Networking,SocialCapitalandGenderRolesintheCotton

SysteminBenin

GuirguissouMaboudouAlidou

Thesiscommittee PromotorProf.DrA.NiehofProfessorofSociologyofConsumersandHouseholdsWageningenUniversityCo‐promotorDrJ.K.KampenAssistantprofessor,EducationandCompetenceStudiesGroupWageningenUniversity OthermembersProf.DrC.Leeuwis,WageningenUniversityProf.DrE.B.Zoomers,UtrechtUniversityDrM.Dekker,AfricanStudiesCentre,LeidenDrC.P.J.Burger,WageningenUniversityThis research was conducted under the auspices of the Graduate School ofWageningenSchoolofSocialSciences(WASS).

Networking,SocialCapitalandGenderRolesintheCotton

SysteminBenin

GuirguissouMaboudouAlidou

Thesissubmittedinfulfilmentoftherequirementsforthedegreeofdoctor

atWageningenUniversitybytheauthorityoftheRectorMagnificus

Prof.DrM.J.Kropff,inthepresenceofthe

ThesisCommitteeappointedbytheAcademicBoardtobedefendedinpublic

onWednesday17September2014at11a.m.intheAula.

GuirguissouMaboudouAlidouNetworking,SocialCapitalandGenderRolesintheCottonSysteminBenin,202pages. PhDthesis,WageningenUniversity,Wageningen,NL(2014)Withreferences,withsummariesinDutchandEnglish ISBN978‐94‐6257‐063‐4

ToAmirath,Redouane,andWakirath

i

Acknowledgement

Theseedforthismore‐than‐four‐yearsresultwassowndecadesagobymyilliterateparentstowhomsincerethanksareduefortheiropen‐mindedness.Beingthesecondoftheonlytwosonsofmyparents,andgiventhatmyelderbrotherwasmilesawayfrom home to pursue his secondary education, I was not destined for westerneducation.Iaccompaniedmyfatherinfarming,whichcouldhardlybecombinedwithgoingtoschool.BecauseIgotboredstayingathomeand,especially,havingmyfriendsleavemeforschooleverytimeweplayedtogether,Idemandedtobesenttoschoolaswell. My father accepted my demand. After these friends dropped out for variousreasons,my fatherwascriticized forhavingsentallhissons toschoolandwas toldthathislastdayswouldbehard.DespitethepressuremybrotherandIexperienced,we did not drop out from school. Indeed, our father’s strong opposition swiftlychangedfrominitiallycajolingtopunishmenttodeterusfromleavingschooluntilweunderstood thebenefit of studying. “Iwould likeyou to studyas farasyou can”,wewererepeatedlytoldbyourlatefather.Ourmotherhadnoauthoritytocoerceus,but,instead,shewouldusetheproceedsfromfood‐processingtopayforourschoolneeds.Thededicationandstrongwillofourfather,whoatthetimehadbeensoboldastosendallhissonstoschool,andthesacrificesofourmother,pavedthewaytothisPhD.

Therealisationof this thesiswasalsopossible thanks touncountablecontributions,which cannot all bementioned here. The daily sharing of happiness andwoes anddiscussionsonall kindsof subjectsand issueswith fellowPhDs, colleagues, friends,countrymates, and acquaintances constituted thepetals on the rosewhich coveredandprotectedmeagainstthethornsthatwerepartofthisPhDjourney.Thatiswhythefollowinglistofcontributionscannotbeomitted.

I would like to express my gratitude to the Netherlands University Foundation forInternationalCooperation(NUFFIC)thatprovidedthefundingofthisPhD.Completingitwouldnothavebeenpossiblewithoutthatfinancialsupport.IamalsogratefulfortheJuniorResearcherGrantthatshoulderedthecostsofattendinganimportantPhDWorkshopinDenmark.

Apartfromthefinancialmeans,astudiousmind,stimulationandguidancetorealizeaPhD, also social and affective support are needed to nurture the academic work.Professor Anke Niehof provided me with the academic, psychological, social, andfinancial support needed during the PhD journey. Even when I could not rewardmyself formaking progress, shewas the one fromwhomwords of encouragementcame,thusrevivingmymotivationtomeetthechallengesandhopetoovercomethedifficult times. I am genuinely honoured to have completed this PhD under the fullsupportofProfessorAnke,towhomIamdeeplyindebted.RoyJordaancomplementedthis support, forwhich I amprofoundlygrateful. Inaddition tohelpingwitheditingmyworkandsupplyingmewithacademicmaterials,hewouldarrangesomerelaxingactivitytorechargemybatterieswhenhedetectedsignsofintellectualfatigue.

Mysincereappreciationforhisavailabilityandscientificrigourgoestoco‐promotorJarlK.Kampen.WhileIhavesomeregretstohavemethimabitlate,Iamalsogladtohave started this PhD journey with Hester Moerbeek as the co‐supervisor.Unfortunatelyshecouldnotcontinue it, inspiteofherdeterminationtodoso.Withtheconsecutivesupportofboth,Icouldenhancetheanalysisofmyquantitativedata,which was crucial for correctly interpreting the qualitative data and the data

ii

interfaces. It was a real pleasure to have experienced such diverse ways ofsupervising.

IwouldliketothanktheentirestaffoftheSociologyofConsumptionandHouseholdsGroupforthestimulatinginteraction.IvaluethecriticalcommentsofChizuSato,Hiljevan derHorst, Gerda Casimir, and Carja Butijn onmywork.My thanks go toHedyMunro for her solicitousness, always smilingly helping me with the manyadministrativeobstaclesonefacesintheNetherlands,toMargaretvanWissenforherhelp in the layout and formatting of the book and Carry Vleeming for heradministrativesupport.ProfessorLisaPrice, thefirstWageningenacademicImet inBenin,linkedmetoprofessorAnkeNiehof.IwillneverforgetthegreattimesIsharedwith Hom, Melissa, Eirik, Mariam, Gisella and Paul Pieters, Christina, Zacharia,Theresa,Jarkyn,Feng,andStefanWahlen.ThankyouforspicingthingsupduringmystayinWageningen.PosthumousthanksareduetothelateDrEkaeteUdong,aclosecolleague,whoselifewasbrutallytakenwhenshefinallycouldenjoyherPhDdegree.IalsowanttomentionVitaliyPopov,OmidNorozy,andRafatSobohandaspecialthankgoestoAlphonseSingboanfLaurentGlinfortheirpsychologicalandmaterialsupport.IwanttothankeverysinglememberofthewarmBeninesecommunityinWageningenwithwhomIsharedtheupsanddownsofaPhDjourney.Inparticular,Iwouldliketomention Augustin Kouevi, Laurent Glin, Eugène Agbicodo, Souleimane Adekambi,Florent Okry, Nicodeme Fassinou, Yann Madode, Latifou I. Aboubacary, WaliouYessoufou, Euloge Togbe, Djalal Arinloye, Jonas Wanvoeke, Rose Fagbemissi, andNathalieKperafortheunforgettabletimewespenttogetherinWageningen.

Iammostgratefultothepeoplewhohelpedwiththefieldworkanddatacollection,namelyNinguiGomaN’Douro,OrouYérimaPaul,Abou‐MoumouniIssifou,andAdamN.Loukoumanou.WithoutLatifouBelcoIwouldnothavegainedaccesstoinvaluablesecondarydata.Without thecotton farmersandtheir leaders,whooftenabandonedtheirworkto listentomeandprovidemeuseful information, Icouldnothavedonethe research. I alsowant tomention the support and encouragementofMamoudouBaba, Yacoubou Boni, Abdul‐Ganiou Alassane, Sacca Lafia and other relatives whoneverstoppedinquiringaboutmyprogress.IowespecialthankstoOudouZakaryandhis family in Italy and Martha Amina Aruna and her family in London for theirhospitality during my stay in Europe. I am indebted to Leonard Hinnou for hisadministrative support and for helpingmewith familymatters duringmy absencefromhome.IamalsogratefultomycolleaguesoftheAgriculturalPolicyAnalysisUnit(PAPA)andthestaffmembersoftheNationalAgriculturalResearchInstituteofBenin(INRAB)fortheirassistance.

The greatest sacrifice for the accomplishment of this PhD journey came from mybeloved spouse Falilatou,who courageously andpatiently took care of our childrenWakirath,RedouaneandthelastbornAmirath,whoispresentlyexactlyasoldasthisPhDjourney.TheyallacceptedthesacrificesforlongerthanIshouldhavedemandedfromthem.Ithankthemfortheirlovingunderstandingandencouragement.

iii

TableofContentsAcknowledgement .......................................................................................................................................i 

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

List of Tables .............................................................................................................................................. vi 

List of Figures ............................................................................................................................................ vii 

Chapter 1    General Introduction ................................................................................................................ 1 

1.1  Motivation and background ................................................................................................................. 2 

1.2  The country of Benin: some macro‐level data ...................................................................................... 3 

1.3  Problem statement and research questions ......................................................................................... 5 

1.4  Theoretical and conceptual framework................................................................................................ 7 1.4.1  Structure, actor and agency ........................................................................................................ 8 1.4.2  Further conceptual clarification ................................................................................................ 10 

1.5  Study design and methodology .......................................................................................................... 12 1.5.1  Study design .............................................................................................................................. 12 1.5.2  Methods of data collection and analysis .................................................................................. 14 1.5.3  General description of the instruments .................................................................................... 16 1.5.4  Ethical issues and methodological challenges .......................................................................... 17 

1.6  Outline of the thesis ........................................................................................................................... 19 

Chapter 2    Changing Farmers’ Perceptions in Benin’s Cotton Production: From White Gold to Pest .......... 21 

2.1  Introduction ........................................................................................................................................ 22 

2.2  Data sources and analysis .................................................................................................................. 24 

2.3  Historical overview of cotton production and marketing in West Africa ........................................... 26 2.3.1  Cotton production and European settlers ................................................................................ 26 2.3.2  Cotton production from the Colony of Dahomey to the Republic of Benin ............................. 27 

2.4  Resistance to cotton production: From colonial to present times ...................................................... 31 2.4.1   Forced work as a common labour policy in colonial times ....................................................... 31 2.4.2  Post‐Independence resistance in Benin .................................................................................... 33 

2.5  The prime time of cotton: 1990–2005 ................................................................................................ 34 2.5.1  The cotton sector reforms ........................................................................................................ 34 2.5.2   Institutional development and linkages ................................................................................... 36 2.5.3  Farmer organisations: From farm to cities ............................................................................... 38 2.5.4  Major flaws in the reforms ....................................................................................................... 39 

2.6  Farmers’ responses ............................................................................................................................. 41 

2.7  Farmers’ perspectives on the revival of cotton production ................................................................ 43 2.7.1  Farmers’ perceptions of cotton revival ..................................................................................... 43 2.7.2  Measures for the revival of cotton production ........................................................................ 44 

2.8   Conclusion ...................................................................................................................................... 45 

Chapter 3    The Factors Behind the Emergence and Collapse of Cotton Networks in Benin during the 2000s ........................................................................................................................................................ 47 

3.1  Introduction ........................................................................................................................................ 48 

3.2  Historical background......................................................................................................................... 50 3.2.1  Cotton farmer networks in Benin: From agency to associations .............................................. 50 

iv

3.2.2   Naming cotton networks: The war of words ............................................................................. 52 

3.3  Research hypotheses and methods .................................................................................................... 55 3.3.1  Research hypotheses ................................................................................................................ 55 3.3.2  Study design and sampling ........................................................................................................ 55 3.3.3   Data collection methods .......................................................................................................... 56 

3.4  Findings from the qualitative research ............................................................................................... 58 3.4.1  Emergence of breakaway networks .......................................................................................... 58 3.4.2   From weak ties to linkages ........................................................................................................ 60 3.4.3   From strong ties to weak ties .................................................................................................... 60 3.4.4   Preliminary conclusions based on the qualitative research ...................................................... 61 

3.5  Findings from the quantitative research ............................................................................................. 66 3.5.1   Operationalization of concepts ................................................................................................. 66 3.5.2   Empirical results of the survey ................................................................................................. 68 

3.6  Discussion ........................................................................................................................................... 75 3.6.1  Push factors ............................................................................................................................... 75 3.6.2  Pull factors................................................................................................................................. 76 3.6.3  Social capital and cotton leadership ......................................................................................... 77 

3.7  Conclusion ........................................................................................................................................... 78 

Chapter 4    Generating Resources and Conflicts: Profitability and Social Cohesion in Benin’s Cotton Sector ....................................................................................................................................................... 81 

4.1  Introduction ........................................................................................................................................ 82 

4.2  Theoretical perspective ....................................................................................................................... 83 

4.3  The cotton system ............................................................................................................................... 84 4.3.1  The socio‐cultural setting of the cotton zone ........................................................................... 84 4.3.2  The first cotton elite: The rise of school dropouts .................................................................... 85 

4.4  Cotton production and collective action ............................................................................................. 86 

4.5  The rise of conflicts and violence ........................................................................................................ 88 4.5.1  Typology of conflicts in cotton farmer organisations................................................................ 88 4.5.2  The first uprising: “Orou Bagou” vs. “Orou Sori” ...................................................................... 90 4.5.3  Political interference and the advent of traders and intellectuals in cotton organisations ...... 91 4.5.4  The Altruist and the Saint .......................................................................................................... 92 4.5.5  The brothers‐in‐law ................................................................................................................... 93 

4.6  The decay of “caution solidaire” and the demise of collective action................................................. 94 

4.7  Conclusion: Killing the goose that lays golden eggs ........................................................................... 95 

Chapter 5    Gender Roles in Cotton Production and Management of Cotton Organisations in Benin .......... 97 

5.1  Cotton production in Benin and the division of labour ....................................................................... 98 

5.2  Gender and leadership ...................................................................................................................... 100 

5.3  Research design and data management .......................................................................................... 101 

5.4  Results ............................................................................................................................................... 103 5.4.1  Gendered profiles of cotton farmers and the division of labour ............................................ 103 5.4.2   Women and cotton organisations ........................................................................................... 104 5.4.3   Life histories of women leaders .............................................................................................. 106 

5.5  Discussion and conclusion ................................................................................................................. 111 

Chapter 6    Rural Households’ Responses to the Cotton Crisis ................................................................. 115 

6.1  Introduction ...................................................................................................................................... 116 

v

6.2  Livelihoods and households .............................................................................................................. 117 

6.3  Research area and data sources....................................................................................................... 120 

6.4  Results and discussion ...................................................................................................................... 122 6.4.1   Demographic and household characteristics .......................................................................... 122 6.4.2  Land allocation to cotton ........................................................................................................ 127 6.4.3  Structure of cotton farmers’ income ...................................................................................... 128 6.4.4  Importance of cotton for strengthening the household’s asset base .................................... 129 6.4.5   Adapting to and coping with the cotton crisis ........................................................................ 130 6.4.6  Effects on intra‐household resource allocation and gender roles .......................................... 136 

6.5  Conclusion ........................................................................................................................................ 137 

Chapter 7    General Discussion and Conclusions ...................................................................................... 141 

7.1  Shedding light on the research questions and hypotheses ............................................................... 142 7.1.1   The emergence of breakaway networks ................................................................................. 142 7.1.2  The decline of social cohesion and the squeeze of collective action ...................................... 143 7.1.3  Livelihoods reconstruction after the demise of cotton production ........................................ 144 7.1.4  Answering the main research question .................................................................................. 146 

7.2  Theoretical and conceptual reflections ............................................................................................ 146 7.2.1   Agricultural (and institutional) liberalisation .......................................................................... 146 7.2.2  Networking and social capital: building trust within organisations ........................................ 147 7.2.3    Gender leadership in rural organisations ............................................................................... 148 7.2.4  Rural household and livelihood adaptation ............................................................................ 149 

7.3  Conclusion and future research ........................................................................................................ 150 

References .............................................................................................................................................. 153 

Appendix 1: Focus‐group discussion guide ............................................................................................... 167 

Appendix 2: Semi‐structured questionnaire ............................................................................................. 169 

Summary ................................................................................................................................................ 177 

Samenvatting .......................................................................................................................................... 181 

About the Author .................................................................................................................................... 185 

Completed Training and Supervision Plan ................................................................................................ 187 

   

vi

ListofTablesTable1.1  Theprocessofdatacollection..................................................................................15 Table2.1  HistoricalphasesofcottondevelopmentinDahomeyandBenin............30 Table2.2  Log‐linearanalysisoffarmers’viewsontherevivalofcotton production.........................................................................................................................43 Table2.3  Rankingofreformsactionsenvisionedbycottonfarmers..........................44 Table3.1  Listoftheactivenetworks(2009–2011surveydata)...................................53 Table3.2a  Crosstabulationoffarmers’levelofeducationbyagebygender............68 Table3.2b  Characteristicsofthesample(metricvariables).............................................68 Table3.3  Farmer'sleadershipstatusbystatusinatomisation.....................................69 Table3.4  Paireddifferencesofmeansranksofperceivedpushfactorsfor

atomisation.......................................................................................................................70 Table3.5  Meanranksofperceivedpushfactorsforatomisation.................................70 Table3.6  Paireddifferencesofmeansranksofperceivedpullfactorsfor

atomisation.......................................................................................................................71 Table3.7  Meanranksofperceivedpullfactorsforatomisation...................................72 Table3.8  Paireddifferencesofthemeanranksofperceivedsourcesofsupportfor

creatingnetworks..........................................................................................................72 Table3.9  Meanranksofperceivedsourcesofsupportforcreatingbreakaway

networks............................................................................................................................73 Table3.10  Analysisofcovarianceforfarmers’membershipofsocialnetworksby

theirleadershipstatusbygenderandeducation............................................74 Table3.11  LinearRegressionofdeterminantsofthedurationofappointmenton

boards.................................................................................................................................74 Table4.1  Collectiveactioninthestudyarea..........................................................................88 Table4.2  Rankofconflictsincottonorganisationsaccordingtotheirfrequency 89 Table5.1  Comparisonofmen’sandwomen’scottonprofile(t‐testforequalityof

means)..............................................................................................................................103 Table5.2  Cross‐tabulationofgenderandeducation........................................................103 Table5.3  Logisticregressionoffactorsaffectingtheaccesstoleadershippositions

incottonorganisations..............................................................................................106 Table6.1  Characteristicsofsurveyedhouseholds............................................................123 Table6.2  Evolutionofcottonproductionbetween2006and2010..........................128 Table6.3  Structureofincomesourcesofmale‐andfemale‐headedhouseholds128 Table6.4  Improvement/reductioninhouseholds’assetsendowmentswithregard

tocottonproduction(percentageofhouseholds).........................................130 Table6.5  Patternsofadaptationandcopingstrategiesincottonzone....................131 Table6.6  On‐farmincomediversificationamongcottonfarmers..............................132 

vii

ListofFiguresFigure1.1  MapofBeninwiththelocationoftheresearchsites........................................4 Figure1.2  Conceptualframeworkandlinksbetweenhypotheses...................................7 Figure2.1  Locationofthecottonbelt.........................................................................................25 Figure2.2  EvolutionoffibreexportfromFrenchWestAfricancoloniestoFrance

(1910–1930)....................................................................................................................28 Figure2.3  Frameworkoforganisationallinkagesincottonproductionbefore2000

(AdaptedfromCIPB(2008)).....................................................................................35 Figure2.4  Institutionallinkagesincottonsectorafterthereforms(Adaptedfrom

AIC(2005)).......................................................................................................................37 Figure2.5  Developmentofcottonproductionwithregardtopricefluctuations....38 Figure2.6  Nationalcottonproductionshareoftheninedistrictslargestproducers

(2008–2010)....................................................................................................................42 Figure3.1  Thehierarchicalstructureofcottonnetworks.................................................51 Figure3.2  Emergingrelationshipsbetweencottonnetworks.........................................54 Figure6.1  HouseholdandhomesteadarrangementofDari’shousehold................125 Figure6.2  HouseholdandhomesteadarrangementofGounou’shousehold.........126 Figure7.1  Changingmeaningsofcotton.................................................................................147 

1

Chapter1

GeneralIntroduction

ABSTRACT: This chapter provides the background of the research andoutlines its theoretical and conceptual foundation. The overall studydesignthatgeneratedthedataispresented,andthechapterendswithanoutlineofthethesis.

CHAPTER1

2

1.1 Motivationandbackground

What prompted me to undertake my PhD journey about the topic of cottonproductionandmanagement?Stillayoungstudentatasenioragriculturalschool,my academic calendar followed the agricultural production cycle. On a day inDecember1992,whileonholidayinmyvillage,Iwitnessedaratherpatheticevent.Afarmer,whocametoreceivehispaymentfromavillage2kmfromthepaymentpoint,wasdeprivedofhisbike.Hiscottonproductioncouldnotcover thecostsofthe inputs (fertilisers and pesticides) he had used to produce it. Fortunately, thevalue of his bikewas estimated to be roughly equal to the debts. Therefore, thefarmercouldwalkbackhomewithoutbeingsenttojail.Howwouldthisfarmerfeelonhiswaybackhomeandwhatwouldhetellhisfamily?Sincethen,Idevelopedakeen interestto investigatewhatcotton farmersdoget inreturn foralltheefforttheyinvestingrowingcotton.

Fromthe late1980sonwards, therehasbeenagrowing interest in thecottoncrop.Cotton is a critical cash crop and plays a key role in Benin’s economic growth,accountingforanimportantshareintheState’srevenuesandtheincomeoffarminghouseholds. Cotton production has been progressively embedded into subsistencefarming formerly dominated by food crops, transforming it into semi‐subsistencefarming. From having a marginal status during the 1970s and the first half of the1980s,cottongrewinimportanceduringthe1990s,bothwithregardtoareacoveredand incomegenerated, averaging37percentof the total cultivatedarea throughoutthe country, and representing up to 80 percent of rural household incomes(Gandonou,2005;WorldBank,2003).Theincreaseofcottonproductionwentalongwithitsgrowingeconomicimportance,bothfortheStateandforfarmers.Theshareofcottonexportsrepresented75percentof thecountry’s totalexportsduringthe firsthalfof the2000sbefore itdroppedto40percent in2008.Meanwhile, theeconomicweight of cotton in the State fiscal revenues and theGDP rose to 25 percent and 6percentrespectively(WorldBank,2003).Thus,cottonproductionhasshapedBenin’sagriculture, therebygreatlyshapingthecountry’seconomicsystem.Thesedynamicswere nurtured by the emergence of farmer organisations, whose importance forcottonproductionsteadilyincreased.

In this situation, the number of farmer organisations in Benin escalated. Theirreputationspreadduringthecottonboomattheendofthe1980s.Theorganisationselapsed into ahierarchical networkwith apyramidal shape, together creatingwhatcouldbecalleda “cottonsystem” throughwhichcultural capitalwasdevelopedandfarmer leaders were formed. The cotton sector became institutionalised, having itsownrulesandpractices.Indeed,asthemaincashandexportcrop,cottonproductionhas shaped the agricultural sector, the farmers’ movement, and the economy ofBenin.Botheconomicand socialbenefitswerenoticeable at individual aswell as atcommunitylevel,throughthedevelopmentofinfrastructuresandtherelativewealthof households (World Bank, 2003). However, inadequate agricultural policies com‐binedwith themismanagementof theorganisationshave led tothe indebtednessof

GENERALINTRODUCTION

3

cotton farmers, a situation that was aggravated by a drop of cotton prices on theinternationalmarkets and a steady increase of the costs of inputs during the early2000s (Ferrigno et al., 2005;WorldBank, 2003). As a consequence, frombeing thebestorganisedagriculturalsectorinBeninthequalityoftheorganisationofthecottonsectordeclined.Thisisthecontextfortheresearchonwhichthisthesisisbased.

1.2 ThecountryofBenin:somemacro‐leveldata

LocatedinWestAfrica,theRepublicofBeninstretchesover750kmfromthecoastintheSouthtotheNigerRiveratitsnorthernborder(Figure1.1).Administratively,thecountryisdividedinto12provinces:Alibori,Atacora,Borgou,andDongaintheNorth;Collines and Zou in the Centre; and Atlantique, Couffo, Littoral, Mono, Oueme andPlateau in the South.With its 114,763 square kilometre, a projected population ofabouteightmillionsfivehundredpeoplein2012(INSAE(2009),andaGDPpercapitaofaboutUS1,508Dollars (UNDP,2011),Beninranks167outof176ontheHumanDevelopmentIndex(HDI).Povertyisquitewidespreadinruralareas.AccordingtoIFAD(2010),Beninisanagriculture‐dependentcountrywithapovertyincidence of 47.3 percent. During the decade 1998‐2008, the growth in agriculturalemploymentwas33.7percent,whichisthehighest increaseofalleconomicsectors.About54percentoftheactivepopulationengagesinagriculturalactivitiesandagri‐cultureaccounts for35–40percentofBenin’sGDP(FIDA,2005;WorldBank,2007).Performedwithtraditionaltechnologies,Benin’sagricultureismainlyfood‐producingandisdominatedbyabout550,000smallholderfarmerswithanaverageplotsizeof1.7 ha (FIDA, 2005; MAEP, 2011). Food crops such as cereals, legumes, roots andtubersarethemainproducts,playingakeyroleinthedomesticeconomyasfoodaswellascashcrops.However,therevenuestheyyieldarerathermarginal.Itiscottonthatplayedakeyroleasacashcropthroughoutthecountryforthelasttwodecades.Itsshareinthestate’sexportrevenuesremainedhighanditscontributiontotheGDPkeptsteadily increasing from1990to2005,whichexplainsthe interestof theState,privateactors,andfarmersinthecrop.Until recently, cottonwas grown throughout the country, including in unfavourableareas. Itwasextensivelycropped in tenoutof twelveprovinces,onawiderangeofsoils.Basedontheagro‐ecologicalconditionsandvulnerabilitytopestinfestationthatdetermine the technical aspects of production, the national cotton research centre(RCF/INRAB) divides cotton‐producing areas into roughly four large zones. TheSoudan Savannah in the extremeNorth, covering Alibori province and the extremeNorth‐east of Atacora province, comprises the zone that hosts the cotton belt andproduces the highest share of cotton production. The Northern Guinean Savannahzone comprises the provinces of Borgou and Donga, the southern part of Atacoraprovinceand thenorthernpartofCollinesprovince.Although thezone isknownasthefoodbelt, italsoproducesasignificantquantityofcotton.TheSouthernGuinean

CHAPTER1

4

Savannah zone covers the northern regions of the provinces of Zou, Couffo andPlateau.Themorehumidagro‐ecologicalconditionsinthiszonedonotfavourcottonproduction. The Guinean Congolean Zone that covers most coastal provinces alsoproducesaverymarginalquantityofcotton(Sinzogan,2006).So,cottonproductionisconcentrated in the Soudan Savannah zone where a lower population density andmore favourable agro‐ecological conditions in terms of low rainfall and humiditymakethezonethelargestcottonproducer.

Figure1.1 MapofBeninwiththelocationoftheresearchsites

Thepresent researchwas implemented in the fournorthernprovinces and coveredtwoagro‐ecologicalzones.Itstartedinthefourprovincesandprogressivelyfocusedontwoprovinces:AliboriandBorgou.Theseprovincesareinhabitedbymanyethnicgroupswhohavetheirinvestmentinagricultureasthemaineconomicactivity,withcotton as the major cash crop, in common. More than 75 per cent of the cotton

GENERALINTRODUCTION

5

producedinthecountryissourcedfromtheseprovinces.Consequently,allthecottonnetworksconcentratedtheiractionsinthisregion,vyingtoenlistasmanyfarmersastheycouldandtocontrolasmuchproductionaspossible.

1.3 Problemstatementandresearchquestions

Farmerorganisationshave leveredcottonproduction,beingattheforefrontofruraldevelopment policies since the revolutionary period when they were managed forpoliticalpropaganda.Inreturn,theyhavebeenshapedbythecrop,andcottonbecametheir lifeline. Originally village‐based, farmer organisations developed according tothecountry’sadministrativestructure,resulting intheformingofahierarchicalnet‐workuptothenationallevel.Withtheboomincottonproduction,theseorganisationshad a de facto central role in community development, which they assumed andachieved with the financial returns they reaped from cotton production, therebyreplacingtheStateatsomepointsintime.Theydevelopedcollectiveactionatvillagelevelandpromotedsocio‐economicdevelopmentbyprovidingthebasicservicesthattheState failed to supply in the fieldsofeducation,health,andrural infrastructures(World Bank, 2003). They also facilitated “mutual and solidarity practices amongmembers” (Tama‐Imorou et al., 2007: 108).The social cohesion and its corollaryofcollective action developed through cotton networking were extended to non‐members,therebybuildingsocialcapitalatcommunitylevel.Butwiththeimportantresources generatedby cotton,mismanagementpractices crept in andadded to theburden of the bureaucracy of the governmental agenciesmanaging the sector. Thissituation seriously overwhelmed the sector, revealing the inefficiency of Statemanagementofthesectorthroughcostlybodies.

The reforms undertaken from1990 onwards to liberalize the sector and tackle theproblemsofmismanagementandinefficiencydidnotonlyestablishmoredemocraticprocedureswithinfarmerorganisationsbutalsobroughttogetherthevariousstake‐holdersandnewinstitutions,likecottonfarmers,inputssuppliers,cottonginnersandbanks, who together composed a huge and complex stakeholders network. Farmerorganisationsarecentraltothisnetworkbecauseoftheirrelationswithalltheotherstakeholders,whocourtedthemassiduously.Theirleadersbecameveryinfluentialforholdingbothinstitutionalandeconomicpower,andattainedahighsocialstatus.Butwhile their leaders became empowered and enriched, farmers got indebted andimpoverished.Inpractice,manyfarmershadarrearsforuptotwocroppingseasonsor were not paid back completely (Sinzogan et al., 2007). This situation created atense and suspicious atmosphere among farmers andweakened the social cohesionwithin thenetwork(WorldBank,2003).Asaresult,dissidentnetworksemergedasbreakawayorganisations(Sinzoganetal.,2007).Theprocessofbreakingawaystartedfrom the top of the pyramid and progressed downwards, resulting in atomisednetworks.Theensuingcompetitiongeneratedandreproducednever‐endingconflictsthatledtoadecreaseoftrustwithinandbetweennetworks.Additionally,atthesame

CHAPTER1

6

timefarmersexperiencedasharpdropincottonprices,decliningyields,andasteadyincreaseofinputsprices(Sinzoganetal.,2007).Thissituationledtoastrongdeclineof cotton production in the cotton belt. Ultimately, the process resulted in theabandonmentof thecropbythousandsofhouseholds(WorldBank,2003),entailingthelossoftheirmajorlivelihoodsource.Mostofthecottongrowerswhochosetogiveuptheirmaincashcrop,underminedtheirlivelihoodandfacedtheriskofimpoverish‐ment.

Thegeneralobjectiveofthepresentresearchwastogainanin‐depthunderstandingofthedynamicsofthe“cottonsystem”,thatistheinteractionsbetweentheeconomicactivityofcottonproductionandthestructureofsocialrelationsfromcommunitytoindividual level in which the activity is embedded. To this end, the main questionaddressedintheresearchishowfarmers’agencyaffectedtheirorganisations,cottonproduction, and the collective action that evolved around that crop. The researchinvestigated the interactions between individual dispositions and actions (microlevel)andtheinstitutionalmechanisms(mesoandmacrolevel),andtheeffectofsuchinteractions on the livelihood systems of the households (the transformativemechanism).Theinvestigation includedananalysisofcottonproduction inBenin inthecolonialpastandinthecontemporarycontext,exploringhowthereformpolicieshaveshapedthesector,independentlyfrominternationalcontingencies.

Toachievethegeneralobjective,theresearchwasguidedbysub‐questionslinkedtothefollowingoverlappingresearchheadings,whichconstitutedtheentry‐pointsoftheresearch:

Heading 1: Emergence of breakaway networks. This heading addresses thequestion about factors that differentiated leaders from grassroots (or ordinary)farmers in thecreationofbreakawaynetworks.Theobjectivehere is toanalyse thesocial mechanisms leading to the emergence of breakaway networks. It is hypo‐thesized that leadersof breakawaynetworks aremore strongly embedded in socialrelationsthanordinaryfarmers(H1).

Heading2:Thedeclineofsocialcohesionand thesqueezeofcollectiveaction.Thisheadingconstitutesaninquiryintotheeffectsofthedeclineofsocialcohesionincottonproducingcommunities.Theaimistoexaminetheeffectofthedeclineofsocialcohesion thatensued from intra‐and inter‐groupconflicts in cottonproductionandcollectiveaction.Itishypothesizedthatthedeclineofsocialcohesionresultedinthedemiseofcollectiveactionincottonproducingareas(H2).

Heading 3: Livelihoods reconstruction after the demise of cotton production.The issues dealt with under this heading are the shifts in assets endowment andlivelihood portfolios of rural households as a consequence of the demise of cottonproductionandhowtheseshiftsaffectedthegenderrelationswithinruralhouseholds.Thisheadinghelpedtestingthehypothesisthatthelossoflivelihoodsconsecutiveto

GENERALINTRODUCTION

7

the demise of cotton production modified the gender relations within cotton‐producinghouseholds(H3).

Figure1.2displaysthedifferentconceptsandtheoriesusedthroughouttheresearchandtheirinterrelations.Theframeworkisintendedtohelpthereaderunderstandtheconnectionsbetweenthetheoreticalconceptsandthedifferentchapters.

Figure1.2 Conceptualframeworkandlinksbetweenhypotheses

1.4 Theoreticalandconceptualframework

This section presents the theoretical structure of the study and discusses the theo‐retical approaches used to investigate the dynamic relations between cottonproduction, the social capital of individuals and the livelihoods of households. Theresearch attempts to unpack the black box represented by the cotton–peoplerelationships,andtoshowhowfarmers’agencybuiltandrebuiltfarmerorganisations,which to farmers are very relevant institutional structures. Ultimately it aims touncover the social changes resulting from the interactions between people,institutions and technical elements through the meanings attached to them. Morespecifically,theresearchdealswiththedualismstructure–agencyinsocialchangeinthecontextofcottonproduction.Theanalysiswasbasedontwobroadertheoreticalapproaches: Actor‐Network Theory (see Chapter 3 in particular) to highlight thepivotalroleofcotton,andtheactor‐orientedapproach(seeChapters4,5,and6)thathas the human actors’ logic as its point of departure. Chapter 2 combines the twoapproaches.

H1

H3

H2

H3

Individuals’positioninthe

network

Individuals’stockofsocial

capital

Collectiveaction

Householdlivelihoods

Genderpowerrelations

H3

CHAPTER1

8

1.4.1 Structure,actorandagency

Theorizingabouttheconstitutionofsocietyandthereproductionofthesocialorderover timehas beendoneby opposing schools of thought such as functionalismandstructuralism. Their proponents argued about the relevance of the concept ofevolutionfortheanalysisofthestructureandfunctioningofsocialsystems(biologicalanalogies). Despite their differences, however, functionalism and structuralism con‐verged on the pre‐eminence of the social whole over its individual parts (Giddens,1984: 1). In his attempt to find out “how the concepts of action, meaning andsubjectivityshouldbespecifiedandhowtheymightrelatetonotionsofstructureandconstraint”, Giddens (1984: 2) formulated structuration theory. Early structuralfunctionalists “assumed that institutions and social roles mesh together to formpatterned and predictable social arrangements” (Hier, 2005: 7). Social actors wereconsidered to take advantage of themeans given in particular situations to achieveparticulargoals.ForParsons(2005)relationshipsareinteractiveandactorsituationscannot be understood scientifically if they are conceived independently of actors’orientations. To transcend the agent–structure or individual–society opposition,structurationtheoryemphasisestheinterpenetrationandtheinteractionaldynamicsofindividualandcollectiveaction,therebypointingouttheimportantroleofagency.Thisfocusunderliesboththeactor‐orientedapproachandactor‐networktheory,twocontemporary seminal contributions to the study of social action thatwere used inthisresearch.

Theactor‐orientedapproachdealswithbridgingtheindividualday‐to‐daylogicofaction (micro level) and the structural context (macro level) of the society. Theapproach entails “exploring the critical interfaces that depict the points of contra‐diction or discontinuity between the different (and often incompatible) actors’ lifeworlds, includingnot only ‘local’ actors but also ‘intervening’ institutional actors orotherstakeholders” (Long,2001:240).Therefore, itpositions thesocialactorat thecentreoftheaction.Long(2001:13)positsthat:

Social actors [...] must not be depicted as simply disembodied socialcategories(basedonclassorsomeotherclassificatorycriteria)orpassiverecipients of intervention, but as active participants who processinformation and strategize in their dealingswith various local actors aswellaswithoutsideinstitutionsandpersonnel.

Asocialactorcanbeanindividualperson,agroup,anorganisationoraninstitution.Inotherwords,anysocialentity“thatcanbesaidtohaveagencyinthat[it]possessesthe knowledgeability and capability to assess problematic situations and organise‘appropriate’ responses” (Long, 2001: 241). The properties of knowledgeability andcapability imply that a social actor is endowedwith the ability to exercise strategicchoices, thereby suggesting that the notion of agency is central to that of power.Agencyisabout“theknowledgeability,capabilityandsocialembeddednessassociatedwithactsofdoing(andreflecting) that impactuponorshapeone’sownandothers’

GENERALINTRODUCTION

9

actions and interpretations” (Long, 2001: 240). Agency can be conceived as theindividual’s or the organisation’s power to and power over other individuals ororganisations,and“representstheprocessesbywhichchoicesaremadeandputintoeffect”(Kabeer,2005:14).AppliedtothespecificcaseofcottonnetworkinginBenin,both individual cotton farmers – either leaders or grassroots farmers – and entirenetworkscouldstrategicallyinteractwitheachothertomakechoicesintheprocessofbuilding and rebuilding cotton networks. These actors are not foreign to thetumultuous cottonnetworking and its consequences for cottonproduction,which –instead–canbeseenastheoutcomeoftheirvariouschoices(seeChapters4and5).

Actor‐NetworkTheory (henceforthANT)originates fromtheneedforanewsocialtheorytoconnectwithscienceandtechnologystudies.Consideredcontroversialatitsinception, ANT has increasingly gained ground as a sociological approach used invarious social and human sciences (Alcadipani & Hassard, 2010; Law, 1992). ANTsurpassestheactor‐orientedapproachinitsanalysisofsocialinteractionsasitaimsat“accounting for the very essence of societies andnatures”, using someprinciples ofsocialnetworktheoryandfocussingonhumanrelationsinsocialandnaturalworlds(Latour, 1996: 369). In designing the theory, the endeavour was to bring togethernon‐humans (objects of science and technology) and humans, making the former“social‐compatible”too(Latour,2005:21).ThecentraltenetsofANTarethematerialheterogeneity of networks, the symmetry of networks that implies their continuity,and thenetworkontologywhich concerns theessenceof societiesandnatures.Theheterogeneityofnetworkssuggeststhat“society,organisations,agentsandmachinesare all effects generated in patterned networks of diverse (not simply human)materials”(Law,1992:380). ThecoreassumptionofANTisdrawnfromthatpropertyanditstatesthatorderingandinteractionareallessentialandimmutable.

TheoriginalityofANTliesingrantingaroletonon‐humanstoo,withnodifferenceinkindbetweenpeopleandobjects.Thecoreconceptof“actant”wascoinedtodesignatethisrole.Anactantisanythingwithoutfigurationthatmodifiesthestateofaffairsbymaking adifference (Latour, 2005). In otherwords, anythinghumanornon‐human(likecotton in thepresentresearch) thatcausesanactionwithnoneedofaspecialmotivation,neitherofindividualhumanactorsnorofhumansingeneral,isconsideredto bear the properties of an actant. Actants incorporate non‐human, material andtechnical elements into the analytical framework of the social world and not justindividuals (or institutions) with motives, in other words not just the “humanintentionalindividualactor”(Latour,1996:372).Therefore,ANTpresentsanewtypeof network configuration, and it emphasises the linkages between thematerial, thephenomenologicalandthesocialcomponentsofanysocialsituation.ImplicitinANTisthewarningthatoneshouldnotassumethat“materialandtechnicalelementshavenoactive role toplay in social dynamics” (Callon&Law,1997), andANT recommends“exploringsocialeffects,whatevertheirmaterialform,ifwewanttoanswerthe‘how’questions about structure, power and organization” (Law, 1992: 389). For ANT,neither humans nor objects have inherent qualities but they give presence to one

CHAPTER1

10

another through their networking relations, that is their “identity emerges – andchanges – in the course of interaction” (Callon & Law, 1997). Thus, ANT enables adisciplinelikegeographyforexample,ameansof“navigatingthosedualisms,suchasnature/society,action/structure,andlocal/global”(Murdoch,1998:357).

Both individual farmersand farmerorganisationsaresocialactors.Theyarealwaystemptedbyindividualprofitattheexpenseofotheractorsortheentirestructureofwhich they are part. Farmers’ deprival of resources consecutive tomismanagementresulted in tense situations between individual and collective rationality in theproduction and use of public goods. This suggests that the actant, cotton, and theactors,individualsandinstitutions,liaisewithoneanotherinthecottonsystem.Oneofthespecificobjectivesofthisstudyistofindoutthedifferentmeaningsthatwereattributedtocottonthroughtimeandwhatthismeantforeachoftheactors.Indeed,cottonlinksvariousinstitutions,namelycottonfarmerorganisations,inputsuppliersassociations,thesyndicateofcottontransporters,cottonginnersassociations,andthemultiple regulatorybodies.Howdid the systemcomprisedof institutions like thosementionedabove,people(individual farmersandleaders),andthematerialelementcottonbecomeonenetwork?

1.4.2 Furtherconceptualclarification

Investigatingthesocialtransformationsinvolvedbythedynamicsbetweenactorsandstructuresrequired theuseofcertainconcepts thatneedtobeclarifiedbeforehand.The process of building and rebuilding cotton networks reveals some connivancebetweennetworkleadersontheonehandandinputssuppliers,cottonginners,publicofficialsanddecisionmakersontheother.The‘illiterate’farmers(astheyareseenbytheseparties)refertodecisionmakersandotherliterateactorsas“intellectuals”.Bothfarmers and leaders claimed relationships with and support from officials anddecision makers in the creation of breakaway networks. Whereas most of thenetworksoriginatedfromeitheraninputssupplieroracottonginner,someemergedfromthesupportfromprivateactors,who,inturn,werereportedlyacquaintedwithpolitical leaders.Rivalnetworksandgroupsof farmerscompetedfornewmembers,thereby negatively affecting collective action. Therefore, it was posited that newleaders use their social connections to mobilize farmers in creating breakawaynetworks,whichbringsustotheconceptofsocialcapital.

Social capital is a special type of resource,which differs from financial, natural orhuman capital by its location in social relations. Its definition has varied over time.Theconceptanditsoperationalizationisdiverseandmultidimensional,buttwomainperspectivescanbeidentified,onelookingatitseffectsattheindividuallevelandtheotherateffectsatgrouplevel(Lin&Erickson,2008).Invariousdefinitions,trustandnetworksareperceivedasitstwokeycomponents,withafrequentmentionofnormsandreciprocity(Schulleretal.,2000).Thissharedunderstanding ledtoaconsensuson what social capital consists of. Thus, social capital is commonly defined as “the

GENERALINTRODUCTION

11

rules, norms, obligations, reciprocity, and trust embedded in social relations, socialstructures, and societies’ institutional arrangements” (Moser, 2010: 392). Trust,particularly,inducesfairbehaviourandleadspeopletoabidebylawsandregulations,entailing reciprocity. In that sense, trust is a key constituent of social capital thatfacilitates cooperation and makes democracy work (Putnam, 1993). High levels ofsocial capital in a given society or group increase the likelihoodof cooperation andfacilitate collective action (Ahn& Ostrom, 2008). Collective action is any concertedandjointlyconductedactionthatbenefitsanentirecommunitywithoutexclusion.

Lin (2001: 24) defines social capital as the “resources embedded in social relationsand social structure,which can bemobilizedwhen an actorwishes to increase thelikelihoodofsuccessinapurposiveaction.”Thisdefinitionguidedtheanalysisinthisthesis.Socialcapitaldependsonsocialnetworksandtheir interconnectionimpliesarealdifficultytodelineatethem(Lin,2008:58).Asocialnetworkisasocialstructureof relations between individuals or organisations tied by a specific type of inter‐dependencysuchasvaluesandnormsthatimplyreciprocity.While"networkanalysisfocusesontherelationsamongactors,andnotindividualactorsandtheirattributes"(Hanneman&Riddle,2005:4),socialcapitaladdressestheinvestmentofresourcesbyindividualsinsocialrelationswithexpectedreturns(Lin,2001).Thus,socialnetworkspave theway to social capital building and appear to bemore tangible than socialcapital,whichislocatedintherelationsamongpeople(Coleman,1988).

The interactions between social actors in the arena of cotton networking revealwomen as particularly disadvantaged. They appear to be pawns in the male‐dominated bargaining taking place in the context of the cotton system. The malehegemonicpositionisbasedongendermythsandstereotypes.Feminist“critiquesoforganizedmalepowerandtheorganisationalformsinwhichitwasexpressed”gavebirth to gender‐and‐organisations as a scholarly field of study (Acker, 2006). Theconceptofgenderisoftenusedasashortcuttorefertoonlywomen’sdisadvantagedpositioninsociety,whilemencanbeaffectedinthesameway.However,genderisadynamic concept that refers to the socially constructed and culturally embeddedcharacterofmen’sandwomen’s rolesandpositions in society. It is a “standardizedmarker” and one of the “most all‐embracing criteria of the social identity” in allsocieties (Giddens, 1984: 85). Gender often operates through the “unquestionedacceptance of power” (Kabeer, 2005: 14). Gender poses serious problems of powerrelations, specifically the imbalance of decision‐making power between men andwomen in society. Women’s interaction with their husbands at household level ismarked by gender inequalities. These inequalities also influence their interpersonalrelationswithmenoutsidethehousehold.Thepresentresearchalsoinvestigatedhowthe imbalance of power betweenmen andwomen affects women’s involvement incotton organisations on the one hand, and the power relations within householdsundergoing(economic)shocksandinneedoflivelihoodadaptationontheother.

CHAPTER1

12

When people’s means of living are profoundly disturbed, they need to inventalternativewaysofprovidingfortheirbasicneedsandobligations.Long(2001:241)notesthat:

Livelihoods are made up of practices by which individuals and groupsstrive to make a living, meet their consumption necessities, cope withadversities and uncertainties, engage with new opportunities, protectexisting or pursue new lifestyles and cultural identifications, and fulfiltheirsocialobligations.

The strategies of finding alternativesways can be glossed as livelihood adaptation.Livelihoodadaptationoftenconstitutesachallengeduetotheconstrainedcontextsinwhich it occurs. Davies and Hossain (1997: 5) defined livelihood adaptation as adynamic process by which choices made by individuals or groups “either enhanceexisting security andwealth or try to reduce vulnerability and poverty.” Using theconceptsofstructureandagencytoanalyseadaptivedecision‐makingprocesses,Carr(2008: 690) showed that “individual decisions take place in a social context that isbeyondthecontroloftheindividualdecision‐maker.”Thedysfunctioningofthecottonnetworkand theoverall cotton sector induced thedeclineof cottonproductionandweakened farmers’ livelihoods. The consecutive changes in cotton networking areobserved at three interrelated levels: individual, household and community. Thehouseholdisofparticularinterest,becauseitissimultaneouslythesiteofproductionandconsumptionand, consequently, constitutes theappropriateunitofobservationforlivelihoodanalysis(Niehof,2004).

1.5 Studydesignandmethodology

1.5.1 Studydesign

This researchwas conducted from January2009 toApril2011 in the fournorthernprovinces of Alibori, Borgou, Atacora and Donga (see Figure 1.2 above), whichtogetherroughlycoverfouragro‐ecologicalzones.Theareahostsmanyethnicgroups,whoallhaveagricultureastheirmaineconomicactivitywithcottonasthemajorcashcrop.More than80per cent of the cottonproduced in the country is sourced fromtheseprovincesandallthecottonnetworksconcentratedtheiractioninthisarea.

The researchwas set up in twomain phases. The first phase was exploratory andconsistedofaseriesoffocus‐groupdiscussionswithgroupsoffarmersintenvillages.The second phase combined in‐depth interviews with a survey among networkleaders and grassroots farmers using a semi‐structured questionnaire, which waselaboratedfromtheresultsofthefocus‐groupdiscussions.Nationalnetworkleadersfrom nine networkswere purposively selected and interviewed individually. Theseinterviewsweredonealloverthecountry,becausetheseleadersdonotalwayslivein

GENERALINTRODUCTION

13

their communities but in cities far from the farmers they represent. Farmerswereinterviewedinfivevillagesinthetwolargestcotton‐producingprovinces:BorgouandAlibori.

Duringtheexploratoryphase,an institutionalmappingat local levelwasconducted.Datawere collectedabout farmers’perceptionsof theirorganisations, the causesofconflictsandthemotivationofleadersinbreakingaway,usingacheck‐listofissuesasheadings.This firstphaseresulted inan insight intocotton farmersnetworkingandguidedthedevelopmentofthesurveytools.

The second phase was divided into two sub‐phases: interviews with leaders andinterviewswithgrassrootsfarmers.Attheleaders’level,theinvestigationfocussedonthediscourseusedbythenewleaderstocreatetheirnetworks,theevolutionoftheirpersonal social networks, and theirmanagement strategies. At the household level,data were collected on farmers’ socio‐economic characteristics, their motivation tojoin a network, factors influencing the decision to give up cotton production, theeffects of the decreasing profit from cotton production on farmers’ livelihoods andhow farmers and households coped with these effects. Data were concomitantlycollected at village level, where the focus was on conflicts within and betweennetworksandhowtheseaffectedinter‐personalrelationsamongcottonfarmersandcollectiveactionwithincotton‐producingcommunities.

Thesamplingcombinedapurposivesamplingwithrandomselection,andwasdoneinathree‐stepclusterprocess:firstchoosingthevillages,secondthenetworks,andthenfarmers.Thepowerofpurposivesamplingresidesintheselectionofinformation‐richcases for an in‐depth investigation of the phenomenon at study (Clark et al. 2008).Duringtheexploratoryphase,tenvillageswereselectedinthefourprovincesonthebasisofcriteriasuchastheextentofcottonproduction,thelocation,andthepresenceof at least one operating network in the village. The villages selectedwere: Perere,Sekere,andSinendeinBorgouprovince;Bagou,Sori,Kanderou,andWagouinAliboriprovince;BougouandFoumbeainDongaprovince;andDohinAtacoraprovince.Fivevillagescomprisingsixnetworkswereselectedfromthesevillagesforthefocus‐groupdiscussions:Bagou,Wagou,Kanderou,SekereandPerere.Theselectionwasbasedoncriterialikethepresenceofatleasttwonetworksinthevillage,theirleveloforgani‐sation,theextentofcottonproduction,andlocationandaccessibilityofthevillages.

Asfortheindividualinterviews,apurposivesampleofleaderswasfirstconstructed.Thethreetofourmostinfluentialboardmembersofselectednetworkswereselected:thepresident, thesecretary, thetreasury,andthecoordinatoror themanagerwhennecessary.Thus,33 leaders fromninecottonnetworkswere interviewed.Then,115grassroots farmerswererandomlyselected throughasystematicsampling fromthelistsoffarmergroupsmembersoftheselectednetworks.Duetothepooravailabilityofcompletememberlists,alternativedatabaseswereusedfortheselection,suchasinputdistributionlists,cottonpaymentlists,etc.Insomecases,thewomenonthelists

CHAPTER1

14

weresystematicallyselectedtoincreasetheproportionofwomeninthesampleandreduce the gender bias. The combination of the two selection methods yielded asampleof148cottonfarmersfortheindividualinterviews.

1.5.2 Methodsofdatacollectionandanalysis

Thefieldresearchadoptedamixedmodesapproach(Creswell&Clark,2008;Johnsonet al., 2007; Morse & Niehaus, 2009), combining the use of qualitative and/orquantitative methods to collect the data throughout the process. While theirrespectivemeritscancumulate foraccuracyand thescientificvalueof theresearch,their limitations can also add up to hinder the benefits deriving from theircombination.Thesolutionresides in thedesignandthe interpretationof theresults(Scrimshaw, 1990). The data collection comprised, in chronological order, thefollowingpartlyoverlappingandmutuallycross‐cuttingmethods:

Literature search.The literature searchwas done bymeans of a desk study usingqualitative document analysis on published and unpublished sources. An importantobjective of the literature search was the historical reconstruction of the cottonsystem in Benin, also in comparison with other West African cotton‐producingcountries.

