Faso, Mali, and Chad—have requestedthat the Doha round of negotiations on trade liberalizationcontain financial compensation for WCA countries for as longas those Western subsidies remain in place. Brazil also broughta case to the World Trade Organization, claiming that the U.S.subsidies cause a reduction in the world prices of cotton, thusreducing the income of Brazilian cotton growers.

Western cotton subsidies should be abolished, but notmuch attention has been paid to another, perhaps moreimportant, issue. Many African cotton-producing countries,especially in WCA, must reform their cotton sector in orderto allow a greater share of the world price to reach the growersand must foster a policy environment that is conducive tothe promotion of new technologies. For the most part, thecotton sectors of the WCA countries are managed by government-owned parastatals. Competition by private entities islimited—with deleterious consequences for the efficiency ofthe cotton sectors.

Basic Facts about the WCA Cotton Sectors

Cotton is the dominant cash crop in most of West andCentral Africa, with respective cotton sectors sharing a numberof similarities. The industries were pioneered during the1960s and 1970s by the French state-owned companyCompagnie Française de Développement des Fibres Textiles(CFDT)1 in conjunction with national state-owned cottoncompanies. Those state-owned companies had a legally protectedmonopsony in cotton buying, and most also had amonopoly on primary processing, marketing, and supplyingof inputs.2

Typically, the state-owned companies would announce abase buying price before farmers planted their cotton, sometimessupplementing that price with a second payment(payable in the following season as a bonus) based on thosecompanies' financial health. Most cotton used to be marketedthrough Compagnie Cotonnière (COPACO), a CFDT subsidiary.The cotton industries also benefited from researchcarried out by the French Agricultural Research Institute orCentre de Cooperation Internationale en RechercheAgronomique pour le Développement (CIRAD).

The performance of the WCA cotton industries has beendescribed as a success story.3 Indeed, between 1970 and1988 WCA cotton yields per hectare grew at 6.1 percent perannum, which compared to 1.9 percent annual growth inworld cotton yields per hectare and implied that, had trendscontinued, WCA yields would have been similar to worldyields by the early 1990s.4 Moreover, between 1970 and2005, cotton production in WCA increased tenfold, from alittle over 100,000 tons in 1970 to one million tons in 2005.The sector's contribution to total merchandise exports in theWCA countries ranges from 25 to 45 percent, while its contributionto GDP ranges from 3 to 6 percent. Moreover, thecotton sector provides income to one million households inthe region.

Yet, the seemingly successful performance of the industriesmasked a number of weaknesses that called into questiontheir long-term sustainability or even their survival.First, the post-1980 production increases reflect solelyexpansion of the area under cultivation. In contrast, the pre-1980 production increases reflected yield increases perhectare, mainly in response to fertilizer use. A growthdecomposition analysis for the 1980–2005 period shows thatcotton yields in WCA countries remained stagnant. Thatcompares unfavorably with the 1.7 percent annual growthrate of global cotton output, which is a reflection of yieldincreases only.

Second, growers in WCA countries received low priceseven when word prices were high. For example, during theearly 1980s, WCA cotton producers were receiving between60 and 70 Communauté financière d'Afrique franc (CFAf)per kilogram for their seed cotton, while the world price ofcotton ranged between the equivalent of 200 and 250 CFAf.5Similarly, following the 1994 devaluation of the CFAf, producerprices paid by the cotton companies were adjustedupwards, but far less than the increase in world price, thusdenying WCA cotton growers the high prices enjoyed bycotton producers elsewhere. Furthermore, econometric evidenceshows that in none of the WCA countries did the pricingmechanism reflect movements in the world price of cotton.6 In other words, the price-setting mechanisms haveentirely ignored world market signals in all WCA countries.That is ironic, considering that the various price formulasused to determine the price to be paid to WCA cotton growersuse as their starting point the world price of cotton.

Third, while the panterritorial pricing mechanism (i.e.,prices being the same in the entire country) common to allWCA countries is a convenient and socially popular incomeredistribution mechanism, in effect it transfers resourcesfrom efficient cotton growers (or growers with transportationand/or location advantages) to less efficient ones. Price controlswithin each country have thus constrained overallgrowth and innovation in the industry by penalizing the mostproductive entities (or areas) of the sector.

Fourth, in periods of price declines most cotton companiesexperienced financial difficulties, which in turn led todemands for fiscal transfers from state budgets, thus puttinginto jeopardy the fiscal position of those countries. Forexample, during the late 1990s the state-owned cotton companyof Mali was in no position to manage the downturn incotton prices, because the stabilization fund, created to setaside a portion of profits from earlier periods of high prices,turned out to be empty, resulting in financial losses of CFAf56 billion ($100 million) to the cotton company. Eventually,the cotton company was bailed out by the governmentthrough budgetary support. Similar bailouts took place inmost WCA countries following the two cotton price collapses—in the mid-1980s, and in the late 1990s and early 2000s.

