Rural Interests and the Making of Modern African States

African Studies Program, University of Wisconsin-Madison Regents of the University of Wisconsin System Rural Interests and the Making of Modern African States Author(s): Catherine Boone Source: African Economic History, No. 23 (1995), pp. 1-36 Published by: African Studies Program, University of Wisconsin-Madison Stable URL: http://www.jstor.org/stable/3601724 Accessed: 08-07-2015 10:18 UTC REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/3601724?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. African Studies Program, University of Wisconsin-Madison and Regents of the University of Wisconsin System are collaborating with JSTOR to digitize, preserve and extend access to African Economic History. http://www.jstor.org This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURAL INTERESTS AND THE MAKING OF MODERN AFRICAN STATES Catherine Boone Department of Government University of Texas at Austin ostpoliticalanalysis hasassumedthatAfricanstates,beingartifacts of colonialism, have no organic links to indigenous societies. Goran den put the case most graphicallyby describing the state as "susin mid-air." One consequence of this is that in African studies, the pended issue of stateformationhas not been taken very seriously.Although many analysts have studied the aggregate growthof the postcolonial state apparatus, or have described similaritiesin the structure and processes of modern African governments, few have explored the social origins of cross-national differences in administrativepractice and in the organizationalconfigurationof state power. What explains differences in the autonomy, strength, and coherence of African states? What explains how political and economic prerogative is distributed throughout the political system?What explains differences in the intrusiveness and intensity of government at the local level? When contrasting colonial legacies do not suffice to explain contemporary variations, political analysts have for the most part been content to attribute observed differences to the ideological preferences of leaders. The problem is that voluntaristicand idiosyncraticexplanationsare not good enough. They do not tell us anything about the societal forces that facilitate or resist the congealing and legitimation of state power. This paper represents an attempt to take up this challenge.' It seeks to explain variation in the political and bureaucratic arrangements that linked state and countryside. In postcolonial Africa, these arrangements constituted the pillars of the state-building project; they were designed to provide the political infrastructure for executing the complex task of governing and taxing export-crop producers. Virtually all of Africa's postcolonial regimes relied upon rural political (oftentimes electoral) support and legitimacy claimed in the name of the rural masses. The irony was that the same agriculturalproducers were doomed to be heavily exploited by the state, for only the product of their labor could finance consumption and investment outside the rural areas. Herein lay one of the chief hazards of postcolonial rule: regimes depended upon the political support of those they exploited the most. Attempts to navigate these treacherous waters found expression in the political institutions and alliances that projected state power into rural Africa. Over time, it became clear that there was wide variation in the staying-power of these arrangements,in their capacity to genAfrican EconomicHistory 23 (1995): 1-36 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 2 CATHERINEBOONE eratelegitimacyforpostcolonialrulers,in theircontributionto "nationalintegration,"and in theirimpact on possibilitiesfor enhancingruralproduction and productivity.Giventhe stakes,it is importantto knowwhy the natureof the linksbetweenstateand countrysidedifferedso much acrosscases. The analysispresentedhereseeksto explaincross-regional differencesin key aspectsof the institutionaldesignof postcolonialstates.It triesto account forvariationin the degreeto whichstatepowerandadministrative prerogative wascentralizedandconcentrated in national-level institutions. Differencesalong these dimensionswere manifestin three areasof administrative design and in rural institutions that access structured to agriculpractice:(a) development turalcreditand purchasedinputs;(b) in stateagenciesthat organizedexportcropmarketing;and (c) in the regional-andlocal-leveladministrative agencies of government.Wherestatepowerwas verycentralized, littleprerogativeover the use of stateresources,andlittleleewayin the interpretation andimplementation of policy,was surrenderedto rural"brokers"or "intermediaries" who drewpoliticallegitimacyfromsourcesof powerandstatusthatlaybeyondthe reachof the state.Wherestatepowerwasveryconcentrated, linksin the administrativechainthat connectedcore and peripherywere few;state agentsgovernedfromthe center,ratherthanfromthe localities;and the presenceof the statein the localitieswas minimal,ratherthan omnipresentand intrusive. The studyfocuseson a set of WestAfricancasesthatprovidestrongcontrasts:Senegal'sgroundnutbasin,the coffee-and cocoa-producingforestbelt of C6te d'Ivoire,and Ghana'scocoa-producingsouth. In all cases, the time period under reviewis the "decolonizationperiod,"definedfor Senegaland C6te d'Ivoireas the early 1950s throughthe mid-1970s, and for Ghana as 1950-1966 (the Nkrumahyears). To accountforvariationin the formand characterof the state-countryside linkage,I do not relyon argumentsabout"stateautonomy"or the purely ideologicaldeterminantsof state design.As a result,in explainingcross-case differencesI do not emphasizecolonialadministrative ideologiesor the contrastbetweenFrenchandBritishadministrative practice.My argumentstresses the extentto whichvariationis explainedby differencesin ruralsocial structure. In ruralAfrica,socialformationsand modes of productiondifferwidely acrossregions.As PeterGeschierehas argued,differentpre-capitalistsocieties offereddifferent"footholds"or "accesspoints"for colonialand postcolonial regimesseekingto extend statepower.2Variationin agrariansocial structure also producedvariationin the capacityof farmingcommunitiesto compete with colonialand postcolonialstatesfor the wealthproducedthroughexport crop production.If Africanstate formationis the story of regimes'effortsto forge the institutionalmechanismsand politicalalliancesto tax and govern export-crop producers, then a great deal of variation in strategy and institu- This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 3 tional outcome,I argue,reflectedthe widelydifferentsituationsthat regimes confrontedon the ground. This argumentcontradictsthe image of Africanstates as "suspended above"ruralsocieties.It suggeststhatstateauthorityand hegemonyhavebeen constitutedin partthrough"fusions"of statepowerwith the societally-based forms of powerthat lie beyondthe directreachof the state,and that are embedded in peasantmodes of production.More generally,this approachemphasizes what is universal ("banal,"as Bayartwould say) in the African experienceof stateformation.In spiteof "imported"originsof the administrative and coerciveapparatus,statepowerin Africais gatheredmuch as it has been elsewherein theworld:thatis, by anchoringit in broaderandmoredeeply rootedhierarchiesof socialdomination. Part I comparesinstitutionaldesign and administrative practicein the three cases studiedhere. It lays out the differencesto be explained.Part II identifiesfeaturesof ruralsocialstructurethatgo farin explainingthesedifferences.This partcomparespeasantsocietiesin Senegal'sgroundnutbasin,the Ivoirianforest,andsouthernGhana.Thediscussionfocuseson two featuresof ruralsocialstructurethat areimportantin differentiating acrosscases,and in explainingvariationin postcolonialinstitutionaloutcomes.PartIII pullsthese elementstogetherin a discussionof the politicsof postcolonialstateformation in the three cases.The finalparagraphsof the paperlay out two hypotheses about the relationshipbetweenruralsocialstructureand the institutionaldesign of modernAfricanstates. I Institutional Design and Administrative Practice: Contrasting Cases This analysisattemptsto accountfor the differencesin administrative design and practicethat emergefrom a comparisonof Senegal'sgroundnut basin,the Ivoiriancoffee-andcocoa-producingzone,and Ghana'scocoa-producingsouth.At the highestlevelof abstraction,the institutionaland political linksbetweenstateand countrysideseem to varyalongtwo dimensions:there are importantdifferencesin the extentto whichstatepoweris centralized, and in the extentto whichit is concentrated. These terms are definedin a way that is relatedto, yet distinctfrom, definitionsfound in much of the currentpublicpolicyliterature.I am using centralizationand concentrationas spatialmetaphorsto describethe defacto distributionof politicalprerogativeand the physicaldistribution/location of state agencieswithinthe nationalpoliticalspace.Thus, I am not using these terms to describe the formal, legal relations between various sub-units of the state, or to describe the formal locus of policy-making authority within the This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 4 CATHERINEBOONE political system. In the analysis here, the terms "decentralization"and "deconcentration" do not connoteanythingaboutaccountability, representativeness,popularparticipation,or democracywithinthe politicalsystem.On the contrary,the purposeof this paperis to describevariationin institutional arrangementsdesignedto subordinateexport-cropproducersto the state. "Centralizationof state power"is a variablethat gaugesthe extent to whichstateagentsgovernthe countrysideby concedingpowerandprerogative to membersof ruralsocietywhose positionand statusderive,in part at least, fromsourcesof powerthatlie beyondthe reachof the state.This variabletaps into the relationshipbetween state officialsand indigenousruralauthorities (such as chiefs,big farmers,ruralnotables,marabouts, etc.) in the localities. whichdevolvepoweroverpersons,prerogative in implementing Arrangements policy,and defactocontroloverthe use of stateresourcesto grassrootsrural leadersworkto "decentralize" the politicalsystem.By this criterion,colonial indirectrule was a form of power-sharingwith indigenousruralauthorities that"decentralized" statepower:local authoritieswhose claimto powerlay in some readingof pre-colonialtradition(andin ruralproductiverelations)exercised wide discretionand autonomyin wieldingcoerciveand taxationpowers delegatedto them by the colonialstate. In more contemporaryterms, state when powerfulsocietalelepowerbecomes more and more "decentralized" ments areallowedto "capture"sub-unitsof the stateand use them with considerableautonomy,therebydiminishingadministrativecoherenceand the authorityof those at the center. "Concentrationof statepower"is a variablethatdescribesthe extentto whichthe regimegovernsthe localitiesby relyingon outpostsof the stateand rulingpartythat are located at the villagelevel.Where state presenceat the local level (andthereforestateinterventionin whatmightappearto be purely local affairs)is minimal,the state apparatuswouldbe describedas "concentrated"at the center.Under these arrangements, therearefew points of contact betweenruralpopulationsand those wieldingstatepowerand allocating state resources.By contrast,the state apparatusis "deconcentrated" where state institutionsare strikinglypresentin localities(as in villagecells of the rulingparty,or state-runproducercooperatives)andwheremultiplestateagencies interveneon the ground,in manydifferentways,in localpoliticaleconomies.Localitiesaregovernedintensively.What havebeen characterized as more or "statist"regimesare activein localitiesin aggressively"developmentalist" these ways.3 a. C6te d'Ivoire In these terms,the institutionaland administrative presenceof the state in the Ivoirianforest zone would be described as centralized and concentrated. Political prerogativewas exercised from the center. In the southern half of the This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 5 country, state agents stationed in the regions and localities enjoyed little autonomy. (The locus of authority was "centralized.")And the regime achieved its political and economic goals without creating an elaborate network of state institutions in the forest belt. (The state apparatuswas "concentrated" at the center.)When contrastedto the other cases examined here, the Ivoirianregime's strategies of economic development and governance in the cocoa and coffee zone look relativelyhands-off. This is what gave Ivoirian economic and political strategies their more "liberal"cast.These featuresof state design were manifest in official strategies to promote coffee and cocoa production, in state regulation of the export-crop marketing circuit, and in provincial administration and the political encadrement(training, organizing) of the peasantry. In promoting smallholder coffee and cocoa production, the regime of Houphouet-Boigny avoided the aggressivelyinterventionistdevelopment strategies pursued by many of its neighbors. Policies affecting access to factors of production (land, credit,purchased inputs) were relativelylaissez-faire.In places like Senegal and Tanzania, by contrast, statist and socialist regimes in the early postcolonial years attempted land reform, engineered settlement schemes, required farmers to use fertilizers,pressured them to cultivate new plant varieties, obliged farmers to change production techniques and sent out scores of extension agents to monitor their progress, and pressured farmers to subscribe loans through the intermediation of rural cooperatives. Direct state controls over access to, and use of factors of production in much of sub-SaharanAfrica led to "the intensive infusion of bureaucraticregulation"into rural society.4 Ivoirian coffee and cocoa farmers did not find themselves in this position. They did not experience heavy-handed state intervention in the production process itself. Producer prices were set by the state, input prices were subsidized, and extension agents passed out cuttings to start new coffee and cocoa