Secondary data collection. This concerned published and unpublished statisticsfrom governmental and non‐governmental sources dealing with cotton production,and data from local cotton organisations. The secondary data fromwritten sourcesincludeddataoncottonproductionbydistrictandnetworkoverthelastdecades,theevolution of cotton networks in terms of membership and coverage of area, theevolutionoftheinputscreditallocationbydistrict,laws,ordersanddecreesthathavebeen regulating the cotton sector, the network‐related documents such as theregistrationnumber,otherreports,andsoon.

Focusgroupdiscussions.Thefocusgroupswereorganisedintenvillagesduringtheexploratory phase. The discussions consisted of interviews with groups of farmerswithnopre‐establishedrelationships,usingacheck‐listofcotton‐relatedissues.Thediscussions were held separately with male farmers and female farmers. This wasdone to facilitate women’s expression and reduce gender bias. Sometimes, thediscussions were organised one after the other, but more often simultaneously. Intotal, there were ten focus groups of men and nine groups of women (due to theunavailabilityofwomeninthetenthvillage).

Survey.Asurveywasconductedamongnetworksleadersandgrassrootsfarmersinfive villages. The surveys targeted the heads of households, either male or female,managingthecottonproductionintheirhouseholdunit.

Oral history, life history and in‐depth interview. These methods were usedthroughout the research process to unravel both the individuals’ and communities’

GENERALINTRODUCTION

15

relationship with cotton production. Narratives on important events and develop‐ments relating to cotton in the Borgou and Alibori provinces were collected andanalysed.Thelifehistorymethodwasappliedinparticulartounderstandwhathelpedwomenleadersachievetheirleadershippositions.

KeyInformants interview.These interviewedweredonethroughoutthe fieldworkwheneverakey informantwasdiscoveredamongextensionserviceagents,networkleaders or ordinary farmers. The key informantswere interviewednot only to giveinsight into issues of their expertise, but also to get their account of stories aboutwhichnowrittendocumentationwasavailable.

Table 1.1 presents the combination of data collection techniques throughout thefieldworkprocessinchronologicalorder.

Table1.1 Theprocessofdatacollection

Phases Periods Datacollected Methods&Techniques

Literaturesearch

Cross‐cuttingbutparticularlyduringMay2008–April2009

Bothqualitativeandquantitative:‐ InsightintoBenincottonsector‐ Statisticsaboutcottonproduction

‐ Deskstudies‐ QualitativeDocumentsAnalysis(QDA)

Exploratoryphase January2009

Qualitative:‐ Institutionalmappingofvillages‐ Insightintocottonproductionandfarmersnetworking

‐ Focus‐groupdiscussion‐ Observation

SurveyOctober2009–April2011

Quantitative:‐ Cottonproduction‐ Networkingtrajectories

‐ InterviewQuestionnaire

‐ Householdsurvey

In‐depthqualitativephase

February2010–April2011

Bothquantitative andqualitative:‐ Lifehistories‐ Networkingexperience

‐ Oralhistory,lifehistory,in‐depthinterview

‐ Observation

Keyinformantinterviews

Cross‐cuttingInsightintoallaspectsofcottonproductionandmanagement

‐ Interview‐ Observation

All interviews,with both individuals and groupswere recorded after obtaining theconsent of the individuals or groups concerned. The interviews were transcribedwhennecessary.

Texts(secondarydata)wereanalysedbyqualitativedocumentsanalysis(QDA),whichisatechniqueusedwhendealingwithunstructuredandunwieldydata.AccordingtoRitchie and Spencer (1994: 176), qualitative data analysis consists essentially of

CHAPTER1

16

“defining,categorizing, theorizing,explaining,exploringandmapping”ofmeaningfulelements among such a big and often messy amount of information. QDA enablesunderstanding internal structures, creating typologies, finding associations, seekingexplanations,anddevelopingnewideas,theoriesorstrategies.

Thequalitativedatacollectedduring the focus‐groupdiscussionsweresubmitted totextual analysis (Fairclough, 2003) to find out how social practices (mobilisingdiscourses and particular behaviour) were related to social structures (emergenceand multiplication of cotton networks). As for the quantitative data, they wereanalysedthroughdescriptivestatisticsandmultivariateanalysesusingSPSS.

1.5.3 Generaldescriptionoftheinstruments

The two basic instruments used to collect the data are the discussion guide for thefocus‐groupdiscussionsessions(seeAppendix1)andthesemi‐structuredquestion‐naire for the individual interviews (see Appendix 2). The emergence of breakawaynetworkswasakeyissueinboth.

Thediscussionguidestartedwiththe institutionalmappingof thevillage(i.e. listingtheorganisationsoperating in thevillage).Thereafter, the importanceof cottonandthe interactionsbetweencottonandnon‐cottonorganisationswerediscussed.Thenthe types of conflicts occurringwithin and between cotton farmers’ networkswereaddressed,includingtheimpactofinter‐andintra‐groupconflictsoncollectiveaction.Farmers’motivesincreatingabreakawaynetworkandfactorsthatdeterminedtheirdecisiontojoinorstaywithanetworkwerelisted.Thediscussionendedbyelicitingsuggestionsofreformactionstobeimplementedfortherevivalofcottonproduction.

Data from the focus‐group discussions was analysed using descriptive statistics toassess the frequency of different factors cited in villages. These factors werecategorised for the semi‐structuredquestionnaire.Thisquestionnairewas first sub‐mitted to network leaders and then to grassroots farmers during individual inter‐views,anddealtwithcottonproductionandhouseholdlivelihoodissues.Respondentswere visited at their place, either at home or on the farm. The order of questionsdepended on the respondent’s answers. This facilitated a fluent interview thatfollowed the reasoning of the respondent in order to obtain an in‐depth under‐standing of the logic behind the respondent’s actions and choices. The interviewswere carried outwhen respondentswere available, but some interviews had to bespreadovertwotothreedays.

The semi‐structured questionnaire consisted of several sections. The first one dealtwith the socio‐economic and socio‐demographic characteristics of the respondents,namely their levelof instruction, the compositionof theirhouseholds.Thenextonefocused on the respondents’ cotton network membership history, including thedescription (when applicable) of how the respondents got board positions. Therespondents’membership of other organisations and their eventual responsibilities

GENERALINTRODUCTION

17

were enlisted in that section too. Subsequently, the relations between therespondents’membership to other organisations and their cotton networkingwereexplored. The sections that followed concerned the push and pull factors toatomisation, the types of conflicts that occurred within the networks, and theinteractions between the multiple cotton networks. Questions in these sectionsincluded themotives for creating breakaway networks, ties that provide the bettersupport in creating networks, etc. Then, issues of collective action within cotton‐producingcommunitiesandtheirimpactonhouseholds’livelihoodwereaddressedinthenextsection.Asectionabouttherespondents’householdlivelihoodfollowedsuit.Questions in this section dealt with the household’s livelihood assets, how thehouseholdsfacedthedeclineofcottonincome,theevolutionofspouses’contributiontothehouseholdbudget,etc.Theseissueswereinvestigatedwiththemaintwocottonproducingperiodsasreference:duringtheprimetimeandafterthedecline.Thisleadtothesectionaboutthelikelihoodofthecottonsectortoreviveandtherespondents’viewsonsomesuggestedreformactions.Thevillage’scharacteristicswithregardtocottonproductionandcottonnetworking(whereapplicable)wereexploredinthelastsection.

Many of these variables were ranked according to farmers’ perception of theirimportance. For the rankingquestions, all variableswere listedbeforehandand therespondentswererequestedtorankthemfrommosttoleastimportant.Theliteraterespondentscouldreadtheoptionsinthequestionnaireandindicatetheirranking.

1.5.4 Ethicalissuesandmethodologicalchallenges

Onthefirstpageofthesemi‐structuredquestionnaireitwasstatedthattheresearchwasapurelyacademicandnon‐commercialactivity,andthat thedatawouldnotbedisclosedtoanythirdpartyapartfromthosedirectlyinvolvedintheresearch.Havingtoldtherespondentthis,theintervieweraskedpermissiontorecordtheinterview.Allnamesusedinthethesisarepseudonyms.

Most of the challenges faced during the data collection process were related tomeetingspecificinformantsanddiscerningrelevantandnon‐exaggeratedstatementsfromtheoften longdiscussionswehadwithsomerespondents.Gettinganappoint‐mentwith private actors other than farmers turned out to be difficult. Specifically,inputs suppliers and ginners were little inclined to make time for an interview.Respondents in thecompaniesweresobusy that theyrarelyreplied toourdemandforaninterview.Theonlyinputssupplierwhofinallyagreedtoaninterviewcategori‐callyrefusedtohaveitrecordedinspiteofourassuranceaboutitsconfidentialityanditsstrictlyacademicuse.Duringtheinterview,heturnedouttobeseriouslycriticalofhiscolleagues(inputssuppliers)anddecision‐makers.

Achallengeweregularly faced iswhatwecalled the trapof theknowledgeable.Wewereusuallyintroducedtofarmersbythelocalleadersoftheirnetworkstoeasethe

CHAPTER1

18

contact.Butmostofthemoftenwentbeyondtheirrole.Theyinterferedbyansweringquestions that were not meant for them and steering the answers of the famers,insteadoflimitingtheirroletoestablishingthecontactbetweentheintervieweeandtheinterviewer.Lessexperiencedinterviewersfallintothetrapoflettingthemselvesbe guided by these leaders during the interview, sometimes hardly succeeding instoppingthem,whichbiasedsomeinterviews.Someleadersdidthis intentionallytoshow off and exhibit their importance, thereby valorising their leadership to thefarmers.Duringtheresearch,wealsobenefitedfromfarmers’enthusiasmaboutbeinglistenedtoontheircoreproblem:cottonissues.Thisadvantagewasadouble‐edgedsword.Onthe one hand, farmers felt itwas time to voice their grievances andwere talkative.Consequently, they made statements that needed to be considered with care. Thisrequired some experience in interviewing, and some standing to challenge thosefarmers who sought to impress the interviewer. On the other hand, many farmersaddressedtheinvestigatorsasdecisionmakers.“Writedowninyournotebook:we,wehave enough of your cotton”, said a farmer to the researcher, warning to abandoncotton production if the ‘intellectuals’ [including the researcher] would not find asolutionfortheirproblems.Dealingwithsuchsituationsisdelicateandneedstacttopreventtheinterviewturningouttobepassionatebutwithlittlefocus.

Furthermore, the victims of conflictswe had to interviewweremore concerned byissues pending at court and their chances of being indemnified for their lostproperties,thanrespondingtotheinquiriesoftheresearch.Thus,avictimofconflictchallengedtheresearcher,askingattheoutsetoftheinterviewwhytheobjectivewasnot to find solution to their predicament resulting from the destruction of theirproperties. Another leaderwhowas interviewed in prison turned around the roles,transforming the interviewer into the interviewee and vice versa. He answeredquestionswithquestions, always trying todischargehimself although the interviewwasnotaboutwhathadlandedhiminjail.Additionally,hespokeloudlyintheprisonhalltoprovethathewasinnocentofwhathewasaccusedof.

Another difficulty was assessing the relative validity of different sources when theinformation was contradictory. This was a problem in most of the case‐studies ofconflictsinwhichactorswhowerepreviouslypartnershadseparatedandstatedtheircasebyusingstrongarguments,totheextentofdenyingthefundamentalthingsthatmade them collaborate in the past. In such caseswe resorted to third partieswhoknewbothpartieswellandthenwetriedtotriangulatetheinformation.

The last challenge that is worth mentioning is an ethical issue. Some respondentsrefusedtohave the interviewrecorded,despiteourassuranceofconfidentialityandanonymity. During the interview, some used lots of quotes of others to refer tosituationsandactorsordigressedatlengthtoavoidtalkingaboutcriticalissues.Butaftertheinterviewwasdeclaredfinishedandtherecorderturnedoff,theyvoluntarily

GENERALINTRODUCTION

19

addressed critical issues provided itwas off the record. The use of the informationcollected in such cases – information thatwe tried to transcribe right afterwe gotback home – poses a serious problem of ethics, since the person interviewed wasclearly against recording the statements. However, the richness of the informationcollectedinthiswaymadeitsusemandatory,thoughwithpreservinganonymity.

1.6 Outlineofthethesis

Thisthesiscomprisessevenchapters.Fiveofthemdealeachwithaparticularaspectoftheresearchproblem.Thesubsequentchaptersareorganisedasfollows.

Chapter2 is a historical overview of cotton production in Benin. It aims at under‐standingthedevelopmentofcottonproductionandhowfromaprofitableactivity itturnedintoanightmarefortheactors.Thechaptershowshowthemeaningofcottonfor farmersandtheStatehasshiftedoveryears,changing fromacolonialsymbol towhitegoldandthentopest.Subsequently,itanalysesthepolicyreformsimplementedto liberalise the sector.Given the shortcomingsof thesepolicy reforms, the chapterconcludes that what is actually going on in the sector are the logical outcomes offlawedpolicies.

Chapter3analyses theemergenceof thebreakawaynetworks incottonproductionthrough the history of cotton networking in Benin and investigates the drivers ofcotton networks disintegration. Using social capital theory and the social networkapproach, the chapter investigates how the breaking away from cotton farmersgroupsresultedinatomisednetworks.

Chapter4explorestheresourcesandconflictsgeneratedbycottonproduction.Usinggroupdynamics theory, the chapter first showshow cotton‐producing communitiesdrewresourcesfromcottonproductiontoavailthemselvesofbasicservicesthroughcollectiveaction.Thechapterusescasestudiestoinvestigatehowinterpersonalandintra‐organisational conflicts disrupted the social fabric in the cotton productionareas.Then,itlooksatconflictsthatoccurredincommunitiesandbetweenindividualsincottonproduction,andanalyseshowthedeclineofsocialcohesion ledtothedis‐ruptionandthefreezeofcollectiveaction.

Chapter5addressesthe issueofwomen’sunder‐representation incottonnetworksandtheconstraintsthefewwomenwhoreachedtheboardsoforganisationsfaced.Itinvestigates the gender roles in the cotton system, linking the exclusion of womenfromboardpositions – regardless ofwomen’s acknowledged contribution to cottonproduction–withgenderstereotypes.

Chapter6examineshowthecotton‐producinghouseholdsadaptedtoorcopedwiththedeclineofincomeconsecutivetothedecreaseofcottonproduction.Itappliesthehousehold livelihood approach to investigate whether women’s increased contri‐

CHAPTER1

20

bution to the household’s provision for daily needs changed the power relationswithinthehouseholdandwhatthemen’sperceptionsabouttheseissuesare.

Chapter7synthetises theoverall findingsbyanswering theresearchquestionsandverifyingthehypotheses.Thenitstressestheaddedvalueofthethesiswithregardtopolicyimplicationsforthecottonsector.

21

Chapter2

ChangingFarmers’PerceptionsinBenin’sCottonProduction:FromWhiteGoldtoPest

ABSTRACT: Industrial cotton production in Benin is intertwined withcolonialism. The pressure on Benin farmers to produce cotton afterIndependencein1960generatedresistanceinsomepartsof thecountry,whichresultedinriotsthatledcottontoacquireanegativeimage.Thanksto various incentives, however, cotton became a "white gold" during the1990swhen itprovidedup to80percentof the incomeofabout45percentof ruralhouseholds.Cottonproductionalsogeneratedan importantamountoffiscalrevenuesfortheState.Butfromthemid‐2000sonwards,incoherent national cotton policies caused farmers to being deprived oftheirincomes.Cottonproductionbecameproblematic,whichresultedinasharp decline of production. This chapter aims at understanding thedevelopmentofcottonproductioninBeninfromcolonialtomoderntimes,with a special focus on farmers’ perceptions. The chapter shows thatfarmersaremorerationalthanoftenassumedandthattheygrowthecropas long as it is a source of livelihood and food security. The results alsoindicatethatliberalisationintheagriculturalsectorcanbeharmfulratherthanbeneficialwhen theState fails toplayacoherent roleduringa shiftfromStatemonopolytoprivateinterest.Keywords:Cottonproduction,liberalization,farmers’perceptions,Benin.

AshortenedversionofthischapterwillbesubmittedtoTheAfricanStudiesReviewunderthesametitle,bytheauthorsMaboudouAlidou,G.,Huijzendveld,F.D.andNiehof,A.

CHAPTER2

22

2.1 Introduction

Becauseofcottonproduction,achildhasnofatheranymore!Becauseofcottonmoney,achildhasnofatheranymore!Achildhasgrownhiscottonandboughtamotorbikewithoutafather’scontribution!Achildhasgrownhiscottonandcelebratedhisweddingwithoutafather’scontribution!Becauseofcottonproduction,achildhasnomotheranymore!Becauseofcottonmoney,achildhasnomotheranymore!Achildhasgrownhiscottonandboughtamotorbikewithoutamother’scontribution!Achildhasgrownhiscottonandcelebratedhisweddingwithoutamother’scontribution!

This song sang by women during one of our focus‐group discussions sessions inBanikoara, the heartland of cotton production, showed the importance of cotton inrural Benin, and some cultural changes in parent‐child relationships. It is about ayoungman’sautonomy indecisionmakingvis‐à‐vishisparents;autonomy thatwasgained fromhis financial independence provided by cotton production. The song isbasedonthetruestoryofayoungboywhodidnotcareanymoreabouthisparents’views,becausebygrowingcottonhebecamecapableofavailinghimselfofwhatwassupposed to be provided by his parents, like the arrangement of hismarriage andmarital gifts. Songs like this and many proverbs and sayings could be heard innorthernBenin,whereruralhouseholdsdependheavilyoncotton,becausethecashfunctionbecamelargelyvestedonit.

FromIndependence in1990to theearly1990s,cottonproductionmultipliedby35,and quadrupled between 1990 and 2005 before dropping again to the 1990 level.Having a marginal status during the 1970s, cotton steadily grew in importance tobecome the backbone of Benin’ economy during the 1990s. Cotton dominatedagriculturalpolicies,beingthemaincashcropandacriticallivelihoodassetuntilthemid‐2000s.The cropwas thengrownbyover350,000 farmers representing35percentofthehouseholdscountry‐wide,anditcovered37percentofthetotalcultivatedarea(Gandonou,2005).Thesectordirectlyor indirectlyofferedasourceof livingtomillionofpeople(AIC,2006).Infact,thecottonindustrystillrepresents60percentofthe country’s industrial sectorwitheighteenginning factories, five textilemills, twofactories for cottonoil extracting, andone factorymanufacturing cotton formedicaluse (AIC, 2008). It accounted for an important share in the State’s revenues and infarm households’ cash earnings, providing up to 80 per cent of rural households’incomes,mainlyintheNorth.Intheperiod1995‐2000,cottonexportsaccountedforabout80percentofthecountry’stotalagriculturalexportsandgenerated25percentoffiscalrevenues(AfDB&OECD,2008),asaresultsofprofoundreforms.

From the early 1990s onwards, the cotton sector underwent many reforms toliberalise it and facilitate the transition from a "monopolistic and centrally

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

23

administered system to a more competitive one" (World Bank, 2003: 59). Thesereformsbroughtnewdynamicstothesectorandplayedakeyroleinmakingitmoreprofitable.Asaresultofallthepolicyactionsimplementedduringthe1990s,andforthatmatterthe1980s,Beninis"thecountrythathaswitnessedthemostimpressiveincreaseincottonacreageoverthelastthirtyyears",offeringtothecountrythemostsignificantjumpinagriculturaloutputs(WorldBank,2003:58).ThecottonsectorinBeninbecamewidelyconsideredasuccessstoryandregardedasapovertyreductionsector(WorldBank,2003).Butthereformshadsomedarksidesaswell. Infact, theliberalization of the sector and the subsequent competition between privatecompanies encouraged the multiplication of dissident organisations among actors(cottonfarmers,inputssuppliers,andcottonginners).Thisrenderedthemanagementofthesectormorecomplexandmadethereformsdifficulttocontrol.

Inaddition,subsidiestocottonfarmersinthedevelopedworldcombinedwiththesesomewhatinternalconstraintstoworsenthepredicamentofcottonfarmersinBenin.Indeed, cotton subsidies are deemed to be the most important trade‐distortinginterventions that depressed the prices on international market. For example, thelossescausedtoMaliandBeninbytheUnitedStates’cottonsubsidiesarehigherthantheaidtheyreceivefromthiscountry(Minot&Daniels,2005).InBenin,a12percentincreaseofpovertyincottonproducingareawaslinkedtothefallofcottonpricesin2001/2002 (UNDP, 2005). However, European subsidies to their handful of cottonfarmersare thehighestperkilogrammeof cottonproduced, andhavebeenpointedoutasmoredestructivetocottonproducersfromfrancophoneAfrica(Cattaneo,2003;Fok,2010;Seuret,2009).Removingthemfromcottonproduction,mainlythosefromtheUnitedStates, still remainsapointofdisputebetween thedevelopedworldanddevelopingcountries(Alstonetal.,2005;Badianeetal.,2002;Minot&Daniels,2005;Pesche&Nubukpo,2004).Thesesubsidiesfurtherburdenedthecottonsector,whichwastransformedintoanightmareforfarmers,leadingtoabandoningtheproductionordevisingadaptationstrategies.

Itshouldbenotedthatcottonproductionwasunequallydistributedoverawiderangeofsoilsthroughoutthecountry.Cottonwasextensivelycroppedintenoutoftwelveprovinces,buttheproductionwasconcentratedintheNorth.Theregionoverlappingthelimitsbetweenthreeofthenorthernprovinces,Alibori,Borgou,andAtacoraformswhatisknownasthecottonbelt,accountingformorethan90percentofthenationalproduction. The central and the southern provinces, which are under moredemographicpressureandhaveamorehumidclimate,contributedtolessthan10percent of the national production during the 1990s (Gergely, 2009: 10), and has nowmovedfromaverylowtoamarginalproduction.

TheaimofthechapteristodescribethedevelopmentofcottonproductioninBeninfromcolonialtomoderntimes,withafocusonthefarmers’perspective.Thechapteris organized into eight sections. After describing the data sources and analysis inSection2.2,thesectionthatfollowspresentsabriefhistoryofcottonproductioninthe

CHAPTER2

24

WestAfricancontext.Section2.4browsesinstancesofresistanceafterIndependence,with occasional references to the situation in colonial times and in other Africancotton‐producingcountries.Subsequently,theoutcomesofthereformsimplementedin Benin are discussed, with reference to neighbouring countries, such asMali andBurkinaFaso. Section2.6examines farmers’ responses to the laterdevelopments incottonproductionthatmadethereformsfailingtoliveuptotheexpectations.Sections2.7dealswithcottonfarmers’perceptionsoftheprospectsofthesector,followedbytheconclusioninSection2.8.

2.2 Datasourcesandanalysis

ThepresentchapterisahistoriographicalresearchofcottonproductioninBeninfromthe early 1900s to 2010. The chapter addresses the question to what extent theliberalisationpolicy implemented in thecottonsectorat thebeginningof the1990scanexplainthecurrentsituationofthesectorinBenin.Itisbasedonaresearchthatcombined an extensive literature review with fieldwork. The data set compriseswritten records, including historical articles about cotton production before, duringand after colonial times, complemented with empirical data about farmers’perceptions on various aspects of cotton production. The research was conductedfrom 2009 to 2011. It used both quantitative and qualitative methods to gatherprimarydataaswellassecondarydata.Theresearchareaspannedthefourprovincesof northern Benin, covering the driest agro‐ecological zones of the country thatconstitute the most suitable areas for cotton production. The research focussedfurtherontheareaknownasthecottonbelt(Figure2.1).

Thefocusgroupdiscussionswereorganisedintenvillages inthreeprovinces,alongwithkeyinformantinterviews.Asurveywasconductedamong148individualfarmersin five villages, covering nine cotton organisations. The sample included grassrootsfarmersand leaders.Forabetterunderstandingof the socio‐politicaldimensionsofcottonproduction,aspecificconflictinthepastwasexplored.Forthecasestudy,keyinformantswereinterviewed.Thus,fiveeye‐witnessesofandactorsintheconflictinthevillageofPerereineasternBorgouandthechiefcommanderofthearmyinthatarea. In spite of its importance and the extent towhich it affected the actors, therewerenowrittenrecordsonthecase.Hence,thestatementsoftheintervieweescouldnotbecross‐checked,buttherewasnoreasontodoubttheplausibilityoftheirstories.Togetabettergraspoftheimplicationsofthereforms,privateactorsininputssupplychainwereinterviewedaswell.

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

25

Figure2.1 Locationofthecottonbelt

Perereisthecity‐centreofthedistrictwiththesamename.ThedistrictisoneoftheeightdistrictsofBorgouprovince, locatedinnorth‐eastofBenin,at theSouthofthepresent cotton belt. The district has a population of about 62,391 inhabitants andstretches up to 2,150km2 (CommunedePèrèrè, 2010). The city centre itself has apopulation of 14,559 including the surrounding villages. Perere’s performance incottonproductionisamongthe lowest intheprovince.Itrangesbetween3,000and4,000tonsayearandcontributesforabout10percentofthedistrict’srevenues,farbehindthecontributionofrootsandtubers(66%)andcereals(14%).Thedistricthasalsoregisteredoneofthehighestratesoffarmersabandoningcottonproduction.

CHAPTER2

26

Thequalitativedatawereanalysedthroughaninterpretativelens.Log‐linearanalysisandt‐testwereusedtoanalysethequantitativedata.

2.3 HistoricaloverviewofcottonproductionandmarketinginWestAfrica

2.3.1 CottonproductionandEuropeansettlers

Industrial cotton production in Sub‐Saharan Africa is linked with colonialism (seeBassett, 2001; Isaacman & Roberts, 1995; Levrat, 2008; Moseley & Gray, 2008;Roberts, 1996), though cotton had been grown and used in handicraft in Africa forhundreds of years, before the contactwith Europeans and the existence ofwrittenrecords (Metcalf, 1987). However, many authors converge on the difficulty ofreconstructing this old African history of cotton, in spite of evidence provided byvariousdisciplinesusingdifferentmaterialsintheirattempttotraceit(Kriger,2005).

Investigations by botanists and archaeologists inWest Africa, for example, suggestthat the crop had beenwidely grown there for its fibre‐bearing properties (Kriger,2005). Handicraft comprising a wide variety of cotton related activities – fromspinning,weaving, dying, tailing and embroidery – developed inWestAfrica beforethebeginningoftheAtlanticslavetrade.Duringtheearlysixteenth‐century,travellerspraisedthehighqualityofMandingaand"SlaveCoast"cottoncloththatwasfoundallalong thewest coast ofWestAfrica (Thornton, 1990). Furthermore, as reported byearly European voyagers and explorers to Africa, distinct forms of cotton, rangingfromrawfibretohand‐woventextiles,wereamongthegoodsthatwerebarteredintheAkanmarketsduringthe1770s(Metcalf,1987).

Before the colonization, Europeans had made a number of ‘ill‐fated’ attempts toproduce plantation‐grown cotton in the West African coastal zones by tradingcompanies, individual merchants and governments (Ratcliffe, 1982). The firstattemptsweremadebythePortugueseandBritish.Asearlyas1500,thePortugueseimportedcottonseedsfromAmericaandtriedtodevelopitscultivationontheislandsof Cap‐Verde and on the Africanmainland, exchanging garments for slaves (Maier,1995).TheBritishwere the first to introduce large‐scale cottonproduction in theircolonies,making theUnitedStates the largestproducerworldwide.Both theFrenchand the British experimented with cotton production in Ouidah (an importanthistorical slave trade fort in Benin) during the first half of the nineteenth century."CottonmustbetoAfricaitsgold‐diggings",saidBuxtonin1852,whowasoneoftheinfluentialproponentsoftheplantosuppresstheslavetrade.JohnDuncan,theBritishVice‐Consul at Ouidah, backed the plan to cultivate cotton in Dahomey. The cottonschemefailedforlackof"familiaritywithtropicalagriculture,Africanfarmingsystemsingeneralorcottoncultivation inparticular" (Ratcliffe,1982:98).Furthermore, theAfrican Aid Society, founded by an American organisation convinced of the cotton‐

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

27

exportingpotentialofWestAfrica, "directed itsmajorefforts toweaning thekingofDahomey from the slave trade by persuading him to grow cotton in its stead"(Ratcliffe,1982:95).Theexperiencegainedfromthesepreviousattemptstopromotecotton production guided the colonial cotton policy that impacted contemporarycottonfarminginWestAfrica.

Nowadays,cottonhasbecomeacommodity“thatissituatedinaninternationalwebofeconomic transfers reflecting historical and contemporary power structures”(Moseley&Gray,2008:3).Thecropnotonlylinksfarmerstotheinternationalmarketbut also contributes to African States having a voice in international negotiations.Cotton is an important African product in the globalized arena. It is featured ininternationaldiscoursesanddebatesonprivatisation,povertyalleviation,agriculturalsubsidies,andsustainabledevelopment.Butunlikegoldoroil,cottongrowsfromtheefforts and sweat of millions of small‐scale farmers, with households and entirecommunitiesdependingon it.Additionally, cottonhas internallygivenpower to thevoiceless and allowed them access to the decision‐making process (Roy, 2010).Externally, it has provided African leaders with a clear agenda on World TradeOrganization rounds of negotiations, showing them as "the heralds of the battleagainstpoverty"(Pesche&Nubukpo,2004).Farfrombeingjustarawmaterial,cottonacts as the medium that links multiple stakeholders from national to internationallevel.Unfortunately,as inmostWestAfricancountries, thecottonsector inBenin isburdened with internal contradictions and conflicts, which make the countrypowerlessintheinternationalmarkets.

2.3.2CottonproductionfromtheColonyofDahomeytotheRepublicofBenin

TheFrenchpolicytodevelopcottoncultivationinDahomeyforexportstartedin1905withsixfieldsofanewvarietycultivatedinthecentreofthecolony(Manning,1982).The results improved throughout the following years, thanks to an increasedknowledgeofprevailingagro‐ecological conditions, therequisitionof labourers,andthe financial incentives given to chiefs to encourage them to promote cottonproduction among local farmers. Notwithstanding the ambiguous success, theexperimentscontinuedandprogressivelyspreadsouthwards,duelargelytothehighdemand for the new variety by local chiefs (Manning, 1982). This led to theconstructionofthefirstcottonmillin1908.TheexporttoFranceincreasedgraduallyfromnine tonsof fibre in1905 to130 tons in1909. From1910 to1924,DahomeyremainedthemostpromisingFrenchWestAfricancolonyforcottonproductionandexport.OnlyTogo,oneofthetrustterritoriesadministeredbyFranceafterWorldWarI, produced more cotton than Dahomey, thanks to the infrastructures left by theGermans(Levrat,2008).Figure2.2comparesthetrendsofcottonexports toFrancefromitscoloniesfrom1910to1930,basedonthesparsestatisticsavailable.

CHAPTER2

28

Figure2.2 EvolutionoffibreexportfromFrenchWestAfricancoloniestoFrance(1910–1930).

Source:CompilationbyauthorsofdatafromBassett(2001);IsaacmanandRoberts(1995);Levrat(2008);Roberts(1996).

ThisrelativesuccessofcottonproductioninDahomeywastheresultofthecolony’sgood quality of the fibre, a dense network of roads, and the "goodwill of farmers"(Adekunle, 1994; Kuba, 2000; Levrat, 2008: 83). The production started to declineafter the creation in 1932 of the Office du Niger in charge of irrigated cottonproduction inSudan(present‐dayMali) (Roberts,1996). Inaddition,moreattentionwasgiventoCôted’Ivoire,wheretheintroductionofanewvarietyandtheincreasedadministrative coercion resulted inmany peaks in cotton exports during the 1930s(Bassett,2001).

The 1929 economic crisis and World War II, however, seriously weakened thegrowingFrenchtextileindustry.TheFrenchtextileindustrysufferedfromanerraticandchaoticsupplyofcottonduringthewar.Until1943,theproductionfromtheWestAfrican colonieswas exported to France for the Vichy Government. After 1943 thecotton supply from the French colonies of Central Africawas diverted to the AllieswhenthecoloniesjoinedtheFreeFranceGovernment(Levrat,2008).IntheaftermathofWorldWarII,anewschemeofcottondevelopmentwasimplementedintheFrenchAfrican colonies. The implementation of the scheme started with the foundation

0

250

500

750

1000

1250

1500

1750

2000

2250

1910 1912 1914 1916 1918 1920 1922 1924 1926 1928 1930

Cotton Export (Tons)

Year

Dahomey

Soudan

Côte d’Ivoire

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

29

between1946and1951ofaresearchinstituteandatextilecompany1.Thegoalofthenew institutions was to reduce the dependency of the French textile industry onsupplies from the United States (Levrat, 2008). Cotton development became theconcern of colonial administrators who, in tandem with cotton companies andresearch institutes tried tomake ita success. InDahomey, thereorganisationof theproductionstartedcuriouslyintheSouthwheretheproductionprovedtobemarginalbecause of inadequate agro‐ecological conditions that resulted in low yields.Apparently,therelativesuccessoftheearly1900swasnotcapitalisedbythecolonialauthoritiesuntilIndependencein1960,whencottonproductionwasextendedtotheNorth by the Compagnie Française pour le Développement des Textiles (CFDT). TheFrenchcottonparastatal ledthecottonsectorinDahomeyuntilthebeginningofthe1970swhenitlostthecontrolintheaftermathsofthemilitarytakeoverin1972.ThepoliticalelitehadaMarxist‐Leninistideologicalorientationandrejectedallthingsthatwere considered colonial symbols, including cotton. Hence, cottonwas left withoutanyincentive.Thecountrywitnessedasteepdropinproductionthatplummetedfrom50,000tonsin1972tolessthan2,000in1978.

Analysing the dynamics of national cotton policies during the years 1895‐2008through the path dependency theory, Kpade and Boinon (2011) identified fourperiodsbasedonthe"criticaljunctures"intheprocess,definedbyMahoney(2001)asthe choice points when a particular option is adopted from among two or morealternatives. These critical junctures are the results of previous policies, includinginternational policies as the present situation of cotton in Benin illustrates. Fiveperiods can be characterised in the development of cotton from its industrialcultivation(Table2.1): i) colonial times, ii) fromIndependence in1960 to1972, iii)the revolutionary period, iv) from the 1990s up to themid‐2000s, and v) from themid‐2000sup tonow.Thedistinction isbasedon thestatusattributed tocottonbyboth farmersandpolicymakers, the incentives for theproduction,andtheresultingproduction.

1TheInstitutdeRechercheduCotonetdesFibreTextiles(IRCT)andtheCompagnieFrançaisepour leDéveloppement des Textiles, founded respectively in 1946 and 1951 to replace theUnion Cotonnière de l’Empire Français, created in 1941 on the ashes of the AssociationCotonnièreFrançaise.

CHAPTER2

30

Table2.1 HistoricalphasesofcottondevelopmentinDahomeyandBenin

PhaseColonialTimes(1895–60)

AftermathofIndependence(1960–72)

RevolutionaryRegime(1972–90):Cottonasasymbolofcolonialism

Primetime(1990–05):Cottonasawhitegold

Thedecline(2005–Now):Cottonasapest

Policies

Developmentofindustrialcottonproduction,basedon:- Experiencefromthemid‐1800s,

- Dahomey’spotentialandcomparativeadvantages:goodagro‐ecologicalconditions,densenetworkofroads,goodwilloffarmers,

- Useofcoercion.

ColonialpolicycontinuedbyFrenchresearchinstitutesandcottonparastatal.- CreationofthefirstfarmerorganisationsinformofcooperativeforcottonintheNorth.

- Firstresistancestocottonproduction.AriotputoffbythearmyintheNorth.

- TheMarxist‐Leninistregimerejectedcottonasasymbolofcolonialism,

- Laternationalisationofproductionstructures,

- Creationin1975ofanetworkofagriculturalextensionheadquartersinthesixprovinces,

- Foundationin1982oftheSociétéNationalepourlaPromotionAgricole(SONAPRA),acorporationforagriculturalpromotion,

- Over‐productionin1986andfirstcottoncrisis.

Productionandmanagementsystemsimproved:- Adventofdemocracyandpoliticalpluralism,

- Completeliberalizationofthesector

- CFAdevaluation,- Empowermentoffarmerorganisationswiththebuildingofanhierarchicalnetwork,

- Cottonshapedagriculturalpoliciesandtransformedintoapoliticalinstrument,

- Stiffcompetitionbetweenprivateactors.

Profoundcrisisandscrambleofcottonproduction:- Over‐regulationandover‐politicizationofthesector

- Riseofconflictsandatomisationoffarmers’networks,

- Shifttoprivatemonopoly,- Highrateofcottonabandonmentandshrinkingofproductionarea.

Production&Pricetofarmers

- Productiongrewfrom135tonsin1914to3,000tonsin1960.

- CottonboughtatFF0.5/Kg(FrenchFranc)until1908,and20FCFA/Kg(US$0.042)in1960.

- Productionsteadilyincreasedfrom3,000tonsin1961to47,000tonsin1972.

- ThepricespassedfromFCFA20/Kg(US$0.004)in1961toFCFA28(US$0.0056)in1972.

- Yieldclimbedupto0.9ton/ha,thankstotheuseofsubsidizedfertilizer.

- Productionfluctuatedbeforestabilisingattheendofthe1980s: Sharpdeclineto1,400tonsin

1978; Jumpto88,000tonsin1985,

andpicto133,000tonsin1987tocreatethefirstcottoncrisisbyover‐production.

- Yieldsincreasedfrom0.9to1.1ton/Ha.

- Pricesdoubledwithintwodecades.

- Productionquadrupledfrom105,000tonsin1990to427,000tonsin2005.

- Yieldstayedaround1ton/Ha.- PricesdoubledduetoCFAdevaluationfromFCFA100/Kgin1991toFCFA225/Kgin1999(US$0.2‐0.425).

- Over‐ginningcapacity:600,000tons#350,000tonsofproduction.

- Productiondeclinedfrom427,000tonsto158,000tonsin2010.

- Yielddecreasedwithbigcontrastamongzones.

- PricesmoreorlessstablearoundFCFA210/Kg.

Source:AdaptedfromAIC(2008).2 FrancdelaCommunautéFinancièred’Afrique(WestAfrica)orFrancdelaCoopérationFinancièreenAfriquecentrale(CentralAfrica),commoncurrencyto14Frenchformercolonies:€1=FCFA655.957(standardexchangerate).

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

31

2.4 Resistancetocottonproduction:Fromcolonialtopresenttimes

2.4.1 Forcedworkasacommonlabourpolicyincolonialtimes

AsBeckert(2005:506)pointedout,"[...]cottonandimperialexpansionwenthandinhand from the beginning, not only for Germany but also for France, Russia, GreatBritain,andevenPortugal,Belgium,andItaly."Theworldwidecottoncrisesin1861‐1865and1901‐1903causedasevereshortageofcottonsupplies,resultinginsoaringprices that threatened thesurvivalof theEuropean textile industry.Thecrisesshedlightontheindustries’criticaldependenceonsuppliesfromtheUnitedStates,whichspurredthemtolookforalternativesourcesofsupply(Onyeiwu,2000;Roberts,1996;Sunseri, 2001). Cotton came tobe seen as a strategic rawmaterial during the earlytwentiethcentury.Thequestforcottonthusbecameanoverridingconcernofcolonialpowers in Africa. The colonial regimes progressively made cotton productionobligatory throughout their territories, with diverse results. In Mozambique, thePortugueseputlocalchiefsinanawkwardpositionbydemandingthemtooverseetheimposition of cotton production and, at the same time,maintain social orderwhenfarmersprotested (Isaacman,1992).TheBritishenforced the “CottonOrdinance” inmostof theircolonies,startingwithUganda in1908(Onyeiwu,2000).Thepolicyofforcedcommunallabourintheproductionofcottonwithanegligiblecompensationtofarmers in 1905, triggered the "Maji‐Maji"3 rebellion in German East Africa, whichcausedhundredsofthousandsofcasualties(Iliffe,1967;Sunseri,2001).TheFrenchschemetoestablishcottonexportsfromAfricastartedintheearly1800sin Senegalwhere it failed.They renewed the schemeduring the1860s, buthad thesame disappointing result (Levrat, 2008; Roberts, 1996). In the late 1890s, the"politiquecotonnière"wasextendedtotheremainingcoloniesofWestAfrica,thistimewithmoreengagement.Francebattledunsuccessfully toestablishcottonproductioninthesecolonies,fromwhereitcouldbarelymeetonetothreepercentofitstextileindustry’sneedsofcottonfibre(Moseley,2002).Amongtheobstaclesinthedevelop‐mentofexport‐orientedcottonproductioninWestAfricancolonieswereignoranceoftropical agriculture and African farming systems, the demand of longstanding anddynamicslocalmarketsincotton‐relatedproducts,andfarmers’disdainfortheexportmarket(Bassett,2001;Roberts,1996).Theobstaclesconstitutedbythelocalmarketsturnedouttobethemostchallengingconstraintsforthecolonialadministrators.Thevitalityoflocalmarkets,wherecottoncouldalwaysmeetthedemandandwassoldathigherprices,madethesupplyforthemetropolitanindustryveryerratic.Todealwiththeseconstraints,twoopposingviewsprevailedamongtheFrenchmaincottonactors.One campwas in favourofpromoting the age‐old traditional cotton cultivationand3Maji‐Majireferstoa'watermedicine'giventorebels,whichwassupposedtohavemagicalpowersthatwouldprotectthemagainstbullits.

CHAPTER2

32

using market incentives to benefit from local dynamics, while the other campadvocatedfortheimpositionoftheindustrialisedcottonproduction,usingirrigationfor a rapiddevelopment and growthof theproduction. This divide iswhatRobertscalls the "two worlds of cotton" (Roberts, 1996). The internal contradictionspreventedcottonactorsfromeffectivelyjoiningtheireffortstoimplementanypolicy,except the use of coercion. Cotton became a forced‐crop4 in all French colonies.Labourwas not only forced but also unpaid or underpaid by the colonial power toincrease theeconomicreturns.The"stickofcolonialismcame tobeas importantasthe carrot of the world market" (Manning, 1982: 156). A cotton cultivationenforcement scheme determined the quota of acres to be cultivated inUpperVolta(Levrat, 2008)5. In Western Sudan, the colonial cotton policy led to riots amongfarmerswhorefusedtoplantthecroportohandovertheirproductiontothecolonialadministration (Roberts, 1996). In Côte d’Ivoire, "repeated attempts to intensifycotton cultivation were never realized in a sustained and satisfactory manner."(Bassett,2001:51‐52).

InthecolonyofDahomey,farmersweremorereceptivetocottonproductionthaninmostoftheWestAfricancolonies,wheretherewasadirectresistance.Theyweresaidto be full of goodwill for cotton production (Levrat, 2008), in spite of the series ofresistance actions6 against the colonial power throughout the colony, which led aGovernor toqualify the situationasDahomean imbroglio.Yet,mostof these revoltswere reactions against the social and economic policies, namely the forcedmilitaryrecruitment, the forced work on roads, the collection of individual taxes, and thetransgressionof customs (Garcia, 1970;Manning,1982).The introductionof cottonstartedstrategicallyfromthemorereceptiveareasintheSouth,whichcorrespondedto the territory of the pre‐colonial kingdom of Dahomey before it was dismantled.Agriculture was there so organised at that time that there existed a minister ofagriculture (Manning, 1982). Nevertheless, local farmers were subjected to forcedlabourthatwaspartofthecottonpolicy.Allthesecontributedtotherelativesuccessofcottondevelopment.

In general, the early colonial timeswere characterised by exports highly correlatedwiththe levelofcoercionexertedbycolonialofficialson farmers like in1930,1938and 1942 in Côte d’Ivoire. The use of coercion, the mode of payment, and themanagement of local markets were continuously debated. The questions anddilemmas centred on cotton were simultaneously about politics, administration,economics,cultureandeverydaysocialrelations(Bassett,2001).

4 “La culturedu commandant” as the cotton cultivationwas named in FrenchWest Africancolonies.5About"lechampcollectifobligatoiredecotonàsuperficieproportionnelleàl’effectifdémogra‐phique(4hapour100habitants)"inHauteVolta. 6TheFrenchfacedafierceresistancefromthe“Holli”people(Southeast),the“Kaba”’sinsur‐rectioninAtacora(Northwest),andtherevoltof“BioGuera”inBorgou(Northeast)until1917.

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

33

2.4.2 Post‐IndependenceresistanceinBenin

After Independence in 1960, the colonial policy of cotton production in Dahomeycontinued under the control of the French CFDT and SATEC. The production wasperformedoncooperativeblocs,rarelyonindividualplots,withsubsidizedfertiliserstopromoteexportthatcouldboosttheState’srevenues.Butforcottongrowers,noorfewimprovementscouldbediscernedregardingthefairnessof theredistributionofcotton revenues. In areas such as Perere, cotton incomes given to chiefswould notalwaysreachtheproducers.Atthattime,hardlyanyonewouldhavecontestedtheuseof the yields by local chiefs. Thanks to Independence, famers started voicing theirgrievanceagainstcottonproductioninMarch1963atPerere,leadingtothefirstopenresistanceagainstcottonproductioninthefledglingRepublicofDahomey.Farmersinthe village refused to grow cotton anymore, braving the warning and threat fromauthorities.Weeksofnegotiationstobringthepopulationtochangeitsminddidnotresult in compliance. Meanwhile, the people prepared themselves in advance torespondtoanymilitaryintervention.Theirapprehensionofamilitaryinterventiontosettle the revoltwas confirmed.Anarmed interventionoccurred lateat thedeadofnight when the city was asleep. A contingent of gendarmes sent to put down theresistancestartedsecretlyarrestingtheleaders.Riotssoonbrokeoutaftertheleadersreactedpromptlybytrumpetingtowakeupthepopulation.Protestersscurriedofftofight,withtraditionalweaponsandtoolsonhand,suchaspoisonousarrows,swords,andmachetes.The confrontationwent fromdawn toduskwith several arrests.Theattempt from villagers to free their arrested peers resulted in two people shot andseriously injured. The confrontation was finally ended by the intercession of theMinisterof Interior,ahalfnative fromthevillage,whowarnedtheprotestersof theriskofopposingtheState.ButinPererethatyear,nocottonplantgrew.

The military attached low importance to the revolt and qualified it as a "simpleoperationtomaintainsocialorder",accordingtothearmycommander.Conversely,thepeople of Perere refer to it as "the cotton war". The event is committed to therepertoire of collectivememory. First, because people consider the resistancewithbarehandsandtraditionalweaponsagainstarmedforcesadaringachievement.Theywithstoodaheavilyarmedcontingentofgendarmesanddeterredthemfromarrestingmorefarmers.Second,asoneofthefirstactsofresistanceagainstthenewauthoritiesafter the colonial regime ended, it raised the people’s awareness of the nature ofpolitical power whoever exercises it. The riots resulted in the migration of someleadersandyoungcontesterstoNigeria, forfearofarrest.Somestayedabroadforashorttomiddletermandothersneverreturnedatall.

For themajority of peoplewho stayed in the village, the revolt and its subsequentrepressionworsenedthereputationofcotton.Bebawasoneofthosepeople.Hewasalso among the leaders who were arrested. Deeply embittered, Beba framed hisfeelinginastatementthatisoftenquoted:”Ifyouseeaplantofcottononmyfarm,theseedhasbeenlaidbyabird,ifbirdseatcotton.”Nobodycouldexpectsuchafarmerto

CHAPTER2

34

cultivate cotton anymore, knowing that birds do not peck at cotton. Cotton wasprogressivelyabandonedinthevillage, infavouroffoodcrops.Itwasgrownonlyinmoreruralneighbouringvillagesuntil theboomoccurredat theendof1980s.Evenso, farmers in Perere remained hesitant about the crop at the outset of the boom,reinforced in their scepticism by some farmers’ disillusionment with the results oftheir cotton season. Indeed, during the early boom, unsuccessful farmers could bedeprivedoftheirbicyclesorotherbelongingswhenthesewereseizedtocompensatefortheirinputsdebts.FarmersfromPererecaughtuplateronduringthe1990s,whencotton regained undeniable profitability. Even the leaders of the former protestmovementstartedgrowingcottonagain,includingBeba.WhenBebawasremindedofhisformerstatement,herepliedthat“thecitymoves,theworddoessoevenmore.”Thestatement is inspired by the Baatombu’s (the autochthonous and dominant ethnicgroup in the region) tradition of moving entire settlements after being hit by anepidemicoradisaster,toexorciseandavertbadfate.Bebaexpressedthelogicofthetypicalcottonfarmer,whochangeshisrhetoricofcottonproductionaccordingtoitsprofitability.

2.5 Theprimetimeofcotton:1990–2005

2.5.1Thecottonsectorreforms

The end of the 1980s witnessed the inefficiency of State interventions in differenteconomicdomains.ThisinefficiencyresultedfromtheinvolvementoftheStateinthemanagement of the sector through costly bodies that overwhelmed the sectorwithdebts. In Benin, the sector functioned as a single channel, with the State companyexertingafullmonopolyuntil2000.TheStatecontrolledtheinputssupply,providedextension services, and controlled the ginneries through the State parastatalSONAPRA(Figure2.3).

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

35

Figure2.3 Frameworkoforganisationallinkagesincottonproductionbefore2000(AdaptedfromCIPB(2008)).

Policyreformstoprivatisethecottonsectorbecamenecessary.Areformisdefinedasa"consciouslychosenchangeinthefundamentalorganizationofasectorandrelatedchangesinthe'rulesofthegame'underwhichstakeholdersoperate"(Tschirly,2009:8). As such, eastern and southern African cotton producing countries underwentreformsintheearly1990sbyeithereliminatingthesinglechannelorbyprivatizingcooperatives of ginners.West and Central Africa countries engaged in that processslowly for subsidies and related‐market distortions reasons (Gergely & Poulton,2009). The issue of reforming the Francophone African cotton sector was debatedbetweentheWorldBankontheonehandandtheFrenchAgencyforDevelopmentandCooperationandtheCFDTontheotherhand.WhiletheWorldBankadvocatedforacomplete liberalizationof thecottonsector tomake itmorecompetitive, theFrenchinstitutionslobbiedforthe"filièreintégrée",thesingle‐strandedrelationshipbetweenproducersandcottoncompanies,regulatedbyStateintervention(Hugon,2005).Thelatter option guarantees sustainable inputs credits but also entails heavy operatingcostsforfarmers.InBenin,aftertwodecadesofmilitaryrulewithasocialist‐orientedsystemofwhich thedrasticeconomicandsocial consequenceswerestillprevailing,the new authorities could not do less than adopting the World Bank option of a

COTTONFARMERORGANISATIONS

THESTATE

Production Farmgatecommercialisation Expressionofinputsneeds DiversservicestoSONAPRA

PurchaseofcottonfromSONAPRA

Ginning Transport Positioningfibreonmarkets PaymenttoSONAPRA

Monopoly Recoveryofinputscredit Purchaseofcottonfromfarmers Paymentoffarmers Paymentofthetransportandfarmersservices

Ginning Fundingofresearchandextension Positioningfibreonmarkets

COTTONGINNERSINPUTSSUPPLIERS

SONAPRA(THESTATERUNPARASTATAL

COTTONCOMPANY)

CHAPTER2

36

completeliberalization.Thechoiceofthatoptionof‘aggressiveliberalisation’wasalsoinfluenced by the World Bank background of the Head of the State at the time(Heinisch,2006).