Fifth, because of their inefficient and inflexible structure,the cotton companies were not sufficiently prepared (interms of improved sales strategies, risk management tools,and adoption of new technologies) to respond to the changingnature of the external environment, especially the downwardtrend and volatile nature of world prices. Those pricechanges reflected technological changes, as well as, to someextent, subsidies by some developed countries (especially theUnited States and the European Union).7 For example, morethan one third of global cotton output is now of geneticallymodified origin. Furthermore, China and India, two developingcountries with high rates of adoption of genetically modifiedcotton have experienced considerable yield gains. Yet,with the exception of Burkina Faso, none of the WCA countrieshas allowed even field trials of genetically modified cottonto assess the likely risks and benefits of such technology.That is unfortunate, because recent research has shown thatthe benefits of fully utilizing biotechnology may be evenhigher than the benefits from the elimination of all cottontrade distortions.8

Finally, the CFAf is fixed against the euro (or the Frenchfranc, FF, prior to 1999). The CFAf exchange rate has beensubjected to only one adjustment since 1948—from CFAf 50to CFAf 100 per FF in 1994. That fixed exchange rate hasoften led to unintended consequences, which is not surprising,given the different structure of the eurozone economiescompared to those of the WCA countries. For example,between 2002 and 2005, the world price of cotton increasedby 20 percent in US$ terms (from US$1.02/kg toUS$1.22/kg) while it declined by 9 percent in CFAf terms(from CFAf 711/kg to CFAf 644/kg). Within the currentpolitical and macroeconomic setting, it is beyond the controlof individual WCA governments to choose the exchange rateregime that is consistent with the structure of theireconomies. That makes the case for reforms even stronger.

Only Limited Reform Efforts

Faced with those constraints, a number of WCA countriesbegan reassessing the structure of their cotton industries.With financial and technical assistance from the donorcommunity, especially the International Monetary Fund andthe World Bank, policy reforms were contemplated duringthe early 1990s to bring the cotton sector back to a sustainabledevelopment path and, ultimately, increase the welfareof the cotton growers.

However, because the reforms were portrayed as ideologicallydriven—that is, forced by the Bretton WoodsInstitutions—they were viewed with suspicion. Not surprisingly,they were subjected to considerable opposition fromthe WCA countries themselves as well as from bilateraldonors.9 For example, in a survey of the cotton sectors ofMali, Burkina Faso, and Benin, Professor Yves Bourdet fromthe Lund University in Sweden described the reasons forsuch opposition as follows:

There are two reasons behind this limited ownership[of reforms by] home government. The first is thestrong opposition on a part of the urban elite andsome farmer associations in cotton-producing countriesto the privatization of the state-owned ginningenterprises, which are at the centre of the network ofinstitutions and actors composing the cotton sector.The second is the opposition of some bilateral donors, in particular France as the main bilateraldonor, to the deregulation of the sector. No doubtthis "lack of enthusiasm" on the part of the homegovernment of cotton-producing countries and somebilateral donors has contributed to the slow pace andmixed outcome of reforms.10

Following the cotton price collapse of the late-1990s,however, it became evident that reforming the cotton industriesby eliminating the monopoly status of the cotton companiesand introducing competition is probably the only feasiblealternative. Yet, despite that understanding and "constructivedialogue," policymakers have been reluctant toengage in serious reform efforts and hence the structure ofthe WCA cotton sectors is not very different from what itwas 30 or 40 years ago. In Chad, for example, reforms arenonexistent. Although the government of Chad announcedthat it would disengage from the cotton sector in 1999, so farit has failed to act accordingly (with the single exception ofthe privatization of the company that makes cotton oil—oneof the by-products of processing of the cotton seed). Factorsbehind the unwillingness to reform include fiscal difficultiesof the cotton company, the lack of ownership of reform bythe government and, more recently, the windfall revenuefrom crude oil that has practically absorbed all capacity andenergy by officials who, otherwise, would have been incharge of the reform process.

Reforms in Benin consisted of three key elements: separationof the various links in the cotton supply chain accordingto the different functions—such as input provision, seedcotton production, transport, ginning, and trading; division ofthe responsibility for handling those functions—except forresearch and training—among a large number of actors; andorganization of the key decisionmaking process (includingissues such as the price setting mechanism and cotton deliverytime) into horizontally organized entities, which must allagree before any sector-wide decision is made. Despitethose, albeit limited, reform efforts, it appears that the performanceof the sector has not improved. For example, therewas a sharp decline in cotton production from 171,000 tonsof cotton fiber in 2004–05 to 82,000 tons in 2005–06.

Some reforms took place in Burkina Faso when the governmentsold part of the state-owned monopoly to privateinvestors. The market is currently structured into three regionalmonopsonies—a dominant state-owned company accountingfor about 85 percent of cotton purchases and two privatecompanies that account for the rest. On the positive side, itshould be noted that Burkina Faso is the only country in sub-Saharan Africa (in addition to South Africa), which is in theprocess of introducing GM cotton. However, the drying up ofthe cotton stabilization fund and the recently revealed €100million debt by the dominant state-owned company, calls intoquestion the sector's long-term sustainability.

