Hemmed In: Responses to Africa's Economic Decline

Thomas M. Callaghy and John Ravenhill, editors

New York

Columbia University Press

1993

Bibliographic Data

8. The Discovery of "Politics": Smallholder Reactions to the Cocoa Crisis of 1988-90 in Côte d'Ivoire

J ENNIFER W IDNER

The February 5, 1989 edition of Fraternité Matin carried a picture of a rural scene and some children carrying a placard that read, "Oui Monsieur Le Président. Nous Vous Suivrons Pour La Lutte Contre Les Spéculateurs!" ₁ The photo caption stated that the children came from Aboisso in southeastern Côte d'Ivoire, where cocoa farmers were suffering from "la chute"-the crisis in cocoa prices and purchasing. The banner supported President Félix Houphouët-Boigny's repeated statements that international speculators had forced the world price of cocoa to new lows to serve Western interests at the expense of Ivoirians.

To a social scientist, the photograph held special interest because it implied the existence of an agrarian "collective consciousness" or awareness of political interest that many Africa watchers have been hard-pressed to find on the continent. It is almost a truism in studies of contemporary Africa that agriculturalists have diverse economic and social interests, are weakly organized to participate in contemporary formal political institutions, and are incapable of influencing policy to their advantage or of lobbying for better government services because of their poverty, isolation, and large numbers. ₂ With a few notable exceptions, cultivators have pressed their interests using what James Scott has called "small arms fire in the class war"-foot-dragging, retreat from official marketing channels, and participation in parallel markets. ₃ The Fraternité Matin photograph conveyed a different view: that farmers might have a common understanding of political interest and the means of organizing collectively to advance their views.

The image coincided with three other events reported in the Ivoirian media that also suggested the beginnings of collective action by rural interests to influence public policy formation. The first of these was the hurried participation of the Minister of Defense, Jean Konan Banny, in a series of meetings with cocoa growers-meetings designed to denounce foreign speculators, cocoa buyers such as the French firm Sucres et Denrées and the American trading company Phillips Brothers, and to rally the farmers to the president's side. After several years of stagnating or declining revenues, the cocoa farmers had grown restless and demanded attention from senior officials. The second was the continual postponement of meetings between the president and a farmers' delegation during the spring of 1989, again a manifestation of increasing farmer dissatisfaction with the government and heightened official nervousness about the potential fallout from a public confrontation. ₄ The third was the important place the treatment of cocoa farmers assumed in the mid-1989 "Days of Dialogue," during which the president met with representatives of a number of different social groups to discuss economic concerns. All three events suggested that Ivoirian farmers were beginning to mobilize for the first time since Independence.

The response departed from the farmers' reactions to earlier, similar crises. Those interviewed in this study recalled a period of comparable hardship during World War II, when the price of cocoa was also low. Farmers could afford to purchase neither adequate food nor other consumer goods. They remember eating cassava with a sauce made not of fish or meat but of green cocoa beans. ₅ Absent money to buy clothes, they unraveled the fabric of old clothing and rewove it to make shirts. Their strategy for survival, they said, was to plant more cassava and hope that the price would go up again.

In the first phases of the current cocoa crisis, farmers in the cocoa zone pursued the same strategy: they planted cassava, trimmed their household budgets, and kept the cocoa bushes in the ground in the expectation that prices would again rise. Their attempts to protect their standards of living revealed barriers to many of the strategies that neoclassical economic theory and common sense had deemed potentially helpful. Some of the responses typical in other parts of Africa-diversification of crop choice and use of kin-based insurance systems-proved unavailable because of new shortages of land, which forced farmers to choose between cocoa and another crop, and the breakdown of extended households.

By 1989, however, farmer behavior had started to change. Gradually, frustration built over the lack of information available for making important investment decisions, over the inability to place trust in local firms and cooperatives, over the predominance of an "informal" guild system in many off-farm industries, and over the procedures the government used to control smallholder participation in the marketing of some crops. All of these loomed as obstacles to diversification, extension of credit, and other short-run and long-run adaptations to higher levels of market-related risk. None were subject to individual control; they required collective action on the part of communities and, in some cases, they required that the government involve itself as "umpire." Farmers turned their backs on the Parti Démocratique de Côte d'Ivoire (PDCI) and took matters into their own hands, sending small delegations to the headquarters of the Ministry of Agriculture and important parastatals in Abidjan and in some cases forming small unions to try to seek policy change.

The involvement of cocoa farming communities in new forms of political participation was not universal, however. Farmers in some parts of the Ivoirian cocoa zone were more likely than others to talk about government policies among themselves, protest to the préfet, travel to the capital to press their demands in the hearing of senior officials, or join regional or national meetings to express concern about their declining incomes. Even within the same region, the propensity to participate varied between villages. Although they faced similar climatic conditions, local income distributions, and economic environments, worked within the same formal systems of representation and administration, and shared demographic characteristics, some villages organized to improve the economic competitiveness of members and supported "militant" farmers while others did not.

To the social scientist, this pattern is especially interesting. We know that African farmers do sometimes organize to influence policy and participate in politics. We know little about how such collective action starts-about the process of interest formation and involvement, the "discovery of politics." Even though the Ivoirian cocoa farmers are just beginning to participate in formal party politics ₆ , the variation in "proto-political" militancy offers a rare chance to study the process by which political and economic identities or "cultures" change.

This research asked two questions. First, how is it possible to explain the pattern of response to the cocoa crisis? The study tried to identify the characteristics that distinguished "participant" from "nonparticipant" villages and make sense of them. Second, how, and under what conditions, does village-level activism begin to broaden and evolve into programmatic participation through formal channels, such as political parties and economic unions?

The study on which this analysis is based took place in the Aboisso region of southeastern Côte d'Ivoire during October and November 1989, and from June to September 1990, about two years after the cocoa crisis began. The 120 smallholders interviewed were randomly selected from several villages, each of which included growers of cocoa, coffee, and oil palms. The pattern of income distribution within each village was similar. Residents numbered farmers who had access only to small plots of land as well as a few farmers with 10 to 25 hectares, some of whom had won the government's annual prize for the smallholders with the highest yields. None of the villages included large land owners or members of the country's political elite.

The discussion has five main parts. The first introduces the existing literature on the "discovery of politics" among farmers in developing areas and outlines the argument. The second describes the dimensions of the crisis and explains the various ways in which the decline in the cocoa producer price affected smallholder standards of living. The third probes the economic adaptations farmers attempted to pursue as a first line of defense against the erosion of their incomes and the reasons for their frustration with these. The fourth explores the origin of variations in patterns of participation between villages. The final section summarizes the findings to date and discusses the process of political incorporation taking place today.

Patterns of Agrarian Response

The literature on patterns of agrarian political participation in Africa is part of a larger body of research and writing on farmer response to income instability induced by climatic variation, changes in tax policy, withdrawal of subsidies, or price fluctuation. This literature has identified three major farmer orientations toward economic "crises" including, first, changes in land tenure arrangements, with a shift away from pure rental or wage relationships toward share contracts instead; second, retreat from farming and reinvigoration of kinship ties as vehicles for movement into other sectors of the economy; and third, adoption of new forms of political participation designed to affect policy or implementation of policy directly. Existing research has little to say about two important questions, however: under what conditions do farmers choose one kind of response over the others? and how does collective action get started in the first place?

Redistributing Risk Between Households and Hired Labor

The economics literature has focused on diversification of income sources and the institution of share contracting as the predominant form of farmer response. A first line of attack for farmers is to diversify their choice of crops, but its utility as a survival strategy is constrained by the type of crops they choose (farmers are less likely to destroy perennial crops than annual crops to adjust to a short-term decline in returns), as well as the availability of land, and ecological conditions.

Share contracts are production relationships in which two or more parties agree to combine their privately held resources in a farming venture and share the output in prearranged proportions. ₇ They become "rational" ways of organizing economic activities when there are multiple sources of risk associated with cultivation and/or information is costly. The starting point of this set of explanations is the recognition that if the only kind of risk farmers experience is production risk, then most of those without land will seek to become wage laborers instead of pursuing ownership, tenancy, or rental arrangements. ₈ By choosing to work for a wage, the cultivator shifts the burden of risk to the owner of the land, who must pay the wage he or she has offered, regardless of the level of return-at least in the short run. At the other end of the spectrum, a rental contract shifts the burden of risk entirely to the "worker" or tenant.

In African settings there are usually multiple sources of risk in the farmer's environment, of course. Newbery and Stiglitz ₉ have pointed out that where most people rely on agriculture for income, ups and downs in producer prices or the costs of critical inputs are likely to affect the labor market as well as the returns received by the landlord. Landlords are likely to reduce the wages they pay, lay off employees, or offer late payments, and potential workers are likely to sell their labor in the informal sector instead. If tenants are unable to reduce risk by diversifying into crops that each respond differently to drought, floods, or price changes-or that require different patterns of labor use or other inputs-share contracting becomes attractive.

The process of share contracting assumes two distinct parties: a landlord who brings land to the transaction and a landless worker, who brings his or her labor. In the models proposed, these are clear-cut roles. In Africa, however, individuals rarely occupy such distinct positions. More often a single person holds multiple roles and behaves as landlord, tenant, and wage laborer at the same time. For example, a producer might allow another community member to farm an inherited plot close to the village, while participating as a tenant in a share contract in an outlying area recently reclaimed from fallow and while devoting a day or two each week to wage labor on government "works" projects. Moreover, relationships become complicated when landlords and tenants are related to one another and have obligations toward each other that are outside the share contract. In these cases, the limited interests and maximizing behavior assumed by economists do not obtain; family obligations modify contractual relations.

Second, the notion of share contracting assumes individual rights of access to land and labor and ability to transfer these freely between individuals. Yet many African farming systems constrain individual rights of access and transfer-and even management patterns. Systems of entitlements vary greatly across the continent and are difficult to categorize, but many prevent transfer of land except to heirs. Indeed, in only four regions of Africa do share contracts stand out as dominant agricultural systems: in the Ghanaian cocoa belt, the Senegambian groundnut region, the cultivation of foodgrains in Lesotho, and production of cotton in the Sudan. ₁₀

Third, even if it turns out to be possible to model these other agricultural systems as functional equivalents of share contracts, it is important to note that bargaining over terms is unlikely to be a "frictionless" enterprise. The terms of share contracts extend in most cases not only to division of labor and output but also to contribution of a variety of inputs ranging from fertilizer to specialized skills. Because of the limits on labor but not on land that have prevailed in much of Africa until very recently, renegotiation of terms has typically benefited the tenant. Only now is that changing. In either case, modifications of community norms regarding appropriate shares requires signaling of the greater bargaining power or need of one of the parties, argumentation to establish the legitimacy of the new claims, reconsideration of the terms of performance and structures for enforcement, and mutual adjustment of parties affected indirectly by the change in terms. In short, achieving a change in the terms of the contract requires collective action on the part of tenants or landlords or both. It requires a "politics" noted in the economic theories only as an unmeasurable "transaction cost."

"Portfolio" Diversification Through the Household

Share contracts are only one of several ways to limit the consumption effects of income instability or economic "shocks." A large proportion of the literature on African agriculture suggests that farmers draw predominantly on kinship-based economic networks. ₁₁ For example, in her study of Yoruba cocoa farmers in western Nigeria, Sara Berry has probed efforts to obtain larger shares of cocoa revenues during and after the drop in yields that occurred in the 1970s. Cocoa prices increased during the 1970s, but labor shortages, deterioration of land quality, and inflation reduced yields and incomes on the one hand and purchasing power on the other. ₁₂ According to Berry, the first casualties of the downward trend were the village cooperatives and associations that had sought to improve terms of trade by bringing pressure to bear on buyers. ₁₃ Berry's interviews suggested that the drop in incomes made it difficult for farmers to support the transactions costs associated with these associations. The farmers found their collective bargaining power vis-à-vis middlemen substantially reduced as cocoa dwindled in economic importance compared to oil.