Thus, Benin led the cotton sector reforms to liberalise the sector in francophoneAfrica.AphasedwithdrawaloftheStatefromthecottonsectorwassetupinafive‐yearplan that transferredcompetencies toprivateactors.The institutional reforms,set up in fourphases, began in1991by the empowermentof farmerorganisations.Thirty‐three village groups and seven district unions constituted the pilot units.Gradually, by 1996, about 1,300 village groups and fifty‐one district unions werecovered.Meanwhile,theFédérationdesUnionsdeProducteursduBénin(FUPRO)wascreated in 1993 (AIC, 2008). This farmer federationwas the first one inWest andCentral Africa. Such a federation was only created in 1997 in Burkina Faso and in2000 inCameroon,whereas theprocess isstillon‐going inMali (Gergely&Poulton,2009). In 1992, the second phase of the reforms was enacted with the entry ofprivatescompanies in the inputs supplyingchain.From1994onwards, thegrowingcotton production convinced decision makers to issue agreements to eight privatecompaniestoinstalltheirginningmillsasthethirdphase.In1999,thefourthandlastphase of the reforms was achieved by the creation of the AssociationInterprofessionelle du Coton (AIC), the first cotton inter‐professional committee inWest and Central Africa. It comprised cotton farmers, inputs suppliers and ginners,andwasexpected toenhance theefficiencyand transparencyof the sector.BurkinaFaso and Mali followed suit in 2006 and 2008 respectively (Gergely, 2009).Complementarymeasuresweretakentostrengthenthereforms.Thefirstconsistedofthesuppressionin2000oftheStatecottonparastatal’smonopolyinthemanagementofcotton.ThesameyearwitnessedthecreationoftheCSPR,theFrenchacronymforAgencyforGuaranteeingPaymentandCreditRecovery.Thisagencywassetupastheregulatoryplatformamongprivatecottonstakeholders.Tokeepcontroloftheprocessof inputs supplying and combat irregularities at the level of farmer grassrootsassociations, the farmer federation created its own body in charge of issuingprocurement to private companies. The implementation of these institutions wasaccompaniedby laws,decreesandorderstopurposefullyregulatetheir interactionswithinthecottoninter‐profession.

2.5.2 Institutionaldevelopmentandlinkages

The liberalisation of the sector and the privatisation of the commercial functionsimpliedthattheroleofStatewaslimitedtolegislatingandregulatingtheinteractionsbetweenprivateactors.Indeed,theprocessbroughttogethermanystakeholdersandnurturedanewdynamic(AIC,2005).Specialisedinstitutionsweresettledtofacilitatethe interactions between actors, resulting in an institutional arrangement andfunctionalrelationsamongthem(Figure2.4).

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

37

Figure2.4 Institutionallinkagesincottonsectorafterthereforms(AdaptedfromAIC(2005)).

The reforms went smoothly and rapidly. The cotton sector in Benin then becamehighlyinstitutionalisedandwashailedasamodelofinventivenessandsuccess.This,atfirst,enhancedtheefficiencyofthesectorandincreasedthereturnstofarmersandtheState.Farmerspraisedcottonproductionandthepositivesocialchangesoccurringinruralareas,wherebotheconomicandsocialbenefitswerenoticeableathouseholdaswellascommunitylevel.TheAfricanpremierexportcrop,the"motherofpoverty"and the most unpopular crop among peasant cultivators during colonial times(Isaacmanetal.,1980;Waller,1997),becamethemotherofwealthinBenin.Alltheseboostedtheproduction,asdisplayedinFigure2.5.

State:Organizerofthelegalframework

Localstakeholders

Cooperative and Economic Organizations

CottonInterprofessionalAssociation

ProfessionalOrganizations

CNDIC(InputsSuppliers’Association)

AIC (Farmers,InputsSuppliersandGinners)

TheStateMinistersandTechnicalDirections

InputsSuppliers

CNPC(CottonFarmerFederation)

CNEC(CottonGinners’Association)

CAGIA(InputsSupplying

Cooperative)

CSPR– GIE(GuaranteeingcreditPaymentandrecovery)

Ginners GV(Producers)Banks

=Financialflux =Products/TechnologiesLegend: =Innovativeconsultativebodiesforregulatingcottonindustry

CHAPTER2

38

Figure2.5 Developmentofcottonproductionwithregardtopricefluctuations

Source:CompilationbytheauthorofdatafromAIC(2007,2008)andunpublishedstatisticsfromtheMinistryofAgricultureofBenin.

Thanks to a regular payment of their produce, farmers’ relationship with cottonremainedstrong.Thetrendsofcottonproductionwereonlyslightlycorrelatedwiththechangesinprices.Eventhedeclineoffarmgatepricesconsecutivetothedropofprices on internationalmarkets at the end of the 1990s did neither affect the pro‐ductionnortheever‐growingfarmerorganisations.

2.5.3 Farmerorganisations:Fromfarmtocities

The most common feature to West and Central African cotton sectors are theflourishing farmer organisations. In all countries, farmer organisations were trans‐formed into ratherpowerfulorganisations thatplayeda critical role in theboomofcotton production (Bingen, 1998; Bonnassieux, 2002). In spite of these similarities,somespecificitiesofthesituationinBeninareworthmentioning.

Fromtheearly1990s, thereformsprovidedanewdynamic to farmerorganisationsand, shortly thereafter, the village‐based groups came to be organised into onehierarchicalnetworkuptothenationallevel.Boththeorganisationsandtheirleadershadtheirinfluencelimitedtoruralareas.Thankstotheboomoftheproductionandthereforms,bothfarmersandtheirorganisationsstartedplayinganimportantroleat

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

0

50

100

150

200

250

300

350

400

450

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

FCFA/Kg & Kg/H

a

x 1000 Ha/Tons

Year

Area (x1000 Ha) Production (x1000 Tons)

Yield (Kg/Ha) Price (FCFA/Kg)

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

39

meetingsandforumsontheimplementationofagriculturalpolicies,thoughtheyhadhardlybeeninvolvedinthedesignofthereforms.Organisationsleaderswerebroughtclosertoexecutivesandpoliticaldecisionmakers.Manydisputesamongfarmersweresettledatcabinet level.Thesecontactsgavetheir leadersasenseof importance,andincreased their awareness that they also held power. The organisations werehenceforth extended from farm to cities. Many leaders became urban dwellers toincreasetheirvisibilityamonginstitutions.Farminghadbeentransformedintoentre‐preneurshipandreacheditsapogeeattheWorldTradeOrganization(WTO)summitat Cancun. The representatives of cotton farmer organisations from Benin, BurkinaFaso,MaliandChad,knownasCotton‐4countries(Minot&Daniels,2005;Woodward,2007), backed up by international NGOs advocating for fair trade, stayed in thelimelightoftradetalksoncottonsubsidiesthroughoutthesummitandlongafter.Thisincreased the confidence of the leaders and contributed to improving the powerbalance between farmers and private actors. The change could be noticed in thestrong position of farmer organisations when controversies about the quality ofpesticidesarose in2005/2006.Farmersvoicedtheirdiscontent largely inthemediato keep the public opinion informed. As a result, the Minister of Agriculture wasousted. But these controversies have persisted and became part of the demise ofcotton.Thefarmers’voicelosteffect,becausetheirorganisationswereweakenedbytheshiftfromtheStatemonopolytoprivateinterests.

2.5.4 Majorflawsinthereforms

Themajorgoalsofthereformspolicieswerecompetitivenessandeffectivenessofthesector.Duetomanyflawsinthereforms,however,theachievementdidnotresultinlastingeffects.Inadditiontothedysfunctionsofthesystemresultingfromthelackofclarityaboutthehierarchicalrelationshipsbetweentheregulatingbodies(AIC,2008),there were among other flaws, an over‐politisation of the sector, an over‐ginningcapacityofthecountry,andanover‐regulationofthesystem.Bythemid‐2000s, thedynamicsofthesectorstartedshowingsomeweaknesses.

Astiffcompetitionaroseamongtheprivateactorsforthecontrolofsomedomainsofthesector.Itstartedintheinputssupplychain,wherefarmerorganisationsatdistrictlevel could choose among themany registered suppliers. Farmer leaders at districtlevel were subjected to too much attention from inputs suppliers looking for theintention topurchase,aconstituentpartof theprocurementcommission.Thechaseafter farmers’ support led to irregularities and unfairness, which were explicitlyattestedbythedeputydirectorofacompanypreviouslyoperatinginthesector:

Wehavehadapainfulexperience.Farmersare,ofcourse, fraudsters,but theother inputs suppliersusually corrupted thenetwork leaders.Thenetworkleadersareallguilty.

CHAPTER2

40

Heconsidershiscompanytohavebeenkickedoutofthesystembythecompanythatacquiredthemonopolyandexertedituntilrecently.Thecontestationsoftheresultsoftenders for inputs supplyworsened the situation.The cotton sector representativesadmittedthat"theinputssupplychain[was]characterisedby[...]repetitiouscontro‐versies and thenon‐compliancewith the legislationpertaining in the supply chain."(AIC,2006:9).Dissidenceoccurredamongtheinputssuppliers’andledtomultiplyingtheir organisations. The types of pesticides multiplied and some were importedwithoutanyagreementandwithdisregardofthehazards involvedintheirhandlinganduse.

The State played an ambiguous role in regulation (Kpade, 2011). For instance, aprivatecottonmillowner,theCEOofMarlan’sCottonIndustry,wasatthesametimeanadvisorof theHeadofState.Hemanagedtogetsuppliedwithcottonoutsidetheofficial system for three years without paying back the farmers. Without anyagreement, he imported and redistributed inputs in the districts under his control,encouraging other inputs suppliers to break the rules aswell. Court sentences andcabinets decisions from the Ministry of Agriculture against his company weresuspended by authorities at the Head of State’s office where he had easy access(Saizonou, 2008). Furthermore, while the country’s production capacity of about350,000tonsofseedcottonwasreasonablybalancedwiththe312,000tonsofginningcapacity,privatecompanieshadbeenissuedlicencestoinstallcottonmillsforabout275,000 tons of ginning capacity. This stepped up the number of cotton ginningfactoriesfromtentoeighteenandbroughttheoverallginningcapacityupto587,000tons(AIC,2006).Thisover‐ginningcapacitytriggeringthescrambleforcottonbytheginning companies. Privates ginners excelled in breaking rules and bribing farmerleaders in order to getmore cotton supplies. These dealswere backed up by theirpolitical acquaintances. Political connections emerged between private companies’owners,decisionandpolicymakers,andfarmers,therebyfrustratinganyenforcementtoabidebytheregulations.Additionally,withthe importantresourcesgeneratedbycotton production and the number of people involved throughout its productionchain,cottonbecameapoliticalinstrumentthatmanypoliticianswantedtocontrol.

The competition between private actors encouraged the emergence of dissidentcotton networks that escalated in number. In the end, it became customary toattribute each of the cotton networks to either a ginner or an inputs supplier,connected with political mentors: the original network (ANPC) to Talon (formerlyinputssupplierandnowalsoginner),AGROPEtoFRUITEXandCSI(inputssuppliersand occasionally cotton ginners), AGROP to Rodriguez (ginner), FENAPROC to DaSilva,aninputssupplierandeminentmemberofthepoliticalpartyoftheMinisterofRuralDevelopmentatthetime,andFENAGROPtotheSONAPRA(theformerState‐runparastatal cottoncompany). Inaddition,until theearly2000s (when theyhad tobenamedascottonorganisations),theorganisationswerenamedwithoutanyreferenceto cotton, even though they were dealing exclusively with cotton, in contrast toneighbouringcountrieswherecotton‐relatedactivitieswerereflectedinthenamesof

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

41

cotton farmerorganisations.Thisencouragedamixofgenres,with farmer’s leadersplaying,theoretically,allkindsofrolesforanycrops.

Ultimately, theeconomicand financial losses causedbyall these flawsamounted toseveralbillionsofCFAattheexpenseoffarmers.Theoperationcostsofthemainthreeinstitutionscreatedforregulatingthesectoramountedto2billionCFA(≈€3million)(Macrae,ND).Theseconstitutedadditionalcostsmainlypaidbyfarmersand,tosomeextent,byginners.Nowadays,thenegativeimpactsofthereformsaredominant(AIC,2007).Although theStatecontinues to issuedecrees,ordersandnewrules, there islittleimprovement.Theincoherenceofpoliciesandthegrowingdiscontentoffarmersarecontinuouslypullingdowntheproduction.

2.6 Farmers’responses

Cotton farmers have been facing multiple constraints for almost a decade: soaringinputs prices, the drop in world cotton prices, the bad quality of inputs, delays inpayment,etc.Forinstance,between2000and2005,thecostsofpesticidesperhectarealoneroseby80percent(Ferrignoetal.,2005)whilecottonpricesatthefarmgaterosebyonly5percent.Inaddition,thelowratioinweightofseedbyfibreofthenewvarietyofcottonthatisbeinggrownforsomeyears,comparedtovarietiespreviouslyused,makesfarmersthelosersofthatinnovation.Theannouncementoftheweightoftheirproductionconvincesthemthattheyare indebtedandwillhaveto findmoneyfromothersourcestopaybackforthe inputstheyused.Thedisappointmentmakesthemfeeldizzyandtheystagger,oftenknockingoverthescaleusedtoweighcottonballots, hence the expression "kilo kurukara" in the local language, an euphemismmeaningtorunoverthescale.“Ifyoustumbleoverthescale,youreffortandsacrificesareyouronlybenefit”,saidawomaninBanikoara.Allthesecombinedtoresultinalowreturn of the investments on cotton production. This low return has a depressingeffect on the production,which is nowbeing squeezed into the cotton belt and thesurroundingareas,comprisingninedistrictsthatstillproducemoreorless5,000tonsayear.TheseareBanikoara,Kandi,Gogounou,SegbanaandMalanvilleintheprovinceofAlibori,KerouandPehuncointheprovinceofAtacora,andKalaeandSinendeintheprovinceofBorgou(seeFigure2.1).AsdisplayedbyFigure2.6,theirshareinareaandproduction is growing at thenational level.During the 2007/2008 season, the ninedistrictsrepresented68percentoftheareacoveredbycottonand83percentoftheproduction,reaching88percentand91percentrespectivelyin2010.

CHAPTER2

42

Figure2.6 Nationalcottonproductionshareoftheninedistrictslargestproducers(2008–2010).

Source:FigureconstructedfromunpublishedstatisticsfromCSPR(2000–2011).

The inadequacy of certain reforms, combined with the multiple productionconstraints led to farmers being deprived of their incomes. Therefore, they devisestrategies to make ends meet. Nowadays, for many cotton farmers in the mainproducingareas, thepurposeofgrowingcotton isshifting fromearningrevenuestoaccessinginputs,mainlyfertilizers,forfoodcrops.Intheabsenceofspecificfertilisersfor food crops, fertilisers received for cotton are diverted tomaize and other foodcrops, or sold for cash.As a consequence, the inputs consumptionprogressesmorethanthecottonproduction.From2000to2006,theratioinputscreditbyseed‐cottonmarketedclimbedfromabout31percentto73percent(Saizonou,2008), implyingvery low economic returns in cotton production. From 2005 to 2010, the inputsconsumptionofthelargestdistrictproducerofcottonhasincreasedinvalueby23percentwhiletheseedcottonproducedhasdecreaseddrasticallyby43percent.In2011,this consumption has slightly decreased by 9 per cent but the production steeplydecreased by 38 per cent. Therefore, the positive effects of cotton on farmers’livelihoodsstartedtodeclineandinsomecasesreversed,astheWorldBank(2003:7)alreadynoticedin2003.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Area Production Area Production Area Production

2007‐2008 2008‐2009 2009‐2010

Percentage of national production

Cotton Seasons

OTHERS

9 COMMUNES

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

43

2.7 Farmers’perspectivesontherevivalofcottonproduction

2.7.1Farmers’perceptionsofcottonrevival

Since the 2005/2006 season, cotton production has been undermined by the badqualityofpesticides,whichtriggeredcontroversies.InspiteofpetitionssentuptotheHeadofState’sofficeby farmerorganisationsandthedismissalofpublicexecutivesincluding aminister, theproblempersistedover the years, leading to awidespreaddiscontentamongfarmers.Farmers’despairabouttheprospectsofcottongotwidelysharedandisreflectedinthesurveyresultsaboutfarmers’perceptionsoftherevivalofcottonproduction(Table2.2).Theseperceptionswerecodedfrom0(pessimistic)to2(optimistic).Pessimisticreactionswereofthekind‘Ihavenosuchhope’or‘No,itisnotpossible’,whileoptimisticreactionswere like ‘Yes’or ‘Ofcourseyes.Evenmorethan before’. The intermediary position was coded 1 (doubtful) and it includedresponsessuchas‘Ifthingschangefromthetop’,‘Ifthey[decisionmakersand/orinputssuppliers]actforthat’,‘Iftherearegoodqualityinputs’,‘OnlyGodknows’,and‘Idonotknow’. Farmers’ views were cross‐tabulated with their leadership status and theirnetwork membership history. Both grassroots farmers and leaders weredistinguished into those who have never changed their network membership(Stayers) and those who were a member of more than one network (Breakers).Grassrootsfarmersarecottonfarmerswhoneverhadanyleadershipresponsibilitiesinthenetwork,asopposedtoleaderswhodidhavesuchresponsibilities.

Table2.2 Log‐linearanalysisoffarmers’viewsontherevivalofcottonproduction

Categoriesofcottonfarmers OverallViewObserved

(N=145)Count Percentage

Stayers

Grassrootsfarmers

Pessimistic 5 3.4

2(2)=1.69n.s

Doubtful 11 7.6Optimistic 4 2.8

LeadersPessimistic 3 2.1Doubtful 6 4.1Optimistic 6 4.1

Breakers

Grassrootsfarmers

Pessimistic 15 10.3

2(2)=1.66n.s

Doubtful 22 15.2Optimistic 8 5.5

LeadersPessimistic 15 10.3Doubtful 34 23.4Optimistic 16 11.0

Note:n.s=non‐significant.

Source:Surveydata(2009–2011).

The results indicate that there is a relationship between leadership status andbreakingaway fromanetwork, indicating that leadersbrokeawaymoreoften than

CHAPTER2

44

grassroots farmers. Farmers’ views depend also on their leadership status in thecotton production system. About 31 per cent of grassroots farmers are gettingdespondentabout the futureofcottonandareclassifiedaspessimistagainst23percentof leaders,whileabout28per centof leadersare classifiedasoptimist againstonly18percentofordinaryfarmers.Thepercentagesofdoubtfulfarmersabouttherevivalofcottonproductionarealmostalikeamongleadersandgrassroots farmers:50 and 51 respectively.With regard to breaking, stayers and beakers do not differmuchintheirviews.About30percentofstayersexpressoptimisticviewsabouttherevival of cotton production against 22 per cent of breakers. The percentages ofdoubtful farmers in the two groups are also alike: about 50 for stayers and 51 forbreakers. The figures show that the optimists constitute a minority. Many of bothleadersandordinaryfarmershaveconsiderablyreducedtheircottonarea.Somehavesuspendedtheproduction(whentheyexpectthecottonenvironmenttoimprove)andothershaveabandonedpermanently.A formersecretaryofa farmerorganisation inthecottonbeltthreatened:“Iftheydonotchangethepesticide,thecurrentseasonwillbetherequiemofcotton.”

2.7.2 Measuresfortherevivalofcottonproduction

Aseriesofactionsthatfarmersconsiderincentivestobeimplementedfortherevivalofcottonwerelistedduringthefocus‐groupdiscussions.Theactionsrangedfromtheprovisionof inputs to thepaymenton timeofmoneyafter thecommercialisationoftheirproduce,includingtheextentoftheStateinterventioninthemanagementofthecottonsector.Theseactionswererankedaccordingtotheirperceivedimportancebyindividualfarmers(Table2.3).

Table2.3 Rankingofreformsactionsenvisionedbycottonfarmers

Rank

GrassrootsFarmers(N=58)

Leaders(N=71) T

1 Improvingtheinputsqualityandsupply

2.4(1.63) 2.9(1.90) ‐1.54

2 Paymentontime 3.2(2.61) 3.0(2.10) 0.403 Loweringinputsprices 4.3(2.53) 3.8(2.45) 1.144 Raisingcottonprices 5.3(2.25) 4.4(2.52) 2.05*5 Eliminatingthejointliability 5.2(2.34) 6.2(2.05) ‐2.47*6 Re‐nationalisationofcottonsector 6.0(2.36) 5.7(2.72) 0.62

7Sharingproductionrisksbetweeninputssuppliersandfarmers 6.2(2.74) 6.8(2.16) ‐1.28

8 Reducingcottonnetworksto1 6.9(2.54) 7.2(2.78) ‐0.739 Restoringrefunds 7.8(1.93) 7.1(2.09) 2.03*

10 Suppressingfarmers’cotton‐relateddebts

7.7(2.21) 7.9(2.12) ‐0.53

*=p<0.05

Source:Surveydata,2009–2011.

CHANGINGFARMERS’PERCEPTIONSINBENIN’SCOTTONPRODUCTION

45

In theoverall classification, improving inputsquality came first, thereby confirmingthe high importance of inputs for farmers. This was followed by payment on time,lowering inputspricesand raising cottonprices.Restoring refundsandsuppressingcottondebtswere classifiedninthand tenth respectively.Thereweredifferences inthe ranking of only three actions between grassroots farmers and leaders. Leadersranked raising of cotton prices and restoring of refundshigher than grassrootsfarmers,t=2.05(p<.05)andt=2.03(p<.05)respectively,whilegrassrootsfarmersranked the elimination of joint liabilityhigher (t = ‐2.47, p < .05),presumably be‐causegrassroots farmers suffered more from theconsequencesof the joint liabilitythantheleaderswhoareinchargeofitsimplementation.Infact,whentheamountofmoney paid to an organisation is not enough for equitable redistribution amonggroupsmembers,theleaderskeepitforawhile,and,accordingtograssrootsfarmers,theyveryoftenputitattheirowndisposal.Refundsweresometimesentirelyusedbyleaders to cover the operating costs of the organisations, and other uncontrollableexpensesoftheorganisation.One additional measure was added, mainly by leaders during the individualinterviews: the suppression of the cotton inter‐professional committee (AIC) andoftheagencyforguaranteeingpaymentsandcreditrecovery(CSPR),twoinstitutionsofwhich farmersareconvincedof theiruselessness.The two institutionsare themainregulatingbodiesinthecottonproductionsystemandtheycostyearlyahugeamountof money to producers, thereby reducing the benefits of cotton production. Theperception of uselessness testifies to the conflicting interests between leaderswhorepresent farmers and the executives of the regulating bodies. Executives of theseinstitutions are encroachingon theprerogatives of farmer leaders, according to thelatter.

2.8 Conclusion

IndustrialcottonproductioninBenindatesfromcolonialtimeswhenfarmersinBeninappeared to bemore collaborative in cottonproduction than in other colonies. JustafterIndependence,incidentsoffarmerresistancegavecottonanegativeimage.Afewyearslater,theregime,withitsMarxist‐Leninistorientedideology,consideredcottonproduction a relic of colonialism, and progressively neglected it in favour of foodcrops. However, the perception of cotton improved during the 1990s when theproduction blossomed. It then became a white gold, regardless of the fluctuatingprices in international markets. The change was made possible by the reformsundertakenfromtheearly1990stoliberalisethesector.Butthesereformshadsomeharmful impacts on the sector, due to flaws in their implementation. This led torecognisethat"theoutcomeofreform[s]clearlyfellshortofexpectationsandresultedinasharpdeclineofthesector’sperformance"(Gergely,2009:35).ThecottonsectorinBeninisactuallyburdenednotonlywithtechnicalproblemsbut,inparticular,withinstitutionalandpoliticalconstraints.Theensuingdecreasedprofitabilityofthecrop

CHAPTER2

46

hasledfarmerstoconsidercottonapest,anditsproductiontodeclinedrastically.Thepurpose of the production has therefore changed. At present, for thousands offarmers,cottonproductionisnolongerawayofmakingendsmeet. Ithasbecomealast resort or, at best, a makeshift crop. Cotton is grown for accessing fertilisers,insteadofearningincomes,trappingfarmersinaviciouscircleof indebtedness.Stillyet,theseconstraintsarenotadequatelyaddressedandtheshifttoprivatemonopolyand the numerous new regulations implemented are not capable of reversing thetrend. Ifcottonisstill important,thencottonfarmersare importantaswellandthisviewshouldbetakenintoaccountindecisionmaking.AsstatedbyMako,athirty‐fiveyear‐oldcottonfarmerinBagou,“cottonwillreviveifonlythey[politiciansanddecisionmakers]listentofarmersandthepoliciesmeetfarmers’needs.”

47

Chapter3

TheFactorsBehindtheEmergenceandCollapseofCottonNetworksinBeninduringthe2000s

ABSTRACT: The liberalisation of Benin’s cotton sector during the 1990sresultedintheemergenceofanoverarchingcottonnetworkwhich,however,soonbroke apart (atomized) into conflicting smallernetworks.This chapterassesses the major reasons for that atomisation of the cotton networks.Extensive fieldwork in Benin’s cotton belt using amixed‐mode approach ofqualitativeandquantitativemethods,indicatedthattheatomisationofcottonnetworks enhanced farmers’ access to leadership positions. Although thequantitative results did not support the qualitative ones, it turned out thatfarmers’decisiontobreakawaydependedforemostonthedegreeandtypeofsocialcapitalthatfarmershadattheirdisposal.Predictably,agreaterstockofsocialcapitalwaspositivelycorrelatedwiththedurationofhavingaleaders’positiononcottonnetworkboards.Keywords:Cottonnetworking;breakingaway;atomisation;socialcapital;Benin

AshortenedversionofthischapterisunderreviewbyAgricultureandHumanValuesas:GuirguissouMaboudouAlidouandJarlK.Kampen.AnempiricalstudyofthefactorsbehindtheemergenceandcollapseofcottonnetworksinBeninduringthe2000s.

CHAPTER3

48

3.1 Introduction

The last two to three decades havewitnessed an increasing interest of agriculturalpolicymakersinfarmerorganisationsinBenin.Farmerorganisationsgrewsteadilyinimportance,alongwithcottonproduction,andbecameinterconnected,elapsingintoalargerandnationwideorganisation.Indeed,cottonproductionhasbeenatthecentreof the institutional dynamics in agriculture and of the farmers' movement. Farmerorganisations became themain instrument for cotton production, as a result of thereforms initiated in the early 1990s to facilitate the transition of the cotton sectorfroma"monopolisticandcentrallyadministeredsystemtoamorecompetitiveone"(World Bank, 2003:59). These organisations played a variety of roles, such asmanaging the production and marketing of cotton, facilitating farmers’ access tocredit,anddevelopingcollectiveactiontopromotefarmers’ inclusion(Tama‐Imorouet al., 2007a). The organisations were hierarchically structured from village tonationallevel,anddevelopedintothefirstnetworkintheagriculturalsectorinBenin.BecausethisnetworkwasbeneficialforbothfarmersandtheState,itwasconsideredanimportantpartnerbybothpublicandprivatedevelopmentagencies.Therefore,itwasatthecentreofruraldevelopmentprograms.However,fromtheendofthe1990sonwards, several conflicts arose within the network that led to its disintegration(atomisation) and the formation of many rival networks, thereby triggeringdegenerativedynamicsincottonorganisations.

Thedevelopmentanddynamicsofsocialorganisationsisacentraltopicinthesocialsciences.Earlyworkonthistopic focussedonthedevelopmentandthedynamicsofsmall groups. During the 1970s, task‐activity behaviour was distinguished frominterpersonalbehaviour.Bothbelongtothe“realmofgroupbehaviour”,inwhichfivecritical stages can be distinguished: forming, storming, norming, performing, andtransforming(Gilleyetal.,2010;ShawandBarrett‐Power,1998:1311;TuckmanandJensen,1977).Thesestagesarenotnecessarilyincremental,giventhattheoccurrenceof any conflict at a specific development stage can force the group to return to apreviousstageforreadjustments.Thoughlargergroupswitnesssimilartrajectoriesofdevelopment,differenttheoriesapplytosuchgroups,suchasnetworktheories.

Thedebatesaboutthedynamicsoflargergroupsandorganisationsfocusparticularlyon social cohesion, social solidarity, and social differentiation. These concepts arewidely acknowledged as important characteristics of group stability, and some ofthem have been used interchangeably. Social cohesion is a most vital feature oforganisational survival and, therefore, is required the success of the group (Moodyand White, 2003). However, cohesion can never be taken for granted, particularlywhen it concerns hierarchical structures like the cotton networks in Benin. Indeed,hierarchyleadstodivergentorientationsbetweensub‐groupswithoppositestatuses(Dovidoetal.,2009).While lower‐statussub‐groupsaremore motivatedforchange,thosewithahigherstatustendtosupportthestatusquo.Thelatterareattractedbytheadvantagesthataccruetoleadershippositionsratherthanbytheresponsibilities

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWORKS

49

pertainingtothesepositions.Thisaffects intra‐groupinteraction,whichissocrucialtothestructuralcohesionofagroup.

Networkapproachestostudiesofsocialstructurefocusonactors’positionsandhigh‐light their relationships with other actors. Networks are sources of social capital.Theirstructureandsizeareimportantfortheircohesionandtheabilityofmembersto cooperate (Moody and White, 2003; Poulsen, 2009). Kilduff and Tsai (2003)identified some fundamental and interrelated concepts in network approaches toorganisations, among which the centrality of an actor and social capital. Centralitydefinesthesocialstratificationofthenetwork.Itdeterminesanactor'saccesstoandcontrolofresourceswithinthenetwork,therebyascertaininghis/hersocialcapital.

Theconceptofsocialcapitalismultidimensionalanddiverselyinterpreted(Kampen,2010; Lin and Erickson, 2008). There are three dimensions of social capital, all ofwhich are located in the structure of social networks and their members: 1) thestructural dimension such as network ties and configuration, 2) the cognitivedimension such as the discourse to build collective identities, and 3) the relationaldimension such as norms andbeliefs (Nahapiet andGhoshal, 1998:243‐244).Threetypes of social capital can be derived from this distinction (Burt, 1992; Cohen andPrusak,2001;Granovetter,1983;Putnam,1995;Woolcock,2001):1)bondingsocialcapital,referredtoasstrongtiesthatgluememberswithingroupssuchasfamiliesorclans,makingthegroupmorethanacollectionofindividuals,2)bridgingsocialcapitalthat links members to outside networks, and are referred to as weak ties, and 3)linkingsocialcapitalthatconnectsanindividualtoothersinapowerposition.Thesedifferenttypesofsocialcapitalarelinkedtoformalorinformalnetworks(Ferlander,2007), completing the classic distinction of strong and weak network ties byGranovetter(1983).Thisdistinctionmeansthatthereisastructuraldynamicwithingroups, which entails interactions between members for access to and control ofresources located therein, for thepurposeof their (re)positioning (Bourdieu,1990).Depending on the stock of resources embedded in members’ social relations, theinteractionscan1)maintainmembersattheirpositioninthenetwork,2)raisetheirposition,or3)suppressthepositionofothermembers.Bourdieuseessocialcapitalasenabling the reproduction of social inequality because it helps strengthening thedominantpositions.

Referring to the specific situation in Benin in the 1990’s (discussed in detail in thenextsection),thehierarchicalshapeoftheoriginalcottonfarmers’networkresultedinanunevendistributionofpowerandresources.Rapidly,within‐networkconflictsarose at all levels of the network in a battle to control these perks (power andresources). This has led to the network breaking apart and ending up in atomisednetworks.Thischapterseekstoelucidatethedrivingfactorsbehindtheatomisationofthe network of cotton farmers in Benin. It will analyse the interactions betweenindividualactorswithinandbetweencottonnetworks.First,thehistoryofthecottonnetworks in Benin will be sketched. This will be followed by a description of the

CHAPTER3

50

researchhypothesesandmethods.Then,theresultsarepresentedanddiscussedandaconclusionisformulated.

3.2 Historicalbackground

3.2.1 CottonfarmernetworksinBenin:Fromagencytoassociations

Thehistoryofwhat is commonly referred to as cotton farmernetworks inBenin isfirmly linked to the experience of farmer organisations from the late 1960s to the1980s. Farmer groups were formed at the village level. With the boom of cottonproduction in the 1990s and the advent of democracy, not only the village‐basedgroups gained more autonomy, they also extended their influence beyond villages.With the liberalisation of the cotton sector, farmer organisations became moreformallystructured,mirroringthecountry'sfour‐leveladministrativedivision7,whichresultedintoahierarchicalandprofessionalnetwork(Figure3.1).The core function of cotton farmer organisations is to supply their members withcottoninputsandmarkettheirproduction.Atthevillagelevel,thereweregrassrootsfarmergroups,which couldbe furtherdivided into sub‐groups to ease themanage‐ment of resources, depending on the size of the village and the volume of cottonproduction. The boards of village groups dealt with farmers’ daily concerns withregard to cotton production. Farmer groups within a district were managed by adistrict union of producers. The district boards collected and centralised villagegroups'demandsforinputs,andmanagedthemarketingandtransportofseedcottonfrom the farmmarkets to ginning factories. They selected and issued the licence toinputs suppliers and delivered the authorisation to the transporters for loadingcotton.Thedistrictunionsformedprovincialunions,whichweremanagedbyaboardelected amongmembers of the district boards. The provincial unions of producersplayedthesamecoordinatingroleasthedistrictunions,managingdemandsforinputsfromthelatteratalargerscale.Atthenationallevel,theseunionsbecameafederationthatconstitutedalargenetworkofcottonproducers.Thisnationalapexwascreatedin1994.Thenationalapexmanagedexclusivelycottonorganisationsfromthetopofthe hierarchy and represented cotton farmers in cotton bodies. It also focussed onlobbyingandadvocacy  for farmers’ interests in thenationaland internationalarena(Koutonetal.,2006).Farmer organisations have depended on cotton levies, so that "producing cottonbecame amore stringentmembership criterion and encouraged other [cash crops]producerstoformtheirowngroups"(Tama‐Imorouetal.,2007b:113). Membership

7Thevillagegroups(GV)formedatthe77districtlevels,thedistrictunionsofproducers.Ineach of the six administrative provinces, the district unions are grouped in the provincialunionsofproducers,whichinturnfederatedinanationalapexcalledfederationofproducers’union.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWORKS

51

guaranteesfarmersaccesstocashcreditandinputssuchasfertilizersandpesticides.In return, the members comply to deliver and sell their production through theorganisation. Control over farmers is exerted through the joint liability of groupmembers enforced by the board of the organisation. A mechanism to enforcecompliance with the system that linked supply of inputs to cotton production andmarketingbetween farmers, inputs suppliers andginnerswas setupandknownasthe "mécanisme de la CSPR". This mechanism was monitored by the cotton inter‐professional association, a body comprising the different institutional actors, toenforcetheregulationsrelatedtothesector.Itreinforcedthehierarchicalstructureoffarmer organisations for a long time, making cotton production for non‐membersimpossible.It should be noted that farmer leaders at all the levels of the hierarchy grew inimportance, both economically and socially. Butwhile the network boardmembersknew each other across districts and provinces because of meetings convened forleaders,thisdidnotapplytograssrootsfarmers.Sharingthemembershipofacottonnetwork rarely involved farmers from different villages in a joint production.Mostmembers of the networks were hardly linked within the cotton networks thatconsequentlycontained"structuralholes"(Burt,1992:65).Hence,grassrootsfarmerswere connected only through the leaders of their networks, with little interactionamongthemselves.Suchnetworksneedahighleveloftrusttofunctioneffectively(cf.Porta et al., 1997). The bottom of this pyramidal organisation was shapeless andlackingnaturalboundaries,inthiswayexhibitingoneofthecorecharacteristicsofasocialnetwork(DegueneandForsé,1994:14;Latour,1997).

Figure3.1 Thehierarchicalstructureofcottonnetworks

CHAPTER3

52

3.2.2 Namingcottonnetworks:Thewarofwords

Bytheendofthe1990s,thecottonfarmernetworkwasburdenedbymismanagementand dysfunctioning. This caused long delays in payments to farmers that couldamount toayear (WorldBank,2003).Asa result,dissidentnetworksemerged thatchallenged the monopoly of the federation (Sinzogan et al., 2007). The process ofatomisation started from the top of the pyramid where access was still highlycontrolled and regulated, and then spilled over into its bottom at village level.Following the movement from the top, the village groups split in many wingsbelongingtodifferentnetworks.Theoriginalfederationcreatedin1994,fellaparttogivebirthin2001tothefirstbreakawaynetwork.Between2001and2006,thesetwoconflicting networks split up endlessly, generatingmore than ten other breakawaynetworksandraisingthenumberofcottonnetworksto14(Table3.1).Althoughthenamesof theseorganisationsdidnotrefer tocottonuntil2005whena legalreformimposedthefocusoncottontotheirnaming,theyhaddealtmostlywithcotton.Ofthebreakawaynetworks,fourfizzledout.Thenineremainingnetworksviedforfarmers'supportuntil2010,whenthecentralgovernmentdecidedtoregaincontrolofcottonproduction. Ironically, this lackofcontrol in thereformprocesshasbeenpraisedasthe"dynamicsoffarmerorganisations"(AIC,2005:8).SimilartotheterminologyusedinChapter2,farmerswhochangedtheirnetworkmembershipwillbereferredtoasbreakers, while those who kept to only one network will be called stayers. Bothcategoriescanstayconstantorchangedependingon theirchoice to joinorcreateanetworklateron.

The process of atomisation stemmed from Gogounou, the third largest district incottonproduction,wherea leadershipconflict ledtothefirstdissidentorganisation.Afterthatitspreadtootherdistricts.Gogounouledthecontestforcreatingbreakawaynetworks in the country thereafter. As a result, no less than six of the total of 13breakaway networks emerged from there. The prominent place of the district ofGogounouinthisprocesswaspointedoutbyfarmersduringfocus‐groupdiscussionsandinterviews,asillustratedbythefollowingstatementatSinende:“Wedidnotknowhowtocreatedissidence.ItisfarmersfromGogounouwhousedtocomeherebynighttocajoleustojointhesenewnetworks.”CottonfarmersfromGogounouarestillproudofthis history and feel honoured to have their district qualified as the laboratory ofcotton networking, an attribute that many of their leaders proudly claim. Theatomisationof networksproceeded alongwith amarked shift in their structure.Byfragmenting, they became smaller andmore diversely structured. Only the originalnetwork exhibited the four‐level structure with boards at village, district, provinceandnationallevel.Thebreakawaynetworksshowedadifferentstructurewithtwotothree hierarchical levels. The village groups were no more on a one‐village basis.Instead,theybecamemulti‐villagebasedandwerethereforecalledproducersgroups.Thetopofthehierarchywasmodified.Itinflatedwithleaders,manyfarmershavingchanged their status and uplifted their positions. Consequently, the shape of thenetworksshiftedfromthe“pyramidalshape”toa“vaseshape”.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWORKS

53

Table3.1 Listoftheactivenetworks(2009–2011surveydata)

N Acronym Description(translated) Creation8 NatureNumberoffarmergroups(2008)

OperationalStatus

1FUPRO(lateronANPC)

FederationofProducerUnions 1994Originalnetwork 2095 Functional

Nationalassociationofcottonproducers 2005

2 AGROPE Associationofproducers’groupsandbusinessmen 2000/2001BreakawayNetwork(BN)

‐ Brokeapart

3 FENAPRA NationalFederationofagriculturalproducers 2003(May2002) BN 360 Functional

4 AGROP–Benin Associationofproducers’groupsOctober2003

(February2003)BN 203 Functional

5 FENAGROP Nationalfederationofproducers’groups April2005(2004) BN 161 Functional

6 AGROP–NV Associationofproducers’groups–NewvisionOctober2006(May2005)

BN 115 Functional

7

AGROP–DEDE/RFC(lateronANACODEDE)

Associationofproducers’groups–Transparency/Fortherehabilitationofcottonsector

December2004(June2004) BN 69 Functional

NationalassociationofcottonfarmersTransparency May2007

8 FENAPROC NationalFederationofcottonproducersMay2005

(December2004)BN 299 Functional

9 UNAPRO–B Nationalunionofproducers 2004(2005) BN 84 Functional

10 UPROCOB Unionofcottonproducers 2006(2005) BN 116 Functional

11 ANCPB Nationalassociationofdistrictsproducers ‐ BN ‐ Fizzledout

12 WEWE–PEPE “Allwhite” 2006 BN 17 Fizzledout

13 URPCB Regionalunioncottonproducers 2006BN(aunionofn°4,5,6,7,8and9)

‐ Fizzledout

14 – NationalsyndicateofproducersofBanikoara ‐ BN Fizzledout

8Yearsinbracketsaredateswhenthenetworksstartedoperatingbeforetheirofficialregistration.

CHAPTER3

54

Dissidence among networks often emerged in the middle of the cotton growingseason.Whenthathappened,breakerswouldhavemarketedtheirproductsthroughanetworkdifferentfromtheonethathadprovidedthemwithinputs.Asaconsequence,mostbreakawaynetworksoperatedforatleastonecottonseasonbeforetheirofficialregistration. For example, during the 2004‐2005 season, dissidents from the threepre‐existingnetworksgenerated threemorenetworks thatmarketed theirproductsseparately, before being officially recognised in 2005. Figure 3.2 pictures the con‐figurationofnetworksbasedonthedateoftheirofficialrecognition.

Figure3.2 Emergingrelationshipsbetweencottonnetworks

Thus,disputesregularlyaroseovertheuseofnetworknamesbetweentheconflictingwings.Thedisputesweresettledbyaddingsuffixestotheacronyms. Illustrationsofsuch practices are the networks derived from AGROPE: AGROP‐BENIN, AGROP‐NOUVELLEVISION,andAGROP‐DEDE.Thesecondsuffixrefersto“newvision”,whilethe third symbolises “transparency”. They all kept the prefix AGROP, the French

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

55

acronym standing for association of producers. By doing so, they aimed to benefitfromthecreditgainedbythefirstbreakawaynetworkAGROPE,whichliveduptotheexpectationsofsomefarmersbeforeitfellapart.

3.3 Researchhypothesesandmethods

3.3.1 Researchhypotheses

Thepresent chapter addresses the factors that affected the interactionbetween the“cotton system” and the structure of production relations in the process ofatomisation. It is intended toshed lighton theprocess throughwhich farmers,whohad greatly contributed to the boom in cotton production through effectiveorganisations, weakened and atomised their cotton networks. Hypotheses wereformulatedbyapplying the theoreticalperspectives to the factors that influence thedynamicsofnetworkingtothespecificsituationofBenin’scottonsector.Thecentralhypothesis is that different backgrounds and attitudes of farmers combined withnetworks characteristics led farmers to make different decisions with respect tostaying within (remnants of) cotton networks (stayers), leaving them to join othernetworks(joiners),orcreatingnewones(creators).Hence,threecategoriesoffarmerswithregardtoatomisationweredistinguished:stayers,joiners,andcreators.Factorsthat influenced the phenomenon of atomisation (backgrounds and motives) weregrouped into push factors andpull factors. Push factors for atomisation are causes,either real or perceived, that forced or persuaded farmers to leave networks. Pullfactors into atomisation are those, either real or perceived, which determinedfarmers'decision to joinothernetworksor tobecoming involved in the creationofnew networks. A third group of factors that could have influenced the process ofatomisationofcottonnetworkscomprisesthesocialnetworksofcottonfarmersintheformofsupportiveties.Indeed,socialrelationsconstitutethevehicleofsocialcapitaland have been revealed to be important in various domains, such as communitydevelopment,jobattainment,andentrepreneurship(MoerbeekandFlap,2008;Nancyetal.,2002;Putnam,1993).Thefollowingframeworkguidedtheresearch.

3.3.2 Studydesignandsampling

Theresearchwasconductedinthemaincotton‐producingprovincesofBenin,BorgouandAliboriintheNorth,fromwheremorethan75percentofthecottonproducedinthe country is sourced. All the registered cotton networks are concentrated in thisregion,vying tocontrolasmuchproductionaspossible.However, theproduction isunequallydistributedwithintheregionandthecottonbelt isquitedistinctive.Mostbreakaway networkswere concentrated in that cotton belt and some exerted theirinfluenceinonlyafewvillages.

CHAPTER3

56

ThefieldworkwasconductedintwophasesbetweenOctober2009andApril2011.Itcombinedquantitativeandqualitativeresearchmethodstoinvestigatetheprocessofcotton networks atomisation. The first phase consisted of an exploratory surveythroughfocus‐groupdiscussionswithfarmers,andkeyinformantinterviews.Duringthe second phase individual interviews were conducted that combined in‐depthinterviewsandsurveyinterviewswithnetworkleadersandgrassrootsfarmers,usingasemi‐structuredquestionnaire.Thequestionnairewassubmittedfirsttoindividualnetworkleadersandthentograssrootsfarmers.

The sampling design combined purposive and random sampling. During theexploratory phase, the sample consisted of ten villages purposively selected andunevenly distributed throughout the provinces. The selection criteria were, amongothers, the levelof cottonproductionand thepresence in thevillageof at leastoneoperatingcottonnetwork.Inthesecondphase,thesamplewasconstructedinathree‐step process: first choosing the village, second the networks, and then the farmers.Four villages andninenetworkswerepurposively selected from the villageswherethefocus‐groupdiscussionshadbeenconducted.Then,33networkleadersatnationalandprovincial levelswere interviewed.Thefirst threeto  fivemost importantboardmembers were selected. According to the networks’ statutes and regulations, theboards shouldhave11‐13members. Inpractice, only the first threepositions – thepresident,secretaryandtreasurer–andthoseof theirdeputiesmatter.About thesepositionsconflictsoccurredmoreoftenbecauseofthepowerandresourcesattachedto them. In the dailymanagement, however, deputies aremostly put aside and leftwith an empty shell. The main position‐holders are omnipresent in order to keepcontrol of resources andopportunities. Therefore,we selected the first threeboardmembers and their deputies where available to participate in the individual inter‐views.Inaddition,ofsomenetworksthecoordinatorsandthemanagerswereinter‐viewed when available. The holders of these positions were added to the samplebecause of their knowledge of networks’ activities and membership development.While coordinators were board members, managers were mostly recruited asemployeesofthenetwork.However,mostmanagersofthebreakawaynetworkswerealso among their proponents. Finally, a systematic samplingprocedurewasused toselect 115 grassroots farmers from the lists of village groups’members of selectednetworkstocompletethesampleoffarmers.Intheend,thisresultedinthecoveringof nine operational networks,which are listed in Table 3.1where their operationalstatusisstated.

3.3.3 Datacollectionmethods

Focus group discussions and open interviews. The focus‐groups discussionsessions were organised in ten villages distributed over four provinces. Thediscussions used a guide of fifteen items that addressed various aspects of cottonnetworking. The guide started with the institutional mapping of the village.Thereafter,theimportanceofcotton,theinteractionsbetweencottonandnon‐cotton

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

57

organisations, and collective action in the villagewerediscussed. Then the types ofconflicts occurring within and between networks were addressed, including theprofile of leaders, and the impact of inter‐ and intra‐group conflicts in collectiveaction. Factors that determine farmers’ decision to break away from and to joinnetworkswerediscussed.Inadditiontothefocus‐groupdiscussions,openinterviewswere carried outwith some key actors of the sector, such as agricultural extensionworkers, employeesof cottonbodies aswell as experienced farmers.Data from thefocus‐group discussions were analysed using descriptive statistics to assess thefrequency of different factors cited in villages. The frequencies indicated theprevalence of the factors and the extent towhich theywere shared throughout thecottonproduction zones. Basedon farmers’ reasoning, the factorswere groupedbysimilarityforthesemi‐structuredquestionnaire.

Questionnaire. A semi‐structured questionnaire was first submitted to networkleadersatnational levelandsecondtograssrootsfarmersandleadersatthedistrictandvillagelevels.Respondentswerevisitedattheirplace,eitherathomeorontheirfarms for face‐to‐face interviews that lasted foraboutone‐and‐half to twohoursonaverage.Tofacilitateafluentinterviewthatflowedaccordingtothereasoningoftherespondents,theorderofthequestionsdependedontherespondents'answers.

Thecontentofthequestionnairewastoahighextentgovernedbytheresultsfromthequalitative research, i.e. the emergence of cotton networks and problems thatoverwhelmed their functioning, and resulting in their breaking apart. The mostimportant causes for breaking away were identified in the previous qualitativeresearch. These are the lack of payment to farmers (which constrained farmers’investment projects), the mismanagement of resources (including unequal re‐distribution of inputs, and themanipulation and corruption of networks leaders byprivate actors), and the expansion of villages (including the subsequent increase ofcottonproduction).Thesehardenedthemanagementofthevillage‐basedgroups.Thequestionnaire was built on these factors, of which the most frequently mentionedweresubmittedtoindividualfarmersforaranking.

Thequestionnaireconsistedofeightsections(seeAppendix2).Thefirstsectiondealtwith the socio‐economic and demographic backgrounds of farmers, such as age,education,maritalstatusofthehouseholdheadandsizeofthehousehold.Otherdataincluded theexperience incotton farming, thehierarchyof incomesources, and theplaceofcottoninagriculturalincome.Dataoncottonproductionduringthelastfiveyearsinareaandtonnagewerealsocollected,followedbythefarmer’scurrentstatusin cotton production (whether s/he was still producing or had abandoned cottonpermanentlyortemporarily).Respondentswerealsorequestedtolisttheirmember‐ship of organisations other than cotton and explain the relations between theseorganisations and cotton networks. A further question concerned the differentpositionstherespondenthadintheseorganisationsandtheirduration.Questionsinthe second section were about the emergence of the network the respondents

CHAPTER3

58

currentlybelongedto: thecontextof itscreation, itssizeatdifferentstages,etc.Thethird section concerned the respondents' networking trajectory, followed by therankingoffactorsthatmighthaveinfluencedthistrajectorysuchaspersonalqualities,thesourcesofsupporttheybenefitedfrom,etc.Thecausesforbreakingawayandthemotives for joining other networks were also discussed. In the fourth section, therelationshipsbetweennetworkproponents(instrumentalinthecreationofnetworks)were explored. The fifth section ended the interview by the ranking of suggestedreform actions that were identified during the focus group discussions. For theinterviewwith grassroots farmers and lower levels leaders, the sections four to sixwereslightlymodified.Thepartonrelationshipsbetweenproponentswas replacedbyquestionsabouttherankingofconflictsoccurringwithinnetworkswithregardtotheirfrequencyandseverity.

Changingnetworkmembershipwasassessedbyfactorstermedperceivedcausesforbreakingaway,whichwererankedaccordingtotheirimportanceforfarmers.Thesetofcausescontainedsevenstatementswhichwererankedfrom1(mostimportant)to7 (least important)byall the respondents.The statementsofwhich the importancefor leavinganetworkhad tobe ranked included "lackofpayment", "manipulation",and so on. Determinants of farmers' decisions to join or create networks weremeasured in the same way. Eight driving factors were put to the farmers asstatementsforranking.Theyincluded"paymentontime"and"expectationofprofit",etcetera. Additional factors that could motivate farmers to join or create were theperceivedsourcesofsupport in joiningorcreatingnetworks.Thesesources, rankedbyfarmersaccordingtotheirrespectivecontributiontojoiningorcreatingnetworks,containedsevenstatementslike"kinshipasasourceofsupport"and"professionalsofcottonasasourceofsupport”.Fortherankingquestions,allthefactorswereputasstatementstothefarmers,whowerethenrequestedtorankthemfrommosttoleastimportant.

3.4 Findingsfromthequalitativeresearch

3.4.1 Emergenceofbreakawaynetworks

The starting point of the atomisation of cotton networks was a leadership conflictabout the control of a district cotton unionwithin the cotton belt. The conflictwasbackedupbylocalpoliticalleadersonthegroundthatthecurrentleadershipdraggedonforyears.Theresultingdissidentnetworkemergedasanacceptablesolutionsincethe country’s political system guarantees freedom of association. This first‐everdissident network put an end to the full control of the former federation over themanagementoffarmerorganisations.Indeed,opposingtheleaderswasriskybecauseitwoulddeprivethedissidentsfromaccessinginputs.Proponentsofthefirstdissidentnetworkexperiencedsuchcoercivemeasuresthattheywereforcedtoresorttonon‐officialsupplierstogetinputsduringtheirfirstyear.Forthattheyeagerlyclaimthat

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

59

“nobody raisedhishand to contest the former leadersbeforewedid so.” From thispoint onwards, the federation could be circumvented by cotton farmers to accessproductioninputs.Farmerswereofferedalternativesolutionstostayingintoremnantnetworks:eitherjoiningotherexistingnetworksorcreatingnewones.Thus,breakingaway and creating new networks became a contest among cotton farmers in thecotton belt, from where more than three quarters of the breakaway networksemerged.