trees. Yet Ivoirian government officials did not attempt to control the distribution of purchased inputs or credit, distribute or redistribute land, or disseminate new technologies. The Ivoirian government eschewed direct control over factors of coffee and cocoa production. These patterns of state penetration of the countryside were reflected in the regime's strategy for extracting rural surpluses. Its central feature was private control, rather than direct state control, over the purchase of coffee and cocoa. The government regulated buying prices and financed the export circuit, but farmers were not required to sell export crops directly to state officials.This meant that in contrast to the export-crop producing regions of much of West Africa, in southern C6te d'Ivoire there was no compulsory network of village-level marketing institutions (cooperatives).5Without compulsory cooperatives,forest zone communities lacked what are, in other contexts, important points of access to state resources and key sites for accumulating power at the community level. In the coffee- and cocoa-producing zone, the regime of This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 6 CATHERINEBOONE Houphouet extracted rural surpluses through the atomized and seemingly apolitical processes of the (regulated) market. If the Ivoirian regime achieved its economic goals without creating an elaborate network of state institutions in the forest belt, then the same holds true for the regime's political goals. In the south, the regime sought to secure civil order, acquiescence to the Houphoudt regime, and acquiescence to the heavy tax burden imposed upon export-crop producers. These political ends were not pursued via an activist strategy of building state institutions to penetrate local communities and link them to the state. Houphoudt's regime never tried to "mobilize the rural masses," penetrate local-level authority structures, or co-opt local-level leaders in the governing apparatusof the state.6No partybureaucratic political machine distributed patronage to buy votes in the countryside,and Houphouet had no use for ideologicalrationalesfor"transformative" grassroots political mobilization. This made the Ivoirian strategy of rural governance quite different from those pursued by its socialist and populist neighbors like Nkrumahist Ghana and Burkina Faso, and also quite different from strategies adopted in other "moderate" African countries that continued to elect parliaments and presidents, such as Senegal, Kenya, and Zambia. In these countries, political machines mobilized local-level political support for regimes. In exchange, material rewards "trickled down" through these same bureaucratic channels. Bureaucratic political machines linked villagers (farmers) to regimes either indirectly (via the intermediation of local notables or strongmen) or directly (thereby bypassing local notables). Meanwhile, regimes tried to harness collective rural consciousness to their cause, and workedto popularizeanimationrurale,ujamaa, socialist, or harambee ideologies. The Ivoirian south was striking in its absenceof ruling-partypoliticking, organization-building, and official consciousness raising.Whathad emerged in postcolonial C6te d'Ivoire was a pattern of postcolonial state formation that contrasted with the norm for much of sub-Saharan Africa in the 1960s and 1970s. This governing strategy had an institutional correlate:there were few official sites, positions, or organizations on the local level that offered direct access to state wealth or power. Local interests had few sites to colonize or capture; provincial political entrepreneurs had little cause to mobilize rural populations to further their ambitions; and political machines were not built to link local networks to national hierarchies.In the rural south, Goran Hyden's description of the African state is apt: the state appeared to be "suspended in mid-air."7 The appearance of "ungroundedness" was related to the extreme centralization of control over the state and parastatalinstitutions that structured Ivoirian political space. It was also related to the concentration of state administrative capacity at the higher levels of the bureaucratic hierarchy, and the consequent "thinness" of state presence in the localities. The unusually high This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 7 degree of administrative capacity that was observed in the Ivoirian provinces was no doubt related to this remarkablycentralized structure of control.8 b. Ghana The regime of Kwame Nkrumah sought to establish centralized control over a state apparatus that reached deep into localities, governing the cocoa belt intensively through a dense network of official institutions that insinuated state power into the micro-level dynamics of local political economies. The striking contrast with the CGted'Ivoire is the statismof the Nkrumahist model as it was elaborated from 1951 onward.9 In the terminology laid out above, institutional design and administrativepractice in Nkrumah's Ghana would be described as "centralized"(that is, driven by attempts to enhance the degree of top-down control at the expense of the autonomy of sub-units of the state) and "deconcentrated" (characterizedby a far-reachingstate presence in the localities, and by dense networks of state institutions at the local level). This strategy of state-building was manifest in the domain of rural development, in official control of the cocoa marketing circuit, and in administrativestructures at the regional and local levels. Although farmers produced the same crop - cocoa - on both sides of the Ghanaian-Ivoirianborder, the role of the state in shaping social relations of production and access to productive resources was very different in the two cases. In the name of promoting cocoa production, the government led by Nkrumah after 1951 undertook vast campaigns to distribute credit to smallholders.Through the loan campaigns, the regime sought not only to displace private sources of rural credit, but also to alter social relations of cocoa production by allowing indebted farmers to redeem cocoa farms that had been pledged to creditors. Once the state established a monopoly over the purchase of cocoa in 1959, producers were organized into compulsory selling cooperatives.These same institutions eventually exercised extensive control over farmers' access to credit, cutlasses, and fungicides in the cocoa belt.The state reached even deeper into the micro-level political economy of the cocoa areas after 1960, when parliament promulgated laws regulating land rents and created national-level state agencies to supervise the use and allocation of stool lands (i.e., lands held by paramount chiefs in the name of the Akan nations) and all revenues accruing from them. The dense network of village-levelproducer cooperatives was the base of the sprawlingbureaucraticapparatusthrough which the Ghanaian government exerciseddirect,monopoly control overthe buying of cocoa. Unlike theirIvoirian counterparts who sold cocoa and coffee to private traders, Ghanaian cocoa growers were obliged after 1959 to sell their crops directly to state agents who manned buying stations in the village cooperatives.Purchasingagents and other cooperative officials owed their positions of power to the ruling party, which This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 8 CATHERINEBOONE appointed and removed them at will.The government thus attempted to maintain centralized control over key institutions within the "deconcentrated" (farflung) state apparatus. Broader strategies and mechanisms for governing the rural areas were similarly"centralized"and "deconcentrated."In the rural areasNkrumah, like Houphouet, sought to secure order, deference to the regime, and acquiescence to taxation. The difference was that the Ghanaian leadership attempted to do this via an activist strategy of institution-building that was designed to penetratelocal communities, micro-managekey politicaland social processes within towns and villages, and link localities directly to the state. Through successive reforms of the institutions of rural government inherited from colonial rule, the regime shifted local balances of power and strengthened its presence and influence at the grassroots.Local councils, which exercised considerable clout in the Ghanaian administrative system, were brought under party control and empowered at the expense of preexisting organs of rural government. Workingtoward the same end, the regime invested tremendous effort in manipulating and coopting chiefs and other provincial notables, thus marginalizing anti-CPP elements while drawing cooperative indigenous authorities into the embrace of the state. As state power was centralized, the position and prerogatives of these individuals became even more dependent upon approvalfrom above and the handing-out of state resources. State power was deconcentrated as the regime multiplied and expanded its "outposts" in the localities. The presence of multiple, partly overlapping party and state institutions imposed a tight web of state power and influence on up-country localities. Nkrumah worked intensively to mobilize the mass of the cocoa-growing population behind the ruling party.The CPP's vote-gathering machine was a hierarchical structure that rested upon village "cells," where state and party agents could use all the institutional resources at their disposal to mobilize the votes of local notables and farmers. In rural districts, these dense networks of local government agencies, producer cooperatives, state marketing institutions, and organs of the ruling party represented multiple and diverse points of access to state power and resources. To win support for the government, CPP agents could manipulate access to local commercial opportunities, to salaried jobs in the village cooperatives,local councils and party branches, and to agriculturalinputs and credit distributed through the cooperatives.'0They could even manipulate the terms on which an individual farmer sold his/her cocoa to the state." In the cocoa belt, the intense electioneering driven by the super-charged engine of the CPP patronage machine contrasted sharply with the "stifling of local political life" on the other side of the Ghana-C6te d'Ivoire border. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 9 c. Senegal My argumentis thatstate-buildingin Senegalunderthe rule of LUopold SederSenghorproduceda stateapparatusthatwasfar-flunganddeeplyrooted in localities(i.e., deconcentrated),in whichcontroloverstatepowerand predecentralizedas the defactoautonomyof sub-units rogativewas systematically of the state increased.12As in GhanaunderNkrumah,the reachof the state extendeddeep into the localitiesin Senegal'sgroundnutbasin. State power to buildup andmanipulateexisting and resourceswereemployedaggressively politicalnetworkswithincommunities. What made patternsof governancein centralSenegalso differentfrom those found in the Ghanaiancocoa belt was the Senghorregime'ssystematic effortsto coopt marabouts and otherlocal notablesinto party/administrative structures.Senegaldifferedfrom Ghanain the relativelygreatdegreeof latitude in implementingpolicyand in distributingstateresourcesthatwas exercisedby the alreadyestablished,indigenousauthoritiesof the groundnutbasin. Nkrumahsought to parcellize,to narrow,and to partiallysecularizethe auchiefs.By contrast,Senghornot thorityexercisedby Ghana's(neo)traditional increased the secular of the only politicalprerogativesof the grandsmarabouts he but also underwrote efforts to maintain the of sacredness groundnutbasin, theirauthority.These of the were state-formation qualities postcolonial process evident in the institutionsand administrativepracticesthroughwhich the Senghorregimesoughtto promoteexport-cropproduction,controlthe commercialcircuit,and governthe groundnutbasin. In Senegal,the postcolonialstate mademembershipin producercooperativescompulsoryfor groundnutfarmers.Throughthese institutions,the state intervenedalmostdirectlyin the productionprocess,and thus playeda virtuallydirectrole in shapingsocialrelationsof production.Producercooperativesin Senegalexercisedeven more controloverthe distributionof key factorsofproduction(credit,seeds,fertilizers) thantheydidin post-1959Ghana. Anotherstate-sanctionedlocal institution,the ruralcouncil,governedaccess to the most basicfactorof production- land.The regimeinterveneddirectly in theagricultural labormarketby actingto curtailthein-flowof migrantgroundnut farmers(navetanes) in 1963, therebyaffectingthe supply,cost, and conditions of accessto rurallabor.Thereis a clearcontrastherewiththe hands-off, role of the Ivoirianstatewhenit cameto structuringaccessto faclaissez-faire tors of coffee and cocoa production.The contrastwith the C6te d'Ivoireextends into the domainof exportcrop marketing.Senegal'sgovernment(like the Ghanaiangovernmentafter 1959), imposed a state monopoly over the exportcircuit. From the perspectiveof thisstudy,whatis criticalhereis thatin Senegal (1) state institutions became the prime arenas and sites of access to key factors of groundnut production, and (2) these institutions tended to fall under the de This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 10 CATHERINEBOONE facto control of already-influentialrural authorities,whose status and authority derived in part from sources that lay outside the state. Senegal and Ghana are similar on the first count. It is the secondfact that differentiatesthis case from that of Ghana. The specificity of the Senegal case was manifest in the producer cooperatives. Rural cooperatives were the conduit of the two-way flow of resources between the state and producers:through them peasants sold their crops to the state, and via the same channel the government distributed credit, seeds, and tools to small farmers. In the groundnut basin, cooperatives tended to fall under the control of already-establishedreligious authorities.They distributed inputs on a patronage basis to their own disciples and clients, and thus used state power and resources to enhance their own status and to expand their personal followings.13Local