Berry writes that the response was to rechannel economic activities and to "turn away from the cocoa sector politically as well as economically." ₁₄ Farmers turned to kinship networks and the revival of descent group activities in an effort to advance the education of a few family members who could, in turn, obtain civil service jobs and channel state resources to their home areas. Farmers formed descent group-based "agencies of progress" to establish investment priorities and to help younger members tap the resources of wealthy or influential relatives. Stratification developed not on the basis of access to labor or to land but on the basis of educational attainment, as both income earning opportunities and ability to direct investment fell to those with the ability to read. Less educated "fathers" increasingly found themselves under the power of their sons. ₁₅

In his study of Hausa farmers in northern Nigeria, Michael Watts similarly found few instances of collective action to reduce the effects of income instability. In the portion of his study that deals with the 1970s, Watts has addressed the impact of climatic variations on income. ₁₆ He has found that accommodation strategies vary first by "class," or economic position of the farmer, and divides Hausa cultivators into three such groups. The first group includes the "big farmers" who are able to hire labor, use the time saved to participate in trade and informal sector activities, invest in cattle (and so diversify their risk to some degree), and accumulate in excess of subsistence requirements. Group two includes the farmers who meet their domestic food demands under normal circumstances and participate in dry-season trade or manufacture. Group three includes the "poor," or those who under normal conditions are barely self-sufficient and rarely have the capital required to enter informal sector activities or to acquire access to land in floodplains suited to dry-season cultivation. ₁₇

Watts suggests that differences in survival strategies employed by these groups generate a process of impoverishment among group three Hausa farmers. To begin with, the stronger bargaining power of the group one farmers constrains the rise of share contracts between group one and group three farmers and colors the terms of these relationships.

[Poor families] . . . generally don't have the political, familial, or economic capital to lubricate the rental process. Second, from the perspective of the landlord, a poor householder is an unattractive tenurial proposition in relation to a middle peasant with a larger gandu labor supply and availability of capital. And third, participation in the market garden sector presupposes some start up for seed, manure, labor, capital equipment, . . . and perhaps tractor hire. 18

Just as Berry sees "exit" or abandonment of agriculture as the dominant strategy of Yoruba farmers facing declining cocoa yields, Watts sees "exit" to wage labor or to the urban informal sector as the strategy of the poor Hausa farmers. Collective action to modify policies is too "costly" compared to uncertain benefits, especially when many of the group three farmers are quasi-clients of wealthier community members and when group one farmers have used their wealth to secure control of local offices.

Political Protest

Other researchers have noted that forms of "informal" political participation or protest on the part of farmers are more common than previously suggested. Naomi Chazan has made the case that social scientists have too often overlooked direct attempts to influence policy because these have not always taken place through official channels, such as political parties. ₁₉ Robert Bates has documented numerous instances of protest accompanying the commercialization of agriculture across the continent. ₂₀ In most of these case studies, the researchers identify a pattern of increasing stratification, perception of impending permanent reduction in welfare, and discovery of the inefficacy of other forms of response to change the terms of the relationship with those who are better off and who control official channels of participation.

For example, C. E. F. Beer has tried to analyze the reasons for a particular form of collective action, resort to political violence as a response to tax changes that reduced real incomes for farmers. Beer has focused on the Agbekoya riots of 1968-69 among cocoa farmers in western Nigeria. ₂₁ Economic and social change during the 1950s and 1960s had produced new elites who sought to turn existing mutual aid societies and cooperatives to their own benefit. Although the national leadership of Nigeria spoke at the time of campaigns to benefit the rural poor, levels of investment in agriculture languished. Toward the end of the 1960s, farmer grievances began to take specific form. The cultivators first objected to tax rates. During this period, the practice was to assess each farmer a flat tax on an estimated income base of 50 Nigerian pounds per year. Many of those in the region fell below this earnings level but local authorities expected all, including those retired from heavy farm work, to pay. The farmers' second concern was corruption among tax collectors, who often demanded bribes, and local councils, which were known to impose extra taxes for special services that rarely materialized. ₂₂

In 1968-69, tensions ran high. During the state governor's visit to one area, poorer farmers marched in protest against tax policies. The response of the governing elites was to call for a crackdown, a tax raid by council clerks before the main harvest, during a season when the farmers had insufficient cash even to bribe the officials. As in the Ghanaian case, the actions highlighted discrepancies in wealth and power and revealed the difficulty of working through official systems of representation. The cultivators organized in revolt, calling for a reduction of the flat tax, abolition of the district councils, and release of all detainees held after the earlier marches. ₂₃ A series of dispersed but organized riots ensued until the tax collectors and an escort of riot police again entered the area and invited the farmers to take their grievances to the government in Ibadan. During the effort to articulate their demands, farmers and police clashed, and 300 people were killed.

Among the elements of a theory to explain the incidence of different types of response one must include, first, a recognition that cultivators usually pursue a series of strategies entirely under their control before leaving agricultural production or seeking changes in policies or institutions. Watts has usefully suggested that cultivators proceed through a series of adjustment strategies, preferring those over which they have complete control and that are consequently associated with relatively low transaction costs to those that require negotiation. If the farmer has anticipated instability it is likely that he will have attempted to limit the effects of price downturns or higher input prices by diversifying crop choice. In most cases, diversification is a measure to protect against weather- and pest-related risks, but it can be used to limit the effects of other forms of risk as well. Its use is far more constrained than social scientists often assume, however, as soil types and ecological conditions often limit severely the kinds of crops that will grow on a plot.

Second, the farmer can seek to reduce expenditures by exploiting family labor to a higher degree instead of hiring a contract worker or labor team, by skimping on labor-consuming soil- and moisture-conservation practices such as ridging of fields and frequent weeding, by cutting down on use of other inputs, or by withdrawing children from school. Third, the farmer can seek sources of supplementary income, either through work on casual labor contracts or by participating in informal sector enterprises such as automobile mechanics. Fourth, the farmer can try to induce other family members to seek additional income by engaging in some form of rural enterprise, such as basket weaving, pottery making, or animal husbandry. Fifth, the farmer can distance himself from obligations to supply food or assistance to members of the extended family, teenagers, or community members. And sixth, the farmer can seek to borrow funds. ₂₄

What, then, precipitates the rise of new forms of participation designed to influence government behavior and policy? Although methods by which farmers confront problems of collective action run through the literature, the actual attention devoted to conceptualizing efforts to build bargaining leverage and supply "solutions" or institutions (and thus the incidence of different types of responses) is limited, as is the effort to consider seriously the influence of formal systems of representation, which appear in most of these accounts as a kind of residual category. ₂₅

This study takes a step toward filling this gap in our understanding. It suggests that new forms of political participation emerge partly in response to changes in formal systems of representation that alter the costs of acquiring information, asking questions, speaking at meetings, sending delegations to speak with officials, and organizing economic interests. In themselves, these changes are not enough to produce collective action among households and villages, however. In many cases, a fundamental change in the understanding of appropriate forms of interaction and of authority relationships-or "political culture" more broadly-is essential. To adopt new forms of political participation requires that an individual have the skills and know-how to use new sources of information, the latitude to fail and try again, and special incentive for making the effort.

In the Ivoirian case, the commercialization of agriculture had reordered the social structures of villages in the cocoa zone and, in some cases, created local "space" for experimentation by destabilizing existing authority systems. It had also brought wealth, and with it, higher levels of formal schooling and greater access to media. It also brought new sources of risk. The commodity price "crisis" of the late 1980s accelerated a process of change in patterns of behavior already underway. It clarified perception of a "class-wide" economic interest and, by creating an economic recession in urban areas, it precipitated the return of highly educated and well-traveled young people to the countryside, bringing new skills to the villages. Finally, and most important, it encouraged those who remained in the cities to try to increase the impact of the increasingly limited funds they could remit to rural relatives, by joining together in associations, prioritizing village needs, providing funds to communities instead of households, and offering special incentives to farmers to submerge intra-household conflicts to broader economic interests.

The Cocoa Crisis of 1987-90

The trigger to the evolution of new patterns of political participation in Aboisso region was the crisis in cocoa production and its conjunction with decreasing returns to coffee and oil palm cultivation. Côte d'Ivoire's "economic miracle" had created problems for farmers unlike those in most African countries, where producer prices for agricultural commodities have typically been too low to make production profitable. In the Ivoirian case, the government maintained relatively high producer prices for its main agricultural exports and sheltered farmers from fluctuations in the world market by setting aside revenues when world prices greatly exceeded the producer price, paying out a subsidy from these revenues when world prices dropped. It established a special stabilization fund, the Caistab, to handle cocoa and coffee marketing and to manage revenues. Under these conditions, production of export crops, especially cocoa, expanded dramatically. Côte d'Ivoire became the world's largest producer and exporter of cocoa and, next to Ghana, the producer of the highest quality cocoa. Beginning in the 1980s, the growth of production began to depress the world market price, however, and by the second half of the decade the government found itself steadily subsidizing production and depleting government revenues in the process. The World Bank and IMF eventually intervened, urging a reduction in the producer price.

A quick look at the relationship between producer prices and the world market price gives the false impression that the Ivoirian government was giving its farmers only a very small percentage of the revenues their commodity commanded on the international market. Because the government was bearing most of the collection, marketing, and transport costs, and because these costs are reflected in the world price, the actual differential between the producer price and the world market price was much smaller than the raw data suggest. By 1988, when farmers were earning 400 FCFA/kilogram, the world market price was only 450 FCFA/kilogram. Because transport and handling charges added about 300 FCFA/kilogram, the total cost of production exceeded the market price, and the government was losing about $1 on each kilogram sold. ₂₆ The government's revenue base disappeared, and Caistab moved to borrow funds abroad, principally from French commercial banks.

The crisis affected urban areas first. Houphou‘t-Boigny imposed restrictions on government hiring, dismantled some of the parastatal organizations, and finally began to remove people from the government payroll. Publicly financed construction came almost to a halt. Differences in income levels between the countryside and the rural areas decreased. According to World Bank estimates, although in 1980 the average per capita income in urban areas was three times the level of rural areas, by 1985 it was only twice as great, and by 1988 it had fallen still farther. ₂₇

The cash crisis also struck the middlemen and shippers who handle Ivoirian cocoa. In an effort to improve efficiency in the cocoa marketing system, Côte d'Ivoire had long used private traders, or traitants, to carry out the actual transactions with farmers and to bring the fermented cocoa beans to port facilities, where it is turned over to licensed shippers or exporters, of whom there are about 40. The shippers pay the traitants. Both traders and shippers are allowed a certain profit margin on the cocoa they handle and are reimbursed for acceptable costs at the end of the cocoa season. In the interim, they finance the marketing and handling of the cocoa out of funds they borrow from commercial banks. When Caistab went into deficit in 1987, however, it was unable to make timely payments to the intermediaries, who, in turn, either failed to pay producers or failed to pay their bank loans (which triggered a liquidity crisis in the banking system). Further, it cut acceptable profit margins for traders, some of whom simply folded their operations.

The farmers themselves constituted the third group affected by the cocoa crisis. They felt the impact initially not through a reduction in the official producer price, which the government maintained until quite late in the crisis, but rather through the increasing unwillingness of the traitants either to offer the official price or to make payment for loads transported. During 1988, there was widespread evidence that traitants were paying as little as 200 FCFA/kilogram-or half the official price. ₂₈ The government warned it would revoke traders' licenses and issue stiff fines for such behavior. Later, it urged that cooperatives (Groupements de Vocation Coopérative, or GVCs) take over responsibility for conveying cocoa to the exporters, although this practice did not spread quickly. Absent funds advanced from shippers, and given the risk of government censure, many traders either resorted to use of promissory notes or declared bankruptcy. Stocks of fermented cocoa began to accumulate in the country's interior.

The government's policy was initially to try to force the world price higher. In 1987, it held production off international markets, surrendering the tax receipts it would have earned from the sale. With the largest world surplus on record and the prospect of increasing production from Côte d'Ivoire in subsequent years, however, most of the cocoa failed to move. By the middle of 1987, the embargo on exports was lifted, and limited sales began again, the government seeking premiums for the high quality of its stock. It succeeded in obtaining limited sales to European confectioners at premium prices, but the effect of the two-tiered pricing system was to push returns to production of poorer grade cocoa still lower.

The lifting of the embargo did not improve the lot of the producers, however. Midway through the 1988-89 season, they still had not received payment for the previous year's crop. The traitants were either unable or unwilling to pay for what they had marketed; many waited for the government to drop the official price, which it seemed likely to do any day. Finally, on June 3, 1989, the government dropped the price for mid-season purchases to 250 FCFA/kilogram, although it suggested that the price for regular season purchases would go back up. In October, it dropped the price again, to 200 FCFA/kilogram, the price the World Bank had urged for several years.