The multiplication of cotton networks went along with an increase of mismanage‐ment,which the leaders tried to coverbyblackingout communication.Particularly,new leaders turned out to behave more selfishly than the former leaders theycontested.Formanyfarmers,however, theatomisationwasexacerbatedbyexternalactorstocottonorganisations.Amongtheseareinputssuppliersandcottonginners,the crucial partners of farmer organisations, as well as political leaders and local“intellectuals”,as thefarmersqualifythelastcategory.Cottonfarmersaccusedtheseactors of connivance and ofmanipulating and corrupting farmers in the process ofatomisation.The accusationsderived from the central role of theboard secretaries,whoselackoftransparencyintheirmanagementofcottonorganisationscontributedto the farmers’ indebtedness. Cotton farmers contested the status of being the onlyproponentsofbreakawaynetworks.Executivesofthecottoncompanies,bothprivateand the State‐owned one, were involved in this as well. Thanks to the growinginfluence of inputs suppliers and private ginners in the wake of the liberalisationprocess, they controlled the cotton production of many networks. In addition todisturbing the cotton supply to the State‐owned cotton company,which previouslyhadthemonopolyoncottonginning,privatecompaniesalsoresortedtobreachingtherules regulating cotton production. To counter the “unfair” competition and securecotton supplies for their ginning mills, executives of the State parastatal companypromotedtheirownnetworks.Takingadvantageoftheirstrategicposition,theyputtogether farmers fromdifferentdistrictswhohardlykneweachotherandcouldnothavecreatedanetworkthemselves.

Aswiththefirstbreakawaynetwork,mostofthelateronesalsoemergedbecauseofintra‐groupconflicts,butthistimeofadifferentnature.Theconflictswereaboutlackofpaymentresultingfrommismanagementofnetworkresources.Consequently,beingpaid on time and getting inputs adequately remained farmers’ main motives forjoining other networks. At the back of the conflicts, however, was the objective offarmerstotakecontroloverthecontestedresourcesformoreprofit.Toachievethis,theyneeded legitimacy,which they sought not only in bondingbut also in bridgingties.Thesearchfor legitimacywasalsoreflectedinthediscourseusedby leaderstoconvince fellow farmers to join them and in the dominant relationships amongmembers of the network's stronghold. The discourse used by leaders included notonly addressing farmers’ needs by promising better management and payment ontime,butalsodenigratingthecontendingnetworksinordertoincreasetrustintheirown.Asforthetypesofrelationships,theydifferedaccordingtothenetworkinvolved

CHAPTER3

60

andthecontextofitsemergence.Whilesomeproponentsofnetworksweresupportedattheoutsetbybondingties(kinandmembersoftheirethnicgroup),othersgottheirfirstsupportfrombridgingtiesorfromlinkingties(peopleinpowerpositions).

3.4.2 Fromweaktiestolinkages

When the first breakaway network (AGROPE) was emerging, the lack of paymentprevailedonasmallscaleinafeworganisationsasaconsequenceofmismanagementpractices, and was looming in others. The proponents of the new network, whooriginatedfromonevillage,extendedtheircampaigntomanyothercottonproducingdistrictsfromwheretheyalsoselectedsomeleadersforboardpositions.Thenetworkwasanhybridorganisationthat includedtransporters, inputssuppliersandginners,hence its denomination as an “associationof producers andbusinessmen”.Throughthe businessmen, the network proponents gained support from political leaders inpower position, who had a great stake in the emergence of new networks. Thiscomplexmembershipmadethenetworkamixofactorswhereonlyfewwererelatedthroughkinshiporfriendship,thoughgeographicalproximitycouldhavealsoplayedarole. “Theonly thing thatbroughtus together is cottonproduction”, saidoneof theproponentswhowasleadinganotherbreakawaynetwork.

A virtual advantage of this lack of initial social tieswas the increased likelihood ofextendingthenetworkthroughoutproducingareas.Eachoftheproponentscouldusehis/her social ties to make cotton farmers join and facilitate the growth of thenetwork.Additionally,thenewleadersstrategicallyadoptedthepracticeofimmediatepayment upon selling the cotton, which in some regions raised enthusiasm amongcottonfarmerstomassivelyjointhebreakawaynetworkfromitsoutset.Mostofthecotton networks referred to by respondents in this research followed this pattern.However, the divergence of the interests (e.g. profiteering on input supply againstmaking a living by producing cotton) inhibited cohesive actionwithin the network.With the increasing financial stakes, leadership conflicts inflated andweakened thenetwork, which ended up falling apart. Mismanagement practices escalated andbecameacriticalcauseforbreakingaway.Thus,mostfarmerslinkedtheemergenceof new networks to mismanagement practices. In all cases, farmers from differentdistricts joined theireffort to “combat thesepractices”and toachieve their commonpurpose:theprofitabilityofcottonproduction.Thesedynamicsshowtheimportanceofstrongtiesfornewnetworks.

3.4.3 Fromstrongtiestoweakties

Onenetwork(UPROCOB)oftheninebreakawaynetworksincludedinthisresearchisknownasanethnicity‐basednetwork.Itemergedinavillagefromtheinitiativeofoneperson. Its prominent leader purposively mobilised farmers from his Gando ethnicgroupand"didall thenecessaryadministrativeprocedures",according tothecurrentpresidentofthenetwork.Whilethenetworkexerteditsinfluencefirstinitsvillageof

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

61

birthandamongthemembersofthisparticularethnicgroup,thenecessityofgettingbroadlyacceptedandbeingconsideredanetworkinsteadofafarmergroupforcedtheproponents to extend the network beyond the village. But still, the districts andvillages targetedwere thosewithpopulationsdominatedby the sameethnicgroup.Although there was no (open) ethnic‐based competition for leadership of cottonnetworks, the leaders of this network used in their discourse the sensitive issue ofethnicity to attract members on the ground of defending their interests. Thiscommunication strategy got them more members. Indeed, the leaders were easilytrustedbythefarmers,whotillthattimehadhardlycomeacrossaleaderfromtheirown ethnic group. In the area, members of this ethnic group are known as hard‐workingfarmers,andtheyareperceivedasbeingatthelowerendofthelocalsocialhierarchy (Jones, 1998; Lombard, 1998). Thus, the success and cohesion of thisparticularnetworkseemstobebasedmainlyonresortingtorelationalsocialcapital,i.e. the ethnic solidarity, and its subsequent homogeneity. Later on, the networkspreadtosocio‐demographically(andethnically)differentvillages.

3.4.4 Preliminaryconclusionsbasedonthequalitativeresearch

The results from the qualitative research allowed for generating a number of pre‐liminary conclusions regarding the reasons for the breaking up of cotton networks.Thesewereusedtoformulatehypothesesthatcould,at leastpartly,betestedinthesurvey research. The internal network dynamics were influenced by severalidentifiable factors, causing and shaping the process of atomisation. Causes thatincreasedthelikelihoodoffarmerstobreakawayfromtheirnetworkwereidentifiedas push factors. These include: the lack of payment, mismanagement of resources,poor communication, intra‐group conflicts, manipulation, the expansion of villages,and imitation. These push factors can be grouped into internal and external ones.Internalpush factors are those related tonetworkmanagement (the first fourpushfactors);externalpushfactorsarethosethatoriginatedfromoutside‐networkactors(the last three push factors). From the emergence of the first breakaway networkonwards, farmershadbeenofferedmorenetworkingoptions.Theycouldeither joinoperatingnetworks,createnewnetworks,orstayintheremnantones.Thefollowingsectionsprovideanoverviewofthesepreliminaryconclusionsandthecorrespondinghypotheses.

Push factors.Delays inpayment forcottonyieldsoften lasted longer thanexpectedandoftenresultedinlackofpayment,whichpushedfarmerstoleavetheirnetworksin searchofabettermanagedone.Therefore, the lackofpaymentshouldpositivelyinfluence the probability of breaking away, given that earning income is the mainpurpose of cotton production.Mismanagement and unsettled conflicts can result inselfish and opportunistic behaviour. Indeed, the mismanagement of networkresources(bothinputsandfinancialoutcomes)ledtodistrustandprobablyincreasedthelikelihoodofopportunisticbehaviour(seeSvendsenandSvendsen,2004:28).Themore resources were mismanaged, the more network members would have been

CHAPTER3

62

inclinedtobreaktherulesasacounter‐reaction.Therefore,mismanagementcanalsobeexpectedtobepositivelycorrelatedwithbreakingaway.Also,intra‐groupconflictsshould positively affect breaking away. Finally, good communication is anotherimportant internal factor that would make members more collaborative and thenetworkmorecohesive.Closecommunicationwithinanetworkallowsenforcementof sanctions against individuals who violate shared beliefs or norms of behaviour(Burt,2000:374).Theextenttowhichnetworkleaderscommunicatedwithgrassrootsfarmersindicatedthequalityoftheirmanagementandwouldhavemadethenetworktighter andmore successful. Oh et al. (2006) correlated the overall effectiveness ofgroups with leaders' ability to connect to various sub‐groups. Additionally, goodcommunicationwouldenhancetheaccountabilityof leadersandthecontrolexertedby their constituents. It has been demonstrated that "organisations that are moremember‐accountablearelikelytodelivergreatereconomicandsocialbenefitstotheirmembers"(Smith‐Sreen,1996).

The liberalisationof the cotton sectorbroughtprogressively togetherprivateactors(inputs suppliers andcottonginners).Theseprivateoperatorsvied for the farmers’favour fordifferentpurposes.Forexample, inputssupplierspermanentlysought forcontracts to supply inputs tonetworks,while cottonginners looked for agreementswith networks in order to be supplied with the cotton they produced. Thiscompetition inflated unfair behaviour, such as manipulation, which is expected topositively influence breaking away. With the multiplication of internal problems,grassroots farmers in search of information from other groups and networkscommunicatedthroughformalandinformalnetworks.Farmersfromdifferentvillagesanddifferentnetworkswouldhavesharedinformationandlearnthowtheotherdealtwith similar problems. Theories on farmers’ behaviour such as adoption of agri‐culturaltechnologieshavecontendedthatcontactiscriticalintheprocessofadoption,given the human tendency to imitate fellows (Rogers, 1995). Therefore, imitationshouldpositivelyaffectbreakingaway.

Inshort,whilebothinternalandexternalpushfactorscanbeexpectedtobepositivelycorrelatedwithbreaking,withregardtonetworkcohesioningeneralandthefarmer’swell‐beinginparticulartheinternalfactorsmaybeassumedtobemoreharmfulthanthe external ones. Hence, when asked to rank the push factors according to  theirimportanceforbreakingaway,itseemsprobablethatfarmerswillrankinternalpushfactorshigher thanexternalones.Basedon this logic, andgranted that the farmers’socio‐economic and socio‐demographic backgrounds may also play a role in therelativeassessmentofthepushfactors,itseemslikelythat:

Hypothesis1.Farmersrankinternalpushfactorsforbreakingawayhigherthanexternalpushfactors.

It will further be inquired whether stayers, joiners and breakers (see section 3.1)rankedthesefactorsinadifferentway.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

63

Pullfactors.Comparabletothedecisiontobreakornot,threealternativeoptionsofchoice corresponded to different motives identified as pull factors: staying in theremnant network, joining another existing (operating) network, or creating a newnetwork. The pull factors were defined as the motives for joining other networks.These factors include:paymenton time, the facilityof getting inputs, trust inboardmembers,thesizeofnetwork,thehopeforboardpositions,theexpectationofprofit,kinship,and the taste fornovelty.Within thepull factors,networkefficiency factorscanbedistinguishedfrompersonalaspirationfactors.

Timely payment of network members, as opposed to lack of payment that pushfarmers to break away, would pull farmers towards better managed networks.Paymentontimeisexpectedtobepositivelycorrelatedwithatomisationofnetworksand to be ranked higher by joiners than by stayers and creators. Like the financialoutcomes, thequestionof inputs supplyhas remained central to cottonproduction.Inputsupplywasalsocrucialtofoodcrops,towhichanimportantpartwasdiverted.Therefore, inputs facilitation is expected to be positively correlated with networksatomisation,andtoberankedhigherbyjoinersandstayersthanbycreators.

Meeting the two criteria above (payment on time, getting inputs) that constitutedfarmers’overridingconcernsatteststothenetworkeffectivenessandwasexpectedtoincrease the trust in itsboardmembers.Trust servesasa lubricant for cooperation(Misztal,1996).With theprevalenceofmismanagement‐relatedconflictswithinandbetweennetworksthathadincreaseddistrust,trustincottonboardmemberswouldincrease from incentives such as effective supply of inputs. Trust would be rankedhigherbyjoinersandcreatorsthanbystayersandpositivelyaffectstheatomisationofnetworks.

Personalaspiration factorsareassumedtopositivelyaffect the likelihoodthatmorenetworks emerge. To get a leadershipposition (hope for boardpositions) and earnmorethanincomefromownproduction,farmerstendedtoengageincreatingbreak‐awaynetworks.Therefore,theexpectationofprofitandthehopeofoccupyingaboardposition shouldbe rankedhigherby creators thanby stayers and joiners.Also, it isexpectedthatinsituationofnoorlittleavailableinformation,farmerswouldprefertojoinnetworkswheretheyknewpeoplewhomtheycouldtrust,suchaskin.

Both network efficiency‐related factors and personal aspiration ones are likely toaffectpositivelytheatomisationbyincreasingthenumberofnetworks.Thelikelihoodof staying within a network, joining another one or creating a new network isexpected tobe affectedby a farmer’s perceptionof thesepull factors. Farmerswhoperceivesomeefficiencyinanexistingnetworkwouldjoinit,whilethosewithhigherpersonalaspirationswoulddecidetocreateanewnetwork.Otherswouldchoosetokeepthestatusquobystayingintheiroriginalnetworks.For instance,while joiners(andstayers)mayvaluenetworkefficiency,creatorsmayassignmoreimportancetopersonal aspiration (e.g. making more profit, becoming a leader) than to network

CHAPTER3

64

efficiency. Due to the membership contest after the emergence of breakaway net‐works, leaders tried to increase the efficiency of their networks. However, due tocreators’ tendency to selfishness, leaders in general facea commondilemma,whichleadsustothefollowingsetofhypotheses:

Hypothesis 2a. Farmers rank network efficiency factors higher than personalaspirationfactors.

Hypothesis 2b. Controlled for socio‐economic and socio‐demographic back‐ground, creators assign higher importance to personal aspiration factors thanstayersandjoiners.

In sum, personal aspiration factors should motivate breakers to create "their ownnetworks"ratherthanjoinexistingnetworks.

Social ties and cotton networking. Creating a breakaway network is a jointendeavour,thecompletionofwhichrequirestheuseofsocialrelationstoextendthenetwork. The stiff competition between breakaway networks also demanded thesupportofexistingsocialtiestomakenewnetworkssuccessful.Tiesinanindividual’ssocial network are believed to be of great benefit for successful networking(Granovetter, 1973). Leaderswould resort tokin, friends, fellow farmers, andothermembers of their social networks to enlarge their cottonnetworks andmake theseprosperous.

Social ties are thus expected to act as pull factors as well, by providing relationalsupport for joining or creating new networks. These social ties can be based onfriendship, kinship, ethnicity and religion, called “strong ties”, or can encompasscottonprofessionals,publicofficials,andpoliticians,whicharecalled“weakties”.Allthese ties contain social resources that correspond to what Wellman and Kenneth(2001:233)termedthenetworkcapital.AccordingtoPlickertetal.(2007:408‐9),"kinare [...] the most likely network members to be densely interconnected, fosteringcommunicationabouttheneedsandnormsofprovidinghelp."

Weaktieshaveagreatercontributioninextendingnetworksthanstrongties,becausetheybridgeactorstootherslocatedindifferentsocialnetworks(Granovetter,1983).Duringthecottonboom,cottonproductionandfarmernetworkshadbecomepoliticalinstrumentsthatmanypoliticalleaderssoughttocontrol.Thus,politicalgroupswereobviouslyimportantsourcesofsupportthatwouldincreasethecreators’efficiencyinthe process. Similarly, political leaders, professionals of cotton (private inputssuppliers and ginners and executives of the  State cotton company) played animportantroletoo.Cottonnetworkleadersneededtointeractwiththeseactorsgiventhat cotton production depends on the inputs being supplied. In return, cottonprofessionalshadagreatstakeinsupportingbreakawayleadersincreatingnetworksasastrategytosecuretheiractivities.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

65

Thoughbothtypesoftiesareimportantfornetworking,itisspeculatedthat:

Hypothesis3a.Cottonfarmersranktheimportanceofthesupportofweaktieshigherthanthesupportofstrongties.

Ifweaktiesareexpectedtoprovidefarmerswithcriticalsupportforthecreationofbreakaway networks, their bridging to new members will be positively correlatedwithatomisation.Therefore,

Hypothesis 3b. Controlled for socio‐economic and socio‐demographic back‐ground,creatorsranktheimportanceofsupportofweaktieshigherthanstayersandjoiners.

Similartothecreationofnetworks,accessingleadershippositionsincottonnetworksrequires themobilisation of resources embedded in an individual’s social network.Indeed,cottonnetworkshavebeenthe largesttypeofnetworksof their typewithinthe cotton‐producing villages and they encompass members of many other socialnetworks.Resortingtorealsocialnetworkswouldincreasethelikelihoodofgettingaleadershippositiononcottonboards,eitherbyelectionorbyunanimousdesignation.AccordingtoLin(2001),actorsinpossessionofvaluedresourcesaremorelikelytobeinvolved in decisions‐making processes. In addition, the "selection of occupants [ofpositions inhierarchical structures] favours thosewhoaresocializedand trained tocarryout[the]rulesandprocedures"(Lin,2001:34).Consequently,themembershipofassociationsotherthancottonwoulddistinguishleadersfromgrassrootsfarmers.

The escalation of leadership conflicts in cotton networks led to never‐ending plotsbetween leaderstooverthrowoneanother.Somesucceededinholdingtheir leader‐shippositionslongerthanothers,eitherwithinthesamenetworkorinnewnetworksthey helped to create. All other things remaining equal, this difference betweenleadersmayresult fromresourcesembeddedintheirsocialnetworks.Indeed,socialnetworks are potential sources of social capital, and resources facilitating action(AdlerandKwon,2000).Leaders’statusintheirsocialnetworksandtheirdegreeofresponsibilitieswouldaccruetotheirstockofsocialcapitalandhelpthemholdontotheir leadership positions on cotton boards. The number of positions occupied andthat of the social networkswould have a cumulative effect on the leaders’ stock ofsocial capital, defined by Burt (2005:4) as "the advantage created by a person’slocationinastructureofrelationships."Thisbringsustothelastsetofhypotheses:

Hypothesis4a.Thelargerthenumberofsocialnetworksafarmerisinvolvedin,themorelikelys/heistobealeader.

Hypothesis4b.Thegreatera leader’s stockof social capital, the longer s/he isappointedatboardpositions.

CHAPTER3

66

3.5 Findingsfromthequantitativeresearch

3.5.1 Operationalizationofconcepts

ThissectiondiscussesthequantitativeoperationalizationofthehypothesesproposedinSection3.4.4.Thesehypothesesweretestedthrougharangeofstatisticalanalysesapplied to thequantitativedatausingSPSS.For thatpurpose,anumberofvariableswereconstructedandcombinedwithfarmers’socio‐economicandsocio‐demographicbackgrounddatatounderstandtheprocessofatomisation.Theemergenceofbreak‐awaynetworkswaspivotaltoalltheinstrumentsusedthroughouttheprocessofdatacollection.Thepushandpullfactorswere

Definition of atomisation. The process of atomisation consists of a sequence ofdecisions made by farmers in their networking trajectory with regard to theirmembership of cotton networks. The first decision of farmers regarded leaving orstaying in a network. Correspondingly, a dependent variable called BREAK wasconstructedtakingthevalues ‘0’ forfarmerswhohavebelongedtoasinglenetworkduringtheirentirecottonnetworklife(stayers),and‘1’whentheybrokeawayfromtheiroriginalnetworkatleastonceduringtheircottoncareer(breakers).

A second decision of farmers related to their choice to either join another existingnetwork or (help) create a ‘breakaway’ network. Ultimately, the process of atom‐isationofcottonnetworksresultedinthreestatuses.Thederiveddependentvariablemeasuresthelikelihoodthatfarmers’choicesincreasethenumberofnetworks,eitherby staying into remnantnetworks, by joiningexistingnetworks, orby creatingnewnetworks.ItisnamedATOMISEandhasthreevalues:‘0’forstayers,‘1’forjoiners,and‘2’forcreators.Thepushandpullfactorsinteractedwithfarmers’socio‐economicandsocio‐demographic backgrounds to shape this process of atomisation. Farmers’perceptionoftheeffectsofpushandpullfactorswasmeasuredbyrankingthemfrommosttoleastimportant.Thefactorswererankedfrom1(mostimportant)to7(leastimportant).Incasefarmerscouldnotdecideontherelativeimportanceoftwocauses,thesecausesmayoccupythesamerank(exaequo).

A thirddecisionof farmers regarded their ambition toobtain a leadershipposition,whichwentalongwith theprocessofatomisation.Thisdecision led toadependentvariablecalledLEADSTATaboutfarmers’leadershipstatus,assigningthevalues‘0’forfarmerswhohadneverbeenappointedasboardmembersand‘1’otherwise.

Operationalization of the research hypotheses. The hypotheses formulated inSection3.4.4were testedas follows.Forhypotheses1,2aand3a, themeanof eachpushandpull factorandofeachsourceofsupportwascalculated.The factorswerethenrankedaccordingtotheirrespectivemeanforapaired‐t‐test.

Tofurtherhypothesis1andcheckfordifferencesbetweenthethree‐leveloutcomesofthe atomisation, an analysis of variancewasperformed.Thedifferenceof themean

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

67

ranks of push factors between stayers ( =0), joiners  ( =1), and creators ( =2)regarding the pth ranked internal cause is equivalent to the simple linear

regression:

, (1)

where referstothemodelconstant(inthisoperationalization, themeanrankfor

causep); denotestheatomisationstatus;and isanunexplainederrorterm.Thehypothesisis(partly)verifiedwhendifferencesexistbetweenthethreecategoriesforat least one internal cause. However, it is still possible that socio‐economic anddemographic backgrounds are the main drivers for ranking the causes. Therefore,observeddifferencesbetweenstayers,joinersandcreatorsmaybecausedbydifferentbackgrounds of the actors in these categories. In the multiple regression analysisperformed,possibledifferencesbetweenstayers,joinersandcreatorswerecontrolledforsocio‐demographic(SD)andsocio‐economic(SE)background:

∑ ∑ , (2)

Hypothesis 1 is confirmedwhen at least one differs from zero. Incidentally, this

linearmodelisequivalenttoanANCOVA(Neteretal.,1996:663).

Totestthehypotheses2band3bthesameprocedureoffollowingupwasusedasforhypothesis 1. An analysis of variance comparing themean ranks of the pull factorsinto atomisation and those of the sources of support between the three‐leveloutcomesofatomisationwasfirstperformed.Ananalysisofco‐variancefollowedsuittodeepen the test of thehypotheses, checkingwhether themean ranksofpersonalaspiration factors and the importance assigned to perceived strong ties by creatorsresultedfromtheirstatusonly,orwereinfluencedbytheirsocio‐economicandsocio‐demographicbackgrounds.

Hypothesis4awastestedthroughananalysisofcovariancecontrastingleaderswithgrassroots farmersregarding thenumberof socialnetworks theyweremembersof,and correcting foranextraneous influenceof their socio‐economic and socio‐demo‐graphicbackgrounds.

Finally,alinearregressionwasusedtotestfortherelationbetweentheleaders’stockofsocialcapitalandthedurationoftheirappointmentoncottonboards(hypothesis4b).Thetotaldurationatleadershippositions ofafarmer isafunctionofhis/herstockofsocialcapitalandthecombinationofhis/hersocio‐economicandsocio‐demographicbackgrounds.Thefunctionhasthefollowingform:

∑ ∑ , (3)

where represents the stock of social capital, which is the combination of thenumberofsocialnetworksofwhichthefarmerwasamemberwiththepositionss/he

CHAPTER3

68

occupied in thesesocialnetworks.Hypothesis4b is confirmedwhenat leastone

differsfromzero.

3.5.2 Empiricalresultsofthesurvey

Description of the sample in term of socio‐economic and socio‐demographicbackgrounds. In total, 148 cotton farmers from nine networks were interviewed,comprising24womenand124men.Withregardtobreaking,thesamplecounted36stayers and 112 breakers. Tables 3.2a and 3.2b present the socio‐demographic andsocio‐economic characteristics of the sample. Table 3.2a is a cross tabulation offarmers’levelofeducationandage(ingroupsof10years)bygender,whileTable3.2bpresentsthefarmers’profilewithregardtothehouseholdfeatures:size,thenumberof workers, the experience in cotton production, the share of cotton income in thetotalagriculturalincome,etcetera.

Table3.2a Crosstabulationoffarmers’levelofeducationbyagebygender

Table3.2b Characteristicsofthesample(metricvariables)

Householdsize

Numberofworkers

Cottonexperience

Shareofcottonincome(2005‐

2010)

Averagecottonarea(2005‐2010)

N 147 147 148 133 147

Mean 16.35 10.27 20.16 45.15 2.60

SD 11.37 8.77 10.29 26.53 2.52

Median 14.00 8.00 20.00 46 2.00

Mode 14 3.50 20 .00 .00

Min. 1 1 0 .00 .00

Max. 100 83.5 46 1.00 14.80

Source(bothtables):Surveydata,2009–2011.

Gender LevelofeducationAgegroups(in10‐yearincrement)

Total<30 30–40 40–50 50–60 >60

Female

Noeducation 1 8 9 2 0 20Primary 0 0 0 0 1 1Secondary&more 0 2 0 1 0 3Total 1 10 9 3 1 24

Male

Noeducation 5 14 17 18 9 63Primary 3 9 1 11 2 26Secondary&more 2 8 14 7 4 35Total 10 31 32 36 15 124

Total

Noeducation 6 22 26 20 9 83Primary 3 9 1 11 3 27Secondary&more 2 10 14 8 4 38Total 11 41 41 39 16 148

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

69

More than50per cent of farmers in the sample has no formal education. Educatedwomenrepresentonly17percentoftheircategory,whileformentheratioeducatedto non‐educated is about 1. Hence, male cotton farmers are significantly moreeducatedthantheirfemalecounterparts.About80percentofthefarmersareintheagesof40to50yearsold,while50percentof thewomenisunder40.Theaveragehouseholdsize is16.35,withanaverageof11workersperhousehold inmaleadultequivalents.Farmers’experienceincottonproductionrangesfrom0to46years,withameanof20yearsandastandarddeviationof10.Theincomegeneratedbycottonduringthefiveyearspriortothesurvey(2005‐2010)wasonaverage45percentoftheagricultural income,whileagriculture is themainoccupationandoften theonlysource of income in the area. For the sameperiod (2005‐2010), the average cottonplot extended from 0 ha to about 15 hawith amean of 2.6 ha. Theminimum of 0representsthosewhoabandonedtheproductionpermanentlyortemporarily.

The process of atomisation resulted in about 20 per cent of stayers, 51 per cent ofjoiners,andabout28percentofcreators(Table3.3).

Theresultsindicateasignificantassociationbetweenleadershipstatusandthethreeatomisation statuses (χ2(1) = 42.32, p < .001). About 58 per cent of stayers in thesample are grassroots farmers against 42 per cent of leaders, while 40 per cent ofbreakersaregrassrootsfarmersagainst60percentofleaders.

Push factors.Themeansof thepushfactors foratomisationpointedout the lackofpayment(M=2.4,SD=1.68)asthefirstpushfactor,followedbymismanagementofresources(M=2.8,SD=1.67).Thelackofcommunication(M=3.5,SD=1.63),intra‐groupconflicts(M=3.9,SD=1.68),andmanipulation(M=4.3,SD=1.89)rankthird,fourth,andfifth,respectively.Expansionofvillages(M=5.5,SD=1.62)andimitation(M=5.6,SD=1.36)occupythelastranks.

Table3.3 Farmer'sleadershipstatusbystatusinatomisation

Statusinbreakingawaynetworks

LeadershipstatusAtomisation

Total Chi‐sqStayers Joiners Creators

StayersGrassrootsfarmers 18 3 0 21

χ(1)=3.29p=.141

Leaders 12 1 2 15

Total 30 4 2 36

BreakersGrassrootsfarmers 0 44 1 45

χ(1)=36.57p<.001

Leaders 0 28 39 67

Total 0 72 40 112

TotalGrassrootsfarmers 18 47 1 66 χ(1)=42.32

p<.001Leaders 12 29 41 82Total 30 76 42 148

Source:Surveydata,2009–2011.

CHAPTER3

70

Thepairwisecomparisonsofthepushfactors(Table3.4)showsignificantdifferencesbetween lack of payment and mismanagement of resources (p < .05). Significantdifferences also exist between mismanagement of resources and lack ofcommunication(p<.01),andbetweenmanipulation,andtheexpansionofvillages(p<.001).Theseresultsprovideenoughevidencetoconfirmhypothesis1.

Table3.4 Paireddifferencesofmeansranksofperceivedpushfactorsforatomisation

Pairpushfactors M(SD) t Diagnos‐ticLackofpayment‐Mismanagementofresources ‐.42(2.40) ‐2.09* Different

Mismanagementofresources‐Lackofcommunication ‐.69(2.55) ‐3.25** Different

Lackofcommunication‐Intra‐groupconflicts ‐.33(2.50) ‐1.57 Exaequo

Intra‐groupconflicts–Manipulation ‐.44(2.79) ‐1.89 Exaequo

Manipulation‐Expansionofvillages ‐1.21(2.75) ‐5.26*** Different

Expansionofvillages‐Imitation ‐.13(2.3) ‐.65 Exaequo

*=p<.05,**=p<.01,***=p<.001

Source:Surveydata,2009–2011.

Theanalysisofvariance(ANOVA)ofthemeanranksoftheperceivedpushfactorsforatomisation(Hypothesis1)revealsinTable3.5thatonlymismanagement(aninternalcause)hasastatisticallysignificanteffectontheoutcomeofatomisation(p<.05).

Table3.5 Meanranksofperceivedpushfactorsforatomisation

Lackofpayment

Mis‐manage‐ment

Lackofcommuni‐cation

Intra‐groupconflicts

Manipu‐lation

Expan‐sionofvillages

Imitation

Stayers(N=27) 2.6 3.3 3.1 3.9 3.7 5.8 5.6

Joiners(N=75) 2.2 2.9 3.7 3.9 4.3 5.2 5.7

Creators(N=41) 2.7 2.3 3.5 3.7 4.6 5.8 5.5

Overall(N=143) 2.4 2.8 3.5 3.8 4.3 5.5 5.6

F(2,142) 1.77 2.99 1.31 0.36 1.82 2.01 0.42

Significance ns p<.05 Ns ns ns ns ns

Correc‐ted

F(7,135) 1.19 0.95 0.95 0.65 2.35 0.79 1.79

Sig. ns ns Ns ns p<.05 ns ns

Note:ns=notsignificant.

Source:Surveydata,2009–2011.

Controllingpossibleobserveddifferencesforsocio‐demographicandsocio‐economicfactors, the corrected model shows a significant difference between the threecategoriesoffarmersonlywithregardtomanipulation(p<.05),whichisanexternalpushfactorforatomisation.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

71

Pullfactors.Thefirstpullfactorforatomisationisrelatedtopayment,similarlyasforpush factors. Payment on time (M = 1.6, SD = 1.04)was followed by the facility ofgettinginputs(M=3.3,SD=1.55)rankingsecond,andbytrustinboardmembers(M=3.7,SD=2.01)at thethirdposition.Onaverage,kinshipwasclassified fourthpullfactor(M=4.8,SD=2.04),andthehopeforboardpositionsfifth(M=5.2,SD=2.14).Thefactorsthatappearedasthe least importantonesweretheexpectationofprofit(M=5.4,SD=1.73)rankingsixth, thesizeofnetworks(M=5.9,SD=1.68)rankingseventh,andfinallythetastefornovelty(M=6.2,SD=1.8).

Thepairwisecomparisons indicatedsignificantdifferencesbetween themeanranksofthreepairedpullfactors(Table3.6).

The difference between the first two pull factors (payment on time and facility ofgettinginputs)ishighlysignificant,showingfarmers’clearpreferenceforpaymentontime. Highly significant differences also exist between trust in boardmembers andkinship.Aborder‐significantdifferenceexistsbetweenexpectationofprofit and thesizeofnetworks(p< .05).Theclassificationaboveandthepairwisecomparisonsofthemeanranksconfirmhypothesis2a.

Table3.6 Paireddifferencesofmeansranksofperceivedpullfactorsforatomisation

Pairedpullfactors M(SD) t Diagnostic

Paymentontime−Facilityofgettinginputs ‐1.66(2.00) ‐9.82*** Different

Facilityofgettinginputs−Trustinboardmembers ‐.40(2.64) ‐1.80 Exaequo

Trustinboardmembers–Kinship ‐1.18(3.08) ‐4.52*** Different

Kinship−Hopeforboardpositions ‐.33(3.16) ‐1.23 Exaequo

Hopeforboardpositions−Expectationofprofit ‐.21(3.02) ‐.82 Exaequo

Expectationofprofit−Sizeofnetworks ‐.53(2.61) ‐2.41* Different

Sizeofnetworks−Tastefornovelty ‐.26(2.70) ‐1.129 Exaequo

*=p<.05,***=p<.001

Source:Surveydata,2009–2011.

Themeansof twoperceivedpull factors foratomisationweresignificantlydifferentbetweenstayers,joinersandcreators(Table3.7).Whiletrustinboardmembers(anetworkefficiencyfactor)wasdifferentlyrankedbythe three categories of farmers (p < .05), the hope for board positions (a personalaspiration factor) was ranked higher by stayers than by joiners and creators whoassignedasimilarranktothat factor(p< .05).Thecontrol forobserveddifferenceswith socio‐economic and socio‐demographic background confirms the differenceabouttrust inboardmembers(p< .05).Thecontrolrevealsanadditionaldifferencebetween stayers on the one hand and joiners and creators on the other handwithregardtotheexpectationofprofit(p<.05).Therefore,hypothesis2bisconfirmed.

CHAPTER3

72

Table3.7 Meanranksofperceivedpullfactorsforatomisation

Pay‐mentontime

Facilityofgettinginputs

Trustinboardmembers

Kin‐ship

Hopeforboardpositions

Expectationofprofit

Sizeofnetworks

Tastefornovelty

Stayers(26) 1.5 3.8 4.8 4.7 3.8 5.2 6.3 5.8Joiners(75) 1.6 3.1 3.7 4.7 5.5 5.4 5.7 6.2Creators(38) 1.6 3.2 2.8 5.2 5.5 5.4 6.0 6.3Overall(139) 1.6 3.3 3.7 4.8 5.2 5.4 5.9 6.2F(2,136) .18 2.06 8.13 .74 7.12 .12 1.31 .71Significance Ns ns p<.001 Ns p<.001 ns ns ns

CorrectedF(7,122) 1.01 1.44 1.92 0.98 1.39 1.88 0.83 1.03Sig. Ns ns p<.05 Ns Ns p<.05 ns ns

Note:ns=notsignificant.

Source:Surveydata,2009–2011.

Table3.8 Paireddifferencesofthemeanranksofperceivedsourcesofsupportforcreatingnetworks

Pairedsourcesofsupport M(SD) T Diagnostic

Professionalsofcotton–Kinship ‐.25(2.89) ‐1.00 Exaequo

Kinship–Friendship ‐.82(2.22) ‐4.33*** Different

Friendship–Publicofficials ‐.08(3.04) ‐.31 Exaequo

Publicofficials–Ethnicity .91(2.91) ‐3.69*** Different

Ethnicity−Politicalgroups ‐.26(2.84) ‐1.08 Exaequo

Politicalgroups–Religion ‐.13(2.76) ‐0.56 Exaequo

***=p<.001

Source:Surveydata,2009–2011.

Socialnetworksof cotton farmers. The overall ranking of the social support tiesyieldsaweaktiethatranksfirst.Professionalsofcotton(inputssuppliersandcottonginners)appeartobethemostimportantsourcesofsupport(M=2.7,SD=1.70)andkinship(M=2.9,SD=1.74)asthesecond.Friendship(M=3.7,SD=1.80)andpublicofficials(M=3.8,SD=2.00)areclassifiedthirdandfourth,respectively.Religion(M=5.1,SD=1.68) is ranked last,precededbypoliticalgroups(M=5.0,SD=1.95)andethnicity (M=4.7, SD=1.63) as, respectively, the fifth and sixth sourceof support.However, the paired comparison reveals differences between only two pairs.Differences of themean ranks between kinship and friendship, and between publicofficialsandethnicityarehighlysignificant(Table3.8).However,theoverallranksofweakandstrongtiesleadtorejectionofhypothesis3a.Significantdifferencesbetweenstayers,joinersandcreatorswerefoundforthemeanranksofthreeperceivedsourceofsupport(Table3.9).

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

73

Table3.9 Meanranksofperceivedsourcesofsupportforcreatingbreakawaynetworks

Cottonprofessionals

Kinship Friend‐ship

Publicofficials

Ethnicity Politicalgroups

Religiousgroups

Stayers(N=27) 3.0 3.1 3.8 4.4 4.4 4.4 4.9Joiners(N=73) 2.8 2.9 3.9 3.4 5.1 5.1 4.9Creators(N=38) 2.3 2.8 3.4 4.3 4.3 5.3 5.7Overall(N=138) 2.7 2.9 3.7 3.8 4.7 5.0 5.1F(2,136‐141) 1.54 .23 1.16 4.38 3.71 1.92 3.33Significance ns ns ns p<.05 p<.05 ns p<.05

CorrectedF(7,130) 0.73 0.69 0.85 2.31 2.26 0.97 1.76

Sig. ns ns ns p<.05 p<.05 ns ns

Note:ns=notsignificant.

Source:Surveydata,2009–2011.

Joiners assign higher importance to public officials (a weak tie) than stayers andcreators(p<.05).Also,whilestayersandcreatorsrankethnicityhigherthanjoiners(p<.05),stayersandjoinersrankreligionhigherthancreators(p<.05).Thecontrolof observed differences for socio‐economic and socio‐demographic backgroundconfirmsthedifferencesbetweenthethreecategoriesforthepublicofficials(p<.05)andethnicity(p<.05).Becauseallfarmersrankprofessionals(a“weaktie”)highest,hypothesis3bisrejectedbythesedata.Totesthypothesis4a,a2(leadership)×2(gender)×3(education)ANCOVAdesignwas performedwith themembership of social networks as the dependent variableandleadership,genderandeducationastheindependentvariables(Table3.10).Thecovariatesareage, thehouseholddependency ratio, the shareof cotton income, theextent of cotton production. The preliminary evaluation of the homogeneity ofregressionslopesindicatesanon‐significantrelationshipbetweenthecovariatesandtheindependentvariables(p=.72). Leadershipissignificantlyassociatedwiththenumberofsocialnetworksfarmersareamemberof (p< .01, η2 = .06), confirminghypothesis4a.Planned contrast revealsthatbeingaleaderslightlyincreasesthenumberofnetworksafarmerisamemberof,comparedtobeingagrassrootsfarmer(p< .05).Also,malefarmershavesomewhathighermembershipof social networks than female farmers (p< .05), and a greatershare of cotton income is associated withmoremembership of social networks (p<.05).

CHAPTER3

74

Table3.10 Analysisofcovarianceforfarmers’membershipofsocialnetworksbytheirleadershipstatusbygenderandeducation 

Parameters F(1,110) B(SE) t Partialη2

Intercept 6.39* 1.85(0.60) 3.10** .06Age 2.94 0.01(0.01) 1.71 .03Householddependencyratio 2.63 ‐0.29(0.18) ‐1.62 .02Shareofcottonincome(%) 7.20** ‐0.97(0.36) ‐2.68** .06Averagecottonplot(ha) 0.54 ‐0.02(0.02) ‐0.73 .01LeadershipStatus(0=GrassrootsFarmers) 7.28** ‐1.53(0.69) ‐2.22* .06Gender(0=Female) 2.01 1.59(0.69) 2.30* .02Education(0=Noeducation) ‐0.29 ‐0.65(0.28) ‐2.36* .05Education(1=Primaryschoollevel) 0.11(0.28) 0.40 .00InteractionLeadership*Education 2.24 .04

CorrectedModel:R2=Square=.29,AdjustedR2=.20,F(13,110)=3.38***,η2=.29

*=p<.05,**=p<.01,***=p<.001

Source:Surveydata,2009‐2011.

Thelinearregressionofthedurationofappointmentatleadershippositionsincottonboardsconfirmstheeffectofsocialnetworking(Table3.11,Hypothesis4b), therebyconfirminghypothesis4b.

Table3.11 LinearRegressionofdeterminantsofthedurationofappointmentonboards

Model1 Model2 B(SE) T ß B(SE) t ßConstant 1.92(4.63) 0.42 0.01(4.49) 0.00

Secondarylevelandmore 2.72(1.18) 2.31* 0.35 2.07(1.15) 1.79 0.27

Indexofsocialcapital ‐ ‐ ‐ 0.52(0.20) 2.61* 0.37

N=71,R2=.12,ΔR2=.10,F(8,62)change=6.80,*=p<.05,

Source:Surveydata,2009‐2011.

Farmers’socio‐economicandsocio‐demographicbackgroundswereagaincontrolledto test for the effect of social capital, whichwas added in the secondmodel of theregression.Whileeducation(abovetheprimarylevel)isborder‐significantinthefirstmodel, it isnot inthesecondmodel.Thestockofsocialcapitalhasasignificantandpositiverelationshipwiththedurationofappointmentonboards.TheR2‐changeof10percentatteststotherelativelyimportantcontributionofsocialcapitaltothemodel.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

75

3.6 Discussion

When distinguishing farmers into breakers and stayers on the one hand, andgrassrootsfarmersandleadersontheotherhand,itwasfoundthatmorethanthreequarters of cotton farmers broke away from their original network at least onceduring their cotton cropping career, either actively or passively. This implies that alargeproportionofcottonfarmersinthecottonbeltwereinvolvedintheprocessofatomisation.The results also show thatmore thanone in two cotton farmershadaleadershippositionat,atleast,oneofthefourlevelsofcottonnetworks.Itcanthenbeinferred that, asa consequenceof theatomisationofnetworks, leadershippositionsarewelldistributedandthatmanyfarmerswereabletogetsuchpositions.However,leadersofcottonnetworksweremorelikelytobreakawaythangrassrootsfarmers.The results further tell us that the creators of new networks aremore likely to beleadersthanstayersorjoiners.Theaforementionedresultsconfirmthatbybreakingaway,farmersincreasedtheirchancesofgettingleadershippositions.

The perceived push and pull factors that influence the dynamics of cotton farmersleaving from and joining networks are related not only to themanagement of net‐works,butalsotosocialties.Theimportanceofthesetiesisstronglysupportedbytheresultsfromthequalitativeanalysis.Thequantitativeanalysis,however,didnotyieldmuchevidenceof a strong relationshipbetween theprocessof atomisationand theperceivedpushandpullfactors.

3.6.1 Pushfactors

The four sub‐groups of causes for breaking away as determined by the pairwisecomparisonimplyavariedperceptionofthepushfactorsforatomisationbyfarmers.Farmersmadeacleardifferencebetweenmismanagementofresourcesandthe lackof payment. Earning income being the main purpose of growing cotton, lack ofpayment,whichbecameacuteasinternalproblemsescalated,affectedfarmersmorethananyothercause.Lackofpaymentpushedthemtobreakaway,principallywhenitclearly resulted from mismanagement. Mismanagement of resources by leadersaffectedboththeinputsusedforcottonproductionandtheincomederivedfromtheproduce. It increased free‐ridingbehaviour fromgrassroots farmerswhosought forcompensation.Intheend,mismanagementburdenedfarmerswithdebts,andblockedtheir access to information. The lack of communication could be perceived in twoways:firstasastrategyofcoveringmismanagementpractices,andsecondasthelackof consideration from leaders. The fact that lack of communication, intra‐groupconflictsandmanipulationwereputintothesamesub‐groupofpushfactorsimpliesthat their effects are correlated and that one factor could result from the other.However,manipulationwashighly ratedby stayers,whoaccusedbreakersofbeingmanipulated by inputs suppliers. The results of the analyses of variance andcovariance, however, attest that all these causes could affect farmers regardless oftheirstatusinbreakingawaynetworks.

CHAPTER3

76

With regard to the overall ranking of the push factors, internal push factors foratomisationappearedtobemoreimportant infarmers’decisiontobreakawaythanexternal factors. Consequently, they have a negative relationship with staying intoremnant networks. However, the differences in the ranks assigned to these pushfactorswith regard to the three‐level outcomeof atomisationarehardly significant,exceptformanipulation.Thehighrankassignedtothisfactorbystayersechoestheirargumentthatmanipulationbyoutsideactorsplaysacrucialroleinpushingfarmersto leavetheirnetworks.The lackofsignificantdifferenceof themeanranksofpushfactorsbetweenstayers,joinersandcreatorsmeansthattheoutcomeofatomisationdid not depend on these factors,whichmay have affected all the farmers similarly.Indeed,allnetworkscomprisedthethreecategoriesofcotton farmers, includingtheoriginalnetworkfromwhichsomemembersbrokeawaybeforetheycameback.Thus,all networks were affected by the phenomenon of breaking away. The analysis ofcovarianceconfirmedthislackofsignificantdifferenceintherankingofthecauses.

3.6.2 Pullfactors

Thelackofanysignificantdifferencebetweenstayers,joinersandcreatorsregardingthe mean rank of payment on time testifies to its similar importance to all thesecategories inpullingthemtowardsothernetworks.Similarly tothepushing forceoflackofpaymentintheprocessofbreakingfromanetwork,paymentontimeexertedastrong pulling effect on farmers. The facility of getting inputs and trust in boardmembers formed a relevant sub‐group. Indeed, the inputs for cotton production aswellasfortheproductionoffoodcropsareofcriticalimportance,andgettinginputsdetermines trust in boardmembers. Kinship, the hope for board positions, and theexpectation of profit, which all belong to the same sub‐group as a result of thepairwise comparison, were rated lower than network efficiency factors. This resultwas confirmed by the overall ranking, which revealed that the three categories offarmers assigned a higher importance to network efficiency than to personalaspiration.

Trustinboardmemberswasthemostsignificantpullfactor,ofwhichthemeanrankconfirmedthedifferencebetweenstayers,joinersandcreatorsafterthecontrol.Also,the hope for board positions and the expectation of profit showed a significantdifferencebetweenstayerswhorankedthemhigherontheonehand,andjoinersandcreatorswho ranked them lower on the other hand. These relations suggest that i)creatorsperceivedthosewhojointheirnetworkastrustingtheboardmembers,andii) stayersperceived thehope forboardpositions as an importantmotive for thosejoining breakaway networks, which is an explanation of joiners’ and creators’behaviour.

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

77

3.6.3 Socialcapitalandcottonleadership

Cotton professionals (inputs suppliers and cotton ginners) turned out to providecritical support in the process of atomisation. This role of weak ties wasacknowledged by farmers throughout the discussions and interviews. Cottonprofessionals were perceived as the most important source of support for theemergence of cotton breakaway networks by the three categories of farmers,regardless of their leadership status. This type of bridge was therefore crucial inbuildingalargecottonnetwork,giventhattheyfacilitateaccesstomoreandfurtherties.Asthecornerstoneofnetworksatomisation,inputssuppliersandcottonginnerstriggeredandmasterminded theprocessuntil they lost control.Kinshipand friend‐ship also played an important role as strong ties in cotton networking, contrary toexpectations. The significant difference between themean ranks assigned to publicofficialsandtoethnicityshowsasimilarperceptionofstayersandcreators,whichisdifferentfromtherankassignedbyjoiners.Thisdifferencesimpliesthatjoinersmighthave been more  oriented by public officials in their movement than stayers andcreators, whereas the latter value ethnicity in expanding their networks. Indeed,publicofficialsareimportantwhencreatingbreakawaynetworksbecausetheyknowtheofficialprocedures,which they canapply inamoreor less flexiblemanner, andbecausetheycanprovidethenetworksandtheirproponentswithvariousmaterials.Thesignificanteffectofethnicityandreligion(strongties)showsthattheyconstitutesecuresourcesonwhichleaderscouldbasetheiraction.However,theoverallrankingdidnotrevealanypre‐eminenceofweaktiesoverstrongties.

The significant and positive relationship of the number of social networks farmersbelongedtowiththeleadershipstatusimpliesthatthemoresocialnetworksafarmerisinvolvedin,themorelikelys/hewillbecomealeader.Thisassociationwasfurthersupported by the results of the regression of the length of appointment on boards.Indeed,beingamemberofasocialnetworkisnotenoughtogenerateresources forinstrumentalaction.Beingaleaderthereinincreasesthevalueoftheseresourcesthatcanbederivedfrominvestmentintothenetworkactivities.Itprovidesopportunitiesof extending one’s social relations, thereby strengthening one’s social capital.According to the strength‐of‐position proposition (Lin, 2001:65), access to socialcapital is linked tohigherpositions in social structures.Thus, the interactionof thenumber of social networks with the number of positions occupied within thesenetworks other than cotton had a significant and positive relationship with theduration of appointment as leaders. The more experience and contacts a farmeraccumulated from his/her social networks, the longer s/he might have stayed inleadershippositions.Therefore,previousleadershipexperienceinsocialnetworksisasimportantasahighermembershipofthesenetworks.Thus,itcanbeinferredthatprevious relations and experience in social networks constitute a resource thatleadersusetostrengthentheirpositionsincottonboards.Thestockofsocialcapitalembedded in an individual’s social relations plays an important role not only in

CHAPTER3

78

building a (new) network (Wellman and Kenneth, 2001), but also in getting intoleadershippositionsandbeingabletoholdontothosepositions.

3.7 Conclusion

This chapter investigated the emergence of breakaway networks in Benin’s cottonproductionandhowthisprocesselapsed intoatomising theoriginalorganisation. Itattempted to combine the strengths of qualitative and quantitative research, butunfortunatelytheresultsdidnotsupporteachother.Thequalitativeresearchaimedtounderstandandidentifyfactorsofatomisation,whereasthequantitativeapproachaimed to find out the relative importance of the factors for atomisation by rankingthemandverifyinganumberofexplanatorymechanisms.

Farmers’agencyandcottonnetworksinfluencedoneanother,shapingtheprocessofnetworksatomisationinBenin.Themajordriversofatomisationwereproblemswithpaymentandproblemswithresourcessuchasthefacilityofgettinginputs(fertilisersand pesticides). Payment‐related issues had been given a primacy in cottonnetworking.Thus,whetherfarmerswerepaidandhowtimelythispaymentwasdonecould push breakers to leave their networks or pull them into other networks,regardlessofwhethertheywerejoinersorcreators.Theresultsfurthersuggestthattheatomisationofcottonnetworksbasicallyrevolvesaroundtwodimensionsofsocialcapital: the structural dimension, through farmers’membership of social networks,andthecognitiveone,relatingtothediscoursethenetworksproponentsresorttoformobilising farmers. The main pattern is that breakaway networks emerge throughweakties.Resortingtostrongtieswasanexception,whichimpliesthattherelationaldimension of social capitalwas ofmarginal importance. In addition, farmers’ socio‐economic and socio‐demographic backgroundmade little difference with regard totheirdecisionofleavingorjoininganetwork.Furtherresultsshowthatamajorityoffarmers at  one point became leaders in new networks, indicating that becoming aleadercouldbeamajorhiddenincentiveforatomisation.Inaddition,agreaterstockof social capital was correlated with the length of leaders’ appointment on cottonnetworkboards.

This type of research is of course prone to theweaknesses of both qualitative andquantitative approaches, i.e. subjectivity and reductionism, respectively. The firstmajor limitation of this researchpertains to the relatively small number of farmersselected for interviewing.Therealityonthegroundwasthatgettingaccessto thesefarmers required considerable logistic effort due their geographical dispersion andthebadruralinfrastructures.Thedegreetowhichthesampleisrepresentativeofthetotalpopulationcan,despiteoureffortstomaximiseit,onlybedeterminedinfutureresearch.Anotherlimitationoftheresearchrelatestothefactthatthesamequestionswereaskedtoallcategoriesoffarmers,whichmadetheanswerstoperception‐relatedquestionsbothanexpressionofone’sownsituationandajudgmentofotherpersons’behaviour. As a consequence, the quantitative findings did not always support the

THEFACTORSBEHINDTHEEMERGENCEANDCOLLAPSEOFCOTTONNETWOKS

79

qualitative evidence. Further research to map the networking trajectory of cottonfarmerscouldyieldmoreinsightintothedynamicspertainingtoBenin’scottonsectorforeffectingapositivequalitativechange.