authoritieswere also able to systematicallysubvert the official intent of state policy by using cooperative resources to enrich themselves, and develop their own landholdings.That they did so with impunity is evidence of defacto autonomy in the use of state resources and prerogative. The argument continues to hold as the focus is broadened to consider larger institutions and more general mechanisms of provincial administration in Senegal's groundnut basin. The Senghor regime relied upon the hierarchically-structured institutions of Senegal's major Islamic confririesas state-sanctioned forces of political order in the groundnut basin. The confrdriesenjoyed the political and spiritual devotion of much of the groundnut-producing peasantry.Through their land pioneeringand "settlementschemes,"they also played a major role in structuring the geographic spread of groundnut production. It would not be a gross exaggeration to say that in the 1960s and 1970s, secular state authority was projected throughthe confririesinto the localities of the groundnut basin, rather than by-passing or competing with them. Observers believed that grandsmaraboutswere able to influence Senghor's appointments to posts in the formal hierarchy of state institutions (the sous-prefects, prifects, and regional governors who served under the authority of the Minister of the Interior) in the groundnut basin.14 So tight was the alliance of secular and religious political leaders, and so closely intertwined were state and maraboutic power, that academic discussion in the 1960s and early 1970s dwelt on the question of which side held the upper hand.'5 Of the three cases considered in this study, the groundnut basin came closest to the colonial ideal of indirect rule. To an important extent, de facto political authority was decentralized,and state power was wielded with considerable autonomy by (neo)traditional authorities. Under Senghor, a great deal of discretion in policy implementation, and real prerogative over the use of state authority and resources, was devolved to local powerbrokers. Senghor's state power by building new party and state institutions regime deconcentrated on the local level, thereby multiplying the number and diversity of points of access to state power and resources. Deconcentration tended to enhance the This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions ANDMODERNSTATES RURALINTERESTS 11 sincetheywereoftenableto use their powerof indigenousruralauthorities, state of control localinfluenceto "capture" agenciesatthelocallevel.Infactit seemedthatthe politicalauthoritiesin Dakarsanctionedsuch a process,forby enhancingthe influenceandpatronagepowersof theirruralallies,theymaintainedan indirectbut effectiveform of controloverthe groundnut-producing called peasantry.This was most obviousat electiontime, when the marabouts to voteforSenghorandtheUPS. upontheirfollowers andcontrastsare the case,thesecomparisons At theriskof overstating representedin a two-by-twotable: Locus of authority Centralized Sites of access to the state Coted'Ivoire concentrated at center deconcentrated Ghana Decentralized Senegal I Relations between State and Countryside: Explaining differences What explainsthese differences?Most of the literaturerefersto the politicalideologiesof postcolonialregimesin explainingvariationsin patternsof ruralgovernance.In these terms,Ghanaand Senegalwouldbe describedas interventioniststatesguidedby socialistideologies.Interventransformative, tion in ruralmarketingcircuitswouldbe attributedto the anti-capitalistelementsof thissocialistideology.C6te d'Ivoire,by contrast,embracedcapitalism and laissez-faire policiesbecause of its economic and politicalconservatism. There arethreeproblemswiththe ideologicalexplanation.First,it is voluntaristin positingthatleadersarefreeto remakestatesandsocietiesin accordance with theirprivatevisions,in the absenceof constraintsimposedby historyor circumstance.In thissense,the ideologicalexplanationis not so much"wrong" The second problemwith the ideologicalexplanaas it is underdetermining. tion, then, is that it simplybegs the questionof why certainregimesadopted particularideologicalrationalesto justifytheir strategies.The thirdproblem followsfromthis,and is obviousin the three-waycontrastdrawnabove:ideology is too blunt an instrument to explain actual patterns of institutional and This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 12 CATHERINEBOONE administrative design; it does not, for example, account for the very striking differences that emerge in the Ghana-Senegal comparison. An alternative explanation for differences in the design and operation of postcolonial states is found in work that points to the institutional legacies of European colonialism. The most popular version of this argument rests on a distinction between British and French rule. It is said that colonial governments in French colonies, like the state in metropolitan France, were highly centralized institutions which governed the countryside through forms of"direct rule" that conceded very little autonomy to provincial agents of the colonial state, whether French or African. The philosophy and strategy of British rule in Africa was very different:the British established forms of "indirectrule" wherein indigenous African authorities would preserve considerable authority and autonomy, and indigenous forms of government would be (for the most part) preserved and upheld.Thus we would expect the institutional legacies of colonialism to produce more deconcentrated and decentralized forms of rule in postcolonial Anglophone Africa, and more centralized and concentrated administrative structures in Francophone Africa. The institutional and political legacies of colonial rule may be just as important as its economic legacies, but the "colonial transplant"argument is not satisfactoryin its most popular form. Obviously it cannot account for variation among (or within, for that matter) territoriescolonized by the same European power, as the contrast between Senegal and C6te d'Ivoire suggests. More generally, much of the rethinking of colonial administrativepractice since the 1960s has shown that expediency drove both Britain and France to rely on improvised versions of "indirect rule" wherever they could. What determined the "directness" of rule was less pre-conceived administrative doctrine than simply the success of the European powers in finding cooperativeAfrican leaders and authority figures (intermediaries and interlocateursvalables) through whom they could effectively govern the regions and localities. This suggests that much of the variation in colonial administrativepractice, and in the structure of"imported" state apparatuses,actually reflected the socio-political realities that the European colonizersconfrontedon the ground. Institutionallegacies of colonial rule are important in explaining postcolonial state structures, but the key point is that (as Bruce Berman has argued) colonial states were themselves shaped by the societies they sought to govern.'6 This means that differencesin agrariansocial structurewould be reflected in the design of both colonial and postcolonial states. The task, then, is to develop this insight in a systematic way, and to use it to decipher configurations of power that crystallized in the postcolonial state. In comparisons of ruralsocial formations in Africa,it is standardpractice to differentiatecases in terms of the extent of commercialization of agriculture and the degree of rural social differentiationthat has resulted from commodity production, and the main export crop that is produced." These variables are This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTS ANDMODERNSTATES 13 commercialized interrelated: forexample,extensively retree-crop producing gions (suchas southernC6te d'Ivoire,southernGhana,southwesternNigeria, centralKenya,southernUganda)tend to show more markedruralsocialdifferentiationalongclasslines.Fromthe perspectiveof the presentstudy,whatis clear is that the unidimensionalfocus on the criticalquestionof ruralclass formation has obscured other aspectsof rural social structurethat prove to be importantin determiningthe governabilityof particularpeasantsocietiesand theircapacityto resistexploitationat the handsof the state. Thus, movingbeyondthe structuralist descriptionsof ruralAfricansocietiesthatappearin mostof thepoliticaleconomyliterature, I defineruralsocioeconomic structurenot only in termsof economicstratificationor class,but alsoin termsof the actualorganizationof production(regionally-specific social relationsof production),the characterof indigenousauthoritystructures,and the natureof groupsolidaritiesand identities.As the workof PeterGershiere demonstratesexplicitly,thereis no need to juxtaposean explanationthat focuses on ruralclassstructureto one thatlooks at these otherfeaturesof rural society,for in a given ruralsocial formationthe actualextent of, and future possibilitiesfor, privateaccumulationand class formationare relatedto the othervariables.'8 My argumenthereis thatthe answersto two questionsaboutruralsocial structureas it existedin the terminalcolonialperiodgo farin accountingfor the variationsin postcolonialstate structurethat are identifiedabove.First, how hierarchical wereindigenoussocieties?Andsecond,wherehierarchies were strong,didindigenouseliteshavethepowerto contestthe colonialstate'sclaims on the ruralsurplus?"' a. C6te d'Ivoire In the context of the three cases studiedhere, peasantsocietiesin the forest zone of CGted'Ivoirestand out for their relativeabsence of political hierarchy,for the diffuseness(broaddispersion)of controlovermaterialand socialresources,andforthe extremeweaknessof the indigenouspoliticalunits that existedundercolonialrule. Peasantsocietiesof the Ivoiriansouthemergedfroma historicalprocess thatworkedagainstthe concentrationof poweroverpeopleandland.Withthe forest exceptionof the Agnikingdomsof the southeast,the sparsely-populated zone of the precolonialerawas not the site of Africanstates.20Thediffuseness of indigenousauthoritywasreflectedin the patternof colonialconquest.In the south-centralregion (Baoulecountry)and the southwest(of the Dida, B&te, and Gouro),the Frenchwereforcedto "pacify"Africanpopulationsby carrying their brutal war from village to village, and hamlet to hamlet. From about 1909 to 1915, the colonizers carried out a widespread program of regrouping and resettlement which gathered populations into new villages along roads cut This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 14 CATHERINEBOONE through the forest. Once control was established, the administrativestrategy of "direct rule" was practiced in near-ideal form throughout much of the south. Finding no compliant chiefs (or no chiefs at all) in many of the administrative districts they delineated, the French appointed loyal Africans - often soldiers, junior clerks, cooks, or interpretersin their employ - as their agents ("chiefs") at the grassroots. "[O]ften [they were] not even members of the ethnic group which they were appointed to control."21 Official chiefs worked as tax collectors and rounded up forced laborers for the colonial authorities. Needless to say, few enjoyed much legitimacy in the eyes of their subjects. Most of the Ivoirian south was thus shaped by a history of precolonial political decentralization, a process of colonial conquest that was violent and extraordinarilydestructive of establishedsocial orders,and "directrule"through agents appointed by the colonial state.22 Peasantization in the forest zone reproduced and accentuated the localized and dispersed structure of authority that tended to characterize indigenous societies in this particularregion. Cultivation of commercial tree crops began in earnest in the southeastern corner of the colony in the late 1920s. Acreages under coffee and cocoa grew rapidly over the course of the next sixty years.As in most of sub-SaharanAfrica, the peasantizationof local populations occurred through the adaptation of precolonial systems of land and labor use. In the Ivoirian forest, existing systems of land use allowed farmers to cultivate whatever virgin forest land they could clear, subject to the permission of lineage elders or village land chiefs. The limiting factor of production throughout the sparsely-populated Ivoirian south was labor. This problem was solved in a way that produced exponential increases in the amount of land under coffee and cocoa. In order to expand their holdings beyond the limit of labor-power available within the household, the Agni of the southeast, who began planting cocoa in the 1920s, relied on migrant labor from poorer parts of the colony. Some short-term laborers workedfor wages, but abusansharecontractswere more common. Under share contracts, migrants farmed land that had been cleared and planted by the household head (sometimes with the help of wage labor).The owner of the trees kept the proceeds of one-half to two-thirds of the crop; what remained was the abusan's. The conversion of wage laborers into abusanwas not uncommon, and in the southeast, many abusanwho proved to be respectable members of the community eventually received land to clear and farm on their own account.Thus, in the oldest cocoa-producing regions of the C6te d'Ivoire, land pioneering and labor influxes fueled the extensive form of coffee and cocoa cultivation that became the defining characteristic of the Ivoirian iconomiede plantation. By 1953, "out of a settled population of 32,000 in the central subdivision of Aboisso cercle,...one-third were strangers.In addition, there were about 20,000 temporary immigrants working on local farms, many of whom settled in the region."23 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 15 This systempushedthe cocoaandcoffeefrontierwestward,creatingethnically-heterogeneous villagesacrossthe southas hardwoodforestfell to make way for smallholderfarming.Enterprisingpioneersfromthe old cocoa zones moved west to obtainland suitablefor colonization;locals in the centerand west began to planttree cropsto earncash;and migrantlaborersworkedfor smallholdersand wereableto clearlandforthemselves.24 