In rural areas, the "Cocoa Crisis" also refers to a more generalized decline in the profitability of other key cash crops. Côte d'Ivoire is a coffee producer, and the Aboisso region has long produced significant quantities of robusta beans. During this same period, international coffee prices were also in decline, partly because of the new favor shown to arabica producers, and Côte d'Ivoire found both the price it received for coffee exports and its quota reduced. In the fourth quarter of 1989, it dropped the coffee producer price to 100 FCFA/kilogram from 200 FCFA/kilogram. Because farmers often produce both coffee and cocoa in combination, the price change reduced still further the incomes of farmers already affected by the cocoa crisis.

To make matters worse, the coffee growers of the region for the most part cultivated older bushes, near the end of their major bearing age. Just before the collapse in prices, the government had initiated a campaign to convince farmers that they would have to prune their bushes heavily, in order to rejuvenate them. Because this pruning (récepage) would prevent the bushes from bearing for some time, it too constituted a "shock" to farmers' incomes. Although the government eventually offered 60,000 FCFA per household to help cover the losses of the farmers who pruned their fields, it is not clear that any of the farmers in this study participated in the program. The program design created a conundrum by not stipulating exactly when payment was to take place-before the farmer cut the field, in which case the farmer might simply take the money and fail to do the work required, or after the process, in which case the government might forget to pay. Uncertainty hindered the expansion of the program and focused the farmers' attention on changes in price rather than pruning requirements.

Finally, the cocoa-growing areas are also oil palm production areas, and returns for this crop were also in decline, although the culprit was partly the change in producer price and partly the withdrawal of fertilizer and input subsidies under the new austerity regime. Palmindustrie, the palm oil parastatal, had responsibility for buying the fruit, and the government set the price. Although domestic demand for palm oil had increased, the world price had steadily fallen during the 1980s, and the nominal producer price had not increased since 1985, when it rose from 19 FCFA/kilogram to 21 FCFA/kilogram. ₂₉ Coincident with the price increase, Palmindustrie had withdrawn most parts of its input subsidy package, limiting the effective price incentive for farmers to switch into oil palm cultivation.

It is not always the case that those who are victims of crisis perceive their situations as armchair analysts do. Subjective perceptions are a product not only of real circumstances but also of local systems of belief, which assign different significances and different priorities to important events. The multiple sources of risk African farmers routinely encounter have prompted one scholar to eschew studies of "risk and choice" for the theme, "coping with chaos." A farming household in Aboisso region encounters many types of risk to its annual income, including, most notably climatic variability, sickness of a household member, and market-related shocks (input shortages and withdrawal of subsidies, as well as price changes). The data collected suggest that during the period 1987-1990, however, smallholders in Aboisso were quite capable of distinguishing shocks of the third type from the others that plagued their lives, however. They were very much aware of the "cocoa crisis," although they preferred to extend the definition of "la chute" to the decline in profitability of coffee, coconut palms and oil palms as well. Of the cocoa producers, 24% identified the reason for diminished returns in the 1987-88 season, when traitants were issuing IOUs and bargaining to pay prices below official levels, as a change in the producer price. During the 1988-89 season, when the government announced the reduction in the official price, fully 96% attributed lost revenue to change in the government producer price policy.

Just before the crisis, farmers had been forced to cope with bad weather. Climatic variability is great in most parts of Africa, and Aboisso is no exception to the rule. In normal years, as part of the forest zone, Aboisso receives substantial precipitation in all months except November through February, when the Harmattan comes. Every so often-and in no predictable cycle or pattern-rainfall declines. If the regular dry season is drier than average or if the rains come a little late, the cocoa, coffee, and palm fruit crops may be unaffected. It is the pattern of rainfall and temperature during the most important part of the growing season that affects yields more than average levels of precipitation.

Although rainfall varied during the years leading up to the "cocoa crisis," causing persistent Western speculation that cocoa and coffee output would fall, only in 1983-84 and 1985-86 did drought constitute a serious problem. Some of the farmers in the region lost their cocoa holdings to brush fires in the 1983-84 season; yields dropped universally. In 1985-86, drought and heat struck both coffee and cocoa plantations during periods of heavy flowering, reducing the amount of fruit set, although rain arrived in time to limit the damage to the palm fruit crop. The perceptions of the farmers coincided with the evidence of rainfall patterns supplied by the local meteorological station. Sixteen percent of the cocoa growers identified the 1983-84 season as one in which returns were reduced by drought. The 1985-86 season proved difficult for "reasons of bad climate" in the view of 36% of the cocoa farmers and 30% of the coffee growers. By comparison, only a small handful of growers in all groups considered the 1987-88 and 1988-89 seasons difficult because of weather problems.

Shortages and other market-related, non-price sources of risk were identified as reasons for reduced returns during this period principally among the oil palm growers. The extension office for cocoa and coffee, SATMACI (Société d'Assistance Technique pour la Modernisation Agricole de la Côte d'Ivoire), had withdrawn input subsidies many years before, when it was clear that to keep world prices high, the country would have to limit production. Only about 4% of the cocoa growers interviewed and 18% of the coffee growers reported regular use of fertilizer. ₃₀ 30 In consequence, few identified fertilizer prices or shortages as important reasons for reduced returns. Those who did use fertilizer (overwhelmingly those who believed producer prices would climb again), complained about the high cost, which, at 3 to 4 50 kilogram sacks per hectare, came to about 18,000-24,000 FCFA ($60 to $80). ₃₁ The oil palm growers, who lost their fertilizer subsidy in the mid-1980s, were far more likely than members of the other groups to indicate rising fertilizer costs as a major component of the crisis, although the problem loomed largest as a source of lower returns in 1985-86 (when Palmindustrie removed the subsidy) and the two subsequent years. By the 1988-89 season, only 9% identified fertilizer costs as a major source of risk; all but 15% had stopped using fertilizer.

If the smallholders of Aboisso perceived the "crisis" as the concurrence of several sources of market-related risk, but especially of producer price variation, what expectations did they have about future changes in price? The overwhelming majority of cocoa, coffee, and oil palm growers believed the government's current price levels only temporary, but there was substantial debate among the farmers in each group about the direction of future changes. The cocoa farmers were evenly divided between those who thought the price would improve again and those who thought the price would drop further. The great majority of the other producers saw rising producer prices on the horizon.

To what did growers attribute price changes? Only a handful of the farmers interviewed understood the pricing system. Most knew only the official price-and the price traitants had asked them to accept during the 1987-88 season. They were unaware of the existence of a separate "world market price," although this information was available in two forms to local officials, députés, and larger planters. The Chambre d'Agriculture, a forum for discussion of agricultural issues, issued a monthly bulletin to subscribers, indicating cocoa and coffee prices on several different markets, as well as the volume of purchases by major international buyers. Although the smallholders did not participate in the Chambre d'Agriculture meetings, they had regular contact with some who did. This information was either not passed on or met with apathy. It was also possible to listen to crop reports and international prices on Ghanaian radio, which most could pick up in at least the eastern part of the region. Awareness of this information source appeared to be restricted to villages that were close to the border and that had members who were comparatively wealthy planters, as opposed to smallholders.

The reasons for the price changes were the source of much discussion in the villages. The government had promoted the view that foreign speculators, the European trading companies, had sought to reduce prices "artificially"-without relation to supply and demand. Most cocoa growers believed that oversupply had created their problem, however. Local economic science tended to downplay the role of non-Ivoirian producers (about whom they knew little) and pinpoint government mismanagement of its economic role as the real cause of oversupply. Farmers spoke of the need to invent uses for cocoa beans so that demand would increase, and turned silent when asked about the role of foreign speculators, indicating either extreme politeness to the interviewer, as a foreigner, or disagreement-more likely the latter. Most expressed concern about the government's ability to manage the economy. They commented that every time the government had urged them to plant a crop in the past, it had created plantations of its own, and the price had fallen. It had done so for copra-and, in the words of one farmer, "everybody knows that government officials own lots of coffee and cocoa, too." ₃₂ Further, farmers were increasingly suspicious about the source of funds for the basilica at Yamoussoukro and spoke more frequently, as the crisis progressed, of the diversion of cocoa revenues to purchase the cathedral and of alleged sales of beans from the president's own plantations while the government allowed their own production to rot in the fields or in warehouses. ₃₃

Economic Adaptations

The first place the crisis made itself felt in farmers' lives was in the standard of living available to households. Retreat to subsistence agriculture was not a possibility in anyone's mind. Cash was essential for the education of children, and despite grumbling that girls made poor field assistants once spoiled by schooling, education was universally prized. At a minimum, the family had to supply 6,000 FCFA per child for books and 3,000 FCFA per child for a uniform. In most cases, schools required some tuition payments as well, but the boarding costs for older children who could not attend school nearby were the most prohibitive charges, and many despaired of the need to pay lodging expenses. Most farmers worked hard to continue payment of these fees, and the women interviewed in the study cooperated in trying to keep children in school by contributing the money from the trade in foodstuffs they would ordinarily have spent on soap, cloth, or jewelry. ₃₄ Depressed incomes also meant more restricted diets, heavy in cassava and low in protein; even rice and yams had become too expensive to use as staples. Indeed, truck drivers complained that the roadside kiosks no longer made foutou with the yams that had been the mainstay only a year or two before. ₃₅ There were fewer purchases of medicines and clothing, and households cut back their use of hired labor by as much as 50% in the early stages of the crisis, later moving entirely to family labor. ₃₆ Family labor was diverted to cash-crop plots, where, in most cases, it was typical to employ workers for wages. ₃₇ Women extended their workdays so that they could spend time tending the cocoa and coffee bushes, in addition to foodcrops. Household tasks suffered, in consequence. Both because of increased workloads and the growing inability of young men to make continuing payments of the brideprice, divorce rates increased, according to some. ₃₈ Scrimping was a stopgap measure with definite limits, however, and most people looked for longer-term solutions to their predicament.

There were a number of possible economic adaptations to income instability. A farmer might borrow on local credit markets or from distant kin to ease the family through bad periods, draw on formal or informal sources of "crop insurance," diversify into production of other sorts of agricultural commodities or into "off-farm enterprises," or participate in share contracts with workers so that workers would bear some of the risks associated with production. Most farmers tried some combination of one or two of these. Because of the characteristics of kinship structure in these villages and imperfections in factor and producers' markets, however, most of these strategies proved difficult to pursue. To work, they required social and political institutions that did not then exist.

Borrowing and Insurance

There were technically four ways a farmer might borrow to support consumption expenditures, as opposed to agricultural investment, during this period. One was to take out a bank loan on the formal credit market, but that was extremely difficult for most farmers. Neither commercial nor development banks were inclined to loan money for consumption, few farmers had the kind of collateral the banks demanded, and the liquidity crisis meant that many banks would dole out cash in extremely limited quantities, at best (by 1990, even in Abidjan it was difficult to cash a check for more than the equivalent of $1000). Even the Caisses Rurales d'Epargnes et de Prêts (CREPs), of which there were three in the Aboisso area, were hard-pressed to provide funds, although two of the farmers, out of 120 interviewed, had borrowed from them.

More readily available, in theory, were three other types of credit transactions: receipt of loans through informal, village- or commune-based credit markets; diversion of funds from investment loans; and borrowing from relatives in urban areas or in other parts of the country, outside the cocoa zone. The statistics suggest that few people borrowed money at all, however. More than three-quarters of the cocoa producers said they had never borrowed during any period of reduced returns during the previous six years. Comparable proportions of other groups said the same. Overwhelmingly, they pointed to two reasons-inability to pay back the money, and since all farmers in the region were suffering reductions of income at the same time, no one really had money to lend. ₃₉ Commented one farmer, "I try never to borrow money because everyone has problems right now." ₄₀

Several respondents spoke specifically about the difficulty of participating in local or regional credit markets-that is, of borrowing from other planters or from merchants. Credit was difficult to obtain because few had funds to lend, and the risk of default was great. Some commented that credit was so scarce on the informal market that they "would not know who to see about a loan." ₄₁ Interest rates were exceptionally high, in consequence. In one village, the going rate of interest was 100 percent. "That's to say, a 5,000 FCFA loan would require reimbursement, within the year, of 10,000 FCFA," reported one village member. Only very rarely did farmers in this region obtain credit from the merchants who bought and sold their crops. On occasion, cassava dealers from Abidjan offered loans, which they would deduct, with interest, from the sales of the farmer's crop. This practice does not appear to have been widespread or to have lasted long because of the high implicit rates of interest charged (as estimated from the low prices offered), if the trader returned to make payment on the crop at all. In the few cases in which a farmer did secure a loan, the collateral used was almost always land, a commodity of increasing value in the Aboisso region. Although it is not clear from the available evidence it appears that in most cases the value of the land far exceeded the amount of money owed. The practice of "pledging" cocoa land was one with historical roots; cocoa farmers in Ghana and in Côte d'Ivoire had often used land as collateral during the price fluctuations of the colonial period. In Ghana, the practice had led to increasing concentration of ownership, but in the year after the Ivoirian cocoa crisis started, increasing concentration of ownership was not yet apparent, or even rumored in the Aboisso region.