81

Chapter4

GeneratingResourcesandConflicts:ProfitabilityandSocialCohesioninBenin’sCottonSector

ABSTRACT:ThischapterisbasedonfieldresearchinthenorthernBenin,WestAfrica. The research aimed to investigate how interpersonal and intra‐organisational conflicts disrupted cotton production and froze collectiveaction.Cottonhasproventobethe lifeline for farmerorganisations,andhasdriven collective action in rural areas. The struggle to control theseorganisations and their economic and social benefits generatedmismanagement and free‐riding behaviour. The greed for resources, in theend,ledtohatred,disruptionofties,andconflictswithinandbetweenfarmerorganisations, which resulted in the decline of cotton production and thefreezeofcollectiveaction.Resultsfromcasestudiesshowthatsocialrelationsbased on kinship and friendship deterioratewhen financial stakes are high,and that cooperation within large groups requires legal sanctions to besustainable.Keywords:Cottonproduction;Collectiveaction;Socialcohesion;Benin

Chapter adapted from Guirguissou Maboudou Alidou and Anke Niehof (2012). GeneratingIncome and Conflicts: Profitability and social cohesion in Benin’s Cotton Sector. TheInternationalJournalofInterdisciplinaryOrganizationalStudies,7(1):33‐46.

CHAPTER4

82

4.1 Introduction

Forthe last twodecades,cottonhasbeenthemost importantcropinBenin,both intermsofthepercentageofthepopulationinvolvedincotton‐relatedactivitiesandtheincome generated by the crop. The boom in cotton production from the end of the1980shas impactedmanyothereconomicdomains in thecountry.Thisboomowednot only to market incentives but also to the dynamics of the organisations thatevolved around cotton production. The most important of these organisations arefarmer organisations, which instilled new dynamics into the sector to make itsuccessful. In turn, farmer organisations have been shaped by cotton farming andhave depended on its levies for their functioning (Kouton et al. 2006). From theresourcesgeneratedbycottonproduction,collectiveactionwassustainedinvillageswherecommunitiesprovidedthemselveswithbasicservices thatwerenotsuppliedadequatelybythestate:education,health,roads,andotherinfrastructure.Thesuccessofthesectorincreasedthevisibilityofcotton‐relatedorganisationsandtheirleaderswhogrew in importancebotheconomically and socially, at local aswell asnationallevel. But cotton as source of income also became a source of conflict, because ofdivergent interests. Cotton interests stirred appetites and attracted new entre‐preneurs in cotton productionwho invaded the organisations under the pretext ofimproving themanagement. This led to conflictswithin and between organisationsandorganisationalbreakdown,andinflatedthenumberoforganisations.Theensuingconflicts disrupted the social cohesion in the organisations and communities, andresulted in the decline of cotton production and the freeze of collective action. Theultimateconsequencehasbeenthelossbymanysmall‐scalefarmers,oftheirmain,oreventheironlysourceofincome.

StudiesinvestigatingthecausesofthedemiseofcottonproductioninBenin,however,have pointed at technical and trade issues. The latter, namely cotton subsidies inindustrialised countries, are associated with the decline of cotton prices on theinternationalmarket,causingimportant lossestoproducersindevelopingcountries.Adebateabouttheremovalofthesesubsidiesopposedthetwogroupsofcountriesininternational forums. The attention focussed on the outcomes of the debate, to theextent of overlooking the dark side of the internal institutional dynamics in cottonorganisations.While in Benin, farmer organisations have been pictured by some asenabling social inclusion (Tama‐Imorou & Wenninck, 2007; Tama‐Imorou et al.,2007), others (e.g. Sinzogan et al., 2007) saw them as pawns caught into heavyinstitutional arrangements and enduring the externalities of global policies. Theproblemof theatomisationof farmerorganisations that subsequentlyoccurredwasnot seen in relation to the inter‐personal and intra‐organisational conflicts thatincreasinglyaffectedcottonfarming.Indeed,farfromundergoingpassivelytheeffectof external contingencies, cotton farmer agency in Benin has impacted the trend ofcottonproduction.Therefore,torevivethesectoracloserlookshouldbetakenatthecriticalissueoftheorganisationaldynamicsofthesector.

GENERATINGRESOURCESANDCONFLICTS

83

The purpose of the present chapter is to investigate how interpersonal and intra‐organisationalconflictshavedisruptedcottonproductionandfrozencollectiveaction.The research was conducted through focus group discussions, key informantinterviews, and in‐depth interviews with 148 farmers including both grassrootsfarmers and organisation leaders. The chapter first discusses the theoreticalperspectiveused in theresearchandthenpictures thecottonsystemintheculturalsettingofthestudyarea.Inthefollowingsections,theemergenceofcollectiveactionin the studyarea isdescribedandananalysis ispresentedabouthowconflicts andviolenceexpandedthroughoutthecottonproductionzonesandweremanipulatedbytradersand intellectuals toentercottonorganisations, resulting in thedisruptionofsocial cohesion and the demise of collective action. Two caseswill be presented tohighlight these processes. The chapter concludes with a brief discussion on theimplicationsforcottonproduction.

4.2 Theoreticalperspective

Theissueofgroupdynamicshasbeenapproachedthroughafocusontheproductionanduseofresourcesandthemanagementof interpersonalrelations.Bothaffectthecohesionof groups, therebydetermining their ability for cooperation. Cohesion andcooperationwithinagroupdetermine its ability toovercome internaldifficulties aswellasthreatsfromoutside.Experimental studies of cooperation stem from the so‐called ‘Prisoner’s Dilemma’model, fromwhich the concepts of the “public goods dilemma” and the “resourcesdilemma”originated.The twoconceptsweregrouped togetheras “socialdilemmas”(Kollock,1998;Komorita&Parks,1995).Asocialdilemmaisdefinedas“asituationinwhich a group of persons must decide between maximizing selfish interests andmaximizing collective interests” (Komorita& Parks, 1995: 190), leading to “tensionbetweenindividualandcollectiverationality”(Kollock,1998:183)intheproductionanduseofpublicgoods.Theuseofpublicgoods,incontrasttotheirproduction,putsactors in the ‘commonsdilemmas’ or a ‘social trap’, a situation inwhich individualsaretemptedbyanimmediatebenefitofwhichthecosthastobesharedbyallusersoftheresource.Theproblemofcooperationinsocialdilemmasposesacriticalissueoftrust.

The role of trust in organisational settings is widely acknowledged by numerousresearches invariousdisciplines.Trust represents “apositiveassumptionabout themotives and intentions of another party [and] allows people to economise oninformation processing and safeguarding behaviours” (McEvily et al. 2003: 92‐93).Misztal (1996: 18) defined it as “the willingness of other agents to fulfil theircontractualobligationthatiscrucialforcooperation”.Trusthastwokeyfeatures:thepsychological and the behavioural (Kramer, 1999). The psychological feature,associated with cognition, entails the trustor’s perceived vulnerability, while the

CHAPTER4

84

behavioural one pertains to the interactions between actors. Trust has threefunctions:theintegrativefunction,thereductionofcomplexity,andthelubricantforcooperation(Misztal,1996).The lubricant functionmakescooperation forcollectiveaction possible and ensures group cohesion. Cohesion, as the sense of belonging ofgroup members, is determined by incentives found through membership. The“collectivity is structurally cohesive to the extent that the social relations of itsmembersholdittogether”(Moody&White,2003:106).

The present research hypothesised that the demise of collective action in cotton‐producing areas resulted from the decline of social cohesion within cottonorganisations. It aims at understanding how the social dynamics of cottonorganisationsfizzled,inspiteofincentivesprovidedbythesectorthatshouldallowanenduringcooperationforthebenefitofthewhole.

4.3 Thecottonsystem

4.3.1 Thesocio‐culturalsettingofthecottonzone

In Benin, the provinces heavily impacted by cotton production are those of BorgouandAliboriintheNorth.Between1990and2010,theseprovincestogetheraccountedfor48to80percentofthenationalproduction(AIC,2008a).Aliboriisconsideredthecotton belt of the nation, while Borgou is the food belt though also producing animportant share of cotton. Together they formed the Benin part of the pre‐colonialkingdomofBorgu,dominatedbytheBaatombu(Jones,1998).

Before the partition of 1898Borguwas a kingdom, stretching fromNorth Benin toNorth‐westNigeria(Kuba,2000).ThetermBorgudesignatesasocio‐cultural,politicalandeconomicspacewithcomplexanddiversepowerrelationsbetweenthedifferentethnic groups (Boesen, Hardung, & Kuba, 1998). It also comprised other ethnicgroups, namely the Boo, theMokole, the Fulani, cattle breeders and herdsmen, theGando,andtheDendi,migranttradersdescendingfromtheWangaraofMali(Brégand,1998).TheBaatombu, theBooandtheFulaniareconsideredthe indigenousgroups,theBaatombubeinglandowners(Jones,1998).Thelargemajorityoftheotherethnicgroups,however,speakBaatonum,thelanguageofBaatombu.

Borgu’s economy was “land‐based, with few resources to attract external trade”(Adekunle,1994:1),thoughitwaslocatedatthecrossroadsofimportantcommercialnetworks between theYoruba andHausa states ofNigeria and the former Songhayempire (Kuba, 1998).Due to colonisation, the kingdomdisintegratedpolitically andeconomically. In addition, urbanization and modernization have transformedBaatombu society and have weakened community cohesion and some practices ofsolidarity.Butevenso,contemporaryBorgouexhibitsthehierarchicalstructureoftheancientBorgu,quite“complexandrigid”withthenobleoraristocraticWassangariatthetop,andcommonerswhoarecultivatorsandhuntersatthelowerend(Lombard,

GENERATINGRESOURCESANDCONFLICTS

85

1960, 1998). In spite of its ethnic diversity, Borgou is socially integrated throughculturalpracticesandprocessesthataimatpeacekeeping.Amongthemisthejokingrelationship, which involves cross‐cousins, kin of alternate generations, differentcraftsmen, and neighbouring ethnic groups. It forms a harmless outlet for socialtensionswithinandbetweengroups(Brydon,2010).Byjokingpeopleavoidopenanddestructiveantagonismdueto jealousyorcompetition”andgive“proofofsolidaritywith every joke they gamely endure” (Schottman, 1998: 159). Neither the socio‐cultural customs and beliefs nor the historical links between the different ethnicgroupshavecompletelydisappeared.TheselinksarealiveacrossbordersandethnicconsciousnessamongtheBaatombusub‐groupsisstillstrong(Akinwumi,1998).Thesocialorganisationofthesocietyinchiefdomsaroundthecentralkingdomstillplaysarole.EachBaatombusocietyhasseverallayersofhistoricaltitlesandpoliticalofficesthat are still locally relevant and conserve their attributes of power (Alber, 2004;Harneit‐Sievers,2002).

4.3.2 Thefirstcottonelite:Theriseofschooldropouts

DuringtheMarxist‐Leninisterainthe1970s,farmerorganisationswerepurposefullypromoted throughout the country by the military regime as a form of villageassociations. These associations were used to infiltrate and control the grassrootspopulations. This logic is reflected in their name: Groupement Révolutionnaire àVocation Coopérative (GRVC) (Maboudou Alidou, 2002). The revolutionarycooperativesservedas thebasis for theprocess that tookplace fromtheendof the1980sonwards,i.e.thenetworkingamongfarmerorganisations.ButwheretheGRVChadalmostnofreedomofactionandwasconstrainedbypoliticalcontrol,thepresentGroupements Villageois (GV) are autonomous. Belloncle, cited by Tossou (1993)contended that theemergenceofnew formsof farmerorganisations coincidedwiththe appeal for a real dialogue between farmers and developers after the ruraldevelopmentapproachesofthe1960sand1970sfailedtoliveuptoexpectations.Theshiftof statusalsobenefited fromtheadventofdemocracywhen the1990NationalConference put an end to the socio‐political unrest of the secondhalf of the 1980s.This policy shift gave the new forms of organisations a level of importance thatpreviousorganisationsneverhad.Italsoledtotheemergenceofanewelite:thatofschooldropouts.

NorthernBeninwascharacterisedbya lowrateof literacyand,apart fromprimaryschool teachers, rarely did intellectuals live in the villages. The first rural literacyprograminBeninstartedintheprovinceofBorgouduringthe1970s,totraintheGVmanagersinreadingandcountinginthelocallanguage.Theacquiredskillswereusedin cotton production to report inputs distribution among members, and in cottoncommercialization (Abou‐Moumouni, 2011). But the skills turned out to beinsufficientwhen cotton production became really important. School dropouts thenstarted taking over the daily management of the GVs. The little literacy they hadconstituted a highly appreciated asset for farmer organisations, where they would

CHAPTER4

86

occupythepositionsofsecretaryontheboard.Lateron,thesecretariesturnedouttobe crucial board members, due to the critical importance of their role in theredistributionofinputs,thecommercialisationofcotton,andpayment.Untilrecently,in many farmer groups the two other important board positions, namely that ofpresident and treasurer,were occupied by virtually illiterate persons. This enabledsecretaries to encroach upon the responsibilities of the president and the treasury.Withtheatomisationofcottonorganisations,thispre‐eminenceofthesecretariesledmemberstoequatetheorganisationwiththesecretary.Themostcommonanswertothequestionaboutwhichorganisationafarmerbelongstois:“IamintheorganisationofSecretaryXorY”,XandYstandingforthenameofthissecretary.

The development of cotton production paralleled the capacity building of ruralorganisations.Leadersoforganisationswereearmarkedforallkindsoftraining.Theiracquired skills would allow them to relate directly to donors and other outsiders,mediating for their populations (Olivier de Sardan, 2005). Thanks to the boomingcottonproductionandthehugeoutputsthesectorgenerated,farmerleadersgrewinimportance,botheconomicallyandsocially,therebybecomingrolemodelsforyoungpeople in the villages. Leaders of cotton farmer organisations formed a distinctivesocial class in their localities. Theywere blessedwith various social and economicprivileges that distinguished them from the others. Inter‐connected from village toprovince, they tended to self‐reinforce the system by putting in place institutionalarrangements, irrespective ofwhatwas stated in statutes and laws. Lack of controlallowedtheleaderstomanagepublicresourcesastheirown.Bystayingontheboardinfinitely,someof the leadersusedthepowertheyheldnotonly for thecommunitybut also for themselves. As a result, their social ascension led to power abuses andmismanagementthatescalatedfree‐ridingbehaviour.

4.4 Cottonproductionandcollectiveaction

There were three constituent functions of the village associations: driving upagricultural production, including food crops; facilitating access to credit and tools;and serving as the basis for the emergence of a cotton elite (Roy, 2010). Thesefunctionsareassumedmainlybyagentsatpositionsofauthority,whoareinchargeofpromoting solidaritywithin the group and collective action for the benefit of groupmembersandthecommunityatlarge.TheancientBorgu ischaracteristicofthecollectiveactionthatstemmedfromcottonproduction.Through a combinationof community‐ andhierarchy‐based approaches(Kiser, 1997) rules and institutions were created within cotton productioncommunities, endorsedby theState’s assistancepolicy topromote collective action.Despite the low level of literacy of the organisations’ leaders, collective actionwaseffectivethereandbenefitted thecommunities,as isstillvisible invillages. Itseemsprobablethatthesuccessofcollectiveactioninthisareawasboostedbycertainsocio‐

GENERATINGRESOURCESANDCONFLICTS

87

cultural values. InBaatombu society, selfishnesswas stronglydetestedandsacrificefor the group highly appreciated. Studies of Borgou’s socio‐cultural characteristicshave revealed the importance of an orientation towards others, “shown to be ofcrucial importance for grasping the dynamic social construction of group identities,and the various power and exchanges relations between populations groups”(Breusers,2000:311‐312).Withregardtothesociologicalargumentthatculturalandstructural factors are the determinants of collective action (Kiser, 1997), it can beinferred that the society bears some of the requirements for the emergence ofcollective action. Communities availed themselves of basic services that were notadequately supplied by the state. Collective action driven by cotton productionconcernedawiderangeofdomainsgoingfromeducationtothemaintenanceofroads,assummarisedinTable4.1.BecausetheprovincesofBorgouandAliboriwereknownfortheirlowlevelsofschoolattendance, collective action first backed up education at local level. Communitiesbuilt schools and recruited teachers who were paid from cotton levies. Thiscommunity support for education became a phenomenon known as “Enseignantscommunautaires”thatspreadthroughoutthecountry.Untilrecently,farmerunionsatdistrict level used to pay the tuition fees for school girls sent to the provincialboarding school. Residences were also built on the university campus to housestudentsfromdistrictslikeBanikoara,GogounouandSinende.Theexampleconvincedother district unions to do the same and inspired students from other cotton‐producing districts to claim similar benefits from the cotton levies their districtearned.In addition, farmer unions built lodging houses inmost of the districtswhere suchinfrastructure and services could hardly be found. Municipal buildings and leisurecentres were built in villages, making the farmers proud, as illustrated by thefollowingquote:“ThisbuildingwherewearemeetingactuallyisnotbuiltbyaMinister;itisbuiltbyourownhoesandeffort.”(Source:Focusgroupdiscussion,Sekere,January2009). This shows how farmers were empowered, availing themselves of basicservicesthataresupposedtobeprovidedby“ministers”.Theseachievementsmadethedeprivedvillages in the regionmoreappealingand lively.Butbecauseofpowerabuse,cottonproductionbecametroublesomeanddeprivedruralpopulationsoftheirpotentialfordevelopment.

CHAPTER4

88

Table4.1 Collectiveactioninthestudyarea

Types InfrastructureLevel(VillageorDistrict)

Socialandcommunityinfrastructure

Agriculturalinputsstores BothSchools(buildingsorentireschools) BothResidenceforschooldirectors VillageStudentsresidenceatuniversity(Cotonou) DistrictWaterpumps/wells BothRuralroads BothHealthcentres BothLeisurecentres BothEquipmentformodernmusicorchestra VillageLodgingcentres DistrictMunicipalbuildings BothPubliclightening BothKerosenesellingpoints VillageFootballfields BothRuralRadio DistrictMosques Village

Payment

Salariesforcommunityteachers BothAssistancetostudents BothAssistancetothemunicipalcouncil BothContributionfortherealisationofinfrastructurebypartners

Both

Maintenance Maintenanceofcommunityinfrastructure Both

Source:Focus‐groupdiscussions,January2009.

4.5 Theriseofconflictsandviolence

Thedynamicsof farmerorganisationshavealwaysbeen idealised,overlooking theirweaknesses that undoubtedly hampered the sustainability of the cotton system.EvidencefromotherWestAfricancottonproducingcountries,however,showedthatcotton farmersplayeda role,directlyor indirectly, in theway theprocessof cottondevelopment unfolded (Bassett, 2001). Far from being passive actors, Benin cottonfarmers have greatly influenced cotton production through their agency. First,positively,bytheircontributiontotheincreaseofproduction,andsecond,negatively,bytheconfrontationalwayintowhichthemanagementoftheirorganisationslapsed.

4.5.1 Typologyofconflictsincottonfarmerorganisations

From the end of the 1990s, the cotton sector in Benin has been overwhelmed byvarious conflicts that undermined the cohesion of cotton organisations. Theseconflicts were identified and ranked by farmers during focus group discussionsaccording to their frequency. The ranks range from 1 (most frequent) to 9 (least

GENERATINGRESOURCESANDCONFLICTS

89

frequent).Table4.2showstherankingoftheconflictsaccordingtofarmers,andthecomparisonofmeanranksbetweengrassrootsfarmersandleaders.

Table4.2 Rankofconflictsincottonorganisationsaccordingtotheirfrequency

NatureofconflictsGrassrootsfarmers

(N=62)Leaders(N=46)

t

Inputsallocation 3.0(2.07) 3.2(2.14) ‐.43Lackofpayment 3.5(2.31) 3.7(2.17) ‐.53Applicationofjointliability 3.7(2.07) 3.7(2.18) .13Mismanagementofresources 4.7(2.27) 3.9(2.11) 2.16*Commercialisationofcotton 4.6(2.07) 4.5(2.07) .19Leadership 5.5(2.39) 4.6(2.32) 1.88Lackofcommitmenttogroupwork 6.5(2.18) 6.2(1.91) .64Outletsforinter‐personalarguments

6.6(2.58) 7.4(1.68) ‐1.97

Paymentofmembershipfees 6.9(1.88) 7.9(1.49) ‐2.84**

N=108;*=p<.05;**=p<.01

Source:Surveydata,2009–2011.

Themeanranksofthefrequenciesoftwotypesofconflictsourcesshowedsignificantdifferences between grassroots farmers and leaders. Leaders ranked mis‐management‐relatedconflictshigher thangrassroots farmers (p<.05). Indeed,of thenine breakaway organisations investigated, seven reportedmismanagement‐relatedconflicts. Also leadership‐related conflicts were ranked higher by leaders than bygrassroots farmers, but the difference was not significant, indicating similarimportanceofthiscausetoallcategoriesoffarmers.

Having a leadership position gave access to material and financial benefits thatimproved the livelihoodsof leaders and enhanced their visibility in the community.This explains why conflicts about inputs allocation are ranked first in the overallranking. This first position ascertains the critical importance of fertilisers andpesticides in cotton production. Controlling inputs gave enormous power to farmerleaders,whocouldusetheseforpurposesevenunrelatedtocotton.Thepaymentofmembership as a cause of conflicts was ranked significantly higher by grassrootsfarmersthanbyleaders(p<.01).Farmersstruggledtocontroltheirorganisationsandthesocialprocessesandpowerplayinvolved.Theseprocesseswereaggravated,how‐ever,bynewtypesofactorswhoinvadedcottonproduction.

CHAPTER4

90

4.5.2 Thefirstuprising:“OrouBagou”vs.“OrouSori”

ThedistrictofGogounouwitnessedthefirst inter‐communityconflictresultingfromthe1990sproductionboom.Theconflictresultedintheemergenceofthefirstcottondissident organisation. It opposed farmers from two villages: Sori and Bagou. Bothvillageswereimportantcottonproducersandparticipatedinthemanagementofthedistrictfarmerunion,onwhichboardBagouheldmorepowerthanSori.AnativefromBagouwas appointed as a vice‐president and destined to take over the lead of theunion.Thepresidentatthetime,whowasthenalsoleadingthenationalfederationoffarmers, was controversial on the grounds of being illiterate, being involved inmismanagement, and extending his ‘reign’ for years. The contestation was led bynativesfromSori,whorequestedhimtostepdown.Sori,themosturbanvillageinthedistrict,hadbeencontestingthepositionofGogounouasthedistrictcapitalforsometime.Therefore,Soricouldnotaccepttobedeprivedofcertainpositionsbythevillageof Bagou, which cotton farmerswere accused of connivancewith the controversialpresident in themismanagement of union resources. Farmers from Sori decided toresolvethematterbytakingcontrolofpropertiesoftheunion.

InOctober1997,tensoffarmersfilledupatruckandinvadedthevillageofBagoutoseize selected properties of the union controlled by the vice‐president, therebytriggeringaviolentconflictbetweenthetwovillages.Apitchedbattleopposedthetwogroupsof farmersatmidday, resulting ina lotofdestructionandhatred. Itwas thefirstandmostviolentconflictbetweenthetwoneighbouringvillagessincetheygrewcotton.Therefore,theopposedpartieswerenicknamed“OrouBagou”and“OrouSori”,referring to the camps of Bagou and Sori respectively, the term “Orou” being theBaatombu’s birth rank name assigned to the first son (Schottman, 2000). Althoughtherewerenocasualties,theprotagonistswanttoforgetabouttheconflictbecauseofitsviolenceandthedireconsequencesitgenerated.Noneofthemeninthemalefocusgroupdiscussionsinthetwovillagesmentionedit;butthewomeninthefemalefocusgroupdiscussionsdid.Theytimidlystresseditsviolenceandreferredustothemenifwe wanted more details. Silencing the conflict appeared to be a “shared denial ofpainful realities” (Cairns&Niens,1999).Cottonhadbecomeanightmare,andsomewonderedwhytheyeverwentsofarintheirdisputes.Atthetimeofthefieldwork,theconflictwasstillpendingatthecourt,likemanyotherconflictsincottonproduction.Nevertheless,thecontesterssucceededincreatingthefirstbreakawayorganisationincotton production, supported by traders and businessmen. This tumultuousexperience that unfolded in the district of Gogounou inspired the claim of itsproducerstobethe“laboratoryofcottonnetworks”.

GENERATINGRESOURCESANDCONFLICTS

91

4.5.3 Politicalinterferenceandtheadventoftradersandintellectualsincottonorganisations

The success story of cotton production contributed to a great extent to improvingpeople’s perceptions of agriculture. Before cottonbecame a critical livelihood asset,agriculturewasnothighlyvaluedandbeingafarmerhadalowersocialstatus.Thankstotheboomincottonproduction,thisstatusimprovedandagriculturebecameanewtype of entrepreneurship. The wealth generated by cotton to formerly deprivedfarmersattractednewfarmers.Withthenumberofpeopleinvolvedinitsproduction,cottonalsobecameaninstrumentthatpoliticalleaderssoughttocontrolforelectoralpurposes. The contestation of cotton “illiterate leaders”, voiced on the ground ofembezzlement, became politically motived. Indeed, two succeeding Ministers ofAgriculture, activists fromdifferentpoliticalparties thatwere strong‐holdingoneofthe largest cotton‐producing districts, supported opposingwingswithin the farmerunion to get control over the organisation for electoral purposes. This strugglebetween local politicians is captured in the following quote from the agriculturaladvisortotheHeadoftheState:

Theprofoundbutunsaidmotiveoftheconflictandtheatomisationofcottonnetworkswaspolitical.Everypolitician in theareawanted tohavethe farmer federationunderhissway.Therefore,theybackeduptheemergenceofdissidentnetworks.

Source:Keyinformantinterview,Banikoara,January2011MeetingstosettlethedisputesconvenedashighasthePresidencyyieldednoorlittleresult,becauseleadersofcottonorganisationsbecameawareoftheirimportance.TotheHeadoftheState,whoaskedthecontestedleadertostepdown,thelatterdaredreplicate:“Ifitweresoeasytostepdown,whydidyourefusewhenyouwererequestedto do so”, referring to the President’s refusal to step down at the 1990 NationalConference to end the country’s socio‐political unrest of the 1980s. To resolve thetense situation, the President askedwhether the contesters could create their ownorganisation, which, according to witnesses, triggered the atomisation of cottonorganisations. Henceforth, the best solution to disputes within an organisation hasbecomethecreationof“one’sownorganisation”.Asaconsequence,thelocalconflictwithin adistrict union resulted in the fragmentationof thenational federation.Theconflictwasalsofuelledbyrivalryaboutafarmer’swifewhowasallegedlysnatchedbyaleader,showinghowsomeleadersabusedtheirpower.Tradersandretiredcivilservants skilfully used the discontent to engage in a struggle for a “bettermanagement” of the organisations. They started practicing cotton production,howevermarginally.Thesenewfarmersweretermed“farmgoers”insteadoffarmersby the leadersof theoriginal federation,arguing that theyweresimplyattractedbytheprofits. Inaddition, inputssuppliersandcottonginnersrelyingon theirpoliticalcontacts, encouraged dissidence among farmers and actively supported the new

CHAPTER4

92

entrepreneurstocreatedissidentorganisations.About29percentofthepromotersofbreakaway organisations in the sample were newcomers in cotton production.Leaders of this category represented about 16 per cent of the total leaders. Theirinfluencecanbegraspedbyobservingtheirlevelofinvolvementintheorganisations.Theyalloccupiedpositionsatthehighestlevelinthehierarchyfromwheretheycouldattend important board meetings and influence the process of decision‐making.Though some had been moonlighting as farmers, many others had never plantedcotton. Some started cultivating cotton from their involvement in theorganisations,growing theminimumnecessary tobe considered a cotton farmer.A retired leadersaid:“IcultivatecottonbecauseIamontheboardofthenetwork.”

Being a cotton farmer was no longer a condition for membership of cottonorganisations,whichsheddoubton the intentionof thenew leaders to improve themanagement of organisations. Their main motive was to gain financially from thecrop. A former federation leader contended that “creating cotton organisations hasbecome a business to cheat farmers and be enriched”. Therefore, the intrusion oftraders and intellectuals is seen by farmers and their former leaders as the mainsource of the trouble the sector and its actors have been experiencing. About theatomisationofcottonorganisationsanditscorollariesoftension,afarmerexpressedhisgrievanceasfollows:

You[intellectuals]aretheonlyonesresponsibleforconflictswithinandbetweencottonorganisations.Ourwoesoriginatefromthetopandthesolutionmaycomefromthetop.Writedownonyournotebook:Wearedisgustedbyyourcotton.

Source:Focusgroupdiscussion,Sekere,January2009

Indeed,pensionersandtradersledandmastermindedtheemergenceofsevenofthenine dissident organisations. But trust usually collapsed after a short collaboration.The promoters started accusing one another of secretly negotiating contracts withinput suppliers and receivingmoney from ginners before the cottonwas produced,thereby subordinating the organisations to the whims of businessmen. The mostfrequentaccusationsweremismanagementofresources,illicitenrichment,badsocialbehaviour, and the denial of legality and a pronounced taste for racketeering. Thevocabularyusedbytheformerproponentswhosoonbecameopponentswasrichwithwordssuchas ‘illiterate’, ‘bandit’or ‘crook’.Thescrambletotakecontrolovertheseorganisationsandtheirresourcesresulted innever‐ending inter‐personaland intra‐organisationalconflicts.

4.5.4 TheAltruistandtheSaint

Anda is a retired agricultural engineer who gave himself the title of “itinerantbenevolentadvisortocottonnetworks”.Heconsidershimselfanexperiencedauthority

GENERATINGRESOURCESANDCONFLICTS

93

oncottonandhewantedthecottonfarmingcommunitytobenefitfromthat.HavingheadedanagriculturalextensionagencyforyearsandhavingworkedintheMinistryofAgricultureandRuralDevelopment,Andastillhasconnectionswithintheministry.Because of that, he took it upon him to provide proofs of embezzlement and toconfrontthe“illiterateleaders”,duringtheconflictbetween“OrouBagou”and“OrouSori”. He worked together with his friend Segue, a retired primary school teacher.AndaandSeguejointlyledthecontestationthatresultedintheemergenceofthefirstbreakawayorganisation.Theywrote the statutes andAnda’s home served asoffice.Theorganisation,however,was ledbyanothernewcomer, a traderwho funded theinitialactivities.Bothfriendsheldapositionontheboard,SegueasthesecretaryandAndaas theadvisor.Butayear later, theycouldnot trust their leaders,whom theyaccusedofracketeering.Consequently,theycreated“theirownorganisation”,alwayssharingrolesontheboard.ButwhileAnda“neverhascroppedaplantofcotton”,Seguegrowscotton.

At the end of their first cotton season, Anda and Segue had a dispute about theredistribution of huge amounts ofmoney, aboutwhich both have a different story.According to Segue, Anda would have claimed a percentage for advising theorganisation. But according to Anda, Segue would have used the resources forpersonalpurposes.WhileSegueconsidershimselfasaintwho“neverhasmisusedonefranc”,Andaassumesheis“hatedforhisrighteousness”.Aftertheyhadmastermindedand headed two dissident organisations, the once “best friends” ended up inexchanging correspondence in which they accused each other of greed and mis‐conduct.Andawasfinallythrownoutofthissecondorganisation.Thetwofriendshadadifferent interpretationoftheexperience.Segueacknowledgedthatthey“werethebestfriendsintheworld”,whereasAndadeniedtheirpreviousfriendship.

4.5.5 Thebrothers‐in‐law

SidiandBoniarebrothers‐in‐law.Boniisafarmer,andSidiatrader.WhenBoniwasapproached to lead a breakaway organisation backed up by executives of the Statecotton parastatal company to “thwart the actions of private companies and securesuppliestotheStatecottonginningmills”,hecontactedSiditohelpspreadingtheneworganisationthroughoutthecottonbelt.AsSidisaid:

Iwas a trader and I did not grow cotton. They involvedme in thisnightmare. Iwasathomewhen theyknockedonmydoor.They saidtheywanted tocreateacottonnetwork thatwouldbeprofitableandasked for my financial support. Who would refuse the promise ofprofits?

Aspromotersoftheorganisation,theysharedtherolesontheboard.Bonibecamethepresident and Sidi the secretary. They vied with contending organisations to getfarmers adhere to theirs. But quarrels about the redistribution of production

CHAPTER4

94

outcomes andother benefits broke the trust that hadprevailed at the outset of theendeavour. Sidi then plotted to overthrow Boni from the lead. With the newlyappointed leader, he signed a deal by which the organisation was engaged toreimburse a private credit he got using his position as secretary. But Bonimade asuccessful come‐back as the president on the board. The credit scheme wasdiscoveredandSidiwassenttojail.Theformeralliesarenowmutuallyaccusingeachotheroffalsification,theft,andmoneygrabbing.

4.6 Thedecayof“cautionsolidaire”andthedemiseofcollectiveaction

The“cautionsolidaire”isacollectivecommitmenttopayforcottoninputsallocatedtogroupsbyalltheaffiliatedmembersandisbasedonthejointliabilitymechanism.Themechanism was first used by the Grameen Bank to overcome traditional bankingpolicies of lending that require physical collaterals and couldnot be applied to dis‐advantaged poor people (Ghatak & Guinnane, 1999; Rai & Sjöström, 2004). Theformulaebecamepopularindevelopingcountriestogivepoorpeopleaccesstocredit.Amongthereasonsmentionedbyscholarsforitssuccessarethereductionof“trans‐actioncosts”,the“peer‐pressure”,andtheuseof“socialcapital”(Ghatak&Guinnane,1999).

In Benin’s cotton production, the caution solidaire allowed farmers access to basicinputs for cottonproductionwithoutany collaterals.Themechanismcontributed totheproductionboombyprotecting farmers fromthe incertitudeofaffording inputs.Thecautionsolidaire,however,wasmisinterpretedandshoweditslimitslaterondueto its misuse. In its implementation, the only requirement to get credit wasmembershipofanorganisation.Thereforetheonlyguarantorsweretheboardsoftheorganisations,andnotindividualfarmers.Inaddition,theorganisationsweresolargethatmembershardlyknewoneanother.Asaresult,allactorsinvolveddrewundulyhuge advantages from the scheme. Inputs suppliersmade profits by over‐supplyinginputstofarmerorganisations,regardlessoftheirrealneeds.Leadersinturnmisusedthestocksofinputsbygivinginputstoacquaintanceswhowereneithermembersofan organisation nor producing cotton. Entire villages and organisations used moreinputs than they needed and could afford. Because the mechanism entails thedeductionofinputscreditbeforefarmersarepaid,cottonrepaymentbecameerratic.Free‐riding behaviour of farmers trying to compensate their arrears increased,rendering the inputs credit recovery more problematic. Henceforth, self‐interestprevailed,attheexpenseofthegroup,inflatingtheindebtednessoforganisations.Asmore positive attitudes and cooperation derive fromhigher levels of trust (Dirks&Ferrin, 2001), it can be asserted that lower levels of trust result in more negativeattitudesandlowerlevelsofcooperationandperformance.Asasolution,thecautionsolidairewas extended from village level to district and province level, a strategytermedthecommunalisationanddepartmentalisationofthecautionsolidaire.

GENERATINGRESOURCESANDCONFLICTS

95

The extension of the caution solidaire aimed at securing the repayment to inputssuppliers, ignoringtheproblemsof farmers.Indeed,withthedepartmentalisationofthe caution solidaire, a village or an organisation could reimburse for an unknownvillage or organisation without the consent of its members. This aggravated theindebtedness in the sector by delaying the payment and increasing inter‐organisational conflicts. Mismanagements affected common resources that weremeant for collective action, leading to their scarcity. Thus, collective action wasdrastically reduced and frozen in many districts. Joint actions by multipleorganisations became impossible, their relationships being dominated by distrust.Leaders mutually accused one other of mismanagement and plotting against theirrespective organisations, providing that fear and greed inhibit behaviour thatcontributetopublicgoods(cf.Komorita&Parks,1995).

4.7 Conclusion:Killingthegoosethatlaysgoldeneggs

Cottonproductionyieldedimportantamountsofresources,whichallowedruralandsemi‐urban communities to avail themselves of the basic infrastructure the Statefailedtosupplythemwith.Theearlystageoftheboomincottonproductionrevivedthehopeofabetter life inruralareas.However,theresourcesgeneratedstirredtheappetite of traders and intellectuals who skilfully took over control of farmerorganisations.Butthesenewcottonentrepreneursturnedouttobemoregreedyandself‐interested than the former leaders. The lack of enforcement of rules combinedwith the decline of some cultural values such as unselfishness, encouraged mis‐managementpracticesthatremainedmostlyunpunished.Asaconsequence,mistrustanditscorollariesoffree‐ridingbehaviourandlackofcooperationincreased,freezingcollective action. Cotton, previously considered themost profitable crop and a cashcow, became troublesome for individuals and communities. The dynamics showedthatboththepsychologicalandthebehaviouralcomponentsoftrustareindispensablefor an effective cooperation within groups. This is easier to realise within smallergroups,wheremembers canexertmore control onone another and feel personallycommitted to their obligations. Hence, enforcing limits to the size of cotton farmerorganisations would improve their functioning and preserve the public goods.

97

Chapter5

GenderRolesinCottonProductionandManagementofCottonOrganisationsinBenin

ABSTRACT: Women's productive roles have nurtured important debates,heuristicaswellaspractical,inthescientificanddevelopmentcommunity.InBenin, women farmers are playing a key role, particularly in cottonproduction,wheretheyareinvolvedthroughouttheproductionprocess.Onlya negligible portion of them, however, is involved in the management offarmer organisations. The present chapter aims to identify factors thatconstrain or enable women's representation in the management of cottonorganisations.Itusessurveydataandthelifehistorymethod.Thelatterwasapplied to two women leaders who are exceptions to the rule, and theirexperience as board members of the organisations was documented. Bothwerecajoledandpressedintotheirposition,andexperiencedasimilarendbybeing ousted. The results suggest that gender myths and stereotypes arecritical in women's involvement in managing organisations. Furthermore,men's motives for involving women in the management appear to bequestionable.Keywords:Cottonfarming;gendereddivisionoflabour;womenleadership;Benin.

Basedonthischapterthefollowingarticlewaspublished:GuirguissouMaboudouAlidouandAnkeNiehof(2013).GenderRolesinCottonProductionandManagementofRelatedOrganisationsinBenin.GenderTechnologyandDevelopment17(3):313‐335.

CHAPTER5

98

5.1 CottonproductioninBeninandthedivisionoflabour

ThepersistenceofgenderdisparitiesinBeninanditsensuinginequalitiesinaccesstoandcontrolofeconomicandsocialresourcesishinderingthecountry'seconomicandsocialdevelopment(WorldBank,2002,2003).Theprevalenceofthesedisparities isin part favoured by traditional codes overlapping with legal dispositions aboutgender, which are still governing the beliefs and behaviour of a large part of thepopulation (Falen, 2008). Women are often "treated as minors with most of theirrightscomingthroughtheirrelationshipswithmen–fathers,husbands,andbrothers"(WorldBank,2002:17).Genderdisparitiesareacuteinruralareas,whereagricultureisthemainsourceofemploymentandwheretraditionalpracticesreproducewomen'ssubordinate position. Indeed, about 80 percent of the active population in Beninengages inagriculture,whichconstitutesan importantsourceofemployment for69percentofwomenand66percentofmen.AccordingtotheWorldBank(2007:320),women'scontributiontoagriculturalactivitiesisatleastasimportantasthatofmen,andtheirshareintheagriculturallabourforceamountedto46percentin2003‐2005.The increase of women's participation in agriculture is the result of changes inproductionpatternssuchastheextensionofcashcrops,namelycotton(WorldBank,2002).

Beforethe1990s, thecottonsectorwasunderastatemonopoly. Itsproductionwasmostlymen‐businessanditsrevenuesweremarginal(Gandonou,2005;WorldBank,2003).Atthattime,womenwereinvolvedincottonproductionbyworkingascasuallabour for men’s cooperatives, the main instrument of cotton production (Albert,1993). From the early 1990s onwards, cotton has become themost important cashcropinBenin, thankstothe liberalizationofthesectorandtheconsecutiveboominproduction.Itprovidedupto80percentofruralhouseholds’incomeanditwasalsoresponsiblefor40‐45%ofthetotalforeignexchangeearningsoftheState(Gergely&Poulton,2009;Hahonou,2011).Cottonconstituted theonlyagriculturalvalue‐chainof which the production was organized into a marketing channel (Nissanke &Mavrotas,2010;WorldBank,2003).Anetworkofinstitutionalstakeholdersmanagedthe sector under the control of an inter‐professional association. Of thesestakeholders, farmer organisations turned out to be one of the most important.Individuallyor as groups, farmerwereproducersof rawcottonand constituted thelargest stakeholders, though qualified as weak (Sinzogan et al., 2007). For themanagementoftheinputsandthemarketingoftheoutputs,farmerswereorganizedfrom village to national level in accordance with the three‐level administrativedivision of the country. At village level there were farmer groups, grassrootsorganisationsthatformedafarmerunionatthedistrictslevel.Districtfarmerunionsformedaunionat theprovincial level, and the latter formeda farmer federationatnationallevel,ahugenetworkoffarmerorganisations.Eachleveloftheorganisationwasgovernedbyaboardwherememberswereelectedorappointed.Membershipingrassrootsorganisation,however,was thebasic criterionof accessing theboardsat

GENDERROLESINCOTTONPRODUCTION

99

anyhierarchicallevel,giventhattheboardsofthesegrassrootsorganisationwerethedepartingpointofthehierarchicallystructurednetwork.

Since the economic significance of cotton increased and cotton became criticallyimportant for rural welfare, its production moved from being primarily men'sbusinesstobecomingaconcernofbothmenandwomen(Baffes,2004;WorldBank,2002). As a consequence, women could have their own cotton plot, independentlyfromthatoftheirhusband.However,onlythosewhowereheadsofhouseholdscouldbe registered individually in the organisations andmarriedwomen operated underthe authority of their husbands (Tama‐Imorou et al., 2007). But still, womenwerevirtuallyabsent fromtheboardsoforganisations.Fromtheendof the1990s, intra‐andinter‐groupconflictsarose,leadingtotheatomisationofthelargenetwork,whichbroke apart intomany conflictingnetworks. Someboards startedhavingwomen asboardmembers.Evenso,womenarestillcriticallyunderrepresentedtherein.Cottonmanagementisoverlymale‐dominated,exhibitingahugediscrepancyincomparisontowomen'simportantcontributiontoitsproduction.Thegendereddivisionoflabourincottonfarmingfollowsthetraditionalpattern:heavytasksformenandthelighterones for women (World Bank, 2002). Tasks requiring physical strength, such asploughing,andactivitiesthatposehealthrisksorrequirespecificknowledge,suchasthe application of pesticides, are considered unsuitable for women. Insecticidestreatmentisdeemedtobeparticularlymoreharmfultowomenthantomen(WorldBank, 2002). Tasks such as sowing, fertilizing and harvesting, are done by all thehousehold members on the main plot, which is prioritized to any personal plot.Additionally, food has to be prepared for the workers, although this is rarelymentionedasacotton‐productionrelatedtask.Takingthisallintoconsideration,theWorld Bank (2002, 2003) estimated that women spend more time in cottonproduction than men and that the time they devote to agricultural work could beunder‐estimatedbyofficialstatistics.

The present chapter analyses the bottlenecks to women's representation in themanagementofcottonorganisationsaswellas factorsenablingtheirrepresentationinthespecificsettingofBenin.Itaddressesthefollowingquestions.Whyarewomenunderrepresented in cotton farmer organisations in spite of their increasingcontribution to cotton production? What is the profile of women who do becomeboardmembersofcottonorganisationsandwhatkindofconstraintsdotheyface?Inthe sections that follow, we introduce the theoretical approach, after which wedescribe thestudyarea,particularlywith regard to thegendereddivisionof labour,and provide information on how the research was conducted. The results arepresentedaccordingtothemethodsofdatacollection.Finallywediscussthefindings,simultaneouslydrawingsomeconclusionsandimplicationsforpolicy.

CHAPTER5

100

5.2 Genderandleadership

Women's reproductive functions have been always more acknowledged than theirproductive roles in society (Sprague & Kobrynowicz, 1999). This has nurturedimportantintellectualdebates,heuristicaswellaspractical,withinthescientificanddevelopment community, leading to theories, approaches, and concepts aboutwomen’spositionandneedsinsociety(eg.Molyneux,1985;Moser,1989,1994).Thus,Molyneux(1985)coinedtheconceptofwomen'sintereststhatpertaintothesharedconcernsofwomenensuingfromtheir femalenature.However, thisconceptglossesover the differences among women, and is different from the concept of genderintereststhatare"thosearisingfromthesocialrelationsandpositioningofthesexesand therefore pertain [...] to both men and women" (Molyneux, 1998). Genderinterestscanbeeitherpracticalorstrategic(Molyneux,1985;Moser,1989).Practicalgender interests are "interests based on the satisfaction of needs arising fromwomen'splacementwithinthesexualdivisionoflabour",suchasaccesstosafewaterfordomesticproduction,whilethestrategiconesrelateto"thoseinvolvingclaimstotransformsocialrelationsinordertoenhancewomen'spositionandtosecureamorelasting re‐positioning of womenwithin the gender order and within the society atlarge" (Molyneux, 1998: 232). The latter, therefore, include gender equity andreducingwomen’ssubordinateposition.

Genderequity isperceived tobebeneficial to thesocietyasawhole, andcanresultfrommodernization,when "socio‐economic development brings systematic changesinpolitical,social,andculturallife"(Inglehart&Welzel,2005:46).Recentprogressingender equity is reported in domains such as economics and politics (World Bank,2012). Gender differences in political participation in Sub‐Saharan Africa may besmaller thanks to institutionalised forms of participation, which constitute "saferarenas for women to participate" (Coffe & Bolzendahl, 2011: 259). However, thisformalequalityandpoliticalrepresentationhidesanexclusionfromleadershiproles.Most organisations may still exhibit gender disparities because of their failure tochange (Ely & Meyerson, 2000). For instance, based on informal practices andsubjective criteria of competence, for a long time it has been conventionallyunderstood that managerial jobs are for men and not suitable for women (Billing,2011; Ely &Meyerson, 2000). In addition,women have a lesser voice thanmen indecision‐makingbothintheprivateandpublicsphereandarepoorlyrepresentedinthepublic sphere (WorldBank, 2012). Such apaternalistic organisational culture isblocking women’s access to leadership positions. Particularly in rural areas indevelopingcountries,women'semancipationseemstobeconstrainedbymythsandcustomsthatarestill thrivinganddeprivewomenoftheirrights,whichunderminesthe achievement of their strategic needs. For thosewho reach leadership positions,theyhavetoconformtomalestereotypesofleadershiptobeabletolead,ortheyhaveto resign (Painter‐Morland, 2011). Among factors that motivate some of them toresignvoluntarilyfromtheirpositionsarethe"exclusionfrommalesocialnetworks"and"valuesclashes".Thus,"organisationsarefailingtoretaintheirexecutivefemale

GENDERROLESINCOTTONPRODUCTION

101

talent because of their paternalistic organisational culture, the poor quality ofmanagement and their inability to accommodate the needs of their top femaleperformers" (Clark & Kleyn, 2011: 203). Quoting Gardner (1995), Werhane andPainter‐Morland (2011: 1) argued that defining leaders as "individuals whosignificantlyinfluencethethoughts,behaviours,and/orfeelingsofothers"mayaffectmembers’ mindsets and results in a “demeaning” authority relationship. Instead,leadershipshouldbeconceptualizedasan interactiveandcomplex interrelationshipbetween various individuals or individuals and leaders of organisations. To betterunderstand the gender role in leadership dynamics within organisations, Werhaneand Painter‐Morland (2011) raised the point of why so few women rise to topleadershippositionswithintheirorganisations.Theanswers lay inpersistentmythsabout leadershipandissuesofthesitewhereleadershipisenacted,thestructuresitrequires,andtheindividualswhoparticipateinit. Insociety,thelowrepresentationcan be self‐perpetuating,withwomenunable to convey their ability to lead (WorldBank, 2012). The present research sought to find out why so few women get intoleadershippositionsincottonorganisationsandwhetherthesepositionsempoweredthewomenwhodid.

5.3 Researchdesignanddatamanagement

ThedatausedcomefromaresearchprojectaboutcottonnetworkingthatwascarriedoutfromJanuary2009toApril2011inthenorth‐easternprovincesofBenin,BorgouandAliborifromwherethebulkofthecottonproducedinthecountryissourced.Asinotherprovincesofthecountry,traditionalcodesgovernthegender‐relatedbeliefsandbehaviourofalargepartofthepopulationintheregion,whichoverlapwithlegaldispositions. The region is characterized by many ethnic groups among which theBaatombuaredominant.Baatombusocietyhasapatrilinealkinshipsystem,virilocalresidence, andapatriarchal culture.Historically,however,womencouldhavemanynobletitlesasimportantasheadofachiefdom9.Fewofthehonoraryonescanstillbeinherited nowadays. Contrary to the achieved titles ofmen, thesewomen titles areascribed.Oneofthemostimportanttitlesisthatof"Yonkogui",theQueenMotherwhochairs the name‐giving ceremony by shaving and bestowing a "baptismal name" onyoung descendants from noble families (Mama Debourou, 2009; Schottman, 2000:94). But still, women's productive functions are constrained by their limitedentitlement to landedproperty. Land is, indeed,mostly community property in thisregion,andagricultural landinheritanceisregulatedbycustomarypatrilinealcodes.Asaresult,women's"rightsarecontingentonstatus"andtheydependuponmenforthelocationoftheirpersonalfields,whichisoftennotthebestland(Gray&Kevane,1999:18;WorldBank,2002).

Theresearchusedbothqualitativeandquantitativemethodsofdatacollection,whichwereappliedinthreemainphases:i)anexploratoryphaseoffocusgroupdiscussions

9KeyinformantinterviewwithBabou,griot(storyteller),Parakou,March16th2011.

CHAPTER5

102

withfarmersintenpurposivelyselectedvillages;ii)individualin‐depthinterviewsinfive villageswith 148 respondents selected fromnine cotton organisations; and iii)key informant interviews that were done whenever a potential key informant wasidentified.Thefocusgroupdiscussionsusedachecklist,whilethein‐depthinterviewswere carried out by means of a semi‐structured questionnaire to simultaneouslycollectquantitativeandqualitativedata.Thequestionnairewasconstructedbasedontheoutcomesofthefocusgroupdiscussions.Itaddressedmorespecificgenderissues,suchasthedivisionoflabourincottonproduction,andwomen'sdependencyontheirhusbandsduringandafterthecottonboom.Partof thedatawascollectedusingthelifehistorymethod(Angrosino,2002;Hagemaster,1992),whichwasappliedtotwowomen leaders.We used their life stories as the primary data source (Shacklock&Thorp, 2005), but complemented them with individual in‐depth interviews for abetter understanding of the full complexity of their lives (Pamphilon, 1999). Issuesaddressedincludedthewomen’sfamilyoforigin,theirownfamily,theirinvolvementincottonproduction,theirtrajectoryasboardmembers,andthechallengestheyfacedaswellasthefavourablefactorsthathelpedthemgettingontheboard.

For the individual interviews a sample of 148 cotton farmers was constructed,combining purposive and random sampling. A total of 33 leaders at the top of theorganisations'hierarchy,ofwhichonlyonewasawoman,waspurposively selectedfor the first round of interviews. The remaining 115 farmers were selected bysystematic sampling from farmer group membership lists at village level. Theseincluded23women,whichraisedthenumberofwomeninthesampleto24(16%).Insearch of women leaders in the organisations, we were informed of three otherwomenwhohadoccupiedboardpositions.Thefirstwasthewifeofaleader,whohadappointedher tobe"inchargeofwomen'saffairs."She left thepositiona fewyearsago,aftershedivorcedherhusband.Theother twowereappointedas treasurers indifferent villages, but both stopped many years ago and could not be reached.Therefore,astheonlyrecentlyactivewomenoncottonorganisationsboards,thetwowomenleadersweretargetedforthelifehistorymethod.