Mali, which has contemplated reforms for quite sometime, reconsidered its reform commitment in July 2004 anddecided to start assessing the pros and cons of the reformprocess in Burkina Faso and Benin instead. In November2005 the government increased its share in the capital of thecotton company (from 60 to 70 percent) and publiclyannounced that reforms will be delayed for several years.

The Way Forward

Admittedly, the global cotton market reflects, in part,rich countries' protectionism. Nevertheless, policymakers inthe WCA countries (as well in other poor cotton-dependentcountries) face a number of challenges.

Reform programs for restructuring the cotton sector toincrease its efficiency remain largely incomplete. Reformsshould become the immediate focus of policymakers. Afterall, even if cotton prices increase either as a result of eliminationof subsidies or as a result of market forces, it will dono good to poor producers if such an increase is absorbed bybankrupt parastatals, debt-ridden cooperatives, or corruptpublic officials unwilling to engage in serious reform efforts.

Moreover, cotton producers face competition from chemicalfibers, especially since technological improvements of theearly 1970s brought the prices of chemical fibers down to cottonprice levels. In an era of globalization and intense competition,cotton producers in many developing countries maywant to explore genetically modified seed technology in orderto compete more effectively with their competitors who havealready embraced such technologies. That, however, wouldentail extensive field trials to develop varieties suitable to localgrowing conditions as well as putting into place the appropriatelegal and regulatory framework—both of which are challengingand time consuming processes.

It has often been argued that poor cotton-producingcountries should engage in domestic value addition, specifically,textiles and clothing. While a successful textile industryis good to have, especially when it comes to employmentgeneration, it is unlikely to improve the welfare of the cottongrowers for a number of reasons. First, regardless of whethercotton is consumed domestically or exported, cotton growerswill receive the same price that is determined by global supplyand demand for cotton. Second, the argument that localcotton will be favored because domestically produced cottondoes not incur transport costs may not be valid since producingtextiles for export markets implies that other types of cottonand chemical fibers must be imported, thus subjectingthe industry to the same bottlenecks that the cotton industrycurrently faces (i.e., high domestic transportation costs,delays at the ports, etc.). Third, producing textiles for localconsumption may also not be profitable, given the proliferationof imports of second-hand clothing.

It is in the interests of the consumers in the rich countriesand growers of cotton in West Africa that American andEuropean cotton subsidies should end. However, the positiveimpact of the end of cotton subsidies on the welfare of WestAfrican cotton farmers will be limited unless it is accompaniedby domestic reforms that should include privatization ofthe state-owned cotton companies and liberalization of thecotton trade.


1. CFDT was renamed Développement des Agro-Industries duSud (DAGRIS) in 2001.

2. In addition to their core activity, which is ginning, the cottoncompanies engage in other pursuits, such as input distribution,provision of research and extension services, and maintenance ofrural roads.

3. See Uma Lele, Nicholas Van de Walle, and MathurinGbetobouo, "Cotton in Africa: An Analysis of Differences inPerformance," in Managing Agricultural Development in Africa,World Bank Discussion Paper no. 7, 1989.

4. For example, world yields during the late 1980s averaged 550kilograms of cotton fiber per hectare, similar to the 460 kilogramsper hectare average achieved by the WCA countries. However, bythe early 2000s, their respective averages were 680 kilograms and425 kilograms of cotton fiber per hectare.

5. The CFA Franc (CFAf) is the common currency of 14 Westand Central African countries, also known as the Franc Zone. Itwas created in 1945, when France ratified the Bretton Woodsagreement. Initially, convertibility with the French Franc (FF)was set at 0.59 CFAf/FF, becoming 0.50 CFAf/FF after the 1948devaluation of the French Franc. In 1958 two zeros were addedto the existing denomination, making it 50 CFAf/FF. In January1994 it was pegged again to the French Franc at 100 CFAf/FF,and in 1999 it was linked to the euro at 656 CFAf/€.

6. More information on such evidence along with statistics canbe found in John Baffes, "Distortions to Cotton SectorIncentives in West and Central Africa," World BankDevelopment Research Group, 2007 (preliminary draft).

7. For a discussion of cotton subsidies and their effect on theworld market in cotton see John Baffes, "The Cotton Problem,"World Bank Research Observer 20 (2005):109–44.

8. See discussion in Kym Anderson, Ernesto Valenzuela, andLee Ann Jackson, "Recent and Prospective Adoption ofGenetically Modified Cotton: A Global Computable GeneralEquilibrium Analysis of Economic Impacts," World Bank PolicyResearch Working Paper 3717, 2006.

9. International Cotton Advisory Committee, "Statement ofFrance" and "Statement of the World Bank," Statements of theInternational Cotton Advisory Committee, 1998, pp. 45–50,50–52.

10. Yves Bourdet, "A Tale of Three Countries: Structure, Reformand Performance of the Cotton Sector in Mali, Burkina Faso andBenin," Country Economic Report no. 2, Swedish InternationalDevelopment Cooperative Agency, Stockholm, 2004, p. 41.

Download the Economic Development Bulletin


John Baffes

John Baffes is a senior economist with the World Bank's Development Prospects Group. This essay is a summary of a larger forthcoming paper entitled "Distortions to Cotton Sector Incentives in West and Central Africa."