An alternative possibility was to borrow investment funds from the traders or parastatal responsible for marketing the cash crop, whether cocoa, coffee, or oil palm fruit, and to divert the loan to consumption. In an earlier era, of course, a farmer would have received credit for purchase of fertilizer, insecticides, herbicides, and tools from the government marketing board, which would then have deducted the loan from the farmer's receipts at harvest time. In a bid to improve the efficiency of its cocoa and coffee marketing system, however, Côte d'Ivoire had placed crop purchases in the hands of private middlemen, or traitants, and the system of credit for agricultural investment had collapsed, as farmers accepted loans, then sold their crops to other middlemen, who would offer a price from which debt repayments were not deducted. Only Palmindustrie, the parastatal responsible for oil palm production, still maintained a system for extending credit, and it offered loans to support school fees in addition to investment funds. Most of the farmers in the sample who said they received credit had accepted loans through Palmindustrie, and Palmindustrie officials at the local factories ₄₂ corroborated the information provided by the farmers. Several smallholders expressed interest in planting oil palms simply so that they might participate in the program, but the parastatal had shrewdly decided that it would not authorize planting of additional plots until 1990.

A third possibility often invoked by cultivators in studies of other parts of Africa was to borrow from members of the extended household living in urban areas or in other regions of the country. This system was less a kind of credit market than a family-based insurance system, to which people paid "premiums" throughout their lives in the expectation that they might call upon resources in time of need. Among the Aboisso cocoa growers, however, this strategy appeared to have little importance, at least in the first year after the effective drop in prices. Although a few farmers borrowed from neighbors, merchants, or, most especially, Palmindustrie, none reported loans from household members.

The key to understanding this choice appears to lie in the low social salience of kinship ties among the region's inhabitants. Where, at the turn of the century, the farmers of Aboisso came principally from two Akan groups, the Agni and the Attié, now they hail from many different regions within Côte d'Ivoire and from far-flung locations in Nigeria, Mali, Guinea, Burkina Faso, Togo, and Ghana. Gone are the days when a single set of descent-based rights and privileges governed access to land, labor, and other critical resources in a village. In their stead are negotiated arrangements, supervised by chiefs and notables often from non-Agni or Attié backgrounds. Gone too are the extensive kinship networks that used to tie immigrants to relatives in the areas left behind. In most cases, immigrants appear to have limited the kin-based claims that could be placed on their resources by cutting themselves off from distant relatives. ₄₃ In times of crisis, of course, the severing of kin ties meant that the cocoa farmers could not draw on the reserves of cousins in other regions, either. Even when sons, daughters, and cousins worked in nearby towns or in Abidjan, they were of relatively little assistance to their rural relatives. The cocoa crisis had affected urban incomes as well, and many were beginning to find that returns to labor on the land were increasingly comparable to returns to labor in the urban informal sector.

In the area covered by the study, there was only one "cluster" of instances in which commercial crop insurance was considered. A traveling insurance salesman had visited one of the villages, and several village members, including the chief, had thought the idea a good one. They paid their initial premiums, but they never heard from the salesman again. Absent government or quasi-public "umpiring" of the market, commercial insurance did not go forward, and in the presence of price-setting by government institutions, futures markets, which would potentially smooth producer price fluctuations and perform much the same function as insurance, failed to evolve.

Diversification of Crop Choice

In response to fluctuating cocoa prices, the farmer could attempt to diversify crop choice or to diversify into nonagricultural enterprises-informal sector businesses. By diversifying, the farmer could ensure that not all elements of his "income portfolio" varied in the same direction at the same time and could thus be relatively certain of receiving a basic level of return. Although this strategy could not and did not protect smallholder incomes when the government suspended cocoa sales and traders failed to pay farmers, most farmers had some inkling that cocoa and coffee prices would fall, in advance of the actual government decision to drop the price to 200 FCFA/kilogram. In the months before the decision the government had "protested too much" that it had no intention of reducing prices, and most farmers in the sample said they anticipated the drop.

Not all farmers pursued diversification. About a third of all coffee growers, a half of all cocoa farmers, and three-quarters of all oil palm growers reported increasing the number of crops cultivated in the wake of reduced returns. Cooperative members were less likely than others to add new crops to their programs. In almost all of these cases, the crops added were food crops, especially cassava. In fewer than 8% of the 120 cases studied did farmers attempt to earn money in nonagricultural enterprises or by participating in wage contracts on nearby plantations.

There were four principal forms of "crop diversification" during the cocoa crisis. The first and most important was increased production of cassava, which could be sold in the towns and in Abidjan. Typically production of cassava and other food crops was the preserve of women, who could use the proceeds from their sales to help finance children's education and to buy things they needed for themselves. With the advent of the economic crisis, men moved into cassava production, working land formerly cultivated by women and opening up new fields along rights of way. As people rapidly discovered, however, the problem was that as more and more farmers grew cassava and other food crops, prices for these commodities fell precipitously, lessening the stabilizing effects of the strategy on standards of living. ₄₄

The second major form of agricultural diversification was to expand fishing in the lagoons, rivers, and swamps of the department. Fish had long been an important part of the local diet. During the early 1980s when drought and bush fires reduced cocoa and coffee harvests, farmers discovered that they could market fish to passersby on the road between Ghana and Abidjan. Truck, taxi, and bus drivers would stop, purchase fish or crabs, and take the produce to the capital or to nearby towns for resale. Demand grew during the mid-1980s, then took off in the first year and a half of the cocoa crisis. Gradually, farmers began to notice that it was harder and harder to catch fish in the most accessible spots, and the price climbed higher. The city dwellers were willing to pay a high price for the fewer and fewer fish available, and, increasingly, Aboisso households began to forego fish in their own diets. They were too expensive to consume. By the end of 1989, women began to report that they had eliminated fish from family meals, except on important occasions or as a splurge once in a while, and when they did include it, it was often tinned fish from the town instead of fresh lagoon fish.

Harvesting of wild kola was another possibility pursued by many cocoa and coffee growers. Kola grew in the forest and could be found interspersed with coffee and cocoa bushes in many fields. Highly valued as a stimulant by Muslims and non-Muslims alike, kola nuts have long had a market in northern areas, and the Sénoufo and Dioula immigrants in the Aboisso region demonstrated particular skill in brokering a new kola trade during the first year and a half of the cocoa crisis. Several farmers interviewed in the study reported that the price for a kilogram of kola had exceeded the price for the equivalent weight of cocoa or coffee in the early months of the crisis and that it was now far more profitable than either beverage crop. Nonetheless, a highly perishable commodity with a limited market, kola was unlikely to support an entire region for long.

A fourth possibility debated hotly among village members was cultivation of rubber. Beginning in the early 1980s, the World Bank and the EEC had put pressure on the Ivoirian government to promote rubber production as a way of buffering farmer incomes from the ups and downs of the highly volatile cocoa market. A second campaign to promote smallholder cultivation of rubber began in 1983, with Bank support. Farmers who could produce deeds or some other evidence of land title, certify that their plots bordered passable dirt access roads, or pistes, farmed plots within 25 kilometers of the processing plant, and demonstrated that they or cosigners were at least 43 years old would receive authorization to plant rubber and to receive loans for fertilizer and other inputs. They would also be reimbursed for their labor in clearing and planting the plots of land. ₄₅ The input packages ended after the initial phase of the program was over, and the farmers of Aboisso deliberated long and hard among themselves about the advisability of planting rubber. Oblivious to the use of synthetic substitutes, some argued that there would always be a demand for tires and that therefore they could not lose if they went ahead. Others commented that every time the government had urged them to plant a crop in the past, it had created plantations of its own, and the price had fallen. They believed that rubber was not going to be a solution to their problems. ₄₆ By the end of 1989, however, only a handful of the farmers in the area had opted to plant rubber trees.

In their debates about rubber the farmers also noted a problem that was an important barrier to most kinds of agricultural diversification: increasing land scarcity. Of the barriers to agricultural diversification where perennial, or tree crops predominate, the new experience of land shortage was probably the greatest. As long as cocoa, coffee, or palm prices might go back up and the farmer believes he or she can earn an average return that exceeds the possible return from other crops, it makes sense to keep the bushes or trees in the ground. Except for kola, it is not possible to maintain other crops on the land while cocoa or coffee are in place; intercropping is not a real possibility. To diversify into other crops, the farmer either has to get rid of his trees or he has to gain access to more land.

Unlike most parts of Côte d'Ivoire, however, the Aboisso region now had a population density that puts pressure on land supply. Immigrants began to arrive in significant numbers during the colonial period, and they now outnumber the Agni and Attié in most villages. Most settled in the area during the mid-1900s, when land was abundant and participation in the cocoa and coffee trades appeared highly attractive. In some instances, young entrepreneurs first worked for Agni farmers and after a period of six years, or so, were given permission to cultivate plots of land on their own. They came in "waves," and most of the elders interviewed indicated a succession of migrations, which often started with an influx of Baule or Abouré from central Côte d'Ivoire, then Dioula and Sénoufo from the north, then people from other countries. The most recent settlers are the Togolese and Ghanaians, and older residents are fond of pointing out, half-jokingly, that in the latter case, Ghanaian women arrived first. As migrant workers became increasingly important in cocoa production, they obtained greater bargaining power vis-à-vis Akan groups and negotiated access to land and later to wives, loosening the hold of the original inhabitants over control of critical resources and weakening their political bases. In a few cases, the colonial government accelerated the cultural diversification of the region by rewarding young men for their military service or their work in road construction with gifts of cocoa land.

With the weakening of Agni power, lineage authority over land came into question. Although the form of the Assassi Tuhô, the Agni system of land-transfer, is retained as an informal ceremony in some parts of the region today, it is no longer even an informal requisite. ₄₇ The system eroded substantially during the middle part of the century and was dealt a final blow in 1971 by the passage of laws abolishing "traditional land tenure," or group rights of access. A land market did not immediately arise, however, and the buying and selling of land is still shrouded in secrecy, although both notables and farmers in all of the villages studied eventually "confessed" to participation in exchanges of land for money. It is possible to purchase land in areas of "classified forest" by approaching the administration of "Eaux et Forêts." These plots currently sell for about 80,000 FCFA (about $265) per hectare. It is also possible to purchase or rent land from other farmers, although officials often deny knowledge of some of these exchanges. The farmers surveyed in this study did participate in land deals and commented that the going rate for purchase of a hectare of land was about 70,000 FCFA at the time they purchased their land, about 15 years ago. ₄₈ More recently, as land has become increasingly valuable, those with land appear to favor rental arrangements. Rental of a hectare for two years generally comes to 40,000 FCFA per hectare. The distinction between renting and buying is often obscure, and many of those interviewed said that the number of cases of land litigation had increased recently as people tried to ascertain the nature of their claims.

Without easy access to land, however, it becomes extremely difficult for the smallholders to diversify their crop choice. This situation places great pressure on the farmer's assessment of future price trends. If, in fact, a commodity is in long-term surplus worldwide, then the farmer is going to face not volatile prices and positive average returns but declining prices and declining returns-and he should cut down his bushes. None of the smallholders interviewed had adequate market information to make that judgment, however, and they emphasized their dependence on government policymakers in making critical assessments.

Diversification into Off-Farm Enterprises

Under these circumstances, one might expect households to seek nonagricultural sources of income-to manufacture stoves or shoes for sale along main roads or in towns, for example, or to sign up with the local labor bureau for work on industrial plantations or road contracts. The study results suggest, however, that in the early phases of the cocoa crisis, 1987-89, only a very few farmers pursued this option. Only 8% of those interviewed indicated participation in informal sector enterprises or wage contracts. This figure corresponds with a World Bank study which found that only 8% or fewer or rural household members in each of several age categories participated in petty trade or manufacturing, regardless of the part of the country. ₄₉ In this instance, again, the structure of existing economic institutions made it difficult for most to modify their income portfolios, in the short run, in a way that would significantly protect their standards of living.