QuantitativedatawereanalyzedusingSPSSStatistics19.0forat‐testandachi‐squaretocomparemenandwomenwithregardtotheircharacteristicsincottonproduction,and a binary logistic regression to assess factors that affected farmers’ leadershipstatus incottonorganisations.Qualitativedatawere interpretedthroughanarrativeanalysis.The lifestoriesof thetwowomenleadersweretranscribedforadiscourseanalysis(Shacklock&Thorp,2005).

GENDERROLESINCOTTONPRODUCTION

103

5.4 Results

5.4.1 Genderedprofilesofcottonfarmersandthedivisionoflabour

The t‐test for equality of means (Table 5.1) and the chi‐square (Table 5.2) ran tocompare cotton farmers’ profile according to gendered‐groups, revealed significantdifferencesbetweenwomenandmenwithregardtoselectedcharacteristics.

Table5.1 Comparisonofmen’sandwomen’scottonprofile(t‐testforequalityofmeans)

VariablesGender

tFemale Male

Farmer’shouseholddependencyratio 1.51(.31) 1.55(.44) ‐.36

Experienceincottonproduction(Years) 14.17(6.98) 21.32(10.49) ‐3.22**CottonproductioninHa(mean2006‐2010) .96(1.13) 2.91(2.60) ‐3.61***Shareofcottonincome(%) 36.10(19.91) 46.66(27.26) ‐1.62Cottonorganisationalmembershiphistory 1.67(.70) 2.15(.78) ‐2.78**Numberofboardpositionsoccupied .29(1.08) 1.02(.95) ‐3.38**Durationofappointmentonboards(Years) .67(2.70) 4.62(4.66) ‐4.02***

Note:**=p<.01,***=p<.001;Figuresinparenthesesarestandarddeviationsofmeans.

Source:Surveydata,2009–2011.

Onaverage,menhadlongerexperienceincottonproduction(M=21.32,SD=10.48)thanwomen(M=14.17,SD=6.98),with themeandifferencesignificantatp< .01.Moreover,men’scottonareawaslarger(M=2.91,SD=2.60)thanthatofwomen(M=.96,SD=1.13).Thisdifferencewashighlysignificantatp<.001.

Table5.2 Cross‐tabulationofgenderandeducation

LevelofEducationGender

2Female Male

Noeducation(0) 20(83.3%) 63(50.8%)

9.05*Primaryschool(1) 1(4.2%) 26(21.0%)

Secondaryschool(2) 3(12.5%) 31(25.0%)

University(3) 0(0.0%) 4(3.2%)

Note:*=p<.05.Percentagesappearinparentheses

Source:Surveydata,2009–2011

With regard to the membership history in cotton organisations, women had beenmembersoffewercottonorganisations(M=1.67,SD= .70)thanmen(M=2.15,SD=.78),withasignificantdifferenceatp<.01.Therewasalsoasignificantdifferenceatp<.01betweentheaveragenumberofboardpositionsoccupiedbymen(M=1.02,SD

CHAPTER5

104

= .95)andbywomen(M= .29,SD=1.08).Ahighlysignificantdifference(p< .001)also existed between the average duration of appointment on boards of men (M =4.66, SD = 4.66) and that of women (M = .67, SD = 2.70). In addition, the cross‐tabulationofgenderandeducationshowedasignificantassociationbetweenthelevelofeducationandgender.

Whenaskingabout thedivisionofwork incottonproduction, theanswersreflectedthe taken‐for‐granted gendered division of labour that underrates women's share.Thus,menspontaneouslyandwidelycitedharvestingasexclusivelyawomen'stask.Ploughing and pesticides treatment were listed as exclusively men’s tasks, whilesowingandfertilizingwereidentifiedassharedtasks.However,toavoiddelaysinthework on their cotton farms, certainwomenhousehold heads are forced to performtasks that were deemed unsuitable and risky for women, such as ploughing andpesticidestreatment.Evenso, therewerealwaysmenclaiminginferiorityofwomento men. During a discussion about the gendered division of labour, a formerorganisationleaderalludedtowomen’sinferiorityasfollows:

Is it not you intellectualswho have decreed thatwomen are equal to usmen?Youignorethecommonfirstinquiryafterabirth.Whichchild[sex]isit?"Tònkuro"[agirl],meaningsomeonewhoisgoingtoleavethehome,or"Tònduro"[aboy],meaningtheonewhoiscomingintothehouse.Howcanyouequatethetwo?

In addition, womenwere less involved in themanagement of cotton production athousehold level compared tomen, and thiswas reflected in the levelof accuracyofcotton‐related information they provided. While men were likely to record cottonproduction data in the area and could specify tonnage over the past years,womenrarelydidso.Thefiguresthelatterwereabletoprovideweremostlytheamountsofmoneytheyreceivedfromtheirproduction.Toshortentheattempttorecalltheexactproduction records, awoman admitted: "I do not go to the cottonmarket. It is ourhusbandwhogoesthereandbringsbackhomewhatisleftafterthedeductionoftheinputscosts."

5.4.2 Womenandcottonorganisations

Women figured in the organisation lists and represented about 20 percent of therandomsample.However, theystill lackedcontrolover theirmembershipandoftendidnotknowthenameoftheorganisationtheybelongedto.Adivorcedwomansaid:"Iammemberof theorganisationofwhichmybrother isamember."Asa resultofthis lackof control, she couldnot freely changeorganisationas longasherbrotherwas amember, though she had been growing cotton for three yearswithout beingpaid.When thebrother finallydecided tostepout, she feltdeeplydisappointedandabandonedcottongrowing.

GENDERROLESINCOTTONPRODUCTION

105

Women's absence from cottonmarketing could generate tension among couples. Acotton farmer swore not to have received the money from the last cotton season,neither forhimselfnor forhiswife.But thewifedidnotbelievehim. "Shehasbeenaccusingmeofhavingspenthermoneywithouttellingher",complainedthefarmer.Awoman farmer had a different experience with her husband. At the end of the2007/08 season, she had quite a good harvest and was expecting some cash.Meanwhile, her husband was indebted. Applying the principle of joint liability fordebtsreimbursementbetweenmembersofthesameorganisation,theboardusedthefinancialoutcomeofthewoman'sproductiontocompensateforherhusband'sdebts,withoutherconsent.Shecouldnotacceptthat"asawoman,onegrowscottontopayforaman'sdebts".Becauseofherhusband'sincapacityorunwillingnesstoreimburse,shedecidedtoseparateandtogobacktoherfamilyoforigin.Therewere,however,married women who grow cotton in shared plots with their husbands. They weremostlynotregisteredincottonorganisationsintheirownname.Aftermarketing,theyreceivetheirshareoftheproceedsfromtheirhusband.Inpolygamoushouseholdsaswe found in a remote village in the heart of cotton belt, the redistribution is doneaccordingtoeachwife'scontributiontotheproductionasestimatedbythehusband.A head of a polygamous household, husband of three wives, assumed that "thiscollectivewayofgrowingcottonismorebeneficialforwivesbecausetheyearnmorethantheycouldhavegotfromaseparateplot."

Thelowinvolvementofwomeninorganisationmanagementwascapturedbyanotherwomaninthefollowingstatement:"Wewomen,wedonotknowtheanswerstoyourquestionsbecausewearenotinvolvedinthemanagementoftheorganisation.Soyoushould ask our husbands." Despite this complaint, a woman stated that "it is notworthwhile to be a boardmember as awomanbecause youwill always be left outfromactivitieswithnoinformation."Thisisbecauseofthelackofconsiderationfrommenwhorarely judgeitnecessaryto involvewomen."Whenawomanis invited", itwould be just symbolic, because "men would agree first among themselves beforeinviting you as a woman", argued another woman, assistant to the councilor of avillage.

The absence of women from cotton boards came out as one of the causes thattriggeredlengthydiscussionswithwomenwheninvestigatingthecausesofconflictswithinandbetweenfarmerorganisations.Indeed,thebinarylogisticregression,withleadership as the dependent variable that takes the value “1” for leaders and “0”otherwise, revealed a positively significant effect of gender (p < .01) on leadershipstatus(Table5.3).

CHAPTER5

106

Table5.3 Logisticregressionoffactorsaffectingtheaccesstoleadershippositionsincottonorganisations

Variables B(SE) Exp(B)

Farmer’shouseholddependencyratio ‐.25(0.64) .78Experienceincottonproduction(Years) .04(0.02) 1.04CottonproductioninHa(meanofthepast5years) .06(0.10) 1.06Cottonorganisationalmembershiphistory 1.01**(0.34) 2.74Gender 3.05**(1.13) 21.07Levelofeducation 1.82***(0.37) 6.18Constant ‐6.29***(1.73) .00

N=148;R2=.56;Model2(6)=79.52;Predictionpower=78.9%;**=p<.01,***=p<.001

Source:Surveydata,2009–2011

Educationandcottonmembershiphistorywerealsosignificantatp<.001andp<.01respectively.Themodel2(6)=79.52, ithadalsoagoodpredictionpowerofabout79%,withanR2=.56.

5.4.3 Lifehistoriesofwomenleaders

The life history method was applied to investigate the life experience of the twowomenleadersaccessedduringtheresearch.

Case1:Ritathe'ironlady'

Birthandchildhood.Rita isa57‐years‐oldwoman, livinginaremotevillageinthenorthwestofBenin.AlthoughRita’sfatherwasextremelypoor,shewassenttoschoolat the age of six. Ritawas still at primary school and nine years old, when – frommoney shehad savedherself – she paid the taxes for her father,whowas arrestedbecausehecouldnotaffordpayingthem.Afterthefatherwasreleased,hedeclared,intears:"Ifitwerepossibletochangethesexofachild,Iwouldhavetransformedyouintoaboy tostaywithme forever."WhenRitawas in the thirdyearofsecondaryschool,her father could no longer afford her education and took her out from school. Ritathenstayedwithareligiousfosterfamily,whereshetookuppettytrade.Sheusedtosell at the church,where shemetherhusbandwhowas thepastor.Thepastorwasamazedbyherdiligenceandhardworkandexpressedhisloveforher.

Ownfamilylife.AfewmonthsafterRitagotmarriedtothepastor,hewasassignedtoa remote villagewhere "lifewashard", according toRita. But her experience of theeconomic hardship she had suffered as a child helped her facing the situation. Shecombined farming and petty trading tomake her family's livingmore comfortable.Therelativewealthofthefamilymadethechurchhierarchydecidetosendthemtwicetoplaceswheretheywouldnothavetheopportunitytocultivateor trade.After thethirdassignment,Ritahadanargumentwithherhusbandwhowouldnotletherfarmanymorebecausehedidnotwanttobemovingconstantly.However,Ritamanagedto

GENDERROLESINCOTTONPRODUCTION

107

acquire a piece of land and started farming, in combinationwithpetty trading. Theresourcesgeneratedbyheractivitiesallowedthemtobuildthehousetheycurrentlylivein."Iamproudoftheseachievements.WehavenosalarybutGodblessedus!",shesaid.

Rita is a mother of twelve children, of whom only her eldest daughter got abaccalaureate, but she refused to continuewith her studies. On the ground of thatrefusalandtheensuingbehaviour,theyhadastrongargumentthatledtoabreachofcontactbetweenthemotherandthedaughter.

Cottonfarmingandrelated‐organisationsactivism.Ritastartedgrowingcottonatthe end of the 1980s, after the third assignment of her husband. In 1998, shewaselected as the treasurer of the district farmer union, "against aman", she proudlyspecifies.From2001onwards,afterRitaresignedfromtheorganisation,shehadledthe implementation of three cotton dissident organisations in the district. She wassuccessivelyappointedasinchargeofwomen'saffairs,deputytreasurer,andgeneralsecretaryintherespectiveboards.Shealwaysbattledforbettermanagementoftheseorganisations.

Inthethirdorganisationshewasamemberof,theprivatecottonginnertheysuppliedwith theirproductionat their secondseasonbrought in severalmillionFCFA to theorganisation.Thefewboardmemberswhocollectedthemoneyfromthebusinessmankept it for themselves and denied having received it. Among the frustrated boardmembers,Ritawastheonlyonewhodaredtoconfrontthemontheissueatstake.Shedeclared them guilty at a meeting. The other male colleagues afterwardscongratulatedher:"Wereallyappreciatedyourcourage,andwewouldlikeyoutojoinustocreateourownorganisation."Butshestraightforwardlynotifiedthemenofherfears, saying that she preferred to stay at home to take care of herself instead ofsufferingforpeople"whowillleavemeoutfromtheredistributionofthebenefits."Butthey promised her that such things would not happen. The initial meeting wasorganizedatherplace,where theydecided tocreateanewcottonorganisationandassigned board positions to the creators. The organisation operated for four yearsbeforeanextraordinaryassemblywasconvened.Ritarefusedtoattendbecauseshewasinformedverylateandhadtotravelfarforthemeeting.Inaddition,shecouldnotleavewhileherhusbandwasontravel.Atthemeeting,partialchangesweremadeinboardmembershippositionsandRitawasoustedfor"unknownreasons".Abouthernothaving reacted tobeingousted,Rita explained: "Idecided tokeepquiet toavoidcontroversies,thoughtheyhavenevernotifiedmeofanyfaultofmineforbeingoustedinthatway."

Beingsuccessivelyamemberofmanycottonorganisationsboards,Ritaattendedthemost importantmeetingsandnegotiationswith inputssuppliers,mostofwhomsheaccusedofbeinginclinedtocorruption.Shesaidthatmanytimesshehadbeenofferedimportantamountsofmoney,directlyorthroughintermediaries.Shefeltquiteproud

CHAPTER5

108

to never have succumbed to the temptation. She also faced all kinds of genderstereotypicalnaming,because"menhaveneveracceptedmeasafullmembercapableofachievingsomething",shecommented.Hercommitmentearnedherthenicknames'ironlady'fromherfellowboardmembersand'crusher'frominputssuppliers."Theythoughtthatbynicknamingme,Iwouldbeupsetandgiveup,butIdidnotmind",shesaid. Rita still recalls how some people tried to discourage and demobilize herfollowers when she promoted the breakaway organisations, saying: "You are silly!Otherwiseyouwouldnot followawoman.Doyouwanttoberuledbyawoman?"Butshe did not give up and many followers trusted her. For that, she judged herparticipationontheboardsasasuccessandcontinuesgrowingcotton.Shesaid:

IfIstopproducingcottonnow,farmerswouldthinkthatIbetrayedthem.They acknowledge my contribution in defending their interests andusuallysay:'Withoutthatwomanwhohaltedotherboardmembers'theft,wewouldhavebeen ina lotoftrouble fora longtime'.Iamgratefulforsuchanacknowledgment.

Membership inorganisations. Rita is also amember ofmany organisations otherthan cotton, on the boards ofwhich she has positions like president (of awomen’sgroup)andtreasurer(ontheboardofthedistrictcivilsocietyorganisation).Forovertenyearsshehaskepther leadingpositioninthewomen’sgroup.Whenitcomestorenew theboardmembers,"women say 'It isyouweprefer'and theydonotwant toreplaceme", she argued.Rita also has a political life, thoughwith no affiliation to aparticular political party. In 2008, she headed an independent list and ranunsuccessfullyforapositionatamunicipalcouncil.

According to Rita, her domestic responsibilities rarely impede her activism. "Myhusbandencouragesme inmyorganisationalactivities [...].He isnot likeothers", shesaid.Replyingtothequestionastowhyshehasbeenappointedonsomemanyboardsandhowshedealswiththeduties,Ritacontended:"Youknownobilityisnotgiven;itisdeserved."

Case2:Gnonthe'libertine'

Birthand familyoforigin.Bornas theeldestof a familyof fivebrothersand foursisters, Gnon is a 37 year‐old widow, living with her siblings, together with theirwidowedmother.Shetakescareofthefamilywiththehelpofherunmarriedyoungerbrother. However, it is rather the brotherwho acts as the head of their household.Gnon dropped out from school when she was in the third grade, because she wasashamedtohavetorepeattheclasswhileherfriendswenttoahighergrade."Itwasdueto ignorance.If itwerenow,Iwouldhavesimplyredonemyclass.",sheregretted.After Gnon dropped out, she was trained as a seamstress. Soon after getting thediploma,shegotmarried.

GENDERROLESINCOTTONPRODUCTION

109

Ownfamily.Gnonmarriedintoalargehouseholdofsevenmarriedbrothers,sharingdomestictaskswithmany"co‐wives".Shegavebirthtofourchildren.Hermatrimonialdutiesdidnotimpedeherorganisationalactivism,inspiteofthegossipgoingonatthetime because of her multiple travels for training. Her husband was regularlyconfrontedwithstatementssuchas:"Youanilliteratefarmer,youhaveadrop‐outwifetravellingeverywhere[...]Makeherstoptheseactivities".Buthewouldanswer:"Onlyherfatherknowshowmuchheexpendedtoeducateher.Iamnotwillingtospoilherjob.Ievenlikehertocontinuebecausewhatshefindsbelongstomyfamily."Thishusbanddiedprematurely,muchtoGnon’sregret.

Afterherhusband'sdeath, sheandallher children returned toher familyoforigin.However,shecontinuescultivatingthelandofherlatehusband,andwhichisownedbyher family‐in‐law.Fiveyearsafter thehusband'sdeath, theeldestbrother‐in‐lawaskedhertoleavethisplotofland,ofwhichtheyieldswereusedtoraiseherchildren.Allegedly, the brother‐in‐law declared: "A woman cannot inherit land from herhusband!"

Cotton farming and the cotton organisation experience. Though Gnon startedproducingcottonwithherhusband,herinvolvementwithcottonboardsstartedafterher husband's death. As awidow and relatively free in a society full of stereotypesaboutunmarriedwoman,Gnonfirstservedasthepersoninchargeofweighingcottoninafarmergroup.Whenthefourthbreakawayorganisationemergedinthevillage,itslocalpresidentcouldnot findaman tobe thesecretary,due toaviolentconflicthewasinvolvedin.HeapproachedGnonandtriedtoconvincehertoaccepttheposition.Sherefused.HeaskedGnon'sparentstointercedeforhim,explainingthathepressedtheirdaughter to 'savehim fromshame'but that she refused.Gnon's fatherobligedher to accept the position, ordering "to go and help him doing what he wanted."AccordingtoGnon,herpresenceontheboardmadethegroupsteadilygrowinsize.Sherecallednewmemberssaying:"Asitisawomanwhoisthesecretaryofthisgroup,thegroup shouldbe trustworthybecausewomenarenot contaminatedbyhypocrisy."However,shealsochallengedscepticswhowerewaitingtoseehowawomanwouldbeabletomanagecottonmarketingandpayfarmers.Sheproudlyadded:"Imarketedcottonasanymancoulddo.Ihadnodebtsandnoarrears."Tothequestionwhatmadeher leave the organisation, she answered: "If a husbandwho courted you does notacceptyouanymore,betterleavehimandgoyourway..."

Indeed,attheendoftheirsecondcottonseason,theorganisationdecidedtoallocatemotorbikes to some leaders.Gnon'swasselectedamong those leadersbecause"ouronlyladyissobraveandhard‐workingthatshedeservesamotorbike",someofhermencolleagueswouldhaveargued.Butafewdaysbeforetheallocationofthemotorbikes,a board meeting was convened by the president. The president would haveintroducedthemeetingstraightforwardly:

CHAPTER5

110

At the point we are now, the government [...] said that it is time forchange. You Gnon, as I see you, you will not be able to do the workpertainedbythatchangeincottonmanagement;itisbeyondawoman'scapabilities.

Replyingtothequestionastowhythisworkwouldbebeyondawoman'scapability,Gnon commented: "Because change has come, our president said [...] and onlymenwouldbemembersof theirboard." Inspiteofherdisenchantment,Gnonrespectfullykneltinfrontofthepresident,sayingwithloquacity:

Daddy,thankyou.Whenyouneededmetoruntheorganisation,Iwasnotanincapablewoman.ButIwilltellyouonething:thisisa'whitepower'[amodernpower].IfitwerethetraditionaltitleIamentitledtoinherit,youwould never have been able to oustmewithout fear for your life.ThankyouandonlyGodwillpayyouback.

Andsheleftthemeeting.WhenGnongotbackhome,shereportedtoherparentswhathadjusthappened:

You remember that the president told me a gourd of honey washangingonandonthedayitwouldbemature,Iwouldbenefitgreatlyfrom it. Today this honey ismature, but he broke the container andpouredallthehoneyonthefloorbeforeIcouldenjoyit.

She explained how shewas just ousted from the cotton board and held her fatheraccountable for that outcome. The father argued that a board position is not aninheritedtitleandthatGodwouldrequitethepresidentforwhathehaddone.

Membership in organisations. Gnon is a very active woman who participates indiverseactivities.Apartfromsewing,sheisthesecretaryofawomen’sgroup,theonlywoman literacy teacher in the village for ten years, and a community volunteer forhealthpromotion.Gon’smembershipinorganisationsotherthancottonstartedafewmonths after she got her diploma,when shewas appointed as the secretary of thegroup without her knowing. But despite being busy with all these duties, she hasmisgivings aboutbeing a 'freewoman'. She feels people's eyesonherbecause theythink that she does not want to marry again to keep her freedom. "I am reallyconcernedaboutgettingahusband,butIamnotlucky",sheconcluded.

Men’sreactiontowomen’soustingMenwerenotinclinedtocommentonthewaythetwowomenleaderswereousted.ThepresidentofRita'sorganisation,whowasoustedtogetherwithher,explainedthathedidnotattendthemeetingbecausehedecidedtostayfarfromcotton,whichhadbecomea troublesomebusiness.He saidhedidnotknowwhyRitawasoustedand

GENDERROLESINCOTTONPRODUCTION

111

was not "willing to commenton it". The newly appointed president argued that theformerboard,ofwhichhewasalsoamember,"didnotreporttonetworkmembers".Asfor the leaderofGnon'sorganisation,he foundonepeccadilloofGnon to justifyherousting:"Youknowsheisnotsocleanmorally."

5.5 Discussionandconclusion

The random selection of women from the list of the organisations reveals thatwomen’s membership of cotton organisations has progressed and that they canactuallyberegisteredasmembersintheirownname.Thiscouldbe(partly)aresultofthe atomisation of organisations with its corollary of competition for membership.Hence, women’s relatively recent investment in cotton productionwhich limits theexpressionoftheirpotentialofproduction.Women’signoranceoftheirorganisationsofmembershipmeansthatsomeofthemmighthavebeenregisteredbytheirhusbandorothermalerelativestoincreasethegroupsize.Theensuingculturallyunderpinneddivisionoflabourhassocialimplicationsbyforcingwomentoadoptandadapttomenhegemonic views, even if they are not aware of that. Deeply entrenchedmyths andstereotypedimagesareinternalizedbywomenaswell,althoughtheyaredemeaningto them. This is what Bourdieu (1989: 18) called “construction carried out understructuralconstraints”,aprocessthroughwhicheventhemostdisadvantagedactorstend toperceive the familiarworldasnatural.This false consciousness justified thewomanfarmer’scontestationofthe(forced)useofheryieldstopayforherhusband'sdebts, on the ground of her "femininity". Such cognitive processes take place ineverydaylife.Thesignificantdifferencesbetweenmenandwomenwithregardtotheexperience in cottonproduction, the areaplantedwith cotton, and the organisationmembership profile (membership history, number of board positions and thedurationoftheappointment)arecertainlyrootedintheseprocesses.Inthisinstance,harvesting is widely considered a women’s task in cotton production, because it ispresumablyeasy. In reality,harvesting is a tiringactivity. It is very time‐consumingand requires concentration. In addition, performing tasks in cotton production hasrevealed ablurredboundarybetween 'heavy and risky tasks' and 'light tasks' sincethe labour constraints force women to perform tasks that are supposedly men’spreserve. As a result, women's participation in cotton production actually runsthroughout the complete production cycle. This shows how women's tasks areunderrated,whilemenarereluctanttodothem.AsimilarobservationhasbeenmadeinUgandawhereasystematicalperceptionalbiasagainstworkperformedbywomen,which thenbecomesde‐skilledand/orde‐valued,undermineswomen’sability tobefairlyremunerated(Lodin,2012).

Inspiteofwomen’somnipresencethroughouttheproduction,theywerefoundtobelargelyabsent fromthecottonmarket,wheretheproduce isweighedandpaymentsare made, thereby lacking the control over the yields. Women were deprived ofessential information and had limited autonomy in decision‐making within cotton

CHAPTER5

112

organisations.Theyacceptedthemoneyfromtheirownproductiongiventothembytheirmalerepresentativesmorebyresignationthanbytrust,asitwasshownbythedisputesbetweenthetwofarmersandtheirrespectivespousesdescribedabove.Thislackofcontrolofwomeninthepublicsphereoriginatesfromculturalpracticesthatstronglyshapewomen'spositionintheprivatesphere,andthataffecttheirstrategicaswell as practical interests. For instance, economic and consumption needsmadeboth Rita and Gnon the de facto heads of their respective households, thoughwithdifferent entitlements.While Rita's role in the household was acknowledged, Gnonwasnotentitledtoheadhers, though in factshewasdoingso.Heryoungerbrotherwasseenastheheadofthehousehold,duetosocialandculturalnormsthatpreventawoman from heading a household where there is a man (Moser, 1994). Also, thesociety at large seems to resist to women's autonomy, as illustrated by thestereotypical reasoning the two women leaders faced in their family and cottonnetworkingexperience.Womenexercising theiragencyhardlymanage toovercomethehurdlesthatblocktheiraspirations.Consequently,whenmeetingpracticalgenderneeds is the overriding concern, meeting their strategic gender needs can be putbeyondwomen'sreachandbecomesmortgaged.

Womenhaveapoorercottonorganisationalmembershiphistorythanmen.Thehigheffect on and positive correlation of cottonmembership history with leadership incottonorganisationsmeans that themore cottonorganisations a farmerhasbeenamemberof,themorelikelys/hewouldbeaboardmember.Theoddsofbecomingaboardmemberwas about 3 for an increase by 1 of the number of organisations afarmerhasbelongedto.Leadershipincottonorganisationswasalsohighlyaffectedbygender,andbyeducationthatwomenwerelessendowedwith.Lessthanonefifthofwomen in the sample had formal education against one of twomen. However, theWorldBank(2012:80)identifiededucationasonethemainfactorstoreducegendergaps in the developing world. Moreover, the positive correlation of education andgender with leadership in cotton organisations implies that their effects had beencumulatively undermining women’s position therein. The odds ratio of 21.07 forgendermeansthatforeachshiftfrom0(female)to1(male),afarmerwasabout21timesmorelikelytobecomealeaderincottonorganisations.Thislikelihoodforonelevel different in educationwas about 6. These illustrate howwomen are excludedfrom cotton organisations. Crenshaw (1991) coined the concept of intersectionalityforsuchinteractions.

TheresearchshowedthatRitaandGnonwereexceptions.Bothexhibitedexceptionalfeatures that many other women lacked, and which might have come into play infavouring the emergence of their leadership. Rita and Gnon were relatively welleducatedwomenintheirrespectivevillages.Athousehold level,bothRitaandGnonwerefortunatetohaveopen‐mindedhusbandswhomtheybothqualifiedas'differentfrom the other men'. Rita’s hardworking caused her family to be moved from oneplace to another by the church hierarchy. As for Gnon, her capabilities wereacknowledged by her late husband who rejected the gossip going on about her,

GENDERROLESINCOTTONPRODUCTION

113

althoughhisattitudeseemsrational inviewof thebenefits for thehouseholdofhiswife's extra‐domestic activities. Men may be inclined to give more freedom towomen's activities when they find them rewarding for themselves. Consequently,picturing gender equity as being inmen's interest could help to have it gainmoreacceptanceamongthemalecommunityinruralareas.

RitaandGnonwerequiteactiveasmembersofandhavingresponsibilities inmanyorganisations, which might have justified why they were among members with arichermembershiphistoryincottonorganisations.Rita'sinitiativeandassertivenessenabledher to confronthermaleboardmateswith their embezzlements.Thus, sheappearedtobeanaturalleaderbyinteractingeffectively.Thisiswhyshewasreferredto by many farmers, including men, for agricultural issues even before she wasappointed as the treasurer of the district union.As a result, she benefited from theordinary farmers' support and, in turn, was concerned about being continuouslytrustedbythesefarmers,towhomshefeltobliged.AsWerhaneandPainter‐Morland(2011:3)stated,"leadershipisnotnecessarilyrestrictedtoindividualsappointedtopositionsofauthority,butisadispersedcapacityoftheorganisationasawhole."Ritaherself concluded the interview with a meaningful statement about the deservedcharacterofnobility.

Thefateofthetwowomenleaderswithintheirrespectiveorganisationsconfirmsthemale stereotypicaldenialofwomen's capabilities to lead, althoughmen individuallyadmittedthesecapabilities.Paternalisticcultureisstilldominant,causingmentodenycertainprinciplesatgrouplevelwhilepersonallyacknowledgingthem.However,thereaction of Rita and Gnon to the unfair way they were ousted uncovers somedifferences of perception between men and women about organisation‐relatedconflicts.Gnon'sreactionshowshowthesamereferencetotraditionalnobletitlesisnegotiable,offeringadifferentperspective.Consideringthepositionontheboardasanelectedandanon‐ascribedtitleimplies,regardlessofhowshewasousted,thatsheunderstoodthatasaboardmemberoneisreplacedbysomebodyelseatsomepointintime. Rita's reaction was to keep quiet and not to stir trouble, while cottonorganisations in Benin have been characterized by a stiff leadership competition.Indeed,mostmaleleaderswereknownfortheirreluctancetoleavetheirpositionsonboards, referring to traditional ascribed titles to justify their behaviour. Thus, thewomen's reaction to group conflicts contrasted radically with that of the men. AsindicatedbyCarli(1989),whilemenfocusmoreondisagreements,womenaremorelikely to voice agreements. The women's presumption that their presence on theboards could reduce intra‐groups conflicts doesnot seem tobe entirelyunfounded,althoughinthiscasethetwowomenleadersweredisempowered.Consequently,"thesocializeddifferencesbetweenmenandwomen"thatarecapturedintheconceptofgender should highlight the value of the female nature for an "effective andmuch‐neededmanagementstyle"(Ely&Meyerson,2000:108‐109).

115

Chapter6

RuralHouseholds’ResponsestotheCottonCrisis

ABSTRACT: Relying on one source of income puts the livelihood system ofrural households at risk. Cottonproduction inBeninhas longbeen the corecashcropofrurallivelihoods,untilthemid‐2000swhenmultipleconstraintsledtothedemiseofcotton.Thepresentchapterinvestigatesruralhouseholds’responsestotheeconomicshockresultingfromthedecreaseofincomefromcotton.Theresultsrevealthathouseholdsdiversifiedtheirsourcesofincomeon‐farm, with food crops gaining increasingly a cash function. However,because the production system heavily depends on cotton for access tofertilisers and other inputs for food crops, farmers continue to grow cottondespite its present low returns. In addition, because of theirmultiple extra‐domesticactivitieswomenseemtobelessvulnerablethanmenwhenitcomesto copingwith livelihood shocks. Hence, their contribution to providing forhousehold needs increases in time of crisis. Further results suggest thatdecisionmaking and certain diversification strategies of young adultsmovefrom the centre of the household to its periphery and loosen householdboundaries.Keywords:Cottonproduction;incomeshortage;livelihooddiversification;ruralBenin

Ashortenedversionofthischapterwillbesubmittedtoasuitablejournalunderthesametitle,bytheauthorsMaboudouAlidou,G.andNiehof,A.

CHAPTER6

116

6.1 Introduction

CottonproductionhasforlongbeenavitallivelihoodactivityforruralhouseholdsinBenin where it constituted a smallholder cash crop for the last three decades (seeChapter2).Itwascriticallyimportanttoruralwelfare,sinceabout45percentofruralhouseholdsdependedoncottonrevenues,whichconstitutedasourceoflivelihoodformorethan325,000ruralhouseholds,i.e.aboutthreemillionsofpeople(Baffes,2004;UNCTAD,2008;WorldBank,2003).Cottonisgrowninrain‐fedlandswithhighuseofagriculturalinputs(Baffes,2004).Duringthe2000s,onethirdofthefarmhouseholdsinBeningrewcotton,allottingabout18percentofthetotalcultivatedareatocotton(Kherallah et al., 2001). In the north‐eastern and central provinces of Borgou andAliborithatarethemaincottonproducingareas,cottonfarmershadonaverage2.62and3.15hectaresofcotton(Kpade,2011).Duringthecottonboomofthe1990s,theaverageacreageallotted to thecropsteadily increased to37and64percentof thetotalcroppedland,respectively(Gandonou,2005;Minot&Daniels,2005).

Duringthe1990s,cottonprovidedupto80percentoftheincometoover45percentof rural households (Minot & Daniels, 2005; World Bank, 2003). The incomegenerated by cotton has the comparative advantage of being received in importantamounts at once,whichprovidesmoney for largeexpendituresas constructionof ahouse,weddings,schoolfees,motorcyclesandcars.Thesefeaturesofcottonrevenuesmadecottonacriticalcropforpovertyalleviationamongpoorruralhouseholds,andcreated a large dependency of those households on cotton incomes for theirlivelihoods(Baquedanoetal.,2008;Woodward,2007).MinotandDaniels(2005:460)found that a 40 per cent fall in the cotton prices resulted in 8 per cent rise of theincidenceofpoverty,resultinginabout334thousandadditionalpeoplefallingbelowthe poverty line. Thus, the cotton sector constituted the backbone of the country’seconomy.However,abouttwodecadesaftertheboomintheproductionstarted,withtheensuingincreasedprosperityofcottonfarmers,signsofcrumblingappeared.

From the mid‐2000s onwards, the positive benefits of cotton have completelyreversed (World Bank, 2003), due to a combination of multiple constraints thatoverwhelmedtheproduction.Amongotherconstraintswerethefallofcottonpriceson international markets, the mismanagement of cotton‐related organisations, andpaymentarrearstofarmers.A40percentdecreaseofpricesresultedina61percentincreaseofpoverty (Minot&Daniels,2005).TheWorldBank(2003)and theUNDP(2005) linked 12 per cent of the increase of the incidence of poverty in the Northduring the early 2000s to the decline of cotton prices on international markets.Therefore, cottonhadbecome an erratic and insecure source of living, especially inthe north‐eastern provinces where an 18 per cent reduction in per capita incomeresulted in an equivalent increase in the incidence of poverty. As a consequence,farmers becamedisenchanted andmassively abandoned cottonproduction, therebynarrowingdrasticallythebasisoftheirlivelihoods.Tothegrowingdisillusionmentofproducers, the State offered no or very few solutions (World Bank, 2003). Thealternatives explored to remedy the country’s high dependency on (conventional)

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

117

cottonandtoreleaseboththeStateandfarmersfromtheirwoeswerediversificationof agriculture and exports and organic cotton production (MAEP, 2011). The first,diversification, seemed an empty shell, since farmers have always diversified theiragriculturalproductiondespitetheirdependencyoncottonforcash.Ruralhouseholdsalways possessed diverse livelihood portfolios to absorb social, economic andenvironmental shocks, implying that especially among poor rural households,diversificationisthenorm(Barrettetal.,2001;IFAD,2010).

Since1996,organiccottonproduction isbeingpromotedbyNGOs inparticularasasustainable solution to the numerous environmental side effects imputed toconventional cotton. Organic cotton production is also used in the lobby againstgenetically modified cotton (Tovignan & Nuppenau, 2004). However, not all actorsbelieve in itspotential.Therelativelyhigherpricecompared to thatof conventionalcottondoes not compensate for its lower yields.Hence, organic cotton turns out toplay a marginal role. Instead, based on their traditional modes of production,householdsdeployvariousstrategiestorespondtothedeclineofcottonrevenuesinordertomakeendsmeet.

Theobjectiveofthepresentchapteristoidentifyfarmers’strategiestodealwiththeshortage of income consecutive to the demise of cotton in the north‐eastern andcentralprovincesofBenin.Howdohouseholdsadapt toor copewith thedeclineofcotton revenues? How did changes in income patterns modify men’s and women’scontributiontotheprovisionfortheirhouseholds’dailyneedsandhowdidthisaffectthe gender relations within the household? The following section of the presentchapterfirstsetsoutthetheoreticalperspectiveoflivelihoodsandhouseholdsbeforedescribing thestudyareaandstudydesign inSection6.3. InSection6.4, theresultsarepresented anddiscussed.The chapter concludeswith a final discussion and theformulationofagriculturalpolicyimplicationsderivedfromthefindings.

6.2 Livelihoodsandhouseholds

Duringthepastdecades,unravellingthelogicoffamilyandhouseholdproductionandconsumptionhasreceived increasingattention fromscientistsandpolicymakers. Inan attempt to uncover the dynamics of urban economies in Africa, two types ofapproacheshavebeendocumented:theinformalsectorapproachesandthesurvivalapproach. The informal sector approaches use three dominant perspectives – thereformist, the institutionalist, and the neo‐Marxist – to shed light on the economicsituation of people classified as active in the informal sector, but these approachesfailed to capture the interplay of simultaneous informal and formal sectoremployment.However,theapproaches“playedacrucialroleindrawingattentiontothe poverty and the plight of people involved in such activities, the employmentpotentialsofthesector,andthecreativityandentrepreneurialskillsofinformalsectorparticipantswhomakeitamidstinappropriatestateregulations”(Owusu,2007:451).Whileformalsectorapproachesfocusmoreoneconomicactivitiesthanonthepeople

CHAPTER6

118

performing theseactivities tomake their living, informal sector approaches capturethe variety of people concerned and their activities. Indeed, the constraints of theformalsectorhaveforcedmanyemployeestojointheinformalsectorasan“income‐supplementingand/orincome‐diversificationstrategy”(Owusu,2007:452).

The survival strategy approach, on the other hand, has themerit of analysing bothpeople’s responses to economic hardship and the context of their decision‐makingprocesses. The survival strategy approach is mostly applied to rural contexts andexclusivelytopoorpeople,whoareattributedarationalityinriskminimization.Ithasbeenwidelyusedtoanalysepeople’sstrategicresponsestoeconomiccrises,showingits value “for exploring the dynamic nature of the environment inwhich livelihooddecisionsaremade”(Owusu,2007:452).Thelimitationofthisapproachstemsfromthe fact thatpeoplewithmoreassetsand, consequently,notpoor, aremoreable to“strategise”thanpoorpeoplewhoare limitedintheirstrategicchoices.This impliesthatanalysingadaptationtoeconomiccrisiscannotbelimitedtosurvivalstrategybutshould be extended to capital accumulation strategies by richer people. In fact, theassumptionofrationalchoiceisonlyjustifiedwhenanindividualhasmultipleoptions.Becauseoftheseconceptuallimitationslivelihoodapproachesgainedmomentum.

Comparedtothepreviousapproaches,livelihoodapproachesaremoreencompassing,sincetheyaredirectedatdiscoveringtheinteractionsbetweenthedomesticunitanditsenvironmentandintegrateinstitutionalistandneo‐Marxistsviews(Owusu,2005).Livelihoodapproachesprovide amodel thathelpsapprehending themanywaysofraisingextraincomethroughadditionalactivities.Theunstableandadverseeconomicconditions in many parts of sub‐Saharan Africa have led households to search foradditional income by engaging in multiple activities (Owusu, 2007). Although thepracticeblossomedamongpubliccivilservantsasoneof thedirectconsequencesofthestructuraladjustmentprogramsduringthe1990s(Owusu,2005),itconstitutesahistorical strategy among vulnerable rural populations (Ellis, 1998; Niehof, 2004).Unfavourablenaturalconditions,suchaslowanderraticrainfall,droughts,poorandinfertilesoils,areconsideredsevereconstraintstotheimprovementofthesituationofrural communities (Frost et al., 2007), forcing them to strategize to improve theirhouseholdlivelihood.Relyingonmorethanonesourceofincomeis,therefore,neithertheprerogativeofpublicsectoremployeesnorofasocialclass.

Livelihood is a complex and “multi‐faceted concept, being bothwhatpeopledo andwhattheyaccomplishbydoingit,referringtooutcomesaswellasactivities”(Niehof,2004: 322). Livelihood refers to “the mix of individual and household strategies,developedoveragivenperiodoftime,thatseekstomobilizeavailableresourcesandopportunities”(Owusu,2007:452).AccordingtoIFAD(2010:52),“thelivelihoodsofpoor rural households reflect on the one hand the opportunities and constraintscharacterizingtheareaswheretheylive[...]andontheother,theirownprofilesandcharacteristicsashouseholds.”

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

119

Thehouseholdconstitutesthe“locus”ofthelivelihoodsystem(Niehof,2004),becauseit provides the immediate context for the strategic management and allocation ofresourcestoprovidefordailyneeds.Asaconsequence,moststudiesaboutlivelihooddiversificationusethehouseholdastheunitofanalysis,whichconceptualisationhaschanged over time. Unitary models previously used to apprehend the householdperceived it as “a collection of individuals who agree over the broad principles ofintrahousehold resources allocation” (Akram‐Lodhi, 1997: 28). But in thesemodels,theinteractionbetweenhouseholdmembersandtheirdifferentpowerpositionsandlogic tended to be glossed over. Alternativemodels of household conceptualisationwere proposed that led the concept to undergo some changes.Niehof (2011) notesthatcontrarytothefamily,thehouseholdhasfluidandadaptableboundariesandthathousehold ties are not underpinned by law, which makes them more fragile andmakes the exit option relatively easy. The collective or joint utility model of thehouseholdsucceededinsheddingmorelightontheinternaldynamicsofhouseholdasunitofconsumption(Maurizio,2007),buttheassumptionofjointutilityfailedtotakeintoaccounttheagencyofhouseholdmembersassocialactorswhofacethedilemmaof cooperationand conflict inhouseholdproductionand livelihoodgeneration (Sen,1990;Wallace,2002).Thebargainingapproachesputthefocusonpowerrelationsinthe distribution and allocation of resources within the household (Agarwal, 1997;Holger Seebens, 2007). Following Rudie (1995: 228), we see the household as afamily‐based co‐residential unit that takes care of resource management and theprimaryneedsofitsmembers.WealsoacknowledgethepointsmadebySen(1990)and Agarwal (1997) that household members may have conflicting interests andunequal power positions,which necessitate bargaining and can result in householdmembersleavingthehouseholdtemporarilyorpermanently.

Conceptualising the household livelihood has led to the emergence of two closelyrelated concepts: diversification and adaptation. Though not synonymous, the twoconcepts are sometimes used interchangeably, both being referred to as strategies.Evidence has shown the strategic role of diversification in rural livelihood systems(Niehof, 2004). Diversification is defined as “the process by which rural familiesconstruct a diverse portfolio of activities and social support capabilities in order tosurviveandtoimprovetheirstandardsofliving”(Ellis,1998:4).Inthisprocess,notonly income but also social institutions, gender relations, and property rights thatcontributetoastandardofliving,areincluded.Diversificationshouldnotbelimitedtoonlysourcesof income,asmanystudiesdo,but includesdiversificationofassetsaswell (Ellis, 1998). Diversification as an individual or a household level strategy isneitheraruralphenomenonnorthepreserveofonlydevelopingcountries(Barrettetal., 2001; Ellis, 1998; Wallace, 2002). It can occur as a purposive strategy or aninvoluntaryresponsetoacrisis.Thelatterisequatedwithcoping:diversificationoutofnecessity,orfor‘bad’reasons(Niehof,2004).Thefirstisdiversificationbychoice,which aims at strengthening the livelihood, accumulation of assets, and riskmanagement. In theeconomicmodelof theruralhouseholds theon‐farmreturns tolabour time can be compared to the returns of off‐farm or non‐farm earning

CHAPTER6

120

opportunities (Ellis, 2000: 292). Hence, livelihood diversification can be realizedthroughfarm,off‐farmandnon‐farmactivities, therebygeneratingdifferenttypesofincomesandassets.Barrettetal. (2001:368‐369)useda three‐wayclassificationofruralhouseholds’sourcesofincomeanduseofproductiveassets:bysector(e.g.farmvs.nonfarm),byfunction(wagevs.self‐employment),andaccordingtospace(localvs.migratory).AbdulaiandCroleRees(2001)foundthatincomeofpoorerhouseholdsislessdiversifiedduetotheir lackofcapital,whichoffersthemfeweropportunities innon‐farmwork.

Livelihoodadaptationisperceivedasadjustmentsofindividuals’orhouseholds’con‐sumptionandproductionpatternsinresponsetoobservedorexpectedeconomicandsocialhardshipsuchasincomedecline.Ellis(2000:298)defineslivelihoodadaptationas the continuous process of “changes to livelihoodswhich either enhance existingsecurity and wealth or try to reduce vulnerability and poverty.” The concept ofadaptationisbroaderthanthatofdiversification,becauseitoffersmoreopportunityfor including activities outside the traditional domains of the targeted group.Adaptation can be by choice (positive) or of necessity (negative), the latter beingconstrained by limited options. Ellis (2000: 290), argued that the distinguishingfeatures of rural livelihood strategies in poor countries are the “maintenance andcontinuous adaptation of a highly diverse portfolio of activities.” The scope fordiversification and adaptation depends on the vulnerability of the individual orhousehold concerned. Vulnerability is dynamic andmultidimensional, capturing theresponse to changes over times. Household vulnerability is determined by externalthreats and internal assets. Consequently, vulnerability “measures the resilienceagainst a shock [and] isprimarilya functionof ahousehold’s assetendowmentandinsurancemechanisms”(WorldBank,2000:139).

Thischapteranalysesthewaysinwhichfarmhouseholdsinthestudyareastrategisetodealwith theeconomicshock to their livelihoodscausedby thedeclineofcottonprofitability.Itfocussesmainlyonon‐farmincomediversificationasthemainstrategyof the rural households, given the scarcity of non‐agricultural income‐earningopportunities.

6.3 Researchareaanddatasources

Theprovinceswherethefieldworkwasconductedcovertwoagro‐ecologicalzonesinthenorth‐easternregionofBenin:the“cottonzone”andthe“north‐centralzone”,alsocalled the “foodbelt”.Therelatively lowrainfall, the typesof soil, and therelativelyhightemperatureinthearea,madeitsuitableforcottonproduction.Throughoutthisregion, households have been always heavily dependent on cotton for theirlivelihoods,duetothehighshareofcottonincomeintheiroverallincome.SiaensandWodon (2008: 174) argued that “cotton producers fared relatively well” over the1990s.Theyfounda7percentreductionintheprobabilityofbeingpoorforcottonproducing households compared to non‐producing ones. Therefore, people are

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

121

reluctanttoabandoncotton.Additionally,inthestudyareatherearefeweralternativeincome‐generatingopportunitiesthaninthesouthernpartofthecountry.

Four villages were selected as the research sites in three of the largest cottonproducing districts: Banikoara, Gogounou and Sinende. Banikoara contributes up to45percenttothenationalproduction,whichmakesthedistricttheheartlandofthecotton belt. Gogounou represents the third biggest producer, while Sinende is thelargest cotton producer within a region that is considered the food belt. In 2004,Banikoara and Gogounou together supplied 64 per cent of the national cottonproduction, while the production of the north‐central zone comprising Sinenderepresented29percentofthenationalproduction(Gergely,2009).Theresearchwascarriedoutduring2009‐2011.Thiswasa fewyearsafter theprecipitousdeclineofcotton production had set in. The relatively huge amounts of subsidies invested torepayvariousdebtsinthesectorhadlittleornoeffectonthistrendthatwastriggeredin2005.Thisledthegovernmenttoenvision(re)managingthesector,whichhadbeencompletelylefttoprivatemonopoliessincethebeginningofthe1990s.

The districts and villages were selected purposefully to cover zones of high andmiddle levels of cotton production, areas of low and high abandonment of cottonproduction,andbothaccessibleandremotevillages.Asrespondentsinthesurvey,asampleof148cottonfarmerswasselectedbysystematicsamplingfrommemberlistsof farmergroups.These farmers,bothmaleand female,werealsohouseholdheads.Prior to the household survey, qualitative datawere collected through focus groupdiscussionsconductedwithgroupsofcottonfarmers.Thehouseholdsurveywasdoneby means of a semi‐structured questionnaire. It addressed household livelihoodissues,suchasthecompositionofthehousehold,productionandmanagementoffoodand cash crops, the place of cotton in the livelihood portfolio during the boom andafter the decline of cotton production, the evolution of household income sourcesduring the same period, and spouses’ contribution to the household needs (seeAppendix 2) for the complete questionnaire). Only the heads of households wereinterviewedabouthouseholdstrategiesandtoprovideinformationonallhouseholdmembers. Household membership included all persons participating in thehousehold’sdailylifeforatleastoneyear.

Foranalyticalpurposes,wedistinguished theprime timeof cottonproduction fromthedecline.Thecottonprimestartedfromtheboomatthebeginningofthe1990sandended in the mid‐2000s. During this period, cotton production had high economicreturns,althoughsomeinputs,suchasfamilylabour,werenottakenintoaccount.Thedecline began in 2005, when farmers experienced ineffective inputs, mainly badqualitypesticides.Thisresultedinlowyieldsandledtoindebtedness.

Although there was a high ratio of male to female respondents in the sample(124/24),itwastriedtoavoidmasculism(Blais&Dupuis‐Déri,2011)orthemistakeof using a male perspective. Farmers’ perceptions of changes in the respectivecontributionsof husbandsandwivestotheprovisionforhouseholdneedsbetween

CHAPTER6

122

the period of cotton prime and the time of the decline were investigated and arepresented by gender. The gender subjectivity in appreciating the spouses’contributions could thusbedocumented. For example, in the appreciationofwives’contributiontohouseholdprovision,themensaidthatduringthecottonprimetheirwives’contributionwasnil,whileaccordingtothewomentheydidcontribute.

6.4 Resultsanddiscussion

Agriculture remains themain economic activity in the study area. Raw agriculturalproductsandsomelocally‐processedfoodsconstitutethebulkoftheitemssoldinthemany rural markets organised in villages. Few local farmers engage in (informal)business activities.Manufactured products aremostly sold by traders coming fromnearby urban centres or by foreign traders who settled in the villages. For theindigenouspopulation,agricultureremainsthecornerstoneoftheirlivelihoods,oftenprovidingtheirtotalincome,withcottonasthemajorcashcrop.

6.4.1 Demographicandhouseholdcharacteristics

Regardingeducation,theliteracyrateamongthemaleheadsofhouseholdsishigherthanthatofthefemaleheadsofhouseholds.Whiletherewerenowomenfarmerswithuniversityleveleducationabout3percentofmeninthesamplehadhighereducation.In both groups however, secondary school level education represented the largestcategory. The difference between men’s and women’s education was significant(p<.05).

Table 6.1 offers an overview of the characteristics of female and male‐headedhouseholdsinthestudyarea.Itshowsthatmale‐headedhouseholdsaresignificantly(p<.001) larger than female‐headed ones. Overall, themean household size is high,which corroborates the general opinion that northern Benin has the largesthouseholds in the country (INSAE,2004).MinotandDaniels (2005) also found thatcotton‐growinghouseholds inBeninhadmoredependants thanhouseholds thatdonotgrowcotton.Thelargerhouseholdsizecouldbeduetotheagriculturalorientationoftheregion,sinceagriculturehasahighlabourdemand.ThenumberofworkersinthehouseholdswascalculatedinmaleadultequivalentsusingtheconversiontablebyNorman(1973).Accordingtothistable,women’scontributiontoagricultural labourforce represents three quarters of that ofmen’s. Based on observation in the field,however, itwasdecided toassignequalweights tomenandwomen.Cotton isverylabour‐intensive and requires the involvement of all able household members,particularly that of women (cf. Chapter 5). In addition to their agricultural work,women also prepare the food for those working on the farm. The success of farmactivities is, therefore, related to theperformance of domestic activities.Moreover,femaleheadsofhouseholdsaresolelyresponsiblefortheir(cotton)farms,onwhichthey have to perform all the work. Male‐headed households have significantly(p<.001) more labour available than female‐headed households. The prevailing

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

123

traditionofpolygynyintheregion,wheremenoftenhavemorethanonewife,evenamong Christians (Falen, 2008; Mandel, 2004), is part of the explanation. Nosignificantdifferencewasfoundinthehouseholddependencyratiobetweenmaleandfemale‐headedhouseholds.