Although arrangementsgoverningaccess to land and labor variedby region and locality,low populationdensitiesthroughoutthe south (and vast expansesof virginforest)meantthat land was not scarcein the economistic fromother sense.Localauthoritiesgrantedusufructrightsfreelyto "strangers" permitpartsof the colonyor beyond.Non-kinlaborsuppliedby "strangers" ted householdsto expand.When the Dida of the center-westbegan to plant coffeein the mid-1940s,for example,migrantlaborerswereincorporatedinto As coffeeproductionbecamewell-estabhouseholdsas "adoptedrelatives."25 lishedin the region,the kindsof indigine-stranger relationsthathad emerged were the in the 1920s and 1930s generalizedacrossthe forest among Agni zone.26Migrantsto the centraland westernforestregionswereauthorizedto clearforestlandfortheirownuse upon arrival,in exchangeforprovidinglabor servicesand/orannualpayments(gifts)to community-levelauthoritiesor lineage headsretaining"moralauthority"overthe land.27 deplantationtended to compromise Expansionof the Ivoirian6conomie the already-limited of authorities who drewtheirsocialand powers indigenous institutionalpowersfromruralcommunities,ratherthanfromthe state.Land pioneeringand influxesof migrantlaborerscreatedvillagesand localitiesthat were not unifiedby common ancestry,myths of origin,or spiritualleaders. Meanwhile,the riseof smallholdercommodityproductionand the plantingof tree cropstendedto breakdown lineages,individualizecontroloverland,and to guaranteea farmer'slanduse rightsforlongperiodsof time.Thiserodedthe prerogativesand socialcontrolsof the community-levelandlineage-basednotableswho exercisedmoralauthorityoverthe land.28 I havedescribedanatomized,sociallyfragmented, andpolitically"acephaWhataboutthe much-discussed"planterbourgeoisie"of the lous"peasantry. Ivoirianforestzone?Fromthe perspectiveof thisanalysis,therearetwo critical aspectsof the Ivoirianplanterbourgeoisie.Thefirstis thatthe accumulationof wealthin the handsof some 50,000 "bigplanters"wasnotaccompaniedby the abilityof the same groupto accumulatepolitical,social,and economiccontrols over the mass of peasantproducers.Big plantersenjoyedrespect and influence,along with the clientelethat statusand money can buy.This was reflectedin the success of theirprominentspokesmenin mobilizingsupport behindthe SyndicatAgricoleAfricain(SAA)and,later,the PartiDemocratique de C6te d'Ivoire(PDCI) in the centerand west. However,no mechanismsof social, economic, and political control over communities and localities were intrinsic to their positions and status as rich planters.29In the forest zone, nei- This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 16 CATHERINE BOONE ther the structureof controloveraccessto productiveresources(land,labor, bases of groupidentities,served credit),nor the ideological-spiritual-material to underwriteandsustainthe politicalpowerof "theIvoirianplanterbourgeoisie"as a classor stratum. In the forest,politicalpowerovercommunitieswas not inherentin the socialrelationsof production.As we haveseen,it was possibleto acquirecontrol overlandwithoutimposingcontroloverpeople.Likewise,the taskof mobilizinglaborwas not accompaniedby the establishmentof durableformsof politicalcontroloverlocalpopulations.Alliesof the Frenchcolonialadministrationwereprobablyableto requisitionforcedlaborersto worktheirland,but this powerevaporatedwhen forcedlaborwas abolishedin 1946. Meanwhile, thanksto influxesof migrantlabor(especiallyafter1946, the period of most rapidexpansionof the plantationeconomy),it was possibleto acquirelabor withoutpoliticallysubordinatinglocalpopulations. Grassroots"politicalculture"or ideologyin the forestdid littleto bolster the politicalstatusor influenceof successfulplanters.Especiallyin the center and west, few had robustclaimsto spiritual-historical legitimacyas leadersof localpopulations.Manycouldprobablyclaimsome chieflystatus,but forreasons mentionedabove,in most of C6te d'Ivoirethis was not sufficientto win the heartsand mindsof the peasantry. Partialexceptionsto this generalizationcan probablybe found in the where the most centralizedprecolonialpolitiesof the Ivoirianforestzone east, werefound. Cocoaproductionandpeasantization werestrongand rapidafter 1920 in and aroundthe Agni/Akankingdomscenteredin Abengourouand Sanwi.Althoughthe historyof Agninationalistpoliticsremainsto be written, it is reasonableto expect that big, prosperouscocoa farmers(includingnotablesof chiefly/royal status)playeda leadingrolein mobilizingvotes,money, and Agni nationalismbehind the PartiProgressiste(PP), which represented the PDCI's eastern,Agni-basedoppositionin the late 1940s.The case of Agni oppositionto the PDCI alsohelpsto provethe generalrule,however.Once the colonialadministration pulledthe plug on the PP,its electoralpowerdrained away.In the 1950s and 1960s,theAgnipowerelitewas easilymanipulatedand undercutby Houphouet.Unlike theirAshanticounterparts,membersof the Agni political-economicelitewerenot ableto mobilizeregionally-based peasant supportfor theirpoliticalcause.Even in the east,then, it seems that the socialrelationsof production,thestructureof controloverproductiveresources, and historical-ideological ties did not providethe Ivoirianplanterbourgeoisie with powerfulmechanismsof controloverlocalpopulations. Inherentin colonialtradeweremicro-levelmechanismsof political-economic controloverpeasanthouseholds.The most obviousof thesewas credit. This source of localized political influence also eluded big planters in southern C6te d'Ivoire. In striking contrast to the situation that prevailed in Ghana, in southern C6te d'Ivoire big farmers did not dominate the local-level commer- This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 17 cializationof exportcrops.Unlike their Ghanaiancounterparts,most of the leadingmembersof the Ivoirian"planterbourgeoisie"were not planter-traders.30Forthe most part,local-levelcommercewascontrolledby Lebanesebuyers who workedeitheras agentsof Frenchbuyinghouses or, in the 1950s, as independentmerchants.It is truethatsomeplantersdid enterthe cropbuying circuitin the 1950s.They owned trucksand operatedas independentagents for the Frenchmerchanthouses. (Supportfromthese planter-traders was an of the as it votes for the elections asset for the mobilized SAA-PDCI important who were it the Yet seems that of 1950s.) heavilyengagedin big planters many the coffee-cocoatradewereregardedasforeignersin the forestbelt.Some were Senegalese,but morenumerousand significantwere"Dioula"northernersor Odiennik6sfarmingin the west (such asYacoubaSylla of Gagnoa,and the FadikafamilyofTouba)."Stranger" statuscompromisedtheirabilityto claim to representthe mass of the coffee- and cocoa-growingpeasantry,and thus werenot a theirpoliticalclout.Forall of thesereasons,Africanplanter-traders powerfulsocialforcein the Ivoirianforestzone. The commercialweaknessof the Ivoirianplanterbourgeoisie,coupled with its lackof structuraleconomic-politicalties to the mass of the peasantry, wouldfatallycompromiseitspoliticalpowervisa visthecolonialandpostcolonial state. As I will showbelow,the contrastwiththe situationprevailingin neighboringGhanawasstriking.In the Ivoirianforest,the riseof a planterbourgeoisie did not reinforceor accentuatepre-existingsocialandpoliticalhierarchies. The accumulationof wealthin the forest surelyincreasedthe status of big planters,but in generalthis wasnot accompaniedby the developmentof concrete,micro-or local-levelmechanismsof politicaland economiccontrolover the behaviorof peasants,or overcommunity-level"collectiveaction." b. Ghana At the householdlevel,the dynamicsof cocoaproductionin Ghanaand CGted'Ivoirewere quite similar.What differedwas the political,social, and economic "superstructure" of peasant export-cropproduction.Ghana is a much-studiedcase, and importantfeaturesof its ruralsocial structureand ruralpoliticsthroughthe 1950s are well known.This accountwill brieflyreview the points that are essentialin describingthe remarkablepoliticaland economic clout of the big Akan/Asantecocoa farmers.These farmersstood atop strongindigenouspoliticalhierarchies(whichwere remoldedand reinforcedby the colonialstate)and exerciseddirectcontroloverthe production and internalmarketingof much of Ghana'scocoa crop. To conquer the territory that became Ghana, the British signed protectorate treaties with chiefs and kings in the coastal zone and in the northern savannah. In the center, the British fought the armies of one of West Africa's This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 18 CATHERINEBOONE most powerful and centralizedstates,Asante (Ashanti),in three wars that ended with Asante's final military defeat in 1901. The well-elaborated political/administrativehierarchies of the Akan states provided a perfect infrastructurefor British-style "indirect rule."The return from exile in 1924 of the Asante king, Asantehene Prempeh II, and Britain'srecognitionof a reintegratedand strengthened Ashanti Confederacy in 1935, were dramatic moves in Britain's strategy of using the indigenous polities for its own ends. This option had not been available to the French in central and western C6te d'Ivoire, because established state hierarchies did not exist in these regions. The rise of a peasant-based export economy in Ghana pre-dated the era of colonial rule. Ghana exported significant quantities of cocoa by the mid1890s, and by 1911 the territorywas the world's leading cocoa supplier.Elaboration of the colonial administrativeapparatusin the 1920s, 1930s, and 1940s played an important role in defining the socio-economic and political effects of the continuing expansion of cocoa production. It seems that within the framework of indirect rule, class formation and commercialization accentuated preexisting socio-political hierarchies by reinforcing the ability of chiefs and indigenous political officials to cash in on the economic changes that were transforming Akan society. Cocoa production first took root in Akwapim, where chiefs in Akim Abuakwa sold large tracts of land to companies of migrant farmers, who subdivided their holdings into individual farms.While the "company land" system took hold in some parts of Akim Abuakwa, in other parts families purchased land. Under both systems, farm labor was mobilized through wage contracts. These new land and labor relations promoted the rise of a nascent rural capitalism.3' Production spread through southern Asante, and in the 1930s a version of the abusa sharecropping system became increasingly prevalent in both Ashanti and the original cocoa-producing areas.32Under this abusa system, migrant farmers did not purchase land, but rather asked local chiefs for permission to cultivate stool lands. Stool lands were "communal lands" which lay under the jurisdiction of the Akan paramount chiefs, who administered them in trusteeship for the nation. Sara Berry describes how the British, eager to restrain the commercialization of land and labor in cocoa-producing regions, persuadedthe chiefsinAsanteandBrongAhafoto prohibitthe saleof landin their domains.... They endorsedthe chiefs'rightto demandtributefromstrangers'who soughtpermissionto cultivate[stool]land ... [C]hiefsin Asanteand neighboring statesdemanded one-third ofthecocoacropastributefromstranger farmers.... In southernGhana,evenmigrants whoboughtlandwereexpectedto paytributeto the local chief,justliketenant'farmersin Asanteand BrongAhafo.3 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 19 All or part of the tributecollectedby local chiefswas claimedby the national (paramount)stool treasuries.In Asante,the paramountchiefreceivedall revenues and then remitteda shareto subordinatechiefs. "Citizens"of particularstools were definedby birth or origin;they enjoyedlinealrightsto land.Unlikemigrants,theycould claimrightsto cultivate stool lands withoutpayingtribute.They could hold long-termusufructthat "can and did become tantamountto freehold,"and such rightswerebought and sold among citizens.4 Citizens'earningsand land sales were subjectto financialclaimsby the chiefsto whom they pledgedtheirpoliticalallegiance, and it seems that these chiefsworkedwith some successto retainauthoritybased economicprerogativesovercitizensandtheirland.In the face of population movements,Asante and Akim Abuakwatried to impose economic extractionson citizenswho hadmigratedto farmlandoutsidetheirdomains.35 of land, stools And in the face of the partial,yet defacto,commercialization or over "freehold" landwithin to absolute reversionary rights sought reimpose theirjurisdictions.36 Landtenureand citizenshipquestionsin colonialGhana fueledintenseintra-stoolpolitickingand rivalriesfor controloverboth people and land. The expansionof the cocoa economyin generalandthe abusasystemin particularreinforcedthe economicpowersoftheAkanstates,reinforcedchiefly/ administrativehierarchies,and enhancedthe extractivepowersof the paramount chiefs.37This was as the Britishwanted it. "Indirectrule"in Ghana in the neounderwrotethe centralization of economicandpoliticalprerogative traditional Akanstates.38The colonialstateencouragedandenabledparamount chiefsto accumulatepowerat the expenseof subordinatechiefs,creatingpoliticalhierarchiesin Akansocietiesthatweremore sharplydefinedthan anythingthat existedin historicalprecedent.39 Small-scale, peasantproductionwasthebackboneof the cocoaeconomy.40 Withinthis contextemergedthe socialstratumthatwouldso decisivelyshape the territory'sfuture,the stratumthatRichardCrookcalledthe "elitenetwork of agro-commercialinterestsso powerfullyrepresentedby the chieftaincy."4' The traditionalpoliticalelite- chiefs,otherofficeholders,and elders- enteredcocoa productionearlyand with the politicaland economicadvantages their positionsconferred.In the Brong-Ashantiareathe politicalelite began cocoaproductionearly,in the 1910s and 1920s.GwendolynMikellwritesthat they"hada headstart.Theywereableto selectextensiveand contiguoustracts of well-situated,fertileland."42Chiefs and other office holderswere able to invest capitalnot only in the developmentof their landholding,but also in transportand trade(andin theirown advancementwithinthe politicalhierarchy).43Manyrichplantersbecamecocoa merchantsin the 1920s, and a powerfulstratumof chieflyplanter-traders and absenteelandholdersdevelopedin the south. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 20 CATHERINE BOONE Mikell arguesthatby the 1930s, stratification in the ruralareashad resultedin two clearlydefinedgroups:one composedof formerand presentahenfo[office holders]andsikafo[wealthy,privilegedpersons]... whosecocoawealthgenerated educationand capitalforfurtherinvestment;andthe othercomposedof ordinary folk whose small cocoa farms,impededby inadequatecapitaland labor,often causedtheir indebtednessto the firstgroup...Pledgingvaluablecocoa land becamethe idealmeansfordealingwithloans.44 In Ghana'scocoabelt,indebtednesstooktheformof landpledging,rather than the mortgagingof cropsto commercialintermediaries, whichwas more often the case in C6te d'Ivoire.Farmspledgedby poor Ghanaianpeasants "mightbe held for severaldecades... therefore,for allpracticalpurposes,such a farmwascompletelyalienated."45 Indebtednesswasa powerfulforceworking to concentratecontroloverlandandpersonsin the handsof the chieflyGhanaian planter-merchant elite, addingmomentumto the processof class formation. Havinga classof capitalistfarmersrisefromthe ranksof the chieflyelite was antitheticalto the purposesand practiceof indirectrule. As Crook explains,"the colonialstate tried to preventtheir consolidationinto a gentry' when it became clear that chiefs were attemptingto amalgamatestool and familyland revenuesutilizingthe legaland politicalpowersaffordedthem by the state."46Tactics employedto thisend47didnot changethe factthatin southern Ghana,officialprerogativesconferredupon neo-traditionalelitesand the economicpowerof an ascendentaccumulating classcouldbe cumulative,symsourcesof strengthin the politicalarena.In biotic, and mutually-reinforcing Ghana,the potent mix of politicalauthorityand economicclout produceda stratumof chieflyplanterswith a greatdealof controloverthe politicalbehavior of peasanthouseholds,as well as a considerablecapacityto mobilizecommunity-levelcollectiveaction.Fromthe 1930sto the 1960s,thisstratumproved itselfwillingandableto lockhornswiththe the colonialandpostcolonialstate. The tightlyinterwovenpoliticaland economichierarchiesof Ghana's "indirectlyruled"cocoa economygaverich chiefsconsiderableleverageover theirsubordinates.Informalpatron-clientrelationshipsweregroundedfirmly in the materialbaseof the chief'scontroloverthe grantingof landuse rights,in indebtedness,and in landlord-tenantrelations.KwameNinsin suggeststhat the prerogativeof determiningwho was a "stranger" and who was a "citizen" ("subject")wasexercisedby localchiefs.In a systemin whichcitizenshipin the traditionalpolityandthe statusof landclaimswere,as Ninsinsays,"inexorably thisgavethe localchief(whomightwellbe one'slandlordand/or intertwined," creditor)the powerto define a person'seconomicand politicalrights.48For peasants, these multistranded relations meant that economic insecurity, and This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 21 insecurityof land tenurein particular,could be manipulatedby politicalauthorities. Economicdependencywas not the onlybond betweenchiefsand their of subjects.Chiefsclaimedreligiousandmoralauthorityas the representatives Africanstates,self-consciousnations,and ancientlineages.This "morallegitimacy"is of coursewhatthe Britishworkedso hardto propup, andto harness for their own purposes.Under indirectrule, a systemthat assumedand was predicatedupon the legitimacyof the neo-traditionalelite, local populations law and judicialpractice.Chiefs at various were subjectto "neo-traditional" levels of the politicalhierarchyadministeredboth civil and criminaljustice, exercisedcoercivepowers,imposedfines,and adjudicateddisputes.49 These ideological,legal,and economicaspectsof chieflyauthoritywere and part parcelof complexforms of social organizationthat, whatevertheir internalstrainsand contradictions,servedin the 1930s through1960s as (in Apter'swords) a kind of "natural"or "ready-made" politicalmachineryfor No wonderthattheConventionPeople'sParty(CPP) mobilizingthepeasantry.50 and its opponents,the NLM and the UP, soughtthe supportof chiefs.They could be trustedto bringin the ruralvote. If Ghana'splanter-chiefs wieldedconsiderablepowerovertheirsubjects, tenants,anddebtors,thentheyalsoenjoyedformsof leverageoverthe colonial statethatmakethemuniquein the contextof thisstudy.The criticalfactis that the indigenouspolitical-economicelite occupiedpowerfulpositionsin the export marketingcircuit.From these positions,they could directlyappropriate and valorizetheirshareof the wealthgeneratedby peasantproducers.Most fromthesepositionsthe cocoa"big-men"couldconfrontthe state significantly, the (and Europeanmerchanthouses)directlyin strugglesto expandtheirshare of the peasantsurplus. Ghanaianmerchantsin the exporttradewereenormouslypowerfulcomparedto theirSenegaleseor Ivoiriancounterparts.Not only did they control strategic positions in the internal commercial circuit, such as credit to andthe buildingof storagedepots,but as earlyas smallholders,transportation, the 1910s a significantgroupexportedcocoa directlyto the UK. This fact is quite importantas an indicatorof the scale of operations;commercialsophistication,and scope of capitalaccumulationof Ghana'slargestplanter-merchants.Accordingto Southall,"inOctober1918 one sourcereportedthatthere were now a total of 292 Africanfirms or individualsinvolvedin the direct exportof cocoa.They wereindependentof the servicesof the expatriatebuyers."57 The growingpowerof the cocoa planter-brokers drewthem into headon confrontationswith the Europeantradinghouses and the colonialstatein the attemptedcocoahold-upsof 1920-21,1922-23,and 1930-31,in the highly successful cocoa marketingboycott of 1937-38, and in the commercial boycott and swollen shoot campaigns of 1948. In the 1930s, African brokers attempted This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 22 CATHERINEBOONE to expandtheirshareof the cocoa surplusat a time of fallingworldprices.52 The Europeanfirmsfoughtback,and in 1937 they concludeda market-sharing agreementdesignedto undercutthe Africanmerchants. It wasprecisely becausethe largercocoabrokerswereutilizingtheirdominancein a largeproportionof the surplusof thecocoaeconomy theruralareasto appropriate in the tradeforthe majorityof the buyingfirms, at a time of decliningprofitability that the lattercame togetherin an arrangementwhose explicitpurposewas to reducethe cost of brokerage." Ghanaianplanters,brokers,andchiefsrespondedto the poolingagreementby organizingandenforcingthe mostsuccessfulfarmers'marketingboycottwaged in all of colonialAfrica.54 The patternof boycottsis an indicatorof the commercialand financial clout of the cocoa elite. It also helps to revealthe extent to which political power, economic clout, and statuswere intertwinedand inter-linkedin the socialhierarchiesof Ghana'sneo-traditionalAkan states.Bigplantersandlargescale brokerswerealso often chiefs,and the cocoa big-mencould wield multiple formsof powerin attemptsto organize,enforce,and mobilizegrassroots participationin the hold-ups. Wealthwas obviouslyan importantpoliticalassetin itself.Importantin explainingthe successof the 1937-38sellingboycottwasthe factthat"brokers had the capitalto buy up and hold the cropsof smallerandpoorerfarmers."55 Politicalauthoritywieldedby the chiefswas anothermajorassetof the cocoa elite in theirconfrontationswith foreignbuyingcompaniesand the state. Recent workon the cocoa hold-upsrevealsthe extentto which the chiefsright down fromNana Ofori-Attato the villagelevelsin EasternProvinceandAshanti were involvedin the formalorganizationand enforcementof the hold-ups.... By early 1948 [therewas a] campaignto boycottexpatriateand Syrianfirms'high priced'imports....Thereis clearevidencethatthe chiefsthroughoutAshantiand the Colony... sidedwiththe boycottandhelpedto enforceit withallthe resources of the NAs.56 RhodaHowardshowsthatchiefsalsoused theirtraditionalpowersof sanction in encouragingthe hold-upof 1937-38;for example,they refusedto perform funeralriteswheretheirsubjectsrefusedto respectthe sellingboycott.But, she adds, "chiefswere not only chiefs;they were also largefarmersin their own right,sometimesmoney-lendersor cocoa-buyerson a largescale."57 If NativeAdministration, relationsof production,and sacredpowersall helpedto undergirdthe powerof the cocoa elite,then so too did the organizational and institutionalresourcesprovidedby the colonialadministrationto cocoaplanter-traders in the wakeof the 1937-38hold-upandWorldWarII. In 1938 the colonial administrationbegan to encourage more activelythe formation of producers' cooperatives, or farmers' unions.s8In the 1940s and 1950s, these quasi-independent marketingunions reinforcedthe economic and politiThis content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 23 cal power of the ruralcocoa elite.The co-existenceand overlappingof neotraditionaland market-basedforms of socialhierarchythat were manifestin other aspectsof Ghanaiansociety also found institutionalexpressionin the unions. As Mikell argues,cooperativebig-men "used polygynousmarriage, kinship,and patron-clientnetworksto enhancethe membershipof the coopambitionsof the large erativesocieties."'5 Unionspromotedthe entrepreneurial by expandingtheiraccessto capital,andby blockingunlicensed planter-traders And as thingsturnedout, the farmers'associabuyers'accessto the market.60 tions also facilitatedcollectiveactionby providingan independentbase for politicalorganizationon the partof the ruralcocoa elite.61 The hierarchicalstructureof peasantsocietyin southernGhana,andthe fact that the ruralsocio-politicalelite occupieda strategicpositionin the exthis casefromthe othertwo studiedhere. port marketingcircuit,differentiate Ghana'sruralelite,unlikeits counterpartsin C6te d'Ivoireor Senegal,was in a positionto contestdirectlythe state'sclaimson the ruralsurplus.I will argue below that these facts of Ghana'sruralsocial structurego far to explainthe state-formationstrategiesof the Nkrumahregime. c. Senegal In Senegal'sgroundnutbasin,ruralsocietywas even more hierarchical than it was in southernGhana.The politicaland economicpowerthat was wieldedby an indigenouselitewas concentratedin the handsof a much narrowersocial stratum.In the case of Senegal,the capacityof the ruralelite to wealthwaspredicateduponits intermediary appropriatea shareof agricultural positionin Senegal'sdefactosystemof colonialindirectrule.Thesefactswould haveimportantimplicationsforthe postcolonialregime'sstate-buildingstrategies. Islamicjihadsin the Sahelin the mid-1800sweakenedthe old statesof the Senegambiaregion,facilitatingFrenchmilitaryconquestin the 1870s.62 Frenchconquestand"pacification" of the territorywasaccompaniedby rapid Islamizationof the defeatedWolofpeople.In the earlypart of the twentieth century,the colonialadministration forgedan allianceof conveniencewithupMuslim and-coming religiousleaders,the Mouride and Tidjane marabouts. The Frenchsupportedthe marabouts, underwritingthe rise of theireconomic andpoliticalpowerin the ruralareasof centralSenegal.Themarabouts, in turn, the of activelypromoted spread groundnutcultivation. They also advisedtheir followersto submitto colonialrule. Withthe supportof the administration, Islamicleadersspearheadedand organizedmassmovementsof agrariansettlementalongnew railroadsthatcut deep into the interior.The religious leaders established groundnut estates of their own with land, credit, and equipment granted to them by the colonial state. These estates were cultivated by their religious disciples who lived aus- This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 24 CATHERINEBOONE expenseduringsevenyearsof nonremuneratedlabor. terelyat the marabout's Aftersevenyears,the marabout gavethe followerland.This pioneeringprocess opened centralSenegalto export-cropproduction,creatingpeasantvillages centeredaroundgroundnutproductionandspiritualallegianceto a marabout. Peasantspaidtithesto theirspiritualguidein cash (earnedthroughthe sale of the exportcrop), in groundnuts,by workingthe marabout's fields,and/orby feedingdisciplesworkingfull-timeon the largeestates.TheIslamicauthorities, for their part, provideddiscipleswith land, and often with credit,seeds, or emergencyrelief.They sponsoredtheirfollowers'children,and they provided communityleadership.Most importantly,perhaps,the Islamicelite provided faithfulfollowersand discipleswith a routeto salvationin the after-life.For all of these reasons,the materialand spiritualties that bound disciplesto their religiousleadersremainedintactovertime. Religiousobligations,new social relationsof production,and groundnutcultivationwerecloselyintertwined.63 Peasantswereheavilydependentuponthe economic(andreligious)prerogativesof the ruralIslamicleaders,andthussubjectto theirpoliticalcontrol. One descriptionof this relationshipwas offeredby RolandPorthres,who was commissionedby the colonialadministrationto analyzethe socio-economic conditionsin Senegal'sgroundnutbasin.Porteresarguedthat the controlof the religiousauthoritiesoverthe groundnutpeasantrywas"total,"andthatthe peasantswerein a stateof "totalsubmission.""Mouridepeasants,he argued, were"themoralandmaterialsubjects"of an exploitativeandrapaciousIslamic elite."Itis to him [themarabout] thatthe peasantsdeliverthe productsof their labor...A new class of serfs,or perhapsit wouldbe betterto say 'slaves,'has emergedbeforeoureyes."''Port res'assessmentlackedsubtletyandqualification, but it is clearthatboth the socialrelationsof productionand the patronclient structuresof the groundnutcommunitiesgave marabouts considerable political,economic,and social clout at the local level.Village-levelholy men, with their own followings,disciples,and clientele,were themselvessubordinate to maraboutshigherup in the well-institutionalized hierarchiesof the Islamicbrotherhoods,or confr&ies. Centralizationandhierarchygavethe aristocratic marabouticelite who controlledthe confririesmuch influence in brokeringpoliticalrelationsbetweenstateandpeasantry,and thusmadethem effectiveagentsof indirectrule.66 In the 1910sthe colonialstatebeganto manageseedstocksin someparts of the colony,in principleto assureanadequatesupplyatplantingtime.Soci6t6s Indig'nes de Pr6voyance(SIP), FrenchWestAfrica'searlyversionof agriculturalcooperatives,were createdto managethis task.67In the early 1930s the state developedthe SIP into a systemfor stockingand distributingseeds on a widespreadbasis.Cultivatorswerenow obligedby lawto joinandpaydues.In an attempt to sustain production and offset declines in productivity, the SIP distributed not only seeds but also tools and credit to farmers. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 25 ThompsonandAdloffarguethat,in the groundnutbasin,SIPwerebuilt in parton the "communalsolidarities"of the Islamicorders.68 Marabouts were and as such were as on often selected SIP communityleaders, they agents the locallevel.They wereamongthe leadingbeneficiariesof the system.Religious leadersenjoyedprivilegedaccessto seeds,credit,andtools,whichtheyused to developtheirownholdings.Justas significantfroma politicalperspectiveis the fact that state effortsto distributeproductivity-enhancing inputs workedto enhancethe religiouselite'spowervis i vis the peasantry.In the groundnut controloverprobasin, the developmentof the SIP expandedthe marabouts' ductive resourcesand reinforcedthe peasants'dependenceon the religious authorities'patronage. The systemof indirectrulethatemergedin centralSenegalwas robust. The Frenchreliedon the Islamicordersto organizethe "colonization" of the groundnutbasinand to ensurethe peasants'acquiescenceto colonialrule.To thisend,the Frenchprovidedthemarabouticelitewiththe economicresources, prerogatives,and politicalbackingthat wererequiredto build the Islamicordersinto wealthyinstitutions,andto centralizepowerin the handsof a narrow elite of grandsmarabouts.69These, withseveralthousandfollowerseach,ranthe hierarchicalconfriries andwereamongthe largestlandholdersin allof Senegal. The structuraldependenceof the Islamicorderson the colonialstatewas reflected in the abilityof the French to manipulatethe successionof grands to the positionof GrandKhalifof the Mourides,and in the success marabouts of the colonial administrationin manipulatingintra-and inter-sectrivalries and intrigues.Controloverthe colonialpursegavethe Frenchthe upperhand in dealingwith the Islamicleaders.In the process,high-rankingmembersof the religiouselite grew wealthyfrom the earningsof theirown estates(from 1931 on, groundnutpriceswere subsidizedby Franceat a rate about 15% aboveworldmarketlevels),the offeringsof theirfollowers(estimatedin the 1960s to totalabout10%of the groundnutharvest),andloansandcashsubsidies from the French authorities.70 Wealthallowedindividualmaraboutsto attracteverlargerpersonalfollowings,andthusincreasedtheirpoliticalweight and that of the brotherhoods. Party competitionin the terminalcolonialperiod enhancedboth the politicalcloutof the Islamiceliteandits dependenceupon statesubsidies.The marabouts'commandof the ruralvotemadethemextremelyvaluablealliesfor the urban-basedpoliticianswho competedfor controloverthe postcolonial state.Donal CruiseO'Brienwrotethat [t]he maraboutsbecamethe politicalagentsof the majorpartiesafterthe rural areaswereenfranchisedafterWorldWarII.The most effectivemeansemployedby the politicalleadersin theireffortsto win the marabouts'favoris directeconomic assistance.In the 1950s,politicsbecameby farthe greatestsinglesourceof revenue for the Mourideelite.7 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 26 CATHERINEBOONE In the groundnutbasin,postwardevelopmenteffortsand partypoliticsboth workedto increasethe flow of state resourcesthroughthe SIP, and thus to enhancethe capacityof the Islamicholy men to distributematerialresources to theirfollowers.These institutionsbecameimportantbuildingblocksforthe ruralvote-gatheringmachinesbuiltby the two partiescontestingthe elections of the 1950s, the Socialists(SFIO) and the Bloc D6mocratiqueS6n6galais (BDS).72In theprocess,stateresourcesbecameincreasingly importantin mainof the bonds economic that to taining dependency helped tie peasantsto the As this the elite. politico-religious happened, abilityof the marabouticleaders to reproducetheir own politicalpowerand economicprivilegebecame ever moredependenton statelargesse,andon the interestof stateofficialsin investing in the Senegaleseversionof indirectrule. The abilityof the marabouticelite to confrontor challengethe colonial statewasalsocompromisedby sharplimitson theirpoweras economicagents, thatis, limitson theirpoweras a "class-like" stratum.Thetop-flightmarabouts wereSenegal'slargestlandowners,and amongthe largestSenegaleseinvestors in urbanrealestateand commerce.They didconstitutean accumulatingelite. Yet theirroom for maneuverin dealingswith the statewas severelycompromisedby the factthattheireconomicprivilegederived fromtheirpoliticalpositions withinthe systemof indirectrule.Eventhe materialrewardsthatflowed fromtheirspiritualauthority(mostsignificantly, volunteerlaborto worktheir fields) were largelydependent upon politicalprivilegesgrantedby thecolonialstate (in this case,powerto organizelandpioneeringandto grantnew landto peasants). Meanwhile,in contrastto the situationin Ghana,colonialtradinghouses and theiremployees(most of whomwereLebanese)establisheda ubiquitous presencein the groundnutbasin and directcontroloverexport-cropbuying. This meansthatbig maraboutestate-ownersandsecularindigenousbusinessmen had extremelylimited access to the prime site of accumulationin the colonialeconomy,the ruralcommercialcircuit.The politicalimplicationsof this are clear:neitherthe marabouticelite nor businessmenallied with the confririescould use control over crop purchasing,storage,and transportas leveragein dealingwith the Frenchtradinghousesor the colonialstate. In colonialSenegal,the ruralelitehad considerablepowerandinfluence overthe export-cropproducingpeasantry.Yet the powerof this elitewaslargely dependentupon alliancewith stateofficialsand accessto stateresources.This severelyconstrainedthemarabouticelite'sinterestin contestingthe state'sclaims to a largeshareof the groundnutsurplus,as well as their capacityto do so. Stateofficials,for theirpart,reapedimportantpoliticaland economicbenefits from theiralliancewith the Islamicconfriries. This helps to explainthe statebuilding strategies of the Senghor regime. The main contrasts in rural social structure that have been described above can be summarized in a two-by-two table: This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 27 Rural elite's capacity to contest state's claims to agricultural wealth High Hieracrchy in Social Structure Low High Low C6te d'Ivoire Ghana Senegal Rural Interests and State Formation In the early years of postcolonial state formation, all rulers sought to forge institutional and political arrangementsthat would ensure rural acquiescence as the exploitation of the peasantry intensified, and as wealth generated by small farmers was transferredto other social groups and to other sectors of the economy. Even with this common goal, there was a great deal of variation in the political alliances and institutional arrangementsthat were forged in the late 1950s and 1960s to link state and countryside.This paper seeks to explain the variation in the degree of (de)centralization and (de)concentration in the institutional apparatuses that projected state power into the rural areas. I am arguing that variation can be largely explained by the fact that rural social structures differed markedly across cases. Peasant societies offered different p6ssibilities for alliance between state and rural elites. They also differed in their capacity to resist the aggrandizingdesigns of postcolonial regimes.These differences had prolonged political consequences, for regimes sought to build state structures that would counter the challenges and seize the political opportunities that they confronted on the ground. Nkrumah, like his counterparts in Senegal and C6te-d'Ivoire, sought to impose political control over export-crop producing regions, and to intensify the exploitation of peasant farmers.To do so, the Ghanaian state established a far-flung and intrusive presence in the localities.The CPP constructed a vast patronage machine that linked rural producers to the state. Local outposts of the ruling party organized peasants at the grassroots, distributed credit and agricultural inputs, and established an official monopoly over cocoa buying. Yet even at the local level, control over resources and political prerogative lay for the most part in the hands of state agents appointed from, and answerable to, the center. In neighboring C6te d'Ivoire, the institutional arrangements linking state and countryside could not have been more different.Houphoubt's This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 28 CATHERINEBOONE "liberal"approach gave the state control over export-crop pricing and regulated the conditions of access to the buying circuit, but export-crop buying and the distribution of credit and agriculturalinputs remained in private hands. The Ivoirian regime discouraged the formation of political organizations on the local level. No political machine mobilized rural voters into the service of the ruling party. My argument is that ideology does not explain this difference. All the elements of Nkrumah's statist, interventionist approach to governing and taxing the cocoa belt were establishedbefore his "turn to socialism" ("turn against capitalism") in 1961. The major CPP strategies that centralized control over state power while extending the state'spresence in the localities (deconcentrating state structure) were implemented in the 1950s. Meanwhile, the contrast between French and British administrativeideology does not explain the difference; this contrast would lead us to anticipate the opposite outcome (i.e. a more "statist"solution in C6te d'Ivoire). My argument is that contrasting strategies of rural institution-building were the political consequence of differences in the structure and dynamics of rural society. In devising means to govern and tax cocoa producers, Nkrumah confronted direct resistance mounted by large-scale cocoa farmers and traders (and supported by much of the peasantry in Ashanti), who contested not only the new state's claim to cocoa wealth, but also Nkrumah's and the CPP's claims to state power. So explosive was this confrontation that, in little more than a decade, it destroyed both the Nkrumah regime and one of sub-SaharanAfrica's most prosperous peasant farming economies. Ghana's cocoa producers were remarkablein the context of decolonizing Africa, not only for their prosperity,but also for their capacity to mount organized challenges to the colonial and postcolonial states.As I have argued above, their capacity for collective action reflected the wealth of big producers, their strong positions in the cocoa-buying circuit, and after 1938 the institutionalization of their positions as community-level economic leaders in the farmers' unions. The ability to mount direct resistance to the state's claims on the rural surplus also reflected the hierarchically-structuredpolitical economies of the Akan states under colonial rule, which gave the rural elite (organized by and around the chiefly planter-merchantstratum) multiple sources of leverage and influence over the behavior of smaller-scale and poorer producers. Nkrumah's strategiesof ruralinstitution-buildingrepresenteda no-holdsbarred attempt to neutralize the capacity of southern Ghana to resist taxation and to contest the hegemony of the regime. To politically subordinate the cocoa belt, the CPP built state structuresin the countryside that were designed to undercut and thereby destroy the stratum of ruralsociety that had been able to organize a direct challenge to the regime. The CPP's efforts to micro-manage political competition and strategic economic exchanges within localities represented an attempt to create new This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 29 community-level powerstructuresthatwouldby-passanddisplacethe oldsocioeconomichierarchies. This was a directattackon the foundationsof the cocoa elite'swealth and authority.In distributingcreditto small farmers,the CPP interposeditselfbetweenthe big farmerand the smallone, betweencreditor and debtor,and between cocoa buyerand client. In distributingcredit and purchasedagriculturalinputs,regulatinglandrents,and organizingthe farmers into grassrootscooperativeslinkeddirectlyto the partyand the state,the CPP sought to create new patron-clientstructuresthat would displaceold ones, and to usurp the patronagepowersof the establishedchieflyplantermerchantstratum.Establishinga CPP cocoa-buyingmonopolyswept away the old farmers'unions and thus not only destroyedan importantpartof the base of the planter-traders, but alsodestroyedone of the orgaentrepreneurial nizationaland socialbases of theirpower.New administrative organsof governmentat the locallevelforcedthe old elite of "indirectrule"to sharepower with party men. This dilutedthe power of the neo-traditionalnotablesand createdincentivesfor them to mortgagetheirfuturesto the new regime.The unitsallowedthe regimeto manipulatethe linesof redrawingof administrative authoritybetweenandwithinthe Akanchieftaincies,and to turn chieflyrivalries to the regime'sadvantage.Extensiveinstitution-buildingand intensive manipulationof localpoliticaleconomieswerestrategiesaimedat subordinatNkrumiststratingthe cocoabeltby "rewiringthe circuits"of ruralauthority.73 egiesdid not arisefromanideologicalcommitmentto abolishingchieflypower: the CPP did not hesitateto allywithand enhancethe powerof neo-traditional eliteswhen this wouldstrengthenthe government'spositionin battlesagainst its rivals.74 WasNkrumahmoreheavy-handedduringthe firstdecadeof CPP rule becausehis regimesoughtto place a tax burdenon ruralproducersthat was heavierthanthe one imposedon theirIvoiriancounterpartsby Houphouetin the 1960s?In otherwords,does the intensityof stateexploitationof peasants, ratherthan ruralsocialstructure,accountfor the differencebetweenthe two cases?There is no completelystraightforward answerto this question,but this line of reasoningdoes not seem to lead to a clearexplanationof the differof the anti-CPPopposition,as was manifestin the rise ence.75Crystallization of the NLM, occurredwhen Ghanaiancocoaproducers'incomeswerestable (1951-1954), duringwhichtime the flow of statepatronageto the ruralareas (in the form of credit,etc.) increased.Beckman(1976) arguesthat the big cocoa producersturned againstthe CPP as earlyas 1952, when it became clearthatthe newgovernmentwouldusethe cocoamarketing boardas a mechanism of state accumulation,ratherthanreturningGhana'scocoa surplusesto the farmers.This suggeststhat the politicalbattlesin southernGhanain the 1950swererootedin a struggleoverthestate'srightto appropriate cocoawealth, rather than in a defensive reaction on the part of a peasantry suddenly victimized by a dramatic increase in the level of state exploitation of the rural masses. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 30 CATHERINEBOONE Civil society in the Ivoirian south was far weaker than it was in southern Ghana, and this reality gave rise to very different state-building strategies in C6te d'Ivoire. Of the cases studied here, the peasantry of the Ivoirian forest zone came closest to resembling the proverbial"sack of potatoes."The fragmented and atomized social structures characteristicof the peasantries of the Ivoirian south meant that capacities for collective action were low. The brilliance of Houphoubt's strategies for governing and taxing the south is that they reinforced this very feature of rural society. The regime built few outposts of state authority in the rural south, makfor a state structure that was "concentrated" in form. In most rural coming munities there were few points of access to state resources. State resources were not used to bolster the powers of established ruralnotables, nor were they used to promote the rise of an alternativerural leadership.Geographical "concentration" of administrative-political-distributiveapparatuses and functions in the capital city had other important implications for politics at the local level.76The state provided little incentive or impetus for political organization or collective action. Observers spoke of the stifling and stagnation of political life at the local level. Michael Cohen arguedin 1973 that the governmentavoided spending money in the localities because state money generates political activity.77BarbaraLewis wrote in 1971 that the PDCI sought to minimize political group formation at the local level." These political strategies were reflected in the highly centralized and concentrated design of Ivoirian state institutions. This logic of rural administration was mirrored in the PDCI's handling of the more expressly economic functions of government in the Ivoirian forest zone. The regime left direct cocoa- and coffee-buying in the hands of a merchant stratum dominated by Lebanese traders,who purchased coffee and cocoa from peasant producers at the farm gate. Relying on private Lebanese traders meant that the regime had also decided to leave the distribution of creditto small farmers in the hands of this same group, for crop-buying and credit are inexorably intertwined. Only the political weakness of the indigenous merchant class (itself revelatory of important features of rural social organization) can explain why the Houphouet government was able to leave the coffee- and cocoa-circuit in foreign hands. It was a shrewd move that strengthened the regime in severalways. First, the private nature of buying and lending transactionsreinforcedthe atomistictendencies inherent in smallholder commodity production. Second, the regime avoided building a hierarchical crop-purchasing apparatus with outposts in the localities, and thus avoided creating organizationalbases in the ruralareasfor political entrepreneurs.(This complemented its broader strategyfor administering the localities.) Third, the state could target rural merchants as the source of the peasantry's exploitation, thereby deflecting attention from its own appropriations of the rural surplus and positioning itself as the defender of peasants' interests. And fourth, the regime ensured that the politically strategicfunctions of export-crop marketing This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 31 and ruralcreditdistributionwouldremainin the handsof a vulnerable"outsider"groupwith no politicalambitionsof its own (otherthan stayingon the good side of powerwieldersin Abidjan).Lebanesetradersdominatedexportcropmarketingandruralcredit,andthishelpedpreemptpossibilitiesthatlarge Africanfarmerswouldparlaytheirwealthinto politicalcloutin the ruralareas. And as politically-vulnerable outsiders,the Lebanesecommunityhad virtually no capacityto fightthe statefora largershareof the ruralsurplus.Theregime's powervisa visthe merchantswasreflectedin the lowpricesthe regimepaidfor the servicesof thisgroup,andthe resulting"efficiency"of the Ivoirianexportcropmarketingcircuit. Houphouetcapitalizedon, and even deepened,the existingpoliticaland economicweaknessesof peasantsocietiesin the south.His politicaland institution-buildingstrategiesmade good politicalsense in the C6te d'Ivoire;they wouldhave beenpoliticalsuicidefor Nkrumah in Ghana. The critical differences betweenthe two caseslie atthe levelof ruralsociety,ratherthanin the personalitiesor ideologiesof the two governments. Senegalrepresentsa thirdscenario,one wherea regimeinstitutionalized a power-sharingdeal with a influentialruralelite. State-buildingoccurredvia the decentralization of statepower.This processdevolvedpoliticalprerogative andcontroloverstateresourcesto Islamicnotableswho "colonized"and"captured"the ruraloutpostsof the regime.Thesearrangements tied the peasantry to the state indirectly,throughthe intermediationof the rural elite, just as similararrangementshad done duringthe colonialperiod.The logic of the alliancebetween state and ruraleliteswas largelythe same before and after politicalindependence.This workedin Senegalbecausethe ruralelite had a powerfulcommandoverthe loyaltiesand politicalbehaviorof the groundnut peasantry(and thus had somethingveryvaluableto offerto the politicalauthoritiesin Dakar),but no economicbaseof its ownthatwouldhaveallowedit to maintainits politicalinfluenceand economicprivilegewithoutdirectand continuousstate backingand support.Senegal'sestablishedruralelite was inexorablytied to andultimatelydependentupon the state,andinstitutionalizing this relationshipwas highlybeneficialfor the Senghorregime. Two hypothesesabout the ruraldeterminantsof state structurein the 1960s and 1970s arisefromthe threecasesstudiedhere.The firstis that the of the state apparatuswas influenced degree of concentration/deconcentration greatlyby the degreeof socio-politicalhierarchyin existingruralsociety.Where societiesweremorehierarchicaland ruralauthoritieswerethereforemore influentialat the local level, the state apparatuswas "deconcentrated," in the sense thatofficialsub-unitsof the stateweremoreubiquitousin the localities. In thesecases,governmentstendedto createmoresitesof accessto statepower and resourcesin rurallocalities. The second hypothesis is that the degree of defacto centralization/decentralizationof power - i.e. the amount of political autonomy conceded to established rural notables - varied as follows: where established rural authorities This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions CATHERINEBOONE 32 were positioned to control an important share of the economic surplus directly, or to appropriateit on their own account (as when they were themselves big farmers or big traders), state managers tried to adopt centralizationstrategies in order to reduce the rural elite's power. By contrast, where powerful rural elites occupied tribute-taking positions that were predicated upon their alliance with the state, and where their privilegeswere in large part propped up by allies in the state, mutually-beneficial arrangements were institutionalized in power-sharing arrangementsthat decentralized control over state resources and official prerogative. Testing these hypotheses across a wider range of cases may generate new insights about how rural social forces shape the institutional configuration of state power in Africa. Notes This paperwaspresentedat the 37th AnnualMeetingof the AfricanStudiesAssociation, RoyalYorkHotel,Toronto,Canada,3-6 November1994.An earlierversionwas presentedat + 50,"a conferenceorganizedby the FrancophoneAfricaResearchGroup/Groupe "Brazzaville de Recherchessurl'AfriqueFrancophone(GRAF),BostonUniversityandthe Centred'Etude d'AfriqueNoire (CEAN)of the Institutd'EtudesPolitiquesat the Universityof BordeauxI, 78 October1994, Boston,Massachusetts.I wouldliketo thankPeterTrubowitz, Kathryn Firmin-Sellers,DwayneWoods,MickMoore,ToyinFalola,RobertVitalis,JenniferWidner, andVivienSchmidtfor commentson earlierdrafts.The materialpresentedheredrawsupon Catherine Boone, MerchantCapital and the Rootsof State Powerin Senegal, 1930-1985 (CambridgeandNewYork:CambridgeUniversityPress, 1992);and CatherineBoone, "Commercein C6te d'Ivoire:Ivoirianization withoutIvoirianTraders," JournalofModern African Studies 31 (1993), 67-92. capitalistdominancethroughthe state:The multifariousrole of 2 PeterGeschiere,"Imposing the colonialstatein Africa,"inWim van BinsbergenandPeterGeschiere,eds., OldModesof Productionand CapitalistEncroachment:Anthropological explorationsin Africa (London, 1985), andNon-Accumulationin Agriculture:Regional 94-143;PeterGeschiere,"Accumulation Variationsin SouthCameroon,"' a paperpresentedat the AfricanSocialHistoryWorkshop, ColumbiaUniversity,31 October1989. it must alsobe deconcentrated.If a systemis 3 Logically,then, if a systemis decentralized, deconcentrated,it is not necessarilydecentralized. 4 Robert Bates, Beyond the Miracleof the Market:The PoliticalEconomyofAgrarian Development in Kenya(Cambridge,1989), 75. ' In the 1960s,90%of all coffeeand cocoawashandled by privatebrokers.In the 1980s, 80% was commercialized throughprivatebrokers,ratherthanthroughcooperatives.See A. Hirschfeld,"Histoire,situation,et perspectivesdu mouvementcoop6rativeen RCI," InformationsCooperatives3 (1975), 60; MarchesTropicaux,n. 2380 (21 juin 1991), 1552. 6 See RichardE. Stryker,"PoliticalandAdministrative Linkagein the IvoryCoast,"in Philip Foster and Aristide R. Zolberg, eds., Ghana and the Ivory Coast:Perspectiveson Modernization (Chicago,1971), 73-102;TessyD. Bakary,"C6ted'Ivoire:l'6tatisationde l'Etat"in Jean- Francois Medard, ed., Etats d'AfriqueNoire:Formation,micanismes,et crise (Paris, 1991), 53- 91; andAlainDubressonand ClaudineVidal,"Loind'Abidjan:Les Cadres,Urbanistesde l'Intirieur- La prefecturedeToumodi,"unpublishedreport(Abidjan:InstitutFrangaisde RechercheScientifiquepourle D6veloppementen Cooperation[ORSTOM],Centre ORSTOMde Petit-Bassam,avril1991). Goran Hyden, No Shortcutsto Progress(London, 1983), 19. Crookarguesthatwhatis "mostdistinctive"aboutC6te d'Ivoirein the Africancontext"isits ability,at a verycrudelevelof comparison,to implementits policies.The exportcropshave been successfullygrownandmarketed;the farmersget paid andreceivetheirinputs;feeder 8 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 33 Administrative roadsget builtand taxescollected"(RichardC. Crook,"Patrimonialism, Effectivenessand EconomicDevelopmentin C6te d'Ivoire," AfricanAffairs88 (1989), 206). in 1966. 9 Nkrumahwas overthrown 1o Bjorn Beckman, Organisingthe Farmers:CocoaPoliticsand National Developmentin Ghana (Uppsala,1976). "1Beckman, Organising,113-15. 12 If the analyticfocuswereon policy-making (the legislativeprocess)andthe formal-legal relationsbetweensub-unitsof the state,as it is inVengroff's1993 analysisof administrative/ politicalreformin Senegalafter1970,then Senegalwouldbe describedin the oppositeterms. AsVengroffpointsout, in Senegalvirtuallyall legislationoriginatesfromthe executive,andthe formalchainof commandin the statehierarchygivesthose at the top full supervisorycontrol who merelyexercisepowersdelegatedto them fromabove.(Vengroff overall subordinates, showsthatsome of this maybe changingin the contextof attemptsto sustainthe processof in Senegal.)See RichardVengroff,"TheTransitionto Democracyin Senegal: democratization In Depth3 (1993), 23-52.The analysishere emphasizesthe The Role of Decentralization," verygreatextentto whichactualpoliticalpracticedivergedfromthismodel,andthe limitsof centralauthoritywhen it cameto imposingpolicies/lawsdraftedby the executive.My depiction of Senegalesepoliciesis consistentwith studiesthatviewthe stateas a politicalmachinein arecooptedinto statestructures, whichpoweris deeplypersonalized,local-levelpowerbrokers and politiciansandbureaucratsareaccordeda greatdealof discretionin implementingpolicy andusingstatemoniesto buildup theirownpersonalclientele. 13 Jean Copans, Les Maraboutsde l'Arachide:La confrbriemourideet lespaysans du Sinigal (Paris, Class and Bureaucrats: 1988), esp.210-12; Donal B. CruiseO'Brien,"Co-operators Formationin a SenegalesePeasantSociety,"Africa41 (1971), 263-77;Donal B. Cruise O'Brien, Saints and Politicians:Essays in the Organisationof a SenegalesePeasant Society (Cambridge,1975);JanHalpern,"Laconfrrie des mourideset le d6veloppementau Senegal," Cultureset Diveloppement3 (1972), 99-125. in Senegal (Cambridge, 1970), 111-16. C. Behrman, Muslim Brotherhoodsand Poblitics '4 Lucy Cruise O'Brien, Saints and Politicians;and 'S See for example Behrman, Muslim Brotherhoods; SocialStructure,the IslamicBrotherhoods,and IrvingLeonardMarkovitz,"Traditional PoliticalDevelopmentin Senegal,"JournalofModernAfricanStudies1 (1970), 73-96. Observersarguedconsistentlythatin spiteof waythingssometimesappeared,the Dakar-based politiciansretainedthe upperhandin this relationship.See Boone,MerchantCapital,103. 