The only significant participation in "the informal sector," outside of agricultural trade, was illegal alcohol production-the tried and true backyard still. Because many of the palms in the area were reaching the end of their productive lives, enterprising young men negotiated with palm owners to get permission to tap the palms for their sap in exchange for cutting down the trees. A farmer who could produce 200 liters of palm wine a day for trade along main roads (with the complicity of the regular police, the customs police, the forest police, and the army, which all staffed roadblocks nearby) would net as much as 1,000,000 FCFA per year, after costs, or a little over $3,000. Of course, one had to be able to secure one's operations from interruption by the authorities or by competitors who might leak information about the still to local officials, and that could cost money. It is probably no accident that the most voluble palm wine producers had secure positions in the village leadership or strong ties with the local administration.

For most farmers, entry into the informal sector was blocked by lack of money, knowhow, and a severe labor constraint. Those interviewed said that to go into business required an apprenticeship. An informal guild system operated in local towns, and it was almost impossible to learn a trade or gain access to critical raw materials unless one worked with a member of the guild first. Apprenticeships cost money, of course, and few farmers had sufficient savings left to pay the fees required. Only one farmer in the study reported working as a mechanic, and he made it quite clear that he could do so because his grandfather, who had also been one, had taught him basic skills. ₅₀ Most viewed entry into petty production and trade as nearly impossible for these reasons.

Participation in wage contracts or the informal sector was also limited by severe labor constraints. Although it is possible to neglect cocoa and coffee bushes for short periods of time without severe deterioration in the field quality, it is still necessary to invest enough labor to prevent catastrophic disease of plants and to produce food for household consumption. In most cases, the farmers interviewed indicated that, with children in school for part of the day, there were simply not enough hands to do the labor required. The loss of an able-bodied worker to another pursuit was not acceptable to them, in large part because they placed no faith in promises of high wage rates or high levels of return. Growing seasons are not divisible goods. Unless the farmer, or a relative, could guarantee high levels of earnings for an agricultural season and could convince the household that these earnings could be used to purchase food equivalent to what his or her labor would have produced on the land for the whole season, off-farm labor seemed a poor gamble indeed. Most respondents replied to continued probing of this issue with a curmudgeonly grumble that, "agriculture was still more remunerative anyway" and comments about the nobility of working the land. ₅₁

Share Contracts

Neoclassical economic theory suggests that another alternative for coping with increased risk, especially with multiple sources of risk, is for the farmer to abandon payment of wages to day laborers, which forces the employer to bear the burden of price variability, for a system in which those who cultivate the land assume some of the risk themselves by receiving a share of the output as payment for their work. This system also limits the need for close supervision of field labor, because the worker's earnings are tied directly to his or her productivity. Such share tenancies have had a long history among the Akan cocoa growers of Ghana and Côte d'Ivoire. The abusa system of the colonial period provided for the division of the output into thirds, and the return of a third to the person who had tended the land in the "owner's" absence. This arrangement operated alongside payment of wages to migrant workers. Sometimes it supplanted wages; at other times it existed in tandem.

With increased market-related risk, one might expect to see an increased incidence of share tenancies in the cocoa- and coffee-growing areas, but such does not appear to be the case-at least not yet. In late 1989, only 8% of cocoa growers reported giving part of their harvests to those who "owned" the land, and only 4% reported receiving part of another's harvest. ₅₂ Those who believed that producer prices for their main crops would continue to drop were slightly more likely than others to participate in these contracts, a finding consistent with the theory. These findings appear to be in line with popular perceptions that abusa-and anything like it-had disappeared by the early 1960s, except in the management of land used for cassava production. Respondents showed little interest or enthusiasm for the system.

Why were share contracts unpopular ways of rewarding labor and allocating risk in the cultivation of perennial cash crops but practiced more widely in the food crop sector during the early stages of the cocoa crisis? The answer probably lies partly in the substantial bargaining power of labor, partly in the increasing shortage of land, and partly in the discovery of new ways of managing a labor force.

At the time of the crisis, affordable labor was still a scarce commodity even in this region of the country. The going wage on nearby pineapple plantations, the most direct competitor for the resource, was between 850 and 1000 FCFA per five-hour day (about $3.00). Few smallholders could afford to pay that much, although they generally agreed that they paid about 600 FCFA per day when they had to hire itinerant laborers. The high value placed on labor meant that workers were in a strong bargaining position vis-à-vis the household heads. They did not have to share the increased risk associated with cocoa, coffee, or oil-palm production. If asked to do so, they could simply turn around and go elsewhere, where their services were equally in demand but where the reward was higher.

For the farmers themselves, share tenancies held out not only the possibility of sharing the risk but also a way to avoid paying wages out of limited and stretched cash reserves. But increasingly the farmers of Aboisso lacked the one resource that could compel participation of workers in these arrangements: land. Under the abusa system, as practiced in this part of Côte d'Ivoire, the tenants who cultivated the land usually received their own plots on the periphery of village fields after about six years. Population density, increased settlement, and classification of some areas as protected forest had produced a land shortage in Aboisso by the time of the cocoa crisis, however. To make matters worse, many farmers had received subsidized input packages (fertilizer, pesticides, plants, for example) for oil palms, for which in return they had only to register the land and present a title to the parastatal. Titles had not been issued to those who admitted to participation in share contracts. Most smallholders had thus lost their ability to make share contracts attractive alternatives to higher wages.

Finally, all of the villages in the area had originated new forms of labor organization during the "mini-crisis" of the early 1980s that reduced the need of farmers to resort to share contracts as means of supervising labor. Several farmers in the study commented that individual workers could prove highly unreliable and required much supervision, which few had time to provide. Beginning in the early part of the decade, most villages had evolved competing "labor teams" of young men and women, who would clear, hoe, seed, weed, and harvest land for a fee, which the members would divide among themselves or, more often, invest in projects or ceremonies. Fees varied according to the size of the group, running from about 16,000 FCFA for a day's work by a team of eighteen to 25,000 FCFA for a day's work by a team of thirty-two. ₅₃ The labor teams were more efficient, in the view of most farmers, because they created peer pressure for hard work and because one always knew to whom to go to complain about a lazy performance. In some villages labor participation was strictly voluntary. In others, a young man or woman who claimed exemption more than three times, whatever the reason, would be fined 1,000 FCFA. Although share contracts remained a way of spreading risk, the labor teams solved the problem of supervision, which loomed more important than distributing the burdens of a temporary downturn in price or increase in costs, in the eyes of many.

Share contracts did come into vogue between older farmers with few children and younger farmers who needed to grow food for sale to pay school fees. These differed from earlier arrangements, however, in that the parties were neighbors from the same village, both with access to land, and the contract involved no agreement to turn over rights of access to the plots.

The Insufficiency of Individual Responses

None of the major forms of economic adaptation could preserve household standards of living during the early phases of the crisis. Even in combination, the strategies barely kept people from despair. For the smallholders of Aboisso, who based their cash incomes on sale of perennial crops, there were four major problems in getting any standard economic strategy to work. One had to do with the emergence of a land shortage at precisely the time when urban job opportunities were few and the government had launched a "return to the country" campaign. Two were rooted in the absence of public- or quasi-public institutions for ensuring the integrity of market transactions. And the fourth was a consequence of the changed kinship structure in the region.

For growers of mature, perennial crops, diversification of crop choice in order to protect against market-related risk requires access to additional plots of land. It is difficult to cultivate alternative crops, especially food crops, on plots heavily shaded by cocoa or coffee bushes or oil palms, especially where roots compete for space. Farmers can cultivate land devoted to annual food crops more intensively, by extending intercropping and succession planting, but there are strict limits on the impact diversification carried out on these plots can have on cash incomes-especially if neighbors have the same idea at the same time. Few smallholders had the cash to buy more land. Some started to cultivate the rights of way between large estates and the main roads or access routes, as estate-owners rarely used this land fully. Most simply found agricultural diversification unworkable.

Those who did try to market surplus food as a short-run strategy encountered a second problem; they found that they could not always trust the traders who came from the city to buy their crops. When the farmers sold on credit, expecting the traders to return with their money after the sale in Abidjan, they rarely saw the proceeds. When they accepted credit from the merchants and agreed to deduct repayment from the revenues the sale generated, they ended up paying usurious interest rates. And they worried about the prices offered, too. Absent an umpire to guard against collusion among merchants, the farmers had no way to trust that the prices offered were fair market prices, and resentment over real or imagined "cheating" boiled right below the surface.

Theoretically, smallholders could take up opportunities in the "informal sector" or establish credit and insurance systems. But these routes, too, were at least partly blocked as a result of efforts by a few to segment and manipulate markets to their own advantage. As in the marketing of food crops, the absence of institutions to enforce contracts and to broadcast market information hampered the creation of insurance programs or credit systems that could have helped farmers through the crisis. As most African governments, the Ivoirian government has concentrated on the management role of the developmental state at the expense of its role as market "umpire." There are few procedures for resolving disputes between traders and farmers-or between farmers-and even fewer organizational vehicles for supporting these. If a dispute arises, the sous-préfet or préfet may choose to step in. These are busy officials, however, with many responsibilities, and few have the time to handle more than an occasional land dispute. Complaints about traders or insurance salesmen are nearly impossible to address, because there are few ways to track itinerant merchants with the resources the departments have available.

Second, the farmers viewed the "informal sector," quite rightly, as anything but the "free market," with low barriers to entry, low capital requirements, and high levels of competition, that Western analysts believed it was. At least in the rural areas, becoming a mechanic in an open-air garage, shoemaking, marketing cloth and other items villagers needed, smithing, and other "professions" required that one find the money to acquire an apprenticeship. Even then, some trades were open only to the sons of those who had practiced these before. And, of course, the returns from taking up one of these vocations would have to equal or exceed the average returns from agriculture, minus the extra wage costs required to replace the field labor lost. Few farmers thought such levels of return likely. The carpenters and vendors evident in the villages were strangers who had temporarily settled in the area from homes far away.

Finally, the smallholders of Aboisso were apparently unable to tap into kinship-based "insurance networks," because they had earlier cut themselves off from relatives in distant regions, in order to limit demands on their own resources. Unlike many parts of Africa, Aboisso is home to people from a wide variety of cultural backgrounds. Many had migrated to the region to take advantage of entrepreneurial opportunities and had severed ties with less enterprising cousins back home in order to preserve their capital from the constant demand to support funerals or weddings. They had done well, compared to more northerly neighbors or compared to relatives in neighboring West African countries, but few had succeeded so greatly that they had paid for a university education of a son or daughter and therefore few had salaried relatives. Even if household members had worked their way into the urban elite, the government's decision to impose a 15 to 40% tax on public and private salaries would have eliminated any savings well-to-do relatives might have provided.

The Rise of a "Proto-politics"

The futility of individual efforts to protect standards of living prompted some of the farmers to join together to communicate their concerns to the administration and to create a larger role for themselves in decision-making about agricultural policy. In the first phase of the crisis, their efforts were clear but limited. The farmers began to try out new forms of participation, to discard those they found wanting, and to sharpen their own understanding of their political interests. They stopped short of organizing the kind of farmers' interest lobby that an unidentified smallholder from the western region suggested necessary in a much-publicized interview with Radio France Internationale. ₅₄ But many took issue, at least implicitly, with the farmer's comment, "We are not happy, but there is nothing we can do against the government." They reorganized the cooperatives that the extension service had constituted in each village and began to turn these to their own ends. They began to send delegations to the préfet and then to Abidjan, where they tried to meet with senior officials of the parastatals and the Ministry of Agriculture. Some began to join farmers from neighboring villages and towns to discuss marketing strategies, either through cooperatives, in the case of cocoa, or between individuals and international buyers, in the case of those who had planted pineapple next to their other crops.

Some farmers did decide to wait out the crisis, however, and for the political scientist, the interesting question is why some began to explore new forms of political participation while others did not. In particular, why did responses tend to vary by village, in the early phases of the crisis? By trying to explain the difference between "active" villages and "inactive" villages, it may be possible to illuminate the foundations of collective action.

"Participant" versus "Nonparticipant" Villages

The typical active village had several characteristics that distinguished it from those whose residents had chosen to cope with the crisis only through reorganization of household consumption and production patterns. These villages were "participant" in three different respects.