Table6.1 Characteristicsofsurveyedhouseholds

DescriptionofvariablesFemale‐headed

(n=24)Male‐headed(123≤n≤124)

2andt

Educationofthehead(level)

Noeducation 20(83.3%) 63(50.8%)

9.05*Primaryschool 1(4.2%) 26(21.0%)Secondaryschool 3(12.5%) 31(25.0%)University 0(0.0%) 4(3.2%)

Householdsize(person) 11.33(5.62) 17.33(11.96) ‐2.40***Labouravailable(maleadultequivalents) 6.75(3.33) 10.96(9.32) ‐2.18***Dependencyratio 1.70(0.39) 1.70(0.54) .096Experienceincottonproductionofthehead(years) 14.17(6.98) 21.32(10.45) ‐3.22**

Ratioofthehead’sexperienceincottonproductionbyhisexperienceinagriculture 0.86(0.16) 1.17(.92) 3.48**

Note:*=p<.05,**=p<.01,***=p<.001.Figuresinparenthesesarepercentagesandstandarddeviationsofmeans.

Source:Surveydata,2009‐2011.

Men have significantly (p < .001) more experience (factor of 1.17) in cottonproductionaspartof theiragriculturalexperience thanwomen.This relativelyhighratioofmen’stowomen’syearsofexperience incottonproductioncouldbegreaterthan the ratio relating to experience in agriculture in general. Experience refers toyearsofautonomyindecision‐makingabouttheproductionsystemandabouttheuseof theyields.Sincecotton isacashcropandthemainsourceof income,youngmenstartgrowingcottonearlyon.Alreadyduringadolescence,boyshavetheirowncottonplots, independently from themainplots of the household,which they fully controlfromthesowingandplantingtoharvestingandtheuseoftheyields.Theyoungmen’sinterestincultivatingfoodcropsusuallystartslater,whentheyareheadsoftheirownhousehold. It is household headship that confers the right to control all householdproduction.Womenhaveonaveragefeweryearsofexperienceincottonproductionaspartofexperienceinagriculture(factor0.86)andtheircontrolovercottonplotsison averagemore recent than that ofmen. Indeed, young girls rarely ownaplot forcroppingbeforetheygetmarried.

Someparticularitiesneedtobementionedhereinordertoshedlightonthedynamicsofsomehouseholdsthatreflecttheoftencomplexstructureofthehousehold inthismodern‐rural context. Multiple arrangements could be identified based on criteriasuchas theplace(s)where thehouseholdmembers live, themode(s)ofproduction,andprocessesofdecision‐makingandhouseholdmanagement.Thiscomplexitygivesthehouseholdasaproductiveunitablurringshapethatoverlapswiththefamilyand

CHAPTER6

124

its holdings and with the household as a unit of consumption (cf. Niehof, 2011).Basically, three types of households can be distinguished in the study area: (i) thenuclearfamilyhouseholdwithonemaleadulthead,hiswife(ves)anddependants,(ii)ahouseholdwherebrothersshareahomesteadthat isheadedbytheeldestbrotherwith limitedautonomy for theyoungerbrothers, and (iii)householdswhere twoormore brothers live at different locations and share some resources, produce jointlybutarerelativelyautonomousinmanagingthenon‐foodoutcomes.Thefollowingtwocasesillustratethevarietyofhouseholdconfigurations.Case6.1:Dariisco‐headingahouseholdof22dependantswithhisyoungerbrother.Therearetwohomesteads,oneinthevillagewherethehouseholdheadisliving,andtheotheronthefarm,whichismanagedbytheyoungerbrother.Thelatterhomesteadwasinstalledtobeclosertothefarmandendedupbecomingapermanenthomestead,withthebrotherlivingtherehavinglittleautonomyofdecisionmaking.Thepracticeofsettinguphomesteadsonthefarmemergedduringthecottonboom.Thegrowingscarcity of land for cotton, the intensity of cotton activities, and the necessity toprotecttheinflammablecropfrombushfires, ledsomehouseholdstopartlytransferthe household to the farm. Dari’s household is an example of a case where theintensityof cotton farmingand thedistance to thehomemade the stay at the farmmuch more frequent and, in the end, permanent. The two components of thehouseholdthencoordinatetheirfunctioning.Whilethehomesteadatthefarmisthecentreofproduction,theoneinthevillageiswheredecisionscomefromoraremadeandresourcesaremanaged.Bothcashcropsand foodcropsarecultivated together.Cotton is producedon the collective farm cared for largely by the younger brother.The resources generated by cotton production are particularly used for biginvestments.Theyieldsofthejointlyproducedcropsarejointlymanaged,butwithapre‐eminent role in decision making of the elder brother. Cotton still remains thehousehold’s main source of income, though its falling profitability and the erraticpaymentofcottonmoneyhaveledthehouseholdtodiversifywithmaizeandrice.Theproduction of these crops demonstrates, however, some exceptions in resourcespooling strategies. Maize for example, which is primarily a food crop with aconsiderablecash function, iscultivated together.Ricethat isconsideredacashandfood crop too but has primarily a cash function, is cultivated separately for privatecashneeds.Inadditiontothecommonfarm,however,theelderbrotherownsasmallcottonplotnearthevillage,forhis“personalpurposes”.Indeed,mostofthehouseholdmembersliveinthevillagewiththehouseholdhead,includingtheyoungerbrother’sschool‐age children.Dari looks after their basic needs, education andhealth, askingassistance from their father only when really necessary. To do this, Dari uses themoneyfromhisprivateplot.

Acommongranaryforfoodcropsislocatedonthefarm,fromwhicheachhomesteadcan freelydraw the foodneeded.Most resources andoutcomes are pooledbecause“dividingtheyieldsofourproductionwillweakenthefamily”,accordingtotheheadof

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

125

thehousehold.Figure6.1portraysthearrangementsofthishouseholdandtheflowofresourcesbetweenthetwohomesteads.

Figure6.1 HouseholdandhomesteadarrangementofDari’shousehold

Case6.2:Thishouseholdof36membersiscomposedoffourbrotherswiththeirsixwivesand26dependants.Thebrotherslivethesamehomestead.Thewivessharethedomestictasks,cookinginturnfortheentirehousehold.ThehouseholdisheadedbyGounou, the eldest brother,who decides to certain extent on behalf of his youngerbrothers. They grow all the necessary food crops together andmaintain a commongranaryto feed themembersof thehousehold.Theircotton farmsareonablockofland,wheretheyallhaveaboundedplot(Figure6.2).

HouseholdGranaries(mainstore)

CoopFold

Housing

Cotton Rice

Smallprivatefarms

Homesteadinthevillage(Headof

household+wives+dependants)

Kitchengarden

Rice

Farmhomestead(Youngerbrother

+wives+fewdependants)

Smallgranary

(provisions)

Housing

Coop

Housing

CHAPTER6

126

Figure6.2 HouseholdandhomesteadarrangementofGounou’shousehold

Thearrangementallowsthehouseholdtojoineffortsingrowingthelabour‐intensivecrop of cotton. The sizes of the respective cotton plots reflect birth rank. The sizedecreasesfromtheeldesttotheyoungestbrother,regardlessofhowhardworkingtheowneris.Harvestingisdonecollectivelybuttheproduceisdisplayedseparatelyandsold accordingly. Each of the wives is attached to her husband with regard to theentitlement to cotton yields. A younger brother cannot have a bigger farm than hiselder brother. The youngest and unmarried brother has no plot within the block.Instead,he isundertheeldestbrother.“It is likethatbecauseastheeldest, Ihavetodealwith theirmatters”, said the eldest. These arrangements allow for aminimumlevelofautonomyforeachbrother.“Eachofushashisprivatefinancialissuestodealwith,togetherwithhiswife(ves).Andmoneyissuesaredifficulttosettlenowadays”,saidthe eldest brotherwho acts as the head of the household. Big investments such asweddingceremonies,orissuesthatengagedirectlythehonourofthefamily,aredealtwith in the group inwhich the eldest has a prominent voice. The decline of cotton

OrchardRice

Kitchengardens KitchengardensStockyard

Homestead(Headofhousehold&brothers+wives+dependents)

Fold

Coop

Granaries&Stores

Housing

Housing

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

127

profitabilityaffectedonly slightly the landdevoted to the crop,whichhas remainedalmost constant from 2005 to 2010. Until now, cotton is the most important cashprovider,which iswhy it is still themain crop in Gounou’s household, although itssharehasdrasticallydeclinedfromabout75percent to40percent.However,moreeffort is now put in cultivating groundnut, rice and maize, to compensate for thereductionofcottonincome.

The two examples above illustrate the central position of kinship ties, thepreservation ofwhich is an important concern for all types of households. Ties arestrengthenedthroughregularinteraction.Onewayofpreservingthesetiesistoworktogetherinfoodproductionandsharethegranary.Thecriterionofco‐residenceinthedefinitionof household thatwasused (see above)doesnotnecessarily imply livingunder one roof.However, the proximity of householdmembers has to be such thatthey share – at least amajor part of – household resources and daily activities, ofwhicheatingfromthesamegranaryisfundamental.Appliedtothehouseholdsinthestudy area, this entails that different domestic units in one homestead or differenthomesteadsinonedomesticunitcan,infact,constituteonehousehold,providedtheyshareatleastatypeofresourcesandjointlymanagethemtoprovidefortheirprimaryneeds.Ahouseholdcan thereforebedefinedasagroupofpeoplewhoeat fromthesame granary, pool the necessary basic resources, and abide to some extent by apowerholder’sdecisionsaboutthemanagementofresources.

6.4.2 Landallocationtocotton

Ofthesampleof148cottonfarmers,aboutthreequarterswerestillproducingcottonagainstonequarterwhodeclaredtohaveabandonedcottonproduction.However,thedynamicsofcottonproductionmakeitdifficultforfarmerstoabandonitcompletely.Indeed,about58percentof thosewhoabandonedcottonproduction(about14percentofthetotalsample)werewillingtoresumeitwhentheconditionsofproductionwouldimprove.Becauseduringthepastthreeyearsincentiveswereimplementedtorevivecotton,e.g.therevocationoftheprivatemonopolyininputssupplyandthesus‐pensionof themaincottonbodies,many farmersarenowresuming theproduction.However,many others remain suspicious of the strategy. These dynamics testify totheeconomicimportanceofcotton,whichplayedacrucialroleinthemonetisationofthe rural economy from the beginning of the 1990s onwards (World Bank, 2003).Despitesthisimportance,adeclineoftheaveragelevelofcottonproductionbetween2006and2010,alongwithanetdecreaseof theaveragecottonplotsizewithin thesame time frame could be observed (Table 6.2). This overall decrease points to ageneral trend of cotton abandonment during 2006‐2010, as a consequence of thetroubleswitnessedbythesectorfromthebeginningofthedecade.

CHAPTER6

128

Table6.2 Evolutionofcottonproductionbetween2006and2010.

Productionparameters N M SD Min. Max.

Household'scottonplotin2006 132 3.67 4.50 .00 30.00

Household'scottonplotin2010 146 2.27 2.37 .00 12.00

Household'scottonproductionin2006 127 5.15 8.03 .00 60.00

Household'scottonproductionin2010 134 2.38 2.53 .00 10.00

Averageareaofcottonplot(2006‐2010) 147 3.30 3.70 .00 30.00

Averagecottonproduction(2006‐2010) 132 4.48 5.71 .00 49.20

Source:Surveydata,2009–2011.

6.4.3 Structureofcottonfarmers’income

Amongthewiderangeofsourcesofincomeoffarmersinthestudyareaarekitchengardens, breeding of animals, processing of agricultural products, food crops trade,(petty) trade such as selling of condiments, handicraft, and salaries and pensions.There was no statistically significant difference between male‐ and female‐headedhouseholdswithregardtotheaveragenumberofsourcesofincome(Table6.3).Thenumberof incomesources ineachcategoryofhouseholds indicates thatanaveragehouseholdhasatleasttwosourcesofincome.

Table6.3 Structureofincomesourcesofmale‐andfemale‐headedhouseholds

ParametersMen

(111≤N≤124)Women

(19≤N≤24)t

Meannumberofsourcesofincome 2.02(.71) 2.13(.68) 0.69

Averagepercentageofagriculturalincomeintotalincome

73.83(31.01) 65.17(24.76) ‐1.26

Averageshareofcottoninagriculturalincome

47.78(26.18) 36.43(19.77) ‐1.80

Note:Figuresinparenthesesarestandarddeviationsofmeans.

Source:Surveydata,2009–2011.

However,whilemencanbe limited in thenumberof livelihoodsactivities,women’slivelihoodportfolioisricher.Inadditiontoagricultural‐relatedactivities,breedingofanimals,andhandicraft,inwhichbothmenandwomenareinvolved,foodprocessingandfoodproductstradingareexclusivelywomen’sactivities.Morewomenthanmenareengaged in small tradeof items likecondimentsorgroceries.Thisgiveswomenmoreopportunitiesofadaptationthanmen.Ofallactivities,agriculturehasremainedthemainincome‐generatingactivity,providingabout74percentofincometomale‐headedhouseholds and65per cent to female‐headedhouseholds (difference is notsignificant). Cotton now represents 36 per cent of agricultural income of female‐headed households and 48 per cent formale‐headed households,making the lattermore dependent on income from cotton than the former. The present figures

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

129

representasteepdeclinefromthesharecottonincomehadbefore2006.However,theaverageshideconsiderabledifferencesbetweenhouseholds.Somestillearnupto100percentoftheirincomefromcotton,whileothersreportedthatcottonprovidedthemnolongerwithincomealthoughtheycontinuegrowingit.Thereasonsforthedeclinearethelesserprofitabilityofthecropandtheproblemsingettingpaidfortheproduce(seeChapter3).

6.4.4 Importanceofcottonforstrengtheningthehousehold’sassetbase

Cottonproductionbroughtmuchwealthandwell‐beingtotheNorthofBeninduringthe1990sandtheearly2000sandwasvaluedas“whitegold”(seeChapter2).Ithasbeencritical forbothmenandwomen,notonly for theregularityof its income,butalsofortherelativeimportanceoftheamountsofmoneyreceivedatonce.Thismadecotton income the means for big investments, such as construction of a house,acquisitionofcarsandmotorbikes,lastingequipmentlikegrindingmills,forweddingceremonies, etcetera. Most of the surveyed households reported improved livingconditions,which they acknowledged to originate from cottonproduction. Concretehouses,roofsofcorrugatedsheets,agriculturalequipment,motorizedtransport,andallsortsofequipmentandappliances,werereportedlyfinancedfromcotton income(Table6.4).Cottonhadbecomethebackboneofthecountry’seconomyandthesourceofruralhouseholds’assets.

Theeconomicgrowthofthecountryduringthe1990sislargelyimputedtothecottonboom.Bycontributingtoassetsandresourcesendowment,cottonhashadenduringeffects for individuals, households, and communities in the North, and, at nationallevel, for the State. An increase of houseswith corrugated roofs, a symbol of “ruralease”, and the construction of community infrastructures are, among others, thevisibleimprovementoffarmers’livingconditionsdrivenbycottonproduction(WorldBank, 2003). Through this substantial contribution to the livelihoods of ruralhouseholds, cotton production shaped people’s life for a long time. Following thesoaringpricesofinputsandloweryieldsfromcotton,hundredsoffarmersabandonedor drastically reduced the area planted with the crop. The World Bank (2003)identifiedlowreturnstolabourasoneoftheriskfactorstoindividualandhouseholdpoverty.About14percentoffarmersreportedhaltingtheconstructionorenhancingthe quality of their house as a consequence of the decline of cotton production.Another 16 per cent of farm households had to reduce their provision for animaltractionandagriculturalequipmentforthesamereason.Theincidenceofprestigiousfuneralsandmemorialceremonies,whichinflatedduringthecottonboomandwereorganisedafterfarmershadreceivedtheircottonincome,hasreportedlydeclinedaswell.Formanyfarmers,thereductionofcottonincomeby50percentmeanthalvingtheirincome.

CHAPTER6

130

Table6.4 Improvement/reductioninhouseholds’assetsendowmentswithregardtocottonproduction(percentageofhouseholds)

Assets AcquisitionduringCottonprime(N=115)

Reductioninquantity/qualityafterthedecline

(N=115)Categories Types

HouseConcrete(orhalf)withcorrugatedsheets 39.13 23.91

Claywithcorrugatedsheets 33.91 6.38

AgriculturalequipmentAnimaltraction 80.87 10.58Tractor 1.74 0.00

Meansoftransportation

Motorbikes 61.74 07.35Cars 2.61 0.00

Homeand"luxurious"equipment

Table,chairs,sofa,etc. 86.96 09.18TV,CDplayer,etc. 75.65 11.50Fridge,generator,grainmill 12.17 7.70

Source:Surveydata,2009–2011.

Orou isa65year‐oldcotton farmer,headingahouseholdof26members inSekere.Thehouseholdearnshundredpercentofitsincomesfromagricultureofwhichcottonrepresents up to 90 per cent of the yearly income. Because of this paramountimportanceofcottontohislivelihood,Oroumaintainsaconstantproductionandheisone of the largest producers in his farmer group. But he resents the problemsoverwhelmingthesectorthatforcedmanyfarmerstoreducetheirlandallocationtocotton.Orou’swoesresultfromthelackofpaymentforhisproducethatsqueezedhislivelihood and reduced the household assets. “All my assets come from cottonproduction, and therefore the deterioration of my living conditions comes from thecottonproblems”,hecontends.Hepointsatanunfinishedbuildingthatwasintendedforhisthreesonsofwhomtwoaremarriedbutstillnotindependent.Clearly,thehighdependencyoncottonhas shown its limitsandnecessitated farmers individuallyaswellascollectivelytodevisenewwaysofgeneratingincome.

6.4.5 Adaptingtoandcopingwiththecottoncrisis

Thecausesofdiversificationaswellas itsoutcomesareaffectedby location,assets,incomes, opportunities and social relations (Ellis, 1998). Three types of livelihoodassets are relevant and critical to rural areas for people’s adapting and copingstrategies.Thesearenatural,socialandhumancapital,inadditiontothesocialnormsandrulesregulatingtheaccesstotheseassets(Scoones,1998).Theseassetsarethemain instrumentsfordiversificationof incomeandresources.Multiplemotivessuchasthereductionofrisk,theexperienceofshocks,andtheuseofopportunitiesdrivethechoiceofdiversification(Barrettetal.,2001).InthecottonzoneofBenin,thefallofpricesand theerraticpayment for cottonyields resulted ina shortageof incomethatcreatedaneconomicshock.Consequently,bothcottonproducersandthosewho(pretended to)haveabandoned it, have toadapt to and copewith thenewcontext,diversifyingtovariousdegreestocompensateforthereductionofincome(Table6.5).

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

131

Table6.5 Patternsofadaptationandcopingstrategiesincottonzone

Adaptation(Choice) Coping(Necessity)

‐Commercialfoodcropsproduction ‐Reductionofcottonarea

‐Increasingeconomicimportanceofmarginalcashcrops ‐Diversionoffertilisertofoodcrops

‐Emergenceoforganisationsaroundfoodcropsproduction

‐ Morefreedomforwives’entre‐preneurship

‐ Seasonalmigrationofteenagers

Source:Surveydata,2009–2011

The main strategies of adaptation and coping are shifting agricultural productionpatterns,migration, and grantingde factomore freedom towomen, tomitigate theeconomic shocks or to reduce the risk of failure. The changes in agriculturalproduction patterns are exemplified by: (i) the shift of the purpose of cottonproduction,(ii)thegrowingeconomicroleofpreviouslymarginalcashcrops,and(iii)the changing status of food crops that have gained a cash function, andwhichhavebecomeparticularlyprofitableafterthefoodcrisisof2008.However,whetherthesestrategiesaresustainableremainsaquestion.

Adaptationandcoping:on‐farmdiversification.Thefirsttwoadaptationstrategiesare all a diversification by choice, and aim to manage the risk. They consist ofconferring a primary cash function on food crops, such as maize, sorghum, yam,cassava, and cowpea, which in the past were produced exclusively for homeconsumption, andofwhichonly the surpluseswere sold.Rice,whichwasproducedbothasfoodandascashcrop,soybean,andgroundnuthaveseentheirmarginalcashfunctionincreasing(Table6.6).

About69per centof the respondents citeda foodcropas their actual first income‐generating crop, of whichmaize representsmore than half, followed by groundnutandsoybean.Riceandsorghumfollowsuitas income‐generatingcrops.This impliesthat about less thanone third (31per cent)of farmers still has cottonas their firstcashcrop.Maizealsodominatesassecond incomegeneratingcrop, followedbyriceandgroundnut.

CHAPTER6

132

Table6.6 On‐farmincomediversificationamongcottonfarmers

Crops Frequency(%offarmers)1strank 2nd rank 3rd rank 4thrank

Maize 60.9 28.7 4.3 1.7Groundnut 2.6 12.2 16.5 8.7Rice 2.6 13.9 20.0 7.8Sorghum 1.7 3.5 7.0 8.7Soybean 0.9 3.5 10.4 9.6Yam ‐ 6.1 4.3 ‐Cowpea ‐ 3.5 1.7 3.5Cashew‐nut ‐ ‐ 1.7 1.7TOTAL 68.7 71.3 67.0 41.7

Source:Surveydata,2009–2011.

Cultivatingthesecropsprovidesadditional incomeandcontributestocompensatingfor the loss of income from cotton. Thusmaize, sorghum, yam, and to some extentcowpea,previouslyproducedprimarilyforownconsumption,arenowlargelysoldforcash. However, their growth needs fertilisers that until recently were hardlyaccessible for non‐cotton producers. Those who can afford fertilisers on the blackmarket,orthroughtheirsocialnetworkswithintheformalcottonorganisation,havecompletely given up cotton production in favour ofmaize. The production is rarelysold locally but transported to trans‐border and urban markets, though with noguaranteeofhigher returns.Thosewho cannot access fertilisers through the cottonsystemcontinuetogrowcotton,albeitinsmallquantities.Inthisway,theycontinueto receive fertilisers that are subsequently partly or fully diverted to food crops.Therefore, theprimarypurposeof cultivating cottonhas strategically changed fromearning income to accessing inputs. But following the disqualification of farmerorganisations inmanagingcottonproduction,a closecontrolof inputs isexertedbytheextensionservices.Thishasledtostrategicbehaviourtoaccessthecriticalinputsin the context of the decreasing land fertility, showing a combination of adaptationandcopingstrategies.

The strategies to access inputs (coping) aim at increasing the effectiveness of thediversificationofincomesources(adaptation).Inthecaseofafulldiversionofcottonfertilisers to food crops, theyieldsof the food cropsbenefitting from thatdiversionare expected to cover the inputs debts.However, this does not alwayswork out asplanned.Wheneverthereisanover‐productionofmaizeandconcomitantlowprices,or when the agricultural season experiences torrential rains or another naturalcontingency occurs, farmers become heavily indebted. Up to half of the maizeproducedmaybeused to payback suchdebts. There is a high covariate risk (Ellis,2000)betweenthesealternativecashcrops,duetothefactthattheyallbelongtothefarm sector. When fertilisers are partially diverted, farmers drastically reduce thecottonsurfacetotheextenttheyexpecttoproduceenoughtocovertheinputsdebts.Butthisstrategyisnotalwayssuccessful.Ifcottondoesnotbenefitenoughattention

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

133

toyieldwell,farmershavetousethetotalproductionofthefoodcroptopayfortheinputs,therebyreducingthereturnsofthealternativecrop.

Youth male migration as non‐farm diversification. Out‐migration has beenidentified as an important strategy for improving livelihood systems in Africa. Instimulatingout‐migration,factorssuchasthereductionofemploymentprospects,theincreaseinpoverty,andthepressureonnaturalresourcesareintertwinedwithsocialdynamics (de Haan, 1999; de Haan et al., 2002; Francis, 2002;Mosse et al., 2002).Rural‐urbanmigrationhas longbeen identifiedas thedominantmigration flowandwasusedtoexplainthegrowthofAfricancities.Butaflowintheoppositedirectioncan be observed as well (Beauchemin, 2011). For example, Bigsten and Kayizzi‐Mugerwa (1992) found that urban households may let members migrate to ruralareas and to remit food to themembers left behind.However,migration is rarely afirst choice or preferred option. Farmers resort to migration when the availableopportunities of diversification on‐farm do not result in any prospect of improvingtheirwell‐being.

Rural‐urban(orsemi‐urban)migrationwasobservedinthevillageofSekere(districtof Sinende), where seasonal migration of teenagers has become an emergentphenomenonandde factopartof the livelihoodsystem.On‐farmdiversificationhasfailed to live up to farmers’ expectations of earning income to improve theirlivelihoods. Foods crops, of which the production has shifted in purpose, cannotgenerateenoughincometomeetallneeds,particularlythosearisingfromthemodernaspirationsoftheyouthtohavemotorbikesandmobilephones.Toberelievedfromthis pressure, some households let their boys temporarymigrate to Nigeria. Aboutnine per cent of surveyed households had up to three members who were onmigration simultaneously. The migration to Nigeria is nurtured by clandestinenetworks of traffickers10, who may have no previous direct connection with theseteenagerswhomtheyrecruitbychance.

Two types of decision‐making were observed that shaped the migration of youngpeople in Sekere,where about 62 per cent of the heads of households interviewedreported to have had (a)migrant child(ren). The first andmost commonpattern ismigrationastheoutcomeofthemigrant’spersonaldecision,ashappensinabout78percentofthecases.Afarmerdescribedhowtheyleaveasfollows:

10ThankstoastoponmywaybackfromSinendein2010,Icameacrossasceneofdisputeinthe village of Guinagourou at the Nigerian border. A car driver carrying a group of eightteenagers,whoturnedouttooriginatefromSinende,wasarrestedbyvigilantes.Noneoftheteenagers gave any name of who contacted or helped them. Instead, the older one,approximately15yearsold,saidtheywererelativesandfriendsandarguedtheywereonavisit to theirmother.Thedriversucceeded toescapeand lefthis car,drivenalongwith theteenagerstothenearestpoliceoffice.

CHAPTER6

134

They usually leave the village in small groups by night and theirabsence isnoticedthefollowingmorning.Sometimes,itisaftermanydays that the parents become convinced that their child(ren) havemigrated.

The second pattern is joint decision‐making and concerns those who inform theirparents of their decision to migrate. Because informing is different from gettingpermission, two sub‐types can be distinguished. The first one includes the caseswhere parents accept the proposal against their will, because they cannot affordgivingtheirchildrenwhattheywant.Therefore,theyacceptwithresignationandcanbequalifiedaspassiveactorsintheprocess.Thesecondsub‐typediffersfromthefirstby the fact that parents discuss the issue with their children and advise them.Althoughnotprovidingthemeansfortransportation,inthiscasetheparentshaveanactiverole.

Inallcases,themigrationwasatrans‐bordermovementthatledtheyoungmigrantstoNigeriaforawhile.Theaveragedurationofthemigrationissixmonths.Duringthisperiod,parentsareusually ignorantabout thewhereaboutsof theirchildrenanddonotknowwhenandwhetherthechildrenintendtoreturn.Theyarecommonlyknowntowork primarily in agriculture and, in the secondplace, in construction. Probablytheyareinvolvedinillegalemploymentinhigh‐riskdomainssuchasminingaswell.At the end of their stay, young migrants reappear at the parental home in smallgroups, just as when they left, each of them riding his motorbike. They travelovernight on rural roads like smugglers, to avoid being arrested by customofficerswhowouldfleecethem,requestingsubstantialamountsofmoney.

Themotorbikes they returned with are exhibited at the edge of the village squarewhereparties are organised in the evening andwhere youngmigrants stage rallieswith their motorbikes. This display of their acquisitions contributes to nurturemigration. Thus, the principalmotive behind this temporarymigration of youths isacquiringamotorbike,whichintheirviewhelpsthemtomaintainaprospectforlife.The boom in the cotton sector had engendered aspirations that could no longer berealized after its decline. To keep their expectations ofmodernitywhile living in aruralarea,theyoungpeoplewanttocultivateamodernlifestyle,whichforcesthemtomigratetemporarily(cf.Ferguson,1994;1999).

Themigration of youth in Sekere differs fromboth circular and seasonalmigration(Hampshire,2002;Rogalyetal.,2002).Unlike in thecaseofseasonalmigration, theyouthmigrationfromSekerecantakeplaceatanytimeoftheyear.Themigrantssendnoremittancesandhavenocontactwith theirparents till theyarebackhome.Thismigration isalsonotcyclical. Instead, thereproductionof theprocess is fedbynewmigrants.Successfulandexperiencedmigrantsrecounttheiradventurestotheirpeerswho thendecide to try it themselves,usingornot thesame itinerary. Incontrast tocommonly known processes of migration in which the flow is maintained by anetworkofsocialrelations(cf.deHaan,1999), inthiscasethetemporarymigration

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

135

revolvesoutofestablishednetworks.Althoughitdoesnotprovidethehouseholdwithdirectcash,itconstitutesadiversifiedsourceofincomeifassumedthatitistheheadof household’s duty to supply the youngmemberswith assets such asmotorbikes.Whenthisrequirementisnotmet,out‐migrationbecomesanoptionforboystoaffordtheirneeds.Althoughtheirabsencefromhomeforseveralmonthsimpliesadeficitoflabour,theirreturnhomeinpossessionofwhattheywanted,releasesnotonlythembutalsotheparentsfromapressureonthehouseholdmeans.Therefore,thisstrategysupportsthehousehold.Inaddition,thedivergenceofinterestsbetweenchildrenandparents, and theshiftofpowerrelationsbetween them, implies that this temporarymigrationinvolvesimplicitbargainingthroughtheprocessesofdecisionmakingfromwhichittakesplace(cf.Agarwal,1997;Sen,1990).

Whiletheshortageofincomeandsocialpressurearepushactors,thedemandfortheservices that can be delivered with the acquired asset constitutes a pull factor forcomingback.Theuseofthemotorbikesismanifold.Inadditiontothepsychologicalfeeling of achievement theybring to the owners, themotorbikes are used for “taxi‐moto”,animportantcontemporarymeansoftransportationinthecountry.Taxi‐motonowadays isan income‐generatingactivity thatcangeneratesubstantialamountsofmoneytoitpractitioners.Itensureseasytrafficinruralareaswhereothermeansoftransportation are not always available. The migrants also help the householdmemberstravellingbymotorbikewhennecessary.

Institutionaldiversificationaslivelihoodadaptationstrategy.Cottonorganisationshave lost theirpre‐eminentpositionofyesteryear,being incompetitionwith farmerorganisations that focus on other crops, which emerged from the quest for newopportunities of added value. Their emergence also benefited from the support ofNGOs that are not interested in cotton production, apart from those lobbying fororganic cotton. Commercial as it is, cotton production is highly regulated and fullycontrolled by economic lobbies that do not leave any space for NGOs. Thus,organisations (pretending to be) specialised in the production of maize, rice,groundnut, cashew‐nut, cassava, soybean, or beekeeping, emerged in the villages tocompetewiththedecliningcottonorganisationsthatwereoverwhelmedbydebtsandmismanagement.Pre‐existingassociationsthatwereontheedgeofcollapseforlackoftimeorlackofinterest,suchasthoseoftraditionaldancersandcraftsmen,benefittedfrom the development. The membership of cotton organisations made farmersfamiliarwithorganisations.Beingamemberofanorganisationbecamepartof theirlives.Farmerswhowereinvolvedinmanagingcottonorganisationsbegantousetheirexperiencetostructureandleadtheneworganisations.

It can be concluded that although cotton production has declined and cottonorganisationsarefading,cottonhascontributedtobuildinghumancapital,aresourcethat farmers use to organise themselves in other areas of production. The neworganisations, therefore, contribute to improving farmers’ livelihoods, given thatparticipatinginawidearrayofassociationsincreasestheabilitytosmoothlivelihoodfluctuations (Grootaert et al., 2002). In addition, leading an organisation gives

CHAPTER6

136

opportunities of mediating between donors and NGOs on the one hand and fellowfarmers on the other (Olivier de Sardan, 2005). It also entitles leaders to managecommon resources of which the use can be diverted from public needs to privatepurposes,ashappenedincottonproduction.Althoughfarmersarecomplainingaboutthelowincome‐generatingcapacityofthesediversifiedsourcesofincome,nothavingtodealwithcottonanymoresatisfiessomeofthem.Asafarmersaid:

Forsure,thedeclineofcottonproductionisnothurtingme,aslongasmaizeisprofitable.Instead,wewillbenefitfromthissituationbecausetheextensionagentswillstopdisturbingus.

6.4.6 Effectsonintra‐householdresourceallocationandgenderroles

Thedeclineofcottonproductionhasmodifiedtheshareofmen’sandwomen’scon‐tribution to provision for household needs and expenditures. Men’s higherdependencyoncottonincomeledthemtobeingmoredeprivedofalternativeincomesourcesthanwomen.Asaconsequence,theirpurchasingpowercomparedtothatoftheir wives has been drastically reduced. This decreased purchasing power ofhusbands is automatically reflected in their contribution to the household needs,whichisnotwithoutconsequenceforthespousalrelationship.Asawifesaid:

During theprimeof cotton,wegavemoney toour school childrenwithoutany concern of being paid back by our husbands.Butnow, because of thedeclineof cottonproduction,wearealwaysquarrellingwith thehusbandsaboutmoney.

Table6.7 Evolutionofwives’contributiontohouseholdneeds(%)

Women’sperspective(N=24) High Average Lowtonegligible

SchoolingfeesDuringthecottonprime 21.7 43.5 34.8

Afterthedecline2005 26.1 56.5 17.4

ClothingDuringthecottonprime 59.1 18.2 22.7

Afterthedecline2005 68.2 18.2 13.6

HealthcareDuringthecottonprime 9.1 9.1 81.8

Afterthedecline2005 31.8 13.6 54.5

DailyfeedingDuringthecottonprime 22.7 22.7 54.5

Afterthedecline2005 40.9 22.7 36.4

Source:Surveydata,2009–2011.

Women’s purchasing power, however, is sustained because of the wide range ofincome‐generatingactivitiestheyperformduringalongperiodoftheirlife.WomeninBeninareeconomicallyveryactiveandstarttoearntheirownincomeasearlyasat

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

137

theageof15andcontinuingtoworkintotheir70s(Mandel,2004;WorldBank,2002).Therefore, their contribution toproviding for thehousehold’sbasicneeds increasesrelativetomen,whohavefeweropportunitiestoearnincome.Table6.7presentsthechanges in the contributionofwives to theprovision forhouseholdneeds from thewives’perspective.

Thedemiseof cottonhas resulted inwomenhavingmore freedomtoperform theirown income‐generating activities. But although wives contribute more to theprovisionforhouseholdneeds,theyarenotasempoweredascouldbeexpected.Theprocess of decision‐makingwithin the household is still unbalanced. Husbands stillhaveapre‐eminentvoicebecauseofculturalpracticesandsocialbeliefsthatnotonlyimposeonthemtoprovidetheirwivesandchildrenbutalsoendowthemwithrightsover thewives. Although to the young generation thismay seemold‐fashioned, theculturallyunderpinnedsubordinationofwives tohusbands turnsout tobe thriving.Thepaternalisticculturethatsubordinateswomentomenandtheunbalancedpowerrelations in decision making between men and women within the household arereinforced by the unequal allocation of resources to enable men to perform their‘naturalduty’asproviders.Themarriedwomenwhowereinterviewedsaidthattheirautonomy to decide on matters without referring to the husband is not positivelyaffected by their increased contribution to the provision for household needs. Theyalways refer to their husband for their own resources allocation. Amale head of acotton‐farminghouseholdstronglybelievedthat:“It isnotbecauseawifecontributestoclothingorfeedingthechildrenathomethatsheandherhusbandwillhaveanequalstatus.”ThestatementshowshowstrongculturalvaluesongenderinnorthernBeninstill are. However, there is also bargaining within households, which is neither aformal negotiation of rights and entitlements nor an open conflict for control ofresources. It isapermanentrepositioningofhouseholdmembersformutualrespectandself‐esteem.Andrarelydomenintheareapubliclyacknowledgethesuperiorityofwivesovertheirhusbands(seeChapter5).Nevertheless,theimportantparttakenbywivesinhouseholds’economiclifeconfirmsthatwomenactasashockabsorberintimesof economichardship (Molyneux,2002), andentails a tacithandoverof somespaceofpowertowomenwithinhouseholds.

6.5 Conclusion

“The household is a complex phenomenon. In rural areas such as northern Beninwhere patrilineal kinship ties are strong, the household follows the contours of thejointfamilyinwhichbrothershavelimitedautonomytomanageanduseincomefortheir own family independently from the joint‐family household. This attests to theimportance of kinship as the fundamental element in establishing a household andhelpsgainingasenseofhowthepoolingofresourcesisfarmoreimportantthanthelocation of the homesteads. The household feature of co‐residence can also bestrategically reduced to sufficient proximitywhen, as in this case, the household isdispersedovermorethanonehomesteadtoincreasetheefficiencyofproductionand

CHAPTER6

138

management of resources. In the area, people refer to this phenomenon as ‘splitfamilies’.Therefore, thecorecriteria forbeingahouseholdmemberareeating fromthesamegranaryandabiding–toacertainextent–bytheprominentroleindecision‐makingoftheeldestbrotherwhoisconsideredthehouseholdhead.Householdsinthestudyareaarelarge,withoftenmanymarriedadults,whichcharacteristicisasignofwealth. This situation contradicts the perception that relying mostly on farm andagricultural labour is a characteristicofpoor ruralhouseholds that tend tobe largeandhaveahigherdependencyratiothannon‐poorhouseholds(IFAD,2010).

Thedeclineofcottonproductionandtheensuingshortageof income,resourcesandlivelihoodassetshavebroughtaboutshiftsindecisionmakingonthediversificationofsources of income. When an economic shock occurs, farmers first tend to devisestrategies by using the resources available on the farm. They only look beyond thefarmwhenon‐farmopportunitiesdonotturnouttobeanadequateresponsetotheshock.Bothon‐farmandnon‐farmdiversificationtookplacewithinhouseholdsinthestudy area. But the heads of households control on‐farm diversification because oftheir control over land allocation to household members. Off‐ and non‐farmdiversificationthataimsatcomplementingon‐farmincomecanthereforeescapetheircontrol.Timeisallottedtoextra‐domesticactivitiesbywivestoacertainextent,andyoungdependantscandecidewhenandwheretomigratetomeettheirneedsaswell.Thismadethesecategoriesofhouseholdmembers lessvulnerabletotheshockthanmen, thereby increasing their contribution to theprovisionof householdneeds andexpenditures.

Inspiteof thewiderangeofactivitiesperformedbyruralhouseholdstomaketheirliving,thelowaveragenumberofincomessourcesatteststotherelativelypoorliveli‐hood portfolio at household level, or, in other words, to the precariousness of thelivelihood systems. Food crops are of paramount importance in diversificationstrategies and they actually outnumber the traditional cash crops in providingincomes to ruralhouseholds.Mostnotablymaize, ofwhich consumptionhas grownsteadily,hasgainedacashfunctioninadditiontoitsfoodfunctionandisnowrankedhigher than cotton with regard to its ability to generate income. But this doublefunction as food and cash crop and the covariate risks existing betweenmaize andotheralternativecashcrops limit itsability to fullyreplacecotton.The incomefromcotton still is themain on‐farm cash revenue formany farmers. As a consequence,cottonisgrownasacopingstrategytomakethediversifiedcropsmoreprofitable,inspiteofitsdecreasingimportanceinprovidingincome.Therefore,whenfarmersrelyheavily on one crop for their livelihoods, they can always invent new ways ofcultivatingittomaketheir living,evenwhenit isnotfinanciallyprofitableanymore.Tomake the food crops profitable and to reduce the dependency on cotton, policymakers should givemore attention to the supply of inputs for promising food‐and‐cashcropsandtoimprovingtheirmarketingoptions.

Thepluralityofhouseholds’responsestothecottoncrisisinBeninexplainswhythereisno idealtypeofdiversification.Diversificationisdynamicandthismakespossible

RURALHOUSEHOLDS’RESPONSESTOTHECOTTONCRISIS

139

for one household to adopt more than one diversification strategy. However, twostrategies can basically be distinguished in the study area: first, on‐farmdiversificationofincomesources,and,second,allowingmorehouseholdmemberstodecideuponandengageinearning(additional)income.

141

Chapter7

GeneralDiscussionandConclusionsABSTRACT:Thepresentchapterdiscussestheoverallfindingsoftheresearchpresentedinthepreviouschapters.Itaddressesthecentralresearchquestionand thequestionsandhypotheses subsumedunder theheadings in the firstchapter.Thisisfollowedbyatheoreticalreflectionbasedontheoutcomesoftheresearch.Thechapterendswitha lookattheimplicationsforpolicyandfutureresearch.

CHAPTER7

142

7.1 Sheddinglightontheresearchquestionsandhypotheses

Networking, social capital, and gender roles in the cotton system in Benin are ofinterest to the State, farmers and private actors because these issues influence notonlytheeconomicprofitabilityofthesectorbutespeciallybecausetheyconstitutethesocialdynamicsthatdrivethecottonsystemandshaperurallivelihoodsinthecottonbelt. Themain question addressed by the research is how farmers’ agency affectedtheirorganisations,cottonproduction,andthecollectiveaction thatevolvedaroundtheir activity. With cotton as the central focus, the interfaces between individualdispositionsandactions(microlevel)andtheinstitutionalandpoliticalmechanisms(meso and macro level), and their effects on household livelihood systems wereinvestigated.Thedimensionsofthemainresearchquestionwereframedunderthreeheadings: theemergenceofbreakawaynetworks, thedeclineof social cohesionandthe squeeze of collective action, and livelihoods reconstruction after the demise ofcotton production. Subsumed under these headings research questions andhypotheseswereformulated(seeChapter1).Theactor‐orientedapproachandactor‐networktheorywerethemainpillarsof thetheoretical frameworkthatwasusedintheresearchandgenderwasacross‐cuttingtheme,eventhoughinrelationtosomeissuesmore pronounced (e.g. Chapters 5 and 6) than in others. In this part of thechapter the research findings will be discussed, aligned according to the threeheadingsthatwereformulated.Attheendofthissectionwecomebacktothemainresearchquestion.

7.1.1 Theemergenceofbreakawaynetworks

Breakawaynetworks started emerging froma local conflictwithin a district farmerunion, before extending to national level, and resulted in atomised networks. Theprocessofcottonnetworksatomisationlastedlessthanonedecade.Itconcernedbothgrassroots farmers and network leaders, and resulted into more than 12 officiallyregisterednetworks.Theprocessrevolvedaroundpushandpull factors,proceedingby constantmovements of breaking away fromnetworks, joining, or creating othernetworks, thereby yielding three categories of cotton farmers: stayers, joiners, andcreators.Pushfactorstoatomisationarecausesthatmotivatedfarmerstoleavetheirnetworkswhilepullfactorsguidedtheirchoicesofactionsandtheirnewnetworkofmembership.Themostinfluentialpushfactorswithregardtotheiroverallrankingarethe lackofpayment, themismanagementof resourceswithinnetworks, intra‐groupconflicts, and manipulation of farmers by outside actors. However, only mis‐management of resources andmanipulation of farmers by outside actors exhibitedmeanranksofattachedimportancethatweresignificantlydifferentbetweenstayers,joiners and creators. When controlled for socio‐economic and socio‐demographicbackground,onlymanipulationhadasignificanteffecton farmers’decisiontobreakawayfromnetworks.Asforthepullfactors,thepaymentontime,thefacilityofgettinginputs,andtrustinboardsmembers(network‐efficiencyfactors)wereoverallrankedhigher than kinship, the hope for board positions, and the expectation of profit

GENERALDISCUSSIONANDCONCLUSIONS

143

(personalaspirationfactors).However,onlytrustinboardmembers,andthehopeforboardpositionsexhibitedmeanrankssignificantlydifferentbetweenstayers,joiners,and creators. When controlled for socio‐economic and socio‐demographic back‐ground,trustinboardmembersandexpectationsofprofithadasignificanteffectonfarmers’ decision and made a difference between staying in, joining or creatingnetworks.

Inthisprocess,investigatedinChapter3,farmerswerebackedupbyvariouskindsofrelationships, in particular by theprofessionals in cottonproduction such as inputssuppliers and cotton ginners, whose support was ranked high by all categories offarmers.Yet,supportfrompublicofficialsandfrommembersofthesameethnicgroupmembersalsohadasignificanteffectonstayers,joiners,andcreatorswhencontrolledfor socio‐economic and socio‐demographic background. The significant effects ofpublic officials and ethnicity show that strong ties as well as weak ties constitutesecuresourcesofsupportonwhichfarmerscouldbasetheiractions.Asforthepublicofficials, they provide breakaway network proponents with support on the officialprocedures about creating and/or managing cotton networks. The role of socialnetworks and the derived social capital in cotton networking are attested by thepositiverelationshipbetweenleadershipandthenumberofsocialnetworksfarmershadbelonged to.Thoughbreakingawaygavecotton farmersmoreaccess to leader‐shippositions,thosewhohadmoremembershipinsocialnetworkswereprivileged.Hence, the hypothesis postulating that leaders of breakaway networks are morestronglyembeddedinsocialrelationsthanordinaryfarmersisverified.

However, leadership in cotton organisations was correlated with gender andeducation,afieldwherewomenlagbehind,whichaddstotheentrenchedmythsandstereotyped images about women’s inferiority vis‐à‐vis men. Hence, women aretrapped in the intersection(Crenshaw, 1991)of factors thatcreateagenderedsocialorder,ofwhichthemainfeatureremainstheexclusionofwomenfrompublicspheres.Thissocialordercreatesafalseconsciousnessamongwomenthatmakesthemaccepttheirposition,aprocessthatRidgeway and Correll (2004: 511)referredtoasthe“socialrelationalcontext.”Thisexplainswhywomenareexcludedfromcottonorganisationboards in spite of their key position in the cotton production system at householdlevel.

7.1.2Thedeclineofsocialcohesionandthesqueezeofcollectiveaction

As Chapter 4 shows, most of the infrastructures realised in the cotton‐producingcommunities during the 1990s and the early 2000s were achieved before theatomisationofcottonnetworks.Therelativelystrongcohesionprevailingwithintheorganisationsnurturedcollectiveactionthroughwhichcommunitiescompensatedforthe deficiencies of the central State, availing themselves of basic facilities thatbenefitted non‐cotton farmers as well. Hence, through collective action cotton

CHAPTER7

144

networks produced public goods for the benefit of their entire communities (cf.Kollock,1998).Furthermore, thesocialandeconomicprivileges that cameout frommanaging cotton resources enhanced the social status of leaders and made theseleadersarolemodelatvillagelevel.Thisimprovedtheoverallperceptionofthevalueof agriculture and aroused the interest of new actors to invest in cotton either byfarmingorbysupplyinginputs.

However, collection action faded when the networks disintegrated due to mis‐management and conflicts increased. Breaking away from networks first steeredcompetition and mismanagement practices, which in turn generated paymentproblemsandmoreconflicts.Indebtednessandtheaccumulationofarrearsthereforetransformed cottonproduction into a nightmare.Additionally, the despair of cottonfarmers resulted in thedeclineof trustwithin andbetweennetworkmembers.Thetemptation of free‐riding increased the tension between individual and collectiverationality,leadingtothediminutionofthepublicgoods.Whereoncenetworkleadersdecidedonrelevantactionstoundertakewithcottonresourcesforthebenefitofthecommunity, now concerted action between leaders of contendingnetworks becameimpossible. The demise of collective action ultimately resulted from this chain ofreactions. This thesis revealed the intertwined relations between trust, reciprocity,cooperation, and collective action.Ordinary farmers felt fooledby their leaders andresorted to free‐riding to compensate. Distrust can, therefore, be equated to thedisappearanceofthebasicfunctionsoftrust:theintegrativefunction,thereductionofcomplexity, and cooperation. Distrust among members of the networks created asituation inwhich no collective action could be initiated. Therefore, the hypothesisthatthedeclineofsocialcohesionresultedinthedemiseofcollectiveactionincottonproducingareasisverified.

WhatdeservestobesingledouttoofromChapter5isthesalientgenderdifferenceinthe settlement of network conflicts. Men and women proved to have a differentperspective on modern as compared to traditional power. As apparent from thereactionoftheprofiledwomentotheiroustingfromcottonboards,theperceptionofthelegitimacyofpowerisgendered.Referringtothesametraditional,ascribedtitlesthataholdercanonly losewhencommittingan immoralact,womentendtoacceptthatthiscannotbethecasewithmodernpower.Contrarily,mendonotdifferentiatebetween the twosourcesofpower.Although in thecaseofmodernpoweraccess isachievedbyelectioninsteadofascription,menwouldtendtoignoretheexpirationoftheirtermasifaboardpositionwereanascribedtitle.Thisdifferenceofperceptionsmadethesettlementofconflictseasierwithwomenthanwithmen,withthe formeracceptingtherulesofthegame,evenattheirownexpense,betterthanthelatter.

7.1.3 Livelihoodsreconstructionafterthedemiseofcottonproduction

Whileacknowledgingthehouseholdasthelocusofthelivelihoodsystem,theresultspresentedinChapter6makeclearthatthehouseholdfollowsthecontoursofthejoint

GENERALDISCUSSIONANDCONCLUSIONS

145

familyandthatintheresearchareapatrilinealkinshipisthefundamentalprincipleinhousehold composition and decision‐making. A household often comprisesmarriedbrothers and their families. Its members are strategically located in different,sometimesdistantplacesfortheefficientproductionandmanagementofresources.Inspite of the spatial separation, specific types of resources are pooled. While somemanagementautonomyisgrantedtoindividualadults, intheendallmembersabideby the authority of one among them, usually the eldest brother. The feature of co‐residencethatispartofclassicdefinitionsofthehousehold(e.g.Rudie, 1995)provedto be of minor importance. In the study area, eating from the same granary andpatrilinealkinshiparethecoreconstituentsofthehousehold.

Cotton producing households proved to exhibit a relatively large asset basis andamountofwelfare.SiaensandWodon(2008:174)concludedthat“cottonproducersfared relatively well” over the 1990s, and they found a 7 percent reduction in theprobability of being poor for cotton‐producing households compared to other ones.Additionally, during the prime of cotton production when it still deserved thequalificationofwhitegold, cottonnurturedyoungpeople’s autonomyvis‐à‐vis theirparents, and indirectly lightened for the latter the burden of proving for the youngpeople’s needs.Thedeclineof cottonprofitability brought about a livelihood shock,for which farmers envisioned institutional and technical remedies. Cotton farmerslong to see the cotton sector revive, but the regains of cotton production are veryerratic and the sector still continues to languish. The continuous decline of cottonreturnstriggeredtheadaptationcapacityofthesehouseholdsthroughthreetypesofstrategies: changing agricultural production patterns, migration of male youth, andgrantingmorefreedomtowomen.Thesestrategiesarenottheprerogativeofanyonetypeofhousehold.However,thecentreofthedecisionmakingaboutagivenstrategyvarieswithinthehousehold.Intheshiftinagriculturalproductionpatternscottonisalways included in thestrategies,and thehouseholdheadsarethedecisionmakers.This shows that a long‐term dependency on cotton productionmakes it difficult toabandon cotton in spite of its lost profitability. But this first strategy fell short ofexpectations and, in particular, failed tomeet the ‘modern’ needs of young people.Youth migration then emerged as an additional strategy. Migration concernsexclusivelyyoungmen,whodecideon thisandalmost impose theirdecisionon thehead of their household. Granting more freedom to women in the households toundertake more extra‐domestic income‐generating activities is a solution thatfollowedfromwomen’s increasingcontributiontothehouseholdincomewhentheirhusbands’incomefromcottondeclined.Becauseofthetraditionalimbalanceofpowerbetween husbands and wives within the household, however, women’s increasedeconomic contribution did not notably improve their bargaining power and theyremained dependent on their husbands for decisionmaking. Hence, the hypothesisthat livelihood losses consecutive to the demise of cottonproductionwouldmodifythegenderrelationsincotton‐producinghouseholdshadtoberejected.

CHAPTER7

146

7.1.4 Answeringthemainresearchquestion

Theanalysisof theprocessofatomisation fromthemicro to themacro levelallowsanswering themain question of the research thatwas stated in Chapter 1, namely:How did farmers’ agency affect their organisations, the cotton production, and thecollectiveactionthatevolvedaroundthatcrop?