16 At the sametime, colonialruleinstitutionalized the powerof certainelementswithinAfrican societiesat the expenseof others,partlyredefinedinterests,and obviouslycreatedmultiplenew avenuesfor accumulatingpowerandwealth. and Los 17 See for exampleRobertBates,Marketsand States in TropicalAfrica(Berkeley Angeles,1981);GoranHyden,BeyondUjamaain Tanzania(Berkeleyand Los Angeles,1980); and the "planterbourgeoisie"literatureon C6te d'Ivoire,whichincludesSamirAmin,Le du capitalismeen C6te d'Ivoire(Paris, 1967); and Bonnie Campbell, "The Social, ddveloppement Political,and EconomicConsequencesof FrenchPrivateInvestmentin the IvoryCoast," Ph.D. thesis,Universityof Sussex,November1973. See also SaraBerry,FathersWork andNon-Accumulation." for 18 Geschiere,"Accumulation theirSons.Accumulation,Mobility,and ClassFormationin an ExtendedYorAibd Community (Berkeleyand Los Angeles,1985). theirshareof the economicsurplusdirectly,or did they 19 That is, did ruralelitesappropriate of the colonialstate? do so indirectly,via the intermediation 20 The Westwas the domainof the societiesof the Dida, B6te, widelydispersed,"acephalous" and the and Guro.The Center,Baouli land,was characterized by "pronouncedstratification, of power" accumulationof a substantialsurplus,allwithoutan concomitantcentralization Productionand SocialFormation:TheBauleRegionof Chauveau,"Agricultural (Jean-Pierre in MartinA. Klein,ed., PeasantsinAfrica: in HistoricalPerspective" Toumodi-Kokumbo Historicaland ContemporaryPerspectives(Beverly Hills and London, 1980), 145). Meanwhile, the eastwas the site of Agnikingdoms(Morounou,Indenie,Sanwi,Bondoukou)thatsigned treatieswith the French,and thus survivedthe colonialconquest.Of these,only Sanwiwas strongenoughto contestthe authorityof the postcolonialstate.Houphouetcrushedthis attemptimmediatelyandthoroughly. 21 Aristide R. Zolberg, One-PartyGovernmentin the Ivory Coast (Princeton, N.J., 1964), 21-22. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions 34 CATHERINEBOONE 22 Zolbergprovidesqualifierswhichhelp to supportthispoint:"Insome waysit is appropriate to speakof French'directrule'...[Yet]in practice,whereverthe Frenchfounda clearly differentiated, functioningtraditionalpoliticalsystem,the county [canton]wasmadeto coincidewith the territorialboundariesof the chiefdom.This was the case,for example,in KiembaraCountyof Korhogocercle[of the north],amongthe S6noufo[of the north],and of AboissoandAbengourou.The latterwas dividedinto throughoutthe Agni-dominantcercles threecountieswhichcorrespondedto Ndi6ni6,B6tti6,and Diab6Statesrespectively. There,the traditionalrulersbecameofficialchiefsas well"(Zolberg,One-Party 53; see also Government, 21-22). 23 24 Zolberg, One-PartyGovernment,41-42. The colonialstateplayedan importantrolein determiningthe pace of this process,but the state cannottakecreditfor its scopeor the particularmodalitiesof smallholderexpansion. the guaranteedsaleof Ivoirian Taxation,the buildingof the railandroadinfrastructure, commoditieson Frenchmarkets,the politicalunificationof the territory,the use of a single of socialandeconomiclife all contributedto the currency,andthe increasingmonetarization expansionof cash-cropfarming.Meanwhile,the colonialauthoritiesencouragedthe colonizationof the westby exemptingBaouleand Malink6pioneerfarmersfromforcedlabor. In 1932,the territoryof HauteVoltawas attachedto the C6te d'Ivoire,doublingthe labor de plantation. supply to the Ivoirian c&onomie 25 RobertM. Hecht,"TheTransformation of LineageProductionin SouthernIvoryCoast," Ethnology:AnInternationalJournal of Culturaland SocialAnthropology23 (1984), 270. See also BarbaraLewis,"Land,Property,andPolitics:RuralDivo at the Fin deRegime," paperprepared for the 1991 meetingof theAfricanStudiesAssociation,St. Louis,MO, 23-26 November 1991. 26 Jean-Pierre ChauveauandJean-Pierre Dozon, "Aucoeurdes 6thniesivoiriennes...L'6tat"in Emmanuel Terray, ed., L'EtatContemporainen Afrique(Paris, 1987), 283. 27 See Lewis,"Land,Property, and Politics." of LineageProduction;" Lewis,"Land,Property,and 2s See Hecht,"TheTransformation Politics;"andJeanMarcGastellu,"Droitd'usageet propri6t6priv6e"in E. Le Bris,E. Le Roy, et F. Leimdorfer, eds., EnjeuxFonciersenAfriqueNoire (Paris, 1992), 269-80. de C6te d'Ivoire(PP), with 29 This is probablyless truein the east,wherethe PartiProgressiste the help of the colonialadministration, mobilizedoppositionto the PDCI. who was a planter-professional (a doctorin the civilservice)seems 30 Houphouet-Boigny, moretypical. andBrokersin the Gold 31 PollyHill, as reportedby RogerJ. Southall,"Farmers,Traders, Coast Cocoa Economy," CanadianJournal ofAfrican Studies 12, 2 (1978), 193. See also Sara Berry, No Conditionis Permanent:The Social Dynamics ofAgrarian Changein Sub-SaharanAfrica (Madison,1993), 107, 111;GwendolynMikell,CocoaandChaosin Ghana(NewYork,1989), 71. the trendtowarda morecapitalist 32 The shiftto the abusasystemhad the effectof "reversing class structure" (Jean Marie Allman, The Quillsof the Porcupine:AsanteNationalism in an Ghana[Madison,1993], 37). Allmanarguesthatthis trendprobablycontinuedinto Emergent the 1950s.GarethAustinfocuseson the importantissue of the eclipseof wage-labor in southernAshantiin the 1930s.See GarethAustin,"The relationshipsby sharecropping c. 1916-1933,"Journalof emergenceof capitalistrelationsin southAsantecocoa-farming, AfricanHistory28 (1987), 259-79. On the colonialadministration's attemptsto restrainthe commercialization of landand labor,see alsoBeverlyGrier,"Contradiction, Crisis,and Class Conflict:The Stateand CapitalistDevelopmentin GhanaPriorto 1948,"in IrvingLeonard Markovitz, ed., Studies in Class and Powerin Africa (NewYork, 1987), 27-49; Richard C. Crook,"Decolonization,the ColonialState,and Chieftaincyin the Gold Coast,"AfricanAffairs 85 (1986), 88; and Anne Phillips, TheEnigma of Colonialism:British Policy in WestAfrica (London,1989). On usufructaryrightsunderGhana'scommunallandtenuresystem,see KwameA. Ninsin,"TheLandQuestionsincethe 1950s"in EmmanuelHansenandKwame A. Ninsin, eds., The State, Developmentand Politicsin Ghana (London, 1989), 165. 33 107, 111. On the lastpoint,see Rathbone'sdescriptionof Akim Berry,No Condition, Abuakwa (Richard Rathbone, Murderand Politicsin ColonialGhana [New Haven, 1993], 59). It seemsthatovertime,tributetendedto becomea moredirecttax or "rent."Mikellreported thatthe AshantiConfederacydoubledthe rateof cocoatributefor the 1950-51 cocoa season This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions RURALINTERESTSAND MODERN STATES 35 fromone-halffarthingto one farthingper tree (CocoaandChaos,154). On the rateof cocoa rent or tributeoverthe period 1913-1930s,see Austin,"Emergenceof CapitalistRelations," in all of 268. GarethAustinarguesthatthe eclipseof wage-laborrelationsby sharecropping southernGhanain the 1930sreflectedthe powerof northernmigrantlaborers,who wereable to pressurelandownersto givethem a betterdeal,ratherthanthe initiativesof the colonial administration. See his "Emergenceof CapitalistRelations,"275-79.This argumentrepresents an importantdeparturefromthe one advancedby Phillips,Enigma,and GeoffreyKay,The PoliticalEconomyof Colonialismin Ghana (London, 1972). Rathbone, Murder,56. 34 5 David E. Apter,Ghanain Transition (Princeton,1972, secondrevisededition),257-63. See also Mikell, Cocoa and Chaos, 162. 89. CrookcitesAkyem(Akim)Abuakwaas an example.On See Crook,"Decolonization," that case, see also Rathbone,Murder. 37 On AkimAbuakwa,see Rathbone,Murder,57-58. 38 Under these systemsof landtenureand landpioneering, expansionof the cocoa economy becomesthe mainforcedrivingchieftaincypolitics.Stoolsdisputedcontrolovercertainlands, rivalchiefsdisputedclaimsto the allegianceof peoplein certainlocalities,and chiefsat various in the nameof the stool. levelsof the hierarchydisputedthe divisionof surplusesappropriated of the AshantiConfederacyin 1935, which 39 One examplewouldbe the restoration reimposedAshantirule overBrongAhafo.A secondexamplewasthe legalframeworkof indirectrulein AkimAbuakwa,whichcentralizedeconomicandpoliticalpowerin the handsof Akimofficials.OforiAtta andthe Nana OforiAttaII, muchto the resentmentof lower-ranking Britishsawcentralization as integralto the modernizationof the AkimAbuakwastate.For an accountof this process,see Rathbone,Murder,esp.60-62. 40 The Nowell Commissionreportin 1939 statedthat 60%of the farmsin Ashantiwereunder one acre (Mikell,Cocoaand Chaos,99). Presumablymost householdsin this categoryfarmed severalsmall,non-contiguousfarms. 98. 4~ Crook,"Decolonization," 36 42 43 Mikell, Cocoa and Chaos, 93-94. Mikellnotes that the selectionof chiefswas influencedby wealthor influence:"nowwealthy persons openly competed for stools"(Cocoa and Chaos, 132). 44 Mikell,Cocoa and Chaos, 95. 45 Mikell,Cocoa and Chaos, 96. 88. Crook,"Decolonization," An exampleis tighterregulationof stoolrevenues,as institutedin 1944 ordinanceswhich establishedstool treasuriesthatweresubjectto semi-annualauditsby the colonialstate.The 1944 Ordinancesaimedat the comprehensive reformandmodernizationof the Native Authorities.See Crook,"Decolonization," 83-4; MaxwellOwusu,UsesandAbusesof Political Power(Chicago,1970), 200. 48 Ninsin,"The LandQuestion." Ordinancein 1910 madeNativeTribunals 49 Amendmentof the NativeJurisdiction "compulsorycourtsof firstinstancein mattersnot fallingclearlyunderBritishlaw" (Rathbone,Murder,60). 50 Apter,Ghana, 340-41. Traders,"195. 5 Southall,"Farmers, 52 RhodaHowardarguesthata drivingforcebehindthe 1937-38boycottwas a groupof wealthyfarmersand coastaltraders"whowantedto be ableto shipcocoa directto the EuropeanandAmericanmarketswithoutgoingthroughEuropeanmiddlemen"(Rhoda Howard,"Differentialclassparticipationin anAfricanprotestmovement:The Ghanacocoa 46 47 boycott of 1937-38," CanadianJournal ofAfrican Studies 10 (1976), 471). 46-47. Traders,"186, see also 197-202.See also Beckman,Organising, 53 Southall,"Farmers, On the Cocoa-Buying(Pooling)Agreementitself,see Howard,"Differentialclass 474-76. participation," classparticipation," 471-72, 479-80; Crook,"Decolonization;" 54 See Howard,"Differential Traders."On the swollenshootcampaigns,see Rathbone,Murder,196; Southall,"Farmers, Mikell, Cocoa and Chaos, 145. Southall,"Farmers,Traders,"205. Crook,"Decolonization," 94, 96. On this point,see also SaraBerrywho arguesthat"[o]ne factorwhichmayhelp to explainthe greaterfrequencyand effectivenessof [cocoa]holdupsin 5' 5 This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions CATHERINEBOONE 36 Ghanais the differentpositionsof AkanandYorubachiefsin the respectivecocoa economies" (Berry, No Condition,75). See also Mikell, Cocoa and Chaos, 98; Rathbone, Murder, 196-7. 471. Howard,"Differentialclassparticipation," From 1930 onward,the colonialadministration providedan impetusforthe formationof buyers'cooperatives,or farmers'unions.See Austin,"Emergenceof CapitalistRelations,"27273, frn.80. However,before1938, cooperativesboughtless than 3%of the crop (Beckman, 48). In the wakeof the 1937-38holdup,officialsprobablycameto see the Organising, developmentof farmers'unionsas a wayto curbthe powerof the biggestprofessionalbrokers. 17 58 59 6 Mikell, Cocoa and Chaos, 150-51. Beckman, Organising,especially 232. This was obviousin the mobilizationof the pro-CPPvote in 1951, andin the rise of antiCPP politicsafter1952. See Beckman,Organising. 62 Most of the discussionwhichfollowsis drawnfromBoone,Merchant Capital. 63 There aremanyexcellentstudiesof SenegaleseIslam,the marabouts andthe religiousorders, 61 and the rural marabout-disciplerelationship. See Behrman, Muslim Brotherhoods;Donal B. Cruise O'Brien, TheMouridesof Politicaland EconomicOrganisationof an Islamic Senegal" Brotherhood(Oxford, 1971); Cruise O'Brien, Saints and Politicians;Christain Coulon, Le Marabout et le Prince:Islam et Pouvoirau Sinigal (Paris, 1981); Copans, Les Marabouts;and Halpern,"LaConfrnriedes Mourides."Formorerecentworkthat focuseson SenegaleseIslam in greaterDakarand the secondarytowns,see LeonardoA.Villal6n,IslamicSocietyandState Powerin Senegal(Cambridge,1995);M.C. Diop, "Lesaffairesmouridesa Dakar,"Politique Africaine 1 (1981), 90-100. * RolandPorteres, de l'6conomieagricoleet ruraleau Senegal:rapportde la "Am6nagement MissionRolandPortures"(Dakar:GouvernementGeneralde l'AOF,Territoire du Senegal, mars-avril1952), 109. 109. 65 Porteres,"Am6nagement," 66 Copans, Les Marabouts. On the SIP,seeVirginiaThompsonand RichardAdloff,FrenchWestAfrica (London,1958), UnderColonialRule(Evanston,IL, 1968);316311-13, 357-9; MichaelCrowder,WestAfrica 67 17; Jean Suret-Canale, AfriqueNoire:L're coloniale1900-1945 (Paris, 1964): 299-310. 67 Thompson and Adloff, FrenchWestAfrica, 357. investedin the prestigeof the confiiriesby contributingto 69 The colonialadministration projectssuch as the buildingof the GreatMosqueatTouba,the capitalof the Mourides. 70 Meanwhile, politicsin the 1950swouldshowthatthe membersof the Islamicelitewere quitesensitiveto the factthattheirgroundnutearningsreflectedin partFrance'sgroundnut subsidy(grantedin 1931), whichraisedpricessome 15%aboveworldmarketrates. 71 Cruise O'Brien, TheMourides,262. See also Ruth Schachter Morgenthau, PoliticalPartiesin FrenchWestAfrica(London, 1964), 147-9, 152; and Behrman, Muslim Brotherhoods,85-120. 2 Thompson and Adloff, FrenchWst Africa, 313. This phraseis usedby JohnDunn in his descriptionof the politicsbehindthe creationof a separateBrong-Ahaforegionout of most of whathad been (sincethe restorationof the Ashanti Confederacyin 1935) theWesternProvinceof Ashantiandrestorationof the Paramountcy overa reconstitutedAhafodivisionto the Kukuomhene.See JohnDunn, "Politicsin Asunafo" in DennisAustinand RobinLuckham,eds., Politicians andSolidiers in Ghana(London,1975), 195. 74 In "Politicsin Asunafo,"Dunn describeswhatis probablythe most obviousexample. " On the problemof measuringcross-casevariationin the intensityof stateexploitationof the countryside(ie., the problemof measuring"urbanbias"),see for exampleAshutoshVarshney, "UrbanBiasin Perspective," Studies29 (1993), 19. See alsoJennifer JournalofDevelopment Widner,"The Originsof AgriculturalPolicyin IvoryCoast 1960-86,"Journalof Development Studies29 (1993), esp. 28-29. MichaelA. Cohen,"The Mythof the ExpandingCentre:Politicsin the IvoryCoast," 76 SocioJournalofModernAfricanStudies11 (1973), 227-40;J.-F.M6dard,"LaRWgulation 7 Politique," inY.-A. Faur6 et J.-F. M6dard, eds., L'Etatet Bourgeoisieen Crte d'Ivoire (Paris, 1982), 70; Stryker,"PoliticalandAdministrative Linkage." 77 Cohen, "Myth," 241, 242. in CarletonT. Hodge,ed., Paperson the 78 Barbara Lewis,"The Dioulain the IvoryCoast,";' Manding(Bloomington,1971), 293. This content downloaded from 158.143.192.135 on Wed, 08 Jul 2015 10:18:11 UTC All use subject to JSTOR Terms and Conditions