First, villagers had joined together to form decision-making groups that operated parallel to older systems of authority or to structures put in place by the state. Sometimes these informal groups used existing organizations to their own ends, and at other times they operated as ad hoc advisory committees. In either case, they made their influence felt by moving to solve disputes that had barred provision of community services-water towers and classrooms, for example-in neighboring communities. To do so, they had to overcome a problem common to most of the villages in the region.

Generally, villages in Aboisso included a small number of Agni or Attié families, descendants of the original inhabitants of the area, and a larger number of households from diverse cultural heritages. The latter had acquired access to land in the area in one of two ways. Earlier in the century, some had come to the region in search of work and had participated in share contracts, agreeing to care for the cocoa and coffee crops of resident farmers in return for part of the harvest and the promise of land several years hence. After they acquired their own plots, other members of their households joined them. Alternatively, some farmers had received grants of land from the colonial government as reward for military service or for work in the colonial administration.

The Ivoirian government installed two structures of authority within these villages. The first was a PDCI cell or comité that included a president, appointed by the chief or notables, and representatives of different quarters of the village. The purpose of this committee was to represent the village at regional party meetings, transmit messages from party headquarters and from the president of the country, and collect annual fees for identity cards.

The second authority structure was a cooperative organization with the power to choose cocoa and coffee buyers, collect and distribute crop payments, manage the local distribution of fertilizer in some cases, and collect revenues from each member to be used for provision of community services. Although the government had initially tried to promote cooperatives or Groupements de Vocation Coopérative (GVCs) and had been instrumental in their financing and organization in the early years, most of the GVCs of Aboisso had lost their strong ties to the administration or to parastatals and, by the time of the crisis, lived or died on the initiative of village farmers. Where they existed, GVCs provided a forum where farmers could argue about the sources of the crisis, about ways of improving their incomes, and about how they might try to influence the policies the government pursued. At this time, slightly more than half of the cocoa and coffee growers participated in annual cooperative meetings and elections. In many cases, however, the cooperatives had collapsed. In a few, they persisted but lost membership as residents disputed the choice of crop buyers and the use of funds collected and as the parastatal that had helped expand coffee and cocoa production, SATMACI, stopped subsidizing inputs.

Initially, some of the farmers thought about using the cooperatives both to make their views heard (it was easier for a cooperative president to get an appointment with the préfet than for an ordinary farmer to do so) and to appropriate some of the functions that "untrustworthy" intermediaries had previously monopolized. As one might expect, a higher proportion of those who envisioned future declines in producer prices, especially for lower grade coffees, participated than did those who believed that prices would remain stable, at their lower levels, or than those who believed prices might increase. Some of these groups contemplated marketing crops directly to exporters in Abidjan in the next growing season, to avoid problems with middlemen, whom they had grown to distrust, and to capture the traitants' margins for themselves. ₅₅

In most villages of the region, however, demographics and political structures, especially cooperative organization, conspired to produce conflict and disrupt the functioning of these groups. Cooperatives funded provision of public services or public goods by taking a fixed proportion of the revenues each farming household received in a given year and using that money to improve community amenities. Cooperative officers were able to monitor revenues of all farmers who sold their produce through the cooperative organization-and those of any additional farmers who received assistance from SATMACI's extension services. Revenues depended on yields, and typically newer residents generated higher returns on their land than did the Agni or Attié inhabitants, whether because of greater attentiveness to good farming practices, a higher proportion of household members of working age, or settlement on richer soils. The newer residents protested that their hard work meant that they supported a much higher proportion of the costs of community services and public goods than did the original inhabitants. The autochtones replied that as the original inhabitants, whose beneficence had allowed the newcomers to settle in the region in the first place, they were entitled to pay a lower proportion of the costs of community benefits. ₅₆

The first respect in which "participant" villages differed from "nonparticipant" villages was in the degree to which they had resolved or submerged this conflict either by creating parallel authority structures or by modifying the rules of operation of the existing cooperatives. In the "participant" villages, farmers had found some way to agree on an appropriate level of contribution from each member household and until the cocoa crisis depressed revenues greatly these villages had continued to improve water systems and classrooms and convince the government to help grade roads.

These farmers were also quick to defend their achievements. The general collapse of the cooperatives and the proliferation of new, small, uneconomic organizations in their stead attracted the notice of the Ministry of Agriculture, which tried to intervene to create larger, more viable units. The Ministry communicated its plans to extension agents in the Aboisso region through its newspaper, Terre et Progrés, where it argued that the management problems of the cooperatives had four components: the failure of the GVCs to repay loans extended by the BNDA and the consequent unwillingness of the BNDA and other credit sources to finance investment or trade by cooperatives, the tendency of private haulers and government security agents to impose high official and unofficial transport costs on the GVCs, eliminating profit margins in consequence, the inability of the factories to receive, process, and store the crops as they arrived, forcing return of truckloads to village storehouses, and "laziness, nepotism, and indiscretions" on the part of GVC officials. ₅₇ Its analysis overlooked the relationship between farmer revenues and provision of public benefits and missed one of the major reasons for cooperative collapse. Its solution in many cases violated the arrangements the "participant" villages had forged. Farmers in these villages were predictably angry and made their displeasure known through delegates they sent to meetings with government officials. ₅₈

The second characteristic common to the "participant" villages but not shared by the others was the creation of "delegations" to represent grower interests locally and in Abidjan. By the end of the first year of the crisis, a new kind of participation had emerged. Dispensing with party representatives, farmers in several villages constituted delegations to take their complaints to the appropriate ministries or traveled to Abidjan on their own behalf. This step was no small accomplishment. It meant breaking with established systems of authority relations mentally, first. In other regions of the country, acceptable behavior toward people in power required that farmers with grievances take their complaints to the sous-préfet or to the préfet. To go above that level, unless invited, was unheard of. Second, it meant debating with others to define the main issues and frame a list of demands. Third, it meant spending money to travel by bus to Abidjan. Fourth, it meant negotiating bureaucratic red tape that would daunt the most clever of academics. ₅₉

An older farmer from one of these villages who had lived through the crisis of the 1940s commented on the village's experience in constituting a delegation to go to Abidjan. He said that in February 1990, when the transporters and other groups went on strike in the capital, some of the farmers in the region thought they should strike too. One person assembled some other farmers from neighboring villages, but when the men met they decided it would be difficult to carry out a successful strike. They decided to go to Abidjan instead. In May, they went to the headquarters of Palmindustrie and to the Caisse (Caistab). They were well received, he said. They urged the officials to raise producer prices and to offer "primes d'encouragement," or monetary incentives for maintaining their plantings during hard times. The officials made lots of promises but did not act on them. The farmers of this village would consider sending another delegation to act on their behalf, he said. ₆₀

The interviews with farmers in "participant" and "nonparticipant" villages revealed that the former were far more likely to travel to meetings and seats of government to express their concerns than the latter, even though all of these villages had relatively easy access to road transportation. Senior parastatal employees in the region and in Abidjan confirmed reports that farmers often showed up in their offices, sometimes with advance warning and sometimes not, to seek action on individual complaints and convey information about problems experienced by the village. In some cases, parastatal officials had memorized the names and case histories of individual farming households, so often did the chef d'exploitation pay a visit to the headquarters.

The third difference between patterns of participation in these communities lay in media use. Residents of the "active" villages followed local and national news to a greater degree than did farmers in neighboring, "nonparticipant" communities. Asked about recent local meetings among grands planteurs to discuss marketing strategies and about meetings of the regional Chambre d'Agriculture, farmers in "participant" villages were far more likely to correctly identify the forum and the topics discussed than farmers in "nonparticipant" villages. None of the farmers in any of the villages surveyed had participated directly in those gatherings, which were generally the preserve of farmers with larger holdings or of growers of different kinds of crops. All had relatives of friends who traveled throughout the region during the week and could potentially return with news, however. Only those in the "participant" villages recalled the events or knew, in some cases, that the organizers of some of these meetings had urged creation of new systems of representation that would allow smallholders in the villages to send a delegate to future gatherings.

Further, the "militant" villages all had bamboo cinemas-VCRs and television sets attached to small generators or car batteries hooked up in series. The films obtained for show in these bamboo theaters rarely had anything at all to do with agriculture. Indeed, the favorites were the kinds of films that cause scholars to invoke the specter of "cultural imperialism" out of despair at the images displayed. ₆₁ With a television set, however, it was also possible to catch the national news, in its local language edition, and the Wednesday evening agricultural special. Although government-controlled, this medium nonetheless allowed viewers some knowledge of the concerns of farmers in other parts of the country. It offered exposure to the outside world less available to the residents of the other villages.

"Exposure to media" was an endogeneous variable too. The images the news conveyed may also have proven instrumental in the choice of the "delegation" as a form of political participation, instead of demonstrations or boycotts. Television watchers would have been familiar with the president's tactic of calling "days of dialogue" during times of crisis and of inviting representative groups to Abidjan to air their views in front of him. Just weeks before the farmers led their community delegations to the Ministry of Agriculture, the president had met with other, larger groups, whose members had not minced their words in voicing their complaints. It is likely that the form of participation the farmers chose was selected precisely because they had seen others use it and witnessed the president's acceptance of it. Although an exceptional form of participation, the delegation was sanctioned where other kinds of activities were not.

Non-economic Incentives for Collective Action

In his study of patterns of collective action among villages in south India, Robert Wade found that ecological and economic variables explained the emergence of "village republics" in some cases and not in others. ₆₂ For example, communities at the remote ends of irrigation systems had greater economic incentive to organize than did those closer to the source of water and were more likely to fit the "active" pattern than the "inactive." The villages studied in the Aboisso region did not fit that model, however. All were similarly situated with regard to water, climate, soil fertility, transport, and availability of land. For the sources of the pattern observed, it is necessary to look to noneconomic variables.

The villages in which farmers organized and went to Abidjan to deposit themselves on the doorstep of the Minister of Agriculture were distinctive from others in three main respects. First, their chiefs did not appear to exercise strong control over the notables. In one case, the chief was extremely sick and unable to participate in village affairs. In another, the chief had no evident interest in governing, or at least seemed in practice to advocate an extremely laissez-faire approach to the job to which his neighbors had elected him. It is likely that the absence of the usual lines of authority in these two villages forced residents to solve problems among themselves, without the invocation of hierarchical authority, and that this experience encouraged them to think more flexibly about authority relations generally.

It is important to note that the legitimacy of the second center of political authority in the village, the PDCI, had also faded. Rates of participation in PDCI meetings remained high. As a way of hearing the latest from Abidjan and as social events, the gatherings are still attractive. The "discussion" was usually unidirectional, in the view of the participants, however. The role of the representative was to transmit the views of the party leadership and the president's words of wisdom, which some of the local representatives viewed as a sacred trust. Farmers may have voiced grievances about government policy, but the party had no clear mechanism for transmitting these, according to villagers.

More irksome, in the farmers' view, was that the party collected money but gave nothing in return. In displays of indignation that rarely occur in front of a visitor, village notables consistently chastised the PDCI representatives for trying to convey the impression that the farmers saw a return on the money spent on the identity cards. In more northerly regions, it was fairly common for villagers to assume that World Bank projects and aid-supported wells or infrastructure were PDCI contributions. The farmers of Aboisso labored under no such illusions, and they took their representatives to task for what they perceived as the party's failure to represent their interests in these times of crisis.

Just two months before the 1989 interviews, the party had also taken a step that was highly unpopular with many rural residents. A wave of violent crime had swept the capital in June, July, and August 1989. Growing fear of urban bandits, and, behind the scenes, the president's concern about the possibility of factional fighting between members of the growing Lebanese population, led the government to order the army out of the barracks and onto the streets of the capital to help police the city. It also installed the army along major rural roads, to supplement the customs police, gendarmerie, forest rangers, and other officials already on patrol. It lacked the money for stepped up security, however, and so, with great fanfare, the PDCI contributed substantial funds toward the cause-funds drawn from rural residents' already stretched pocketbooks. Because transport costs were already high as a result of the "indélicatesses de certains agents de contrôle sur les routes," ₆₃ to use the Ministry of Agriculture's phrasing, the farmers were none too pleased by the PDCI's proposal.

The second source of variation in patterns of participation was demographic. In the "participant" villages the "notables" were younger and more diverse in their cultural backgrounds than in the other villages. In most cases, how they became "notables" was not at all clear, except that they took an active interest in local affairs. Most were farmers who had participated in cooperatives and watched them fall apart or who had signed on to one or another of the government campaigns to diversify crops, only to have subsidy packages removed after they joined. It is likely that their experiences in these programs had given them a keener sense than their neighbors of which policies were inappropriate or poorly framed. To the extent that smallholders learned what their interests were by experimenting and trying to make their farms work, these people were among the most "educated," even if few had any formal schooling at all. Because of the loose authority structures in these villages, these younger and especially committed farmers were able to gain some control over their affairs within the village and build confidence in their views.

Some of these farmers had higher levels of education or experience than was the norm for the region, as well. In the mid-1980s, the government had initiated a campaign to return youth from the cities and towns to the land, on the grounds that the rural labor force was aging and needed the ideas and muscle-power of the young men and women who had left home. During the crisis, youth began to return to the countryside in increasing numbers, bringing both skills and problems. In contrast to most of the farmers surveyed, these return migrants could read and write-and speak French. One very successful farmer from an area adjacent to the study site commented that in his village, the labor invested in organizing the community and in maintaining the cooperative generated sufficient returns only if the farmers were literate. ₆₄ Illiterate farmers, who could not read the accounting books of the cooperative, only caused trouble, in his view. The "relève paysanne" slowly began to increase the number of farmers with the skills necessary to monitor use of funds within the community.

The influx of young people also created new pressure to increase the returns to labor on the land. A prizewinning farmer on the outskirts of one of the less active communities noted that the dynamics of the household and community were beginning to change, with the need to accommodate the ambitions of the young men who chose to come home. ₆₅ Two of his own sons had returned earlier in the year and already one had joined a delegation to go to Abidjan to protest low prices for oil palm fruit. He noted that there would be "real problems" if yields and prices did not increase and that his sons would probably become more involved in trying to influence government policy.

Third, these villages also differed from the less active or "nonparticipant" communities in that those who had left, students and urban workers, remitted part of their earnings to the community, not just to individual households, and took an active role in determining how that money should be spent. Urban-based village associations are common to both "participant" and "nonparticipant" communities. What varied was the kind of relationship the urban association maintained with the home area. As urban dwellers found themselves facing tougher economic straits, association members often decided to give to the home community as a group and to exert greater control over the use and management of the funds than they had previously. They constituted a new center of authority which further altered the role of the chief and notables but also made the managers of the cooperative or of the informal development association far more accountable for their actions than had been the case. The control over money combined with ability to exert pressure on household members provided instruments for reducing the costs of collective action within the village. Further, these groups reported important information that shaped the diagnosis of economic problems and interests and the selection of appropriate responses.

By way of illustration, the ressortissants of one village had decided that it had to take a stronger role in resolving the conflicts that perpetually afflicted the cooperative and undermined its effectiveness. In the very early years of the crisis, when price trends were still unclear, the members decided to open a headquarters in the village instead of in Abidjan, so that they would all have incentive to return home and monitor the problems farmers faced. They helped the new, young president of the GVC call residents together to meet as a group and then called the notables to witness the general support proposals for management change had among the villagers. Members returned regularly for GVC meetings and intervened when necessary to help resolve conflicts and ensure the wise use of funds contributed. ₆₆

In conclusion, therefore, the greater willingness of farmers in some villages to organize associations to bolster crop sales, lower costs, and provide basic amenities; constitute delegations to advance farmers' interests in negotiations with officials; and monitor local and national news about farming problems had its roots in three conditions. First, the authority for resolving conflicts and managing resources was already decentralized in these villages, partly because the cultural diversity of residents had already attenuated existing forms of central control and motivated farmers to think about their relationships with others in somewhat malleable, "contractual" terms and partly because illness or personality had weakened the influence of central figures, such as the chief. Second, in the "participant" villages, younger men and women, often literate and often with experience of travel through other parts of the country, had returned to play important roles in decision-making, either indirectly, by demanding the chance to realize new kinds of ambitions, or directly, by moving to fill power vacuums. Third, the "participant" villages received help and ideas in organizing their affairs from urban associations which, in economic hard times, increasingly pushed for a role in improving the economic lives of those who demanded resources from them and who controlled an important form of social security-agricultural land. Although the commercialization of agriculture and the process of industrialization and urbanization would probably have gradually yielded these conditions under any circumstances, the cocoa crisis accelerated this process.

The "Discovery of Politics": An Epilogue

By way of conclusion are some reflections on what happened after the period of this research and on the degree to which Côte d'Ivoire's agricultural development was "hemmed in" by the country's position in the international economy.

The Formation of a Farmers' Union

Delegations and strikes do not programmatic politics make. The farmers' early experiments with new forms of political participation stopped short of the formation of a political party or even a broadly based union. In May 1990, the government announced a move to multiparty democracy, which, however, failed to capture the imaginations of most farmers. And, indeed, the major parties made little effort (or could afford little) to influence rural voters. By the end of July 1990, none of the parties' representatives had stopped at these villages, accessible though they were to a main road. Party competition took place in urban centers and seemed a world apart even to farmers who were members of the delegations that had traveled to Abidjan earlier in the year.

That changed in July 1991. A little less than a year after the country's first multiparty elections, the farmers' growing predisposition to participate blossomed into the formation of a union, the Syndicat des Agriculteurs de Côte d'Ivoire (SYNAGCI). The first congress, a meeting of more than 1,000 delegates and hangers-on, took place at Boudépé, a village about 55 miles from the country's capital, Abidjan. A newly-organized opposition political party, the Front Populaire Ivoirien (FPI), had taken a hand in launching this first union congress-a fact state television took pains to point out. Since the legalization of multiparty competition in Côte d'Ivoire at the end of April 1990, FPI leader Laurent Gbagbo had criticized the government of President Félix Houphou‘t-Boigny and the Parti Démocratique de Côte d'Ivoire (PDCI) for transferring funds away from the Caisse de stabilisation, the commodity board charged with stabilizing producer prices and managing the marketing of key export crops. He argued that the government had used revenues from international sales to finance unproductive urban schemes and the Basilica Notre Dame de la Paix, a Florentine cathedral at the president's birthplace, Yamoussoukro. Although Gbagbo helped inaugurate the first congress, he announced to his listeners that the FPI would remove itself from union affairs after the initial meeting. In his view, SYNAGCI should operate independently of all political parties. "There is no such thing as PDCI cotton or FPI cotton," he remarked to his audience. "There is just cotton." ₆₇

A model of participation for SYNAGCI may lie in the unions that have formed in those parts of the agricultural sector where the government has already retreated as a major player and chosen to liberalize markets. The best example is the pineapple growers' union that has flourished in Bonoua, on the edge of the Aboisso region. In the pineapple sector, the presence of many small producers and the absence of a market umpire initially led to overproduction of poor-quality pineapples that depressed prices for Ivoirien produce on the European market and reduced farmer incomes. First larger growers, then smallholders organized to create cooperatives that would coordinate sales and provide some quality control. The cooperatives advanced the farmers a portion of the anticipated proceeds of the sale and paid the rest after transactions were complete. In many cases, however, the farmers ended up with negative net returns, so high were the marketing costs of these organizations. The union moved to work out agreements that would regulate quality but allow individual farmers or small groups of farmers to sign contracts directly with buyers, rather than work through a central company. Although only a few years old, it also started to act as an advocate for farmers' broader economic interests. Already well organized, this group provides a model of participation already closely watched by farmers in the more active villages.

The speed with which union involvement and party-centered political participation spreads in rural areas is likely to depend on the amount of contact between villagers and university students or younger urban relatives who currently constitute the major support and source of energy for all of the new political groups. The année blanche, or year without school, the president declared in the wake of street demonstrations, in March 1990, sent many youth back to their rural families. Although these students are without great exposure to the platforms of the new parties or the speeches of their leaders, they do constitute a link between these movements and the rural areas, or, at the very least, an attentive audience. Their own views are likely to shape those of their age cohort who have chosen to return to the land and to farming and who are the main experimenters in the "participant" villages.

Agricultural Development and Côte d'Ivoire in the World Economy

At their first congress, in July 1991, farmers throughout the country spoke of the need to make government more accountable in its use of resources. They recalled the support they had given Houphou‘t-Boigny in the country's first elections and suggested that the president had forgotten to whom he owed his political success. In their view, political elites had misspent revenues the farmers had generated over the previous years. Instead of paying attention to mechanization of agriculture, crop research, and training "les jeunes" to work the land, civil servants had sought to enrich themselves. If they had formed a union earlier, some said, they would not have suffered so.

In many respects, the farmers were quite correct; their incomes need not have declined so if government had invested funds from cocoa, coffee, and cotton sales more wisely. Like many other countries, Côte d'Ivoire had to accommodate to a world in which new producers of cocoa and coffee had appeared, increasing supply of these crops and lowering the prices countries received from their sale. The international terms of trade for these commodities deteriorated in the period after the "boom" of 1976-77. Yet farmers in southeast Asian countries, such as Malaysia and Indonesia, also poor and also growers of these crops, fared much better than Ivoiriens. Policy choices mattered. It is possible for incomes to remain stable when prices decline, as long as yields improve constantly, as they did in Malaysia and Indonesia, where output grew eightfold in the 1980s. That requires information, research, and improved extension, however.

One of the reasons southeast Asian farmers experienced less deterioration in standards of living (indeed, many saw their welfare improve) was that government had chosen to invest in agricultural research and infrastructure to a greater degree than did Côte d'Ivoire. Beginning in the late 1960s and accelerating throughout the 1970s, Houphou‘t-Boigny's appointees borrowed funds from the Caisse (or against its revenues) to build new public enterprises, which they subsequently managed with the aim of enhancing their own cash-flow. Monies invested in sugar development in the northern areas, in the mid-1970s, disappeared into untraceable construction invoices. The projects were guaranteed to yield negative returns from the outset; European markets were highly protected and the North American industrial countries already faced an oversupply. The main rationale for going ahead with the sucriers was partly to distribute patronage to restless northern regions but also, and more centrally, to let new generations of politicians accumulate resources and reduce the pressure they brought to bear on the elders who had fought for Ivoiriens to share the advantages of the French settlers. Similarly, state funds disappeared at this time into a large program to house civil servants, an operation whose directors managed to benefit political elites and French advisers who had properties to rent or sell above market prices. Both of these types of expenditure reached such scandalous and public proportions that the president eventually stepped in either to dismiss the ministers involved, in the first case, or to prosecute the civil servants in the courts, as in the second instance (the famous LOGEMAD scandal).

In short, the farmers who met at Boudépé had a good grasp of the source of their problem. The state was unresponsive to their long-term interests, its officers seeking short-term individual benefit at the expense of the "goose that laid the golden egg." It wasn't that the country did not invest. The problem was that it invested earnings in the wrong way. One way to try to avoid becoming "hemmed in" in the international economy was to organize to make the governors more accountable for their choice of investment targets. That was precisely what Côte d'Ivoire's farming population decided to do. In the long run, of course, export agriculture could not keep the government's accounts balanced and the standards of living of its citizens on the rise. That would require not only a healthy export agriculture but also a shift out of import-substitution industries into processing of agricultural exports and manufactures. Nonetheless, limiting the drain on rural resources was and remains an important component of that transformation, if the experiences of Malaysia, Indonesia, and the East Asian countries, most of them resource-poor, are any lesson.

---

Note 1: "Yes, Mr. President. We will follow you in the battle against the speculators." Fraternité Matin, Abidjan, February 5, 1989. The research for this chapter, which is part of a larger project, was carried out in Côte d'Ivoire between September 1989 and September 1990. The author wishes to thank Atta Brou No‘l, Thomas Eponou, and Kouadio Yao of the Centre Ivoirien des Recherches Economiques et Sociales (CIRES) for their guidance and kind assistance and the World Bank McNamara Fellowship Program for financial support. The author has also benefited greatly from comments provided by the Harvard University Department of Government Political Development Study Group. Back.

Note 2: For example, see the discussion in Henry Bienen, "The Politics of Trade Liberalization in Africa," Economic Development and Cultural Change, 38, 4 (July 1990): 718. Although recognizing that farmers have employed various means of resistance to state policies over the past several decades, several researchers have documented farmers' efforts to influence policy. This work includes the essays of Robert Bates in Essays on the Political Economy of Rural Africa (Berkeley: University of California Press, 1983) and Beyond the Miracle of the Market (Cambridge: Cambridge University Press, 1989), as well as case studies that include Stephen Bunker, Peasants Against the State: The Politics of Market Control in Bugisu, Uganda 1900-1983, (Chicago: University of Illinois Press, 1987) and Jeffrey Herbst, State Politics in Zimbabwe (Berkeley: University of California Press, 1990). Back.

Note 3: James C. Scott, Weapons of the Weak: Everyday Forms of Peasant Resistance (New Haven: Yale University Press, 1985). Back.

Note 4: Economist Intelligence Unit, Quarterly Economic Review of Côte d'Ivoire, Third Quarter, 1989, p. 6. Back.

Note 5: See the interviews of July 13 and July 14, 1990 in Aboisso region. Back.

Note 6: President Félix Houphou‘t-Boigny announced in the spring of 1990 that Côte d'Ivoire would move toward a multiparty political system. At the time of this writing there are 23 political parties in addition to the Parti Démocratique de Côte d'Ivoire (PDCI). Back.

Note 7: Definition of share contracting drawn from the work of A. F. Robertson, The Dynamics of Productive Relationships: African Share Contracts in Comparative Perspective (Cambridge: Cambridge University Press, 1987). Back.

Note 8: David Newbery and Joseph Stiglitz, "Sharecropping, Risk Sharing, and The Importance of Imperfect Information," in James Roumasset, Risk, Uncertainty and Agricultural Development (New York: Southeast Asian Regional Center for Graduate Study, 1979) p. 315. Back.

Note 9: Ibid. Back.

Note 10: A. F. Robertson, The Dynamics of Productive Relationships: African Share Contracts in Comparative Perspective (Cambridge: Cambridge University Press, 1987) p. 26. Back.

Note 11: There is an extensive literature on this pattern of response. Among the political scientists who have written on this subject, Gšran Hyden is perhaps the best known. See Gšran Hyden, Beyond Ujamaa in Tanzania (Berkeley: University of California Press, 1981). Back.

Note 12: Although not a case of response to fluctuation around a secular trend per se, the drop in yields was corrigible and did not represent a change in a secular trend itself. Back.

Note 13: Sara Berry, Fathers Work for Their Sons (Berkeley: University of California Press, 1985) p. 98. Back.

Note 14: Ibid., p. 105. Back.

Note 15: Ibid., p. 107. Back.

Note 16: Michael Watts. Silent Violence: Food, Famine, and Peasantry in Northern Nigeria (Berkeley: University of California Press), 1983. Back.

Note 17: Ibid., pp. 411-412. Back.

Note 18: Ibid., pp. 437-438. Back.

Note 19: Naomi Chazan, "The New Politics of Participation in Tropical Africa," Comparative Politics (January 1982): 169. Back.

Note 20: Robert Bates, "The Commercialization of Agriculture and the Rise of Rural Political Protest," in Bates, ed., Essays on the Political Economy of Rural Africa (Berkeley: University of California Press, 1983). Back.

Note 21: C. E. F. Beer, The Politics of Peasant Groups in Western Nigeria (Ibadan: Ibadan University Press, 1976). Back.

Note 22: Ibid., p. 164. Back.

Note 23: Ibid., p. 184. Back.

Note 24: The sequence draws on Watts's characterization but is not identical with his. Back.

Note 25: Robert Bates, "Contra Contractarianism: Some Reflections in the New Institutionalism," Politics and Society 16, 2/3 (1988): 387-401. Robert Bates has pointed out that the collective choice literature can offer an explanation of the demand for institutions such as share contracts but it offers no account of the supply. It affords no adequate explanation of the behavior of the political entrepreneurs who negotiate contract terms that approximate Pareto optimal or efficient solutions. The standard response that, given a special or selective incentive, a member of a community will sustain the costs of supplying an institution that is socially optimal or "efficient," or support the transaction costs associated with negotiation of the share contract, does not hold up under scrutiny. Offered a disproportionate share in the outcome, the entrepreneur is indeed likely to establish new contract forms, but these are likely to favor the entrepreneur's interests and result in socially suboptimal outcomes. Bates suggests that only when individuals are uncertain of one another's interests, intentions, and capabilities are they likely to choose terms that move the community toward the "efficient" set of relationships the economics approach envisions. Back.

Note 26: Economist Intelligence Unit, Quarterly Economic Review of Côte d'Ivoire. Second Quarter, 1988, p. 11. Back.

Note 27: Ibid., p. 8. Back.

Note 28: Ibid., Third Quarter, 1988, p. 10. Back.

Note 29: United States Department of Agriculture, IV0003, February 1, 1990. Back.

Note 30: This finding is in line with data collected by the World Bank Living Standards Measurement Study. See Angus Deaton and Dwayne Benjamin, "The Living Standards Survey and Price Policy Reform: A Study of Cocoa and Coffee Production in Côte d'Ivoire," LSMS Working Paper Number 44, Washington, D.C.: World Bank, 1988. Back.

Note 31: Provisional figure. Back.

Note 32: Although the economic logic in these statements seems incorrect upon first examination because Côte d'Ivoire is a price taker in every commodity market except cocoa, the farmers had accurately noted a correlation between government promotion and investment in certain crops and declines in the value of those crops. The example most often noted by the farmers was copra production, or cultivation of coconut palms, which the government had encouraged and in which it had invested extensively. Producer prices had subsequently collapsed, making it unprofitable to bother harvesting the trees. Many farmers were moved to chop down their palms and replace them with pineapples (for which prices have also started to diminish) or other crops. The farmers may be wrong in the way they phrase the problem, but their assessment of the government's ability to pick winners and losers and their general understanding of the relationship between supply and demand are valid. Back.

Note 33: The sales of the president's cocoa during the period when smallholder cocoa was being held off the market to force the world price to increase is documented in Jean-Louis Gombeaud, Corinne Moutout, and Stephen Smith, La Guerre du Cacao (Paris: Calman-Lévy, 1990). Back.

Note 34: The Aboisso region is part of the Akan "goldbelt." Alluvial gold is common, and among the Akan peoples gold jewelry has special significance and remains a tie to the past. Back.

Note 35: Foutou is a staple of the diet in the southern and central parts of the country. It comes in the form of a small, hot loaf made variously of pounded yams, cassava, or plantain-or some combination of these ingedients-and is used to scoop up bits of sauce, or stew. Back.

Note 36: The survey results indicate a sharp drop in the use of hired labor between November of 1989 and July-August, 1990. In this region, farmers typically hired labor on annual contracts, paying food, housing, and medical expenses as well as a flat salary. By purchasing labor in this way instead of relying on family labor, sending children to school and young men and women to the towns, the household could improve its long-term earnings and hedge against the risk that a family member might become ill during a critical work period, forcing abandonment of part of a crop. By 1990, returns and household monetary reserves were so low that farmers were unable to engage workers, typically Burkinabé migrants. Unable to locate sufficient work at remunerative wage levels, the migrant workers left the region. The households still able to hire labor complained that workers were nowhere to be found. Back.

Note 37: For example, see Respondent #61, Group 4, Village 3 and Respondent #93, Village 4, both October 1989. Back.

Note 38: See, for example, the remarks of Respondent #103, Group 6, Village 1, October 1989. Back.

Note 39: Hans Binswanger and John McIntire, "Behavioral and Material Determinants of Production Relations in Land Abundant Tropical Agriculture," Operational Policy Staff Discussion Paper, #17, June 1984. Back.

Note 40: Respondent, Group 1, Village 1, October 1989. Back.

Note 41: For example, see Respondent #99, Village 4, October 1989. Back.

Note 42: There are two factories and industrial plantations in the region, one at Toumanguié, and the other at Ehania. Back.

Note 43: It is difficult to estimate how extensive this practice was and is in this region. When asked to describe the "household," including relatives living in town and in other areas who had provided assistance in cash or in kind, almost all of those surveyed described relatively small family units only rarely dispersed outside of the cocoa zone or towns within that zone. Back.

Note 44: It is important to note that the short-run effect of the collapse of the cocoa and coffee markets was to flood the cities with cheap food, keeping food prices low at precisely the time the government imposed an austerity plan on urban dwellers. Back.

Note 45: See Terre et Progrés, Ministère d'Agriculture, no. 69 (January 1987) and no. 64 (July 1985). Producer prices for rubber were to be indexed to an international reference price, so that the farmer would get a floor price per kilogram, and an additional, variable premium. Back.

Note 46: See note 32. Back.

Note 47: Ministère de la Recherche Scientifique, "Le dynamisme foncier et l'économie de plantation," Abidjan: Publication Inter-Instituts de CIRES, IGT, Gerdat, ORSTOM, octobre 1978, p. 2. Back.

Note 48: This information corresponds exactly with data obtained by Jean-Philippe Colin in his study of the village of Djiminikoffikro in the 1980s. Colin reports a steady increase in land values, as follows:

PERIOD 1945-50 1950-55 1955-60 1960-65 1965-70 1970-75 1975-80

VALUE/HECTARE 1,284 5,236 8,093 22,572 24,860 41,374 70,000

# OF TRANSACTIONS (FCFA) 1 3 5 2 14 12 4

See Jean-Philippe Colin, "La Mutation d'une Economie de Plantation: Contribution à l'analyse de la dynamique des systèmes productifs agricoles villageois en basse Côte d'Ivoire." Montpellier: ORSTOM, 1987, p. 84. Back.

Note 49: Wim Vijverberg. "Nonagricultural Family Enterprises in Côte d'Ivoire." LSMS Working Paper No. 46, World Bank, 1988. Back.

Note 50: Respondent #59, Group 3, Village 2, October 1989. Back.

Note 51: The government had tried for several years to convince Ivoirians of just that. As part of its "back to the country" campaign, it produced a large quantity of buttons that read, "Je Suis Fier D'être Cultivateur!" Back.

Note 52: A puzzling anomaly appeared in the initial analysis of the survey data. A significant 24% of oil palm growers indicated that they received part of another's harvest in return for granting access to land, which, at face value, suggests that share contracts are practiced far more frequently than open-ended interviews suggested. This finding also flew in the face of oral histories offered by village elders about changing land tenure patterns. The leaders suggested that the abusa system had been replaced temporarily by a fifty-fifty split of the crop between the tender and the "owner" and then died out completely, with the introduction of oil palm cultivation. It is likely that the statistics reflect arrangements with palm-wine producers, who cut aging trees and give a portion of their liquor to the owner (if only to keep the owner quiet). Back.

Note 53: Interview with the notables of village #2, November 17, 1989 and interview with the notables of village #4, November 18, 1989. Back.

Note 54: Economist Intelligence Unit, Quarterly Economic Report, Third Quarter, 1989, p. 14. The RFI report was denounced by then Minister of Agriculture Denis Bra Kanon. Back.

Note 55: The government has tentatively authorized such arrangements after the collapse of parts of the cocoa marketing system. Because the traitants' government-determined allowances for marketing costs are widely believed to be below actual costs and because most of the cooperatives have loaned money to the members and cannot finance the trade, it is unlikely that this arrangement will work. Indeed, it may prove a further aggravation. Back.

Note 56: This description draws on interviews with village residents and with the "chef de zone" of SATMACI, who initially identified the problem as the major reason for the collapse of cooperative performance in some parts of his district. Back.

Note 57: Terre et Progrés, no. 69 (January 1987). Back.

Note 58: The farmers from the region maintain that they steadfastly objected to measures supported by representatives of the government and the International Labor Organization, which sent a consultant to help organize the unions. The officials offer a different account of the same events. Back.

Note 59: It is important to note, however, that unlike many Sub-Saharan governments, Houphouët-Boigny's government does not position armed soldiers at the doors to all ministries, and access is remarkably open. Back.

Note 60: See interview July 14, 1990, number 2. Back.

Note 61: The top three were, in order of preference: Rambo (the entire series), Bruce Lee's Kung-fu movies, and Dallas. The connection between the films' content and political behavior was remote, especially because the films were shown in their dubbed French versions while most of the viewers, save the school children, generally conversed in any one of a number of local languages. Back.

Note 62: See Robert Wade, Village Republics: Economic Conditions for Collective Action in South India (Cambridge: Cambridge University Press, 1988). Back.

Note 63: Terre et Progrés, no. 69 (January 1987). Back.

Note 64: See interview of July 14, 1990, number 1. Back.

Note 65: Ibid., number 3. Back.

Note 66: Ibid., number 1. Back.

Note 67: "Big Turnout For New Farmers' Union in Ivory Coast," Reuters News Service, July 29, 1991. Back.