It was found that breaking away from cotton networks enhanced the access toleadership positions and the advantages attached to these of previously ordinaryfarmers. However, it also turned out that social connections are needed to makecotton networks thrive, and that long‐lasting leaders had more social connectionsthansimpleleaders.Accessingthesepositionsontheonehandandholdingontothemontheother,showsthatthestocksofsocialcapitalfarmerscouldmobilisefromtheirsocialnetworksmakethedifferenceamongthem.

Themajorconclusion that transpires fromall this is that farmers’agency is thekeyelementintheprocessofcottonnetworking.Agencyhastwosides.Itmattersthroughits positive (the power to) and negative (the power over) manifestations (Kabeer,2005).Bothsidesofthecoinofagencyarevisibleintheprocessesthatwereanalysedin this thesis.Theprocessof atomisation is theobservableandaggregated resultofindividualfarmers’choicesintheirconstantmovementsofbreakingfrom,joining,orcreatingnetworks. Their choiceshavemodifiednot only the internal organisationaldynamics, but also the goals of the networks. The choicesweakened thepreviouslyprevailingtrustandinhibitedcooperationforcollectiveaction.Theseoutcomesoftheatomisationshowthatintheend,theagencyofindividualswasmoredeterminantinmodifyingthenetworkingtrajectoriesthanthatofinstitutionsandorganisations.

7.2 Theoreticalandconceptualreflections

7.2.1 Agricultural(andinstitutional)liberalisation

Thecottonsector inBeninseemstobe inapermanentflux.Whenthe farmers’pre‐dicamentwearsoffandsignsofprospectsbecomevisible,anothercrisisappears.Sucha cycle of recommencement has led cotton farmers to being disgruntled and worndownbythedespairofmakingthecottonsectorrevive.Themeritofliberalisationhasbeen to help cotton farming gain an entrepreneurial status by putting togetherresources and people previously unconnected (Weik, 2011). Some new governancestructuresemerged.However, theslackregulationof thesectorby theStatedidnotpromote compromise between groups of actors. It is assumed that “optimisingindividuals choose the most efficient institutional arrangement or governancestructureforcarryingouttransactionsinaspecificenvironment”(Slangenetal.,2004:104).Yet,theoutcomesofthestrugglesforcontrollingresourceswerefullofwinner‐losersituations,wherepartiesendlesslyseekingforequilibriumcouldhardlycometoanagreement.

GENERALDISCUSSIONANDCONCLUSIONS

147

ItalsohastobenotedthatcottonnetworkinginBeninhascontributedtobuildingandstrengtheninghumancapitalinruralorganisations,atrendthatalsowasobservedinasimilarcontext inMali (Roy, 2010). Individualsandcommunitiesbecameawareofthebenefitsoforganisingthemselves,alsoaroundothercropsandothercauses,andgainedmuch experience by doing so. As a consequence, after the decline of cottonproduction in the area, organisations flourished to replace the cotton organisationsthat had begun to wane. The permanently changing perceptions about the crop,however,showsthatfarmersaremorerationalthanisoftenassumed,andthattheycontinuegrowingacropaslongastheyfindsomebenefitsinit.

7.2.2 Networkingandsocialcapital:buildingtrustwithinorganisations

Thedynamicsof cottonnetworkingexhibit the featuresdescribedbyactor‐networktheory(ANT).Differentsocialactorswereinterconnectedthroughcottonproductionand marketing. Cotton commanded these interconnections and each of the actorsinvolvedattributedaparticularmeaning to theproduct,dependingon their specificrelationshipwiththecropatacertaintime.Hence,cottonbecameanactantasdefinedbyLatour(2005),changingitsmeaningininteractionwiththeactors.TheserelationsandtheirmeaningsareillustratedinFigure7.1.

Figure7.1 Changingmeaningsofcotton

Inaspatialsense,cottonasanactantcanbeeitheroutput(for farmers), inputs(forginners and transporters), or a means (for the State inputs suppliers and theregulatorybodiesof the cottonplatform). Seen froma temporal perspective, cottonchangedmeaningsthroughouttheyears, frombeingacolonialsymbol(justafterIn‐dependence)tobecomingwhitegold(duringthenineties),andlateron(secondhalfofthe2000stothepresent)beingdeemedapest,inspiteoftheamountofeffortandresources put in to make it revive. The shift of meanings is nurtured by cottonprofitability with its corollary of disagreements, conflicts, and the social divides it

Cottonbodies(Cottonasmeans)

Ginners(Cottonasaninput=rawmaterial)

Farmers(Cottonasanoutput)

Inputsuppliers(Cottonasameans)

Transporters(Cottonasan

input)COTTON

CHAPTER7

148

generatesamongindividualsandbetweenindividualsandinstitutions.Thedynamicsinvolvedbringaboutapowerplaythatunveilsan imbalance in thedecision‐makingprocessaboutthemanagementofcottonandthedirectionofcotton‐relatedpolicies.

An actor‐oriented approach and actor‐network theory together convey themessagethatobjects,eveninanimateandlackingwill,shouldnotbeconsideredasexemptofaction. While persons and their networks “attribute agency to various objects andideas,which, inturn,canshapeactors’perceptionsofwhat ispossible”(Long, 2001: 241), the actant is anything, human or non‐human that induces an action (Latour, 2005).Inaddition,besidestheinteractionsaccountingforthelifeoftheorganisations,cottonleadersdevelopedsocialrelationsthatperhapswouldnotexistiftherewerenocotton.Networks leaders became rolemodels andmaintained a network of leaderswiththecottonsystem,developingsomeculturalcapitalamongthemand“forminganelitestructure”(KilduffandTsai,2003:22).

The complexity of exchange in cotton production and marketing requires trust,particularlyinsituationswherelegalframeworksareweakorfuzzy.Theimportanceoftrustasacriticalinputforsuccessfulorganisationshasbeenestablishedforyears.But the success of organisations is not only determined by trust. The types ofinstitutionsprevailingwithinasocietyinfluencethatsuccessforanyorganisationinthe society concerned,becauseorganisations are always shapedby the institutionalarrangements prevailing in a society. The social structures of a given society are acultural manifestation of that society. Hence, effective organisations emerge insocietieswithintenseinteractions,whilethosesocietieswithpoorinteractionswouldproduceineffectiveorganisations.

7.2.3 GenderleadershipinruralorganisationsInthisstudyitwasfoundthatwomenareadmittedonboardsoforganisationsaslongastheyaccepttoplayfigurativerolesdefinedforthembypaternalisticculture.Whentheystartstandingforsomethingthatreducesthepowerandinfluenceofmen,theyinspirereticenceanddistrustfromtheirmalecolleagues.Men,therefore,seewomen’sinvolvement inmanagement as excluding real participationbydecisionmaking andresourcesdistribution,whichwouldneverimprovegenderequityintheorganisation.Thispointstoanaspectofthegendereddifferenceofmanagementstyleasdescribedby Rosener (2011), who demonstrated that participation reduces the likelihood ofoppositiontodecisionswithinorganisations.Thisstudyshowsthatwomen’sunder‐representationincottonorganisationandtheirlimitedrolesthereinderivefromtheirtraditionalsubordinatepositioninsocietywheremenascribewomenonlydomesticroles, which bear a high potential of being a burden to them (Moser, 1989). Theintertwinedcharacteroflifespheresmakesthatinequalitiesandachievementsinonespherecanbereproducedinanother(Kabeer, 2005).

The struggle of the two women profiled in Chapter 5 shows that some women’sagency iscapableofmodifying toagreatextent theirpositionswithin their families

GENERALDISCUSSIONANDCONCLUSIONS

149

and improve their trajectory within men‐dominated organisations. That agency, asshownby theWorldBank(2012),hasboth intrinsicand instrumental relevance forwomen’squalityoflifeandforactionsthatimprovetheirwell‐being.Giventheinter‐sectionofconstraintsthathinderwomen’sadvancementandempowermentinsocietyat large, this agency is highly required if they “are toplay an active role in shapinginstitutions, social norms, and the well‐being of their communities” (World Bank,2012: 115). Yet, generally, individual men exercise less negative power to curtailwomen’sagencythantheirassociationsdo.At individual level,mendoacknowledgewomen’s values and merits. To break the cycle of the reproduction of genderinequalitiesthereisaneedto:(i)liftupruralwomen’sfalseconsciousness(cognitive)by demonstrating that their disadvantaged position is not natural but sociallyconstructed;and(ii)getthelegitimacyofgenderequitywidelyacknowledgedattheindividual level for its wider acceptance at other societal levels. Addressing thesestrategicgenderneedsofwomenrequires(re)constructinggenderas“acomplexsetofsocialrelationsenactedacrossarangeofsocialpracticesthatexistbothwithinandoutsideofformalorganisations”(ElyandMeyerson,2000:113),andasembeddedinthreesocietaldimensions:individual,interactional,andinstitutional(Risman,2004).

7.2.4 Ruralhouseholdandlivelihoodadaptation

Thefindingsofthisresearchquestiontheclassicaldefinitionofthehouseholdwithapowerfulheaddecidingonallissuesofproductionandconsumptiononbehalfofthedependants.Theprocessesofdecisionmakingencounteredshowdifferentcentresofgravitywithregardtodomesticarrangementsanddecisionmaking.Also,ahouseholdhasanobjectiveandasubjectivedimensionwhichmakes theco‐residencecriterionproblematic.Althoughmemberscanbeobjectivelyseparatedanddispersed,eitherbylivingindifferenthomesteadsorbybeingamigrant,theheadofthehouseholdwouldautomaticallylistthemasdependentmembersofthehousehold.Inspiteoftheoftenlongabsenceofmigrantmembers,theywereconsideredhouseholdmembersbytheheadsofthehouseholdsconcerned,whereasthosepaidworkerslabouringandeatingin the household were not perceived as household members. The same applies tobrothers living in dispersed homesteads,whowere listed as household dependantsand considered themselves asmembers of a householdwhose headwas living at adifferent place. This attests to the importance of the power of kinship relations. AsRudie (1995: 228) noted, “households are small units in which primarily familialloyalties have to come to termswith some requirements for efficient resources.” Inour case, the subjective (and cognitive) dimension of household outweighs theobjective(andspatial)dimension.

Dispersed and spatially separated members of a household may still pool specifictypes of resources for the consumption andother needs of all householdmembers.This is what (Elmhirst, 2008) called the multi‐locality of livelihoods. In a study ofmigrationinruralNepal,Gartaula(2011)foundasimilarsituationofhouseholdssplitup across different geographic locations but interacting through information and

CHAPTER7

150

communication technologies. He referred to this type of households as modifiedextendedhouseholds.

To tackle thedeclineof cotton incomes, cotton‐farminghouseholdsdiversified theirsourcesofincomeandresourcesanddevelopedadaptationorcopingstrategies.Whileadaptationstrategiesareexclusivelyon‐farmstrategies that includegrantinga cashfunctiontoorincreasingthecashfunctionofothercrops,copingstrategiesprimarilyconsist of changing the purpose of production to the crop in crisis. This suggest atemporaldimensiontodiversificationstrategies.Thefactthatcottonisstillpartofthecoping strategies implies that these types of diversification are reversible, and thathouseholds that diversify by coping are highly likely to resume cotton productionwhen the fundamental production parameters improve. In contrast, adaptation canturnouttobeirreversibleifthealternativesourcesofincomeyieldasmuchincomeastheformercropandbearlessconstraints.

7.3 Conclusionandfutureresearch

The importance of cottonproduction toBenin’s economy is undeniable. By shapingthecountry’sproductionsystemoverthelasttwotothreedecades,cottonhasbecomethe backbone of its economy. Decision makers became aware of the necessity toorganise the sector and devised and implemented reformpolicies to effect changesandmake the sector more efficient. These reforms aimed at suppressing the Stateintervention in the sector and granting full autonomy to private actors for jointmanagementwithin theirprofessionalorganisations.Yet, the implementationof theState’s interventions intended to induce qualitative (and quantitative) changes intothe fundamental organisational system of the sector proves to be inadequate. Thereformpoliciesaswellas their implementationhad flawsthathinderedthesector’sefficiency.Among thesewere the lack of coordination and the over‐politicization ofthesector,thesurplusginningcapacityofthecountry,andtheover‐regulationofthesector.WhiletheStatelegislatedforliberalisation,itcontinuessubsidizingthesector,on the grounds of ensuring minimum revenue to cotton farmers. Hence, relativelyimportant amounts of money are invested to support inputs costs and sustain thecotton prices at the farm gate. These subsidies became sources of power play andconflicts. Furthermore, the institutions installed to enforce regulations in the sectorwereintheiractionsnotbackedupbyclearlyformulatedrules.Thehighlyregulatedsystem implemented to control the interactions between actors turned out to beheavy and ineffective. The reforms therefore burdened the sectorwith institutionalandpoliticalconstraintsinadditiontothetechnicalones.

Whatthisthesisshowsisthatduetotherelativelyhighprofitabilityofcottonsector,thenecessityofdiversifyingagriculturehasalwaysremainedagoodintention.Alltheactorsinthecottonproductionandmanagementremain‘addicted’tothecropfortheimportantamountofresourcesitgenerated.Asaconsequence,theStateinterventionsincottonaremouldedbypoliticalmotivesandthetechnicalaspectsareoverlooked.

GENERALDISCUSSIONANDCONCLUSIONS

151

Thesuccessivecrisesthesectorhasbeenundergoingsincethemid‐2000saresomeoftheoutcomesofsuchapoliticalmanagementofaproductivesector.Hence,reformsare being permanently implemented into the sector, the assessment of an on‐goingonefollowedbytheimplementationofnewones.

The present thesis went beyond the aggregate level to investigate the dynamics ofsmallholders’ cottonproduction, providingmicro‐level data about the impact of thedemiseof thecottonsystemon farmers.Taking together the fiveempiricalchaptershighlightsthesaliencyoftheuseofamixed‐methodsapproachforunderstandingthecomplexities of the interfaces between informal and formal structures. Indeed, thefocusonthe interactionbetweenactorsandstructuresofferedroomtoexplorehowindividual cotton farmers navigate through cotton networks in the context ofatomisation. Human agency is always constrained by structures, which implies thenecessity for individuals to integrate them into whatever their options are. At thesame time, we have seen that individual agency through choices and actions alsochangesstructures.

Thedemiseof cottonproductionyieldednegativeaswell aspositiveoutcomes.Themostsignificantnegativeeffectsarethedrasticshrinkingoflivelihoodassetsandthefreeze of collective action. Among the positive outcomes that could be blessings indisguise, is the strengthening of the human capital of farmers. Their struggle forrepresentation provided themwith invaluable experience that could be invested inorganisingtheproductionofothercrops.Itisalsoimportanttomentiontheenlargedmanoeuvringspaceforwomenandthe increasedautonomyofyouth,whichderivedfromtheirincreasedfinancialindependence.

CottonisspecialtoBeninandhasbeenspeciallymanagedfordecadesthroughuniquestructures. But the management of such a critical crop requires transparent andeffectiveregulation.Thefindingsofthisresearchcallforfurtherpoliciesandrulesfortheir implementation on issues such as farmer organisations, gender equity, andlivelihood systems. For their application to be sound and effect lasting changes,however, these policies would need further investigation of insufficiently exploredissues.Below,threesuchissuesareelaborated.

1. Returning to the cotton system, with regard to networks, it is imperative toimpose limits on their size ifmultiple networks are admitted. Members shouldknoweachotherinordertobeabletoexertcontrolovertheirpeers.Thatwouldleavelessroomforthemalpracticesthathavecreatedmountingcompetitionforcontrollingthenetworksandtheirresources.

2. It would be highly relevant to do research on the ways of empowerment ofwomen through their involvement in cotton management. The results of suchresearchcouldfurtherthepossibilityforwomencottonfarmerstoregisterandbeaccountable in theirownname, therebydecidingupon theuseof theirproducewithintheorganisations.

CHAPTER7

152

3. Throughlongitudinalresearch,itcouldbeinvestigatedwhetherthedispersionofhouseholdmembers reduces their vulnerability to shock. Thiswould provide abetter understanding of how the adaptation and coping strategies complementeach other and of how providing cotton‐producing households with specificinputsforfoodcropswouldsustaintheirlivelihoodstrategies.

153

REFERENCESAbdulai,A.,&CroleRees,A. (2001).Determinantsof incomediversificationamongst

ruralhouseholdsinSouthernMali.FoodPolicy,26(4),437‐452.Abou‐Moumouni, I. (2011). L'Alphabétisation dans les Rapports entre Organisations

Paysannes et Etat Local dans la Commune de N'Dali. (Mémoire de DEA),Universitéd'Abomey‐Calavi,Cotonou.

Acker,J.(2006).GenderandOrganizations.InJ.SaltzmanChafetz(Ed.),HandbookofSociologyofGender(Vol.11,pp.177‐194).NewYork:SpringerScience+BusinessMedia,LLC.

Adekunle,J.O.(1994).BorguandEconomicTransformation1700‐1900:TheWangaraFactor.AfricanEconomicHistory(22),1‐18.

Adler,P.S.,&Kwon,S.W.(2000).SocialCapital:theGood,theBad,andtheUgly.InL.E. Lesser (Ed.),Kowledgeand SocialCapital:Foundationsandapplications (pp.89–115).Boston:ButterworthHeinemann.

AfDB,&OECD (2008).AfricanEconomicOutlook2008:BeninOECD (Ed.) (pp. 137‐149).Retrievedfrom:http://www.afdb.org/pls/portal/docs/PAGE/ADB_ADMIN_PG/DOCUMENTS/ECONOMICSANDRESEARCH/BENIN_AEO2008_EN.PDF

Agarwal, B. (1997). ''Bargaining'' and Gender Relations: Within and Beyond theHousehold. Feminist Economics, 3(1), 1‐51. doi: http://dx.doi.org/10.1080‐/135457097338799

Ahn,T.K.,&Ostrom,E.(2008).SocialCapitalandCollectiveAction.InD.Castiglione,J.W. van Deth & G.Wolleb (Eds.),TheHandbook of Social Capital (pp. 70‐100).NewYork:OxfordUniversityPress.

AIC(2005).LesréformesdelafilièrecotonauBénin.Cotonou:MAEP.AIC (2006).Plan de campagne agricole cotonnière 2006–2007 etPerspectives 2007–

2008.Cotonou:AICRetrievedfrom:http://www.aicbenin.org/index2.php?option=com_docman&task=doc_view&gid=25&Itemid=42.

AIC(2007).EvaluationdesRéformesOpéréesdans leSecteurAgricoleauCoursdes25DernièresAnnées:CasdelaRéformedelaFilièreCoton.Cotonou:AIC.

AIC (2008a).Plandecampagneagricolecotonnière2008–2009etPerspectives2009–2010.Cotonou:Retrievedfrom:http://www.aicbenin.org/index2.php?option=com_docman&task=doc_view&gid=25&Itemid=42.

AIC (2008b).Privatisation/LibéralisationdesFilièresCotonnièresAfricaines:Casde laRépubliqueduBénin.Paperpresentedatthe6thAnnualCongressoftheAfricanCottonAssociation(Privatization/LiberalizationoftheAfricanCottonSectorandCompetitiveProduction),Lusaka(Zambia).

Akinwumi, O. (1998). Separatist agitation in Nigerian Borgu: A reflection onadministrative groupings under colonial and post‐colonial government. In E.Boesen,C.Hardung&R.Kuba(Eds.),Regardssur leBorgou:Pouvoiretaltéritédansunerégionouest‐africaine(pp.121–137).Paris:L'Harmattan.

Akram‐Lodhi,A.H.(1997).TheUnitaryModelofthePeasantHousehold:anObituary?EconomicIssues,2(Part1),27‐42.

Alber,E.(2004).GrandparentsasFoster‐Parents:TransformationsinFosterRelationsbetweenGrandparentsandGrandchildren inNorthernBenin.Africa: JournaloftheInternationalAfricanInstitute,74(1),28‐46.

REFERENCES

154

Albert,I.(1993).Desfemmesuneterre:unenouvelledynamiquesocialeauBénin.Paris:Harmattan.

Alcadipani,R.,&Hassard,J.(2010).Actor‐NetworkTheory,organizationsandcritique:towards a politics of organizing. Organization, 17(4), 419‐435. doi:http://dx.doi.org/10.1177/1350508410364441

Alston, J.M., Sumner, D.A., & Brunke,H. (2005). Impacts of Reductions inUS CottonSubsidies on West African Cotton Producers. Oxfam America. Retrieved fromhttp://www.oxfamamerica.org/newsandpublications/publications/research_reports/paying‐the‐price/Paying_the_Price.pdf

Angrosino,M.V.(2002).ConductingaLifeHistoryInterview.InM.V.Angrosino(Ed.),DoingCulturalAnthropology:Projects forEthnographicDataCollection (1sted.,pp.33‐44).Illinois:WavelandPress.

Badiane,O., Ghura,D., Goreux, L.,&Masson, P.R. (2002).CottonSectorStrategies inWest and Central Africa. IMF Working Paper No. 02/173. Working Paper.InternationalMonetaryFund.Washington.Retrievedfromhttp://ssrn.com/paper=880223

Baffes, J. (2004). "Cotton: Market Setting, Trade Policies, and Issues." World BankPolicyResearchWorkingPaper3218(pp.xii+76).Washington:TheWorldBank.

Baquedano,F.G.,Abbott,P.C.,&Sanders, J.H.(2008).TheWestAfricanCottonReformProcess:WhatNeedstobeDone?PurdueUniversity.Purdue.

Barrett, C.B.,Bezuneh,M.,&Aboud,A. (2001). Incomediversification,poverty trapsandpolicyshocksinCôted’IvoireandKenya.FoodPolicy,26(4),367‐384.

Barrett, C.B., Reardon, T., & Webb, P. (2001). Nonfarm income diversification andhousehold livelihood strategies in rural Africa: concepts, dynamics, and policyimplications.FoodPolicy,26(4),315‐331.

Barrett, S.R., Stokholm, S., & Burke, J. (2001). The Idea of Power and the Power ofIdeas:AReviewEssay.AmericanAnthropologist,103(2),468‐480.

Bassett,T.J.(2001).ThepeasantCottonRevolution inWestAfrica,Coted'Ivoire1880–1995(2001ed.).Cambridge:CambridgeUniversityPress.

Beauchemin, C. (2011). Rural–urban migration inWest Africa: towards a reversal?Migration trends and economic situation in Burkina Faso and Côte d'Ivoire.Population,SpaceandPlace,17(1),47‐72.doi:http://dx.doi.org/10.1002/psp.573

Beckert,S.(2005).FromTuskegeetoTogo:TheProblemofFreedomintheEmpireofCotton.TheJournalofAmericanHistory,92(2),498‐526.

Bigsten,A.,&Kayizzi‐Mugerwa,S. (1992).Adaptationanddistress in theurbaneco‐nomy:AstudyofKampalahouseholds.WorldDevelopment,20(10),1423‐1441.

Billing,Y.D. (2011).AreWomen inManagementVictimsof thePhantomof theMaleNorm? Gender,Work & Organization, 18(3), 298‐317. doi: http://dx.doi.org/‐10.1111/j.1468‐0432.2010.00546.x

Bingen, R.J. (1998). Cotton, democracy and development in Mali. The Journal ofModernAfricanStudies,36(02),265‐285.

Blais, M., & Dupuis‐Déri, F. (2011). Masculinism and the Antifeminist Counter‐movement. Social Movement Studies, 11(1), 21‐39. doi: http://dx.doi.org/‐10.1080/14742837.2012.640532

Blumberg,R.L.(1998).Review:Field,Forest,andFamily:Women'sWorkandPowerinRuralLaos.ByCarol J. Ireson.Boulder,Colorado:WestviewPress,1996.Pp.xxiii+285.TheAmericanJournalofSociology,103(5),1457‐1459.

REFERENCES

155

Boesen,E.,Hardung,C.,&Kuba,R.(1998).Introduction.InE.Boesen,C.Hardung&R.Kuba (Eds.), Regards sur le Borgou: Pouvoir et altérité dans une région ouest‐africaine(pp.11‐20).Paris:L'Harmattan.

Bonnassieux,A.(2002).Filièrecoton,émergencedesorganisationsdeproducteursettransformationsterritorialesauMalietauBurkinaFaso.LesCahiersd'Outre‐Mer,Filièresdeproduitstropicaux(220),421‐434.Retrievedfrom:http://com.revues.org/index961.html

Bourdieu,P. (1989).SocialSpaceandSymbolicPower.SociologicalTheory,7(1),14‐25.doi:http://dx.doi.org/10.2307/202060

Bourdieu,P.(1990).Thelogicofpractice(R.Nice,Trans.).Cambridge:PolityPress.Brégand,D.(1998).AnthropologiehistoriquedesWangaraduBorgou.InE.Boesen,C.

Hardung& R. Kuba (Eds.),Regards sur leBorgou:Pouvoir et altéritédansunerégionouest‐africaine(pp.245‐263).Paris&Montréal:L'Harmattan.

Breusers,M.(2000).SocialDiversityandIdentities:ReviewofRegardssurleBorgou:Pouvoiretaltéritédansunerégionouest‐africaine,byE.Boesen,C.Hardung&R.Kuba.1998.EtudesAfricaines.Paris:L'Harmattan.TheJournalofAfricanHistory,41(2),311‐313.

Brydon, L. (2010). Review: Strength beyond Structure: Social and HistoricalTrajectoriesofAgencyinAfrica editedbyM.deBruijn,R.vanDijk&J.B.Gewald(Eds.). Development and Change, 41(1), 180‐182. doi: http://dx.doi.org/‐10.1111/j.1467‐7660.2009.01624.x

Burt,R.S. (2000).Thenetworkstructureofsocialcapital.Research inOrganizationalBehavior,22(0),345‐423.

Burt, S. R. (1992). StructuralHoles: The Social Structure of Competition. CambridgeMassachusetts:HarvardUniversityPress.

Burt,S.R.(2005).BrokerageandClosure:AnIntroductiontoSocialCapital.NewYork:OxfordUniversityPress.

Cairns,E.,&Niens,U. (1999).BookReview:ThePoliticsofDenial.HumanRelations,52(9),1225‐1229.doi:http://dx.doi.org/10.1023/a:1016952112460

Callon,M., & Law, J. (1997). After the Individual in Society: Lessons on Collectivityfrom Science, Technology and Society. The Canadian Journal of Sociology /Cahierscanadiensdesociologie,22(2),165‐182.doi:http://dx.doi.org/10.2307/3341747

Carli, L.L. (1989). Gender differences in interaction style and influence. Journal ofPersonalityandSocialPsychology,56(4),565‐576.

Carr, E.R. (2008). Between structure and agency: Livelihoods and adaptation inGhana’sCentralRegion.GlobalEnvironmentalChange(18),689‐699.

Cattaneo,O.(2003).LesleconsdeCancunpourl'Afrique:Ledifficileapprentissagedela démocratie à l'OMC. Afrique Contemporaine, 2003/4(208), 81‐99. doi:http://dx.doi.org/10.3917/afco.208.0081

CIPB(2008)Etudemonographiquesurlecoton.Février2008,40pp.Clark, D., & Kleyn, N. (2011). Why Do They Leave? Voluntary Turnover of South

African Women Executives. In H.P. Werhane & M. Painter‐Morland (Eds.),Leadership, Gender, and Organization (pp. 185‐207). London and New York:Springer.

Clark,V.L.P.,Creswell,W.J.,Green,D.O.N.&Shope,J.R.(2008).MixingQuantitativeandQualitativeApproaches:AnIntroductiontoEmergentMixedMethodsResearch.InS.N.Hesse‐Biber&P.Leavy(Eds.),HandbookofEmergentMethods(pp.363‐387).NewYork&London:TheGuilfordPress.

REFERENCES

156

Coffe,H.,&Bolzendahl,C. (2011).GenderGaps inPoliticalParticipationAcrossSub‐Saharan African Nations. Social Indicators Research, 102(2), 245‐264. doi:http://dx.doi.org/10.1007/s11205‐010‐9676‐6

Cohen, D., & Prusak, L. (2001). In Good Company: How social capital makesorganizationswork.HarvardBusinessSchoolPress.

Coleman, J. S. (1988). SocialCapital in theCreationofHumanCapital.TheAmericanJournalofSociology,94,S95‐S120.

CommunedePèrèrè.(2010).PlandeDéveloppementCommunalDeuxièmegénération,2011‐2015.Pèrèrè:MairiedePèrèrè.

Crenshaw, K. (1991). Mapping the Margins: Intersectionality, Identity Politics, andViolenceagainstWomenofColor.StanfordLawReview,43(6),1241‐1299.

Creswell, W.J., & Clark, V.L.P. (2008). Designing and Conducting Mixed MethodsResearch.London:Sagepublications.

Davies,S.,&Hossain,N.(1997).Livelihoodadaptation,publicaction,andcivilsociety:areviewoftheliteratureInstituteforDevelopmentStudies.WorkingPaperNo.57(pp51).

DeHaan,A.(1999).Livelihoodsandpoverty:Theroleofmigration‐acriticalreviewof the migration literature. Journal of Development Studies, 36(2), 1‐47. doi:http://dx.doi.org/10.1080/00220389908422619

DeHaan,A.,Brock,K.,&Coulibaly,N.(2002).Migration,LivelihoodsandInstitutions:ContrastingPatternsofMigrationinMali.JournalofDevelopmentStudies,38(5),37‐58.doi:http://dx.doi.org/10.1080/00220380412331322501

Deguene,A.,&Forsé,M.(1994).IntroducingSocialNetworks(A.Borges,Trans.1999ed.).London:SagePublications.

Dirks, K.T., & Ferrin, D.L. (2001). The Role of Trust in Organizational Settings.OrganizationScience,12(4),450‐467.

Dovidio,J.F.,Saguy,T.,&Shnabel,N.(2009).CooperationandConflictwithinGroups:Bridging Intragroup and Intergroup Processes. Journal of Social Issues, 65(2),429‐449.doi:http://dx.doi.org/10.1111/j.1540‐4560.2009.01607.x

Ellis, F. (1998). Household strategies and rural livelihood diversification. Journal ofDevelopment Studies, 35(1), 1‐38. doi: http://dx.doi.org/10.1080/‐00220389808422553

Ellis, F. (2000). The Determinants of Rural LivelihoodDiversification in DevelopingCountries. Journal of Agricultural Economics, 51(2), 289‐302. doi: http://dx.‐doi.org/10.1111/j.1477‐9552.2000.tb01229.x

Elmhirst,R.(2008).Multi‐locallivelihoods,naturalresourcemanagementandgenderin upland Indonesia. In B.P. Resurreccion & R. Elmhirst (Eds.), Gender andNaturalResourceManagement: Livelihoods,Mobility and Interventions (pp. 67‐87).London:Earthscan.

Ely, R.J., & Meyerson, D.E. (2000). Theories of gender in organizations: A newapproach to organizational analysis and change. Research in OrganizationalBehavior,22,103‐151.

Engelen, B. (2010). Beyond markets and states: the importance of communities.International Social Science Journal, 61(202), 489‐500. doi: http://dx.doi.org/‐10.1111/j.1468‐2451.2011.01773.x

Fairclough,N.(2003).AnalyzingDiscourse:TextualAnalysis forSocialResearch (Firsted.).London:Routledge.

Falen, D.J. (2008). Polygyny and Christian Marriage in Africa: The Case of Benin.AfricanStudiesReview,51(2),51‐74.

REFERENCES

157

Ferguson, J. (1994). Modernist Narratives, Conventional Wisdoms, and ColonialLiberalism:ReplytoaStrawMan.JournalofSouthernAfricanStudies,20(4),633‐640.doi:http://dx.doi.org/10.2307/2636977

Ferguson,J.(1999).ExpectationofModernity:MythsandMeaningsofUrbanLifeontheZambianCopperbelt.Berkley:Universityofcalifornia.

Ferlander,S. (2007).The ImportanceofDifferentFormsofSocialCapital forHealth.ActaSociologica,50(2),115‐128.

Ferrigno, S., Ratter, S.G., Ton, P., Vodouhê, D.S., Williamson, S. & Wilson, J. (2005).OrganicCotton:ANewDevelopmentPathforAfricanSmallholders?GatekeeperSeries,(120),25.Retrievedfrom:http://www.povertymonitoring.go.tz/documents/IIED_on_organic_cottons_potential_in_Tz.pdf

FIDA (2005). République du Bénin: Evaluation du Programme de pays. (pp. 65).Cotonou:FIDA.

Fok,M.(2010).CottoninCrisesofFrancophoneAfrica:FatalityorChallengeforMulti‐dimensional Cooperation? In M. Nissanke & G. Mavrotas (Eds.), Commodities,Governance and Economic Development under Globalization (pp. 165‐220).Hampshire,NewYork:PalgraveMacmillan.

Francis, E. (2002). Gender, Migration andMultiple Livelihoods: Cases from Easternand Southern Africa. Journal of Development Studies, 38(5), 167‐190. doi:http://dx.doi.org/10.1080/00220380412331322551

Frost,P.,Campbell,B.,Luckert,M.,Mutamba,M.,Mandondo,A.,&Kozanayi,W.(2007).In Search of Improved Rural Livelihoods in Semi‐Arid Regions through LocalManagement of Natural Resources: Lessons from Case Studies in Zimbabwe.WorldDevelopment,35(11),1961‐1974.

Gandonou,A.J.M.(2005).EssaysonPrecisionAgricultureTechnologyAdoptionandRiskManagement.(PhDThesis),UniversityofKentucky,LexingtonUSA.

Garcia, L. (1970). Lesmouvements de résistance auDahomey (1914‐1917).Cahiersd'ÉtudesAfricaines,10(37),144‐178.

Gartaula,H. (2011).TheOtherSideofMigration inRuralNepal:SocioculturalTrans‐formation and theWomen Left Behind. (PhD Thesis), Wageningen University,Wageningen,theNetherlands.

Gergely, N. (2009). 'The Cotton Sector of Benin' (Vol. World Bank Africa RegionWorkingPaperSeriesNo125,WashingtonDC:TheWorldBank.

Gergely, N., & Poulton, C. (2009). Historical Background and Recent InstitutionalEvolution of African Cotton. In D. Tschirley, C. Poulton & P. Labaste (Eds.),OrganizationandPerformanceofCottonSectorsinAfrica:LearningfromReformExperience(pp.31‐44).Washington:WorldBank.

Ghatak, M., & Guinnane, T.W. (1999). The economics of lending with joint liability:theoryandpractice.JournalofDevelopmentEconomics,60(1),195‐228.

Giddens,A.(1984).TheConstitutionofSociety(1986ed.).Cambridge:PolityPress.Gilley, J.W., Morris, M.L., Waite, A.M., Coates, T. & Veliquette, A. (2010). Integrated

theoretical model for building effective teams.Advances inDevelopingHumanResources,12(1),7‐28.doi:http://dx.doi.org/10.1177/1523422310365309

Granovetter, M. (1983). The Strength of Weak Ties: A network theory revisited.SociologicalTheory,1,201‐233.

Granovetter,M.S. (1973). The Strength ofWeakTies.American JournalofSociology,78(6),1360‐1380.

Gray,L.,&Kevane,M.(1999).Diminishedaccess,divertedexclusion:WomenandlandtenureinSub‐SaharanAfrica.AfricanStudiesReview,42(2),15‐39.

REFERENCES

158

Grootaert, C., Oh, G.T., & Swamy, A. (2002). Social Capital, Household Welfare andPovertyinBurkinaFaso.JournalofAfricanEconomies,11(1),4‐38.doi:http://‐dx.doi.org/10.1093/jae/11.1.4

Hagemaster, J.N. (1992). Life history: a qualitative method of research. Journal ofAdvanced Nursing, 17(9), 1122‐1128. doi: http://dx.doi.org/10.1111/j.1365‐2648.1992.tb02047.x

Hahonou,E.K.(2011).Benin(Vol6).InA.Mehler,H.Melber&K.vanWalraven(Eds.),AfricaYearbook.BrillOnline:ECAS2011/Leiden:AfricanStudiesCentre.

Hampshire,K.(2002).FulaniontheMove:SeasonaleconomicmigrationintheSahelasasocialprocess.JournalofDevelopmentStudies,38(5),15‐36.doi:http://dx.‐doi.org/10.1080/00220380412331322491

Hanneman, R.A. & Riddle, M. (2005). Introduction to social network methods.Retrievedfromhttp://faculty.ucr.edu/~hanneman/

Harneit‐Sievers,A. (2002).Review:ModesofPower inBorgu.The JournalofAfricanHistory,43(2),348‐349.

Hier, S.P. (2005). Structural Functionalism. In S.P. Hier (Ed.), Contemporary Socio‐logical Thought: Themes and Theories (pp. 7‐10). Toronto, Ontario: CanadianScholars'PressInc.

Holger Seebens, J.S. (2007). Bargaining power and efficiency–rural households inEthiopia.JournalofInternationalDevelopment,19(7),895‐918.

Hugon, P. (2005). Les Réformes de la Filière Coton au Mali et les NégociationsInternationales. Afrique Contemporaine, 2005/4(No 216), 203‐225,. Retrievedfrom: http://www.cairn.info/revue‐afrique‐contemporaine‐2005‐4‐page‐203.htm.doi:http://dx.doi.org/10.3917/afco.216.0203

IFAD (2010). Rural Poverty Report 2011: New Realities, New Challenges: Newopportunitiesfortomorrow’sgeneration(pp.317).Rome:InternationalFundforAgriculturalDevelopment.

Iliffe, J. (1967). The Organization of the Maji Maji Rebellion. The Journal ofAfricanHistory,8(3),495‐512.

Inglehart,R.,&Welzel,C.(2005).Modernization,CulturalChange,andDemocracy:TheHumanDevelopmentSequence.Cambridge:CambridgeUniversityPress.

INSAE(2004).Cahierdesvillagesetquartiersdeville(RGPH3).Cotonou:DirectiondesEtudesDémographiques.

INSAE(2009).EnquêteModulaireIntégréesurlesConditionsdeViedesMénages2èmeEdition (EMICoV 2011): Rapport Préliminaire. Cotonou: INSAE Retrieved fromhttp://www.insaebj.org/2012/doc/Publications/Rapport%20pr%E9liminaire%20emcov%202011.pdf.

Isaacman,A.(1992).Peasants,WorkandtheLaborProcess:ForcedCottonCultivationinColonialMozambique1938–1961.JournalofSocialHistory,25(4),815‐855.

Isaacman,A.,&Roberts,R.(Eds.).(1995).Cotton,ColonialismandSocialHistoryinSub‐SaharanAfrica.Portsmouth&London:Heinemann.

Isaacman, A., Stephen, M., Adam, Y., Homen, M.J., Macamo, E. & Pililao, A. (1980).CottonIstheMotherofPoverty:PeasantResistancetoForcedCottonProductionin Mozambique, 1938‐1961. The International Journal of African HistoricalStudies,13(4),581‐615.

Johnson,R.B.,Onwuegbuzie,A.J.&Turner,L.A.(2007).TowardaDefinitionofMixedMethodsResearch.JournalofMixedMethodsResearch,1(2),112‐133.doi:http://dx.doi.org/10.1177/1558689806298224

REFERENCES

159

Jones,R.(1998).TheEthnicgroupsofthepresentdayBorgu.InE.Boesen,C.Hardung&R.Kuba(Eds.),RegardssurleBorgou:Pouvoiretaltéritédansunerégionouest‐africaine(pp.71‐89).Paris&Montréal:L'Harmattan.

Kabeer,N.(2005).Genderequalityandwomen'sempowerment:Acriticalanalysisofthe thirdmillenniumdevelopmentgoal1.Gender&Development,13(1),13‐24.doi:http://dx.doi.org/10.1080/13552070512331332273

Kampen,J.(2010).Onthe(in)consistencyofcitizenandmunicipallevelindicatorsofsocial capital and local government performance. Social Indicators Research,97(2),213‐228.doi:http://dx.doi.org/10.1007/s11205‐009‐9498‐6

Kherallah, M., Minot, N., Kachule, R., Soulé, B.G. & Berry, P. (2001). Impact ofagriculturalmarketreformsonsmallholderfarmersinBeninandMalawi.Vol.1,pp.350.WashingtonDC:InternationalFoodPolicyResearchInstitute(IFPRI).

Kilduff, M. & Tsai, W. (2003). Social Networks and Organisation. London: SagePublications.

Kiser, E. (1997). Review of The Cooperator's Dilemma by Mark Irving Lichbach.AmericanJournalofSociology,102(6),1725‐1727.

Kollock, P. (1998). Social dilemmas: The anatomy of cooperation.AnnualReview ofSociology,24(1),183‐214.doi:10.1146/annurev.soc.24.1.183

Komorita,S.S.&Parks,C.D.(1995).InterpersonalRelations:Mixed‐MotiveInteraction.Annual Review of Psychology, 46(1), 183‐207. doi: http://dx.doi.org/‐10.1146/annurev.ps.46.020195.001151

Kouton, T., Yorou, G., Nouatin, G. & Wennink, B. (2006). Orienting research anddevelopment for cotton production: The national federation of producers'unions (FUPRO) in Benin. In B. Wenninck & W. Heemskerk (Eds.), Farmers'organizationsandagriculturalinnovation:CasestudiesfromBenin,RwandaandTanzania. In B. Shapland & J. Kampen (Series Eds.), Developement Policy &Practice (Vol. 374, pp. 47‐54). Amsterdam: Royal Tropical Institute. Retrievedfromhttp://www.kit.nl/net/KIT_Publicaties_output/ShowFile2.aspx?e=1561.

Kpade, C.P. (2011). Adaptation de la coordination et nouvelles contradictions entreacteursdusystèmecotonauBéninfaceàlalibéralisationéconomique.(ThèsedeDoctorat),UniversitédeBourgogne.

Kpade,C.P.&Boinon,J.P.(2011).DynamiquedespolitiquescotonnièresauBénin:Unelectureparladépendancedesentier.EconomieRurale(321),58‐72.

Kramer, R.M. (1999). Trust and distrust in organizations: Emerging perspectives,enduringquestions.AnnualReviewofPsychology,50(1),569‐598.doi:http://dx.doi.org/10.1146/annurev.psych.50.1.569

Kriger, C.E. (2005). Mapping the history of cotton textile production in precolonialWestAfrica.AfricanEconomicHistory(33),87‐116.

Kuba, R. (1998). Les Wassangari et les chefs de la terre au Borgou: Une histoired'intégrationmutuelle.InE.Boesen,C.Hardung&R.Kuba(Eds.),RegardssurleBorgou:Pouvoir etaltéritédansune région ouest‐africaine (pp. 93‐119). Paris:L'Harmattan.

Kuba,R.(2000).ReviewofCommerceCaravanieretRelationsSocialesauBénin:LesWangara du Borgou. By Denise Brégand. Collection Sociétés Africaines etDiaspora.Paris:L'Harmattan,1998.Pp272.TheInternationalJournalofAfricanHistoricalStudies,33(1),201‐202.

Latour,B. (1996).Onactor‐network theory:A fewclarifications.SozialeWelt,47(4),369‐381.doi:http://dx.doi.org/10.2307/40878163

Latour, B. (1997). On actor‐network theory: A few clarifications. Retrieved fromCibersociologiawebsite:http://cibersociologia.com/web

REFERENCES

160

Latour,B. (2005).Reassembling theSocial:An Introduction toActor‐Network‐Theory.NewYork:OxfordUniversityPress.

Law, J. (1992). Notes on the Theory of the Actor‐Network: Ordering, Strategy andHeterogeneity.SystemsPractice,5,379‐393.

Levrat,R.(2008).LecotonenAfriqueOccidentaleetCentraleavant1950:UnexempledelapolitiquecolonialedelaFrance(Firsted.).Paris:L'Harmattan.

Lin, N. (2001). SocialCapital:ATheory of Social StructureandAction. Cambridge&NewYork:CambridgeUniversityPress.

Lin,N.(2008).ANetworkTheoryofSocialCapital.InD.Castiglione,J.W.vanDeth&G.Wolleb(Eds.),TheHandbookofSocialCapital (Firsted.,pp.50–69).NewYork:OxfordUniversityPress.

Lin,N.&Erickson,H.B.(2008).Theory,MeasurementsandtheResearchEnterpriseonSocialCapital. InN.Lin&H.B.Erickson (Eds.),SocialCapital:An InternationalResearchProgram(1sted.,pp.1‐24).NewYork:OxfordUniversityPress.

Lodin,J.B.(2012).IntrahouseholdBargainingandDistributionalOutcomesRegardingNERICAUplandRiceProceedsinHoimaDistrict,Uganda.Gender,TechnologyandDevelopment,16(3),253‐278.doi:http://dx.doi.org/10.1177/0971852412459425

Lombard, J. (1960). La vie politique dans une ancienne société de type féodal: LesBaribaduDahomey.Cahiersd'ÉtudesAfricaines,1(3),5‐45.

Lombard,J.(1998).Lemodèlesoci‐politiquedespeuplesduBorgoudanslessociétésd'AfriqueNoire.InE.Boesen,C.Hardung&R.Kuba(Eds.),RegardssurleBorgou:Pouvoir et altérité dans une région ouest‐africaine (pp. 27‐37). Paris:L'Harmattan.

Long,N.(2001).DevelopmentSociology:Actorperspectives(1sted.).LondonandNewYork:Routledge.

MaboudouAlidou, G. (2002).Le travailau seindesorganisationspaysannes: casdesgroupements féminins de la Commune de Ouaké. [Unpublished master thesis],Universitéd'Abomey‐Calavi,Cotonou.

MAEP. (2011). Plan Stratégique de Relance du Secteur Agricole (PSRSA). Cotonou:Ministèredel'Agriculture,del'ElevageetdelaPêche.Retrievedfrom:http://www.insaebj.org/2012/doc/Publications/PLAN_STRATEGIQUE_DE_RELANCE_DU_SECTEUR_AGRICOLE.pdf.

Mahoney,J.(2001).Path‐DependentExplanationsofRegimeChange:CentralAmericain Comparative Perspective. Studies in Comparative InternationalDevelopment(SCID),36(1),111‐141.doi:10.1007/bf02687587

Maier, J.E.D. (1995). Persistence of Precolonial Patterns of Production: Cotton inGerman Togoland 1800‐1914. In A. Isaacman & R. Roberts (Eds.), Cotton,ColonialismandSocialHistory inSub‐SaharanAfrica (pp.71‐95).Portsmouth&London:Heinemann.

Mama Debourou, D. (2009). La société baatonnu du nord‐Bénin: son passé, sondynamisme, ses conflits et ses innovations, des origines à nos jours. Essai surl'histoire d'une société de l'Afrique Occidentale. (Thèse de Doctorat d’Etat),Universitéd'Abomey‐Calavi,Cotonou.

Mandel, J.L. (2004). Mobilitymatters: women's livelihood strategies in Porto Novo,Benin. [Gender, Place& Culture].Gender,Place&Culture,11(2), 257‐287. doi:10.1080/0966369042000218482

Manning,P.(1982).Slavery,colonialismandeconomicgrowthinDahomey,1640‐1960.Cambridge[etc.]:CambridgeUniversityPress.

REFERENCES

161

Maurizio,M.(2007).Householdintertemporalbehaviour:Acollectivecharacterizationandatestofcommitment.ReviewofEconomicStudies,74(3),857‐895.

McEvily,B.,Perrone,V.&Zaheer,A.(2003).Trustasanorganizingprinciple.Organi‐zationScience,14(1),91‐103.

Meredith, M. (2005). The State of Africa: A History of Fifty Years of Independence.London:TheFreePress.

Metcalf, G. (1987). A microcosm of why Africans sold slaves: Akan consumptionpatternsinthe1770s.TheJournalofAfricanHistory,28(3),377‐394.

Minot, N. &Daniels, L. (2005). Impact of global cottonmarkets on rural poverty inBenin.AgriculturalEconomics,33(3),453‐466.

Misztal,B.A.(1996).Trust inmodernsocieties:thesearch forthebasesofsocialorder(Firsted.).Cambridge:PolityPress&BlackwellPublishersLtd.

Moerbeek, H., & Flap, H. (2008). Social resources and their effect on occupationalattainment through the life course. In N. Lin & B. H. Erickson (Eds.), SocialCapital:An InternationalResearchProgram (First ed.,pp.133‐156).NewYork:OxfordUniversityPress.

Molyneux, M. (1985). Mobilization without emancipation? Women’s interests, thestateandRevolutioninNicaragua.FeministStudies,11(2),227‐254.

Molyneux,M.(1998).Analysingwomen'smovements.DevelopmentandChange,29(2),219‐245.doi:http://dx.doi.org/10.1111/1467‐7660.00077

Molyneux, M. (2002). Gender and the silences of social capital: Lessons from LatinAmerica. Development and Change, 33(2), 167‐188. doi: http://dx.doi.org/‐10.1111/1467‐7660.00246

Moody,J.&White,D.R.(2003).Structuralcohesionandembeddedness:Ahierarchicalconceptofsocialgroups.AmericanSociologicalReview,68(1),103‐127.

Morse,J.M.&Niehaus,L.(2009).MixedMethodDesign:PrinciplesandProcedures(Vol.4).California,USA:WalnutCreek,Inc.;LeftCoastPress,Inc.

Moseley,W.G. (2002). Review: The Peasant Cotton Revolution inWest Africa: Côted'Ivoire, 1880‐1995.Annals of theAssociation ofAmericanGeographers,92(4),801‐803.

Moseley,W.G.,&Gray, L. (Eds.). (2008).HangingbyaThread:Cotton,Globalization,and Poverty in Africa. Athens, Ohio, Uppsala: Ohio University Press and TheNordicAfricaInstitute.

Moser,C.(2010).Movingbeyondgenderandpovertytoassetaccumulation:evidencefrom low‐income households in Guayaquil, Ecuador. In S. Chant (Ed.), TheInternationalHandbook of Gender and Poverty: Concepts, Research, Policy (pp.391‐398).CheltenhamandNorthampton:EdwardElgarPublishingLimited.

Moser, C.O.N. (1989). Gender planning in the third world: Meeting practical andstrategicgenderneeds.WorldDevelopment,17(11),1799‐1825.

Moser, C.O.N. (1994). Gender Planning and Developement: Theory, Practice andTraining.Routledge:LondonandNewYork.

Mosse, D., Gupta, S., Mehta, M., Shah, V., Rees, J.F. & Team, K.P. (2002). Brokeredlivelihoods: Debt, labour migration and development in tribal Western India.JournalofDevelopmentStudies,38(5),59‐88.doi:http://dx.doi.org/10.1080/00220380412331322511

Murdoch, J. (1998). The spaces of actor‐network theory.Geoforum, 29(4), 357‐374.doi:http://dx.doi.org/10.1016/S0016‐7185(98)00011‐6

Nahapiet, J. & Ghoshal, S. (1998). Social capital, intellectual capital, and theorganizationaladvantage.TheAcademyofManagementReview,23(2),242‐266.

REFERENCES

162

Nancy, J.,Ruth,S.,&Lundy,M. (2002).The importanceofsocialcapital inColombianruralagro‐entreprises.CAPRiWorkingPaper,No.26.

Neter, J., Kutner, M., Wasserman, W., & Nachtsheim, C. (1996). Applied LinearStatisticalModels.Boston:McGraw‐Hill/Irwin.

Niehof,A.(2004).Thesignificanceofdiversificationforrurallivelihoodsystems.FoodPolicy,29(4),321‐338.

Niehof,A. (2011).Conceptualizing thehouseholdasanobjectofstudy. InternationalJournal of Consumer Studies, 35(5), 488‐497. doi: http://dx.doi.org/10.1111/‐j.1470‐6431.2011.01026.x

Nissanke, M. & Mavrotas, G. (Eds.). (2010). Commodities, Governance and EconomicDevelopment under Globalization (1st ed.). Hampshire, New York: PalgraveMacmillan.

Norman,D.W.(1973).MethodologyandProblemsofFarmManagementInvestigation:ExperiencesfromNorthernNigeria:MichiganStateUniversity.

Oh,H.,Labianca,G.,&Chung,M.H.(2006).AMultilevelModelofGroupSocialCapital.TheAcademyofManagementReview,31(3),569‐582.

Olivier de Sardan, J.P. (2005). Anthropology and Development: Understanding Con‐temporarySocialChange.London:ZEDBOOKS.

Onyeiwu, S. (2000). Deceived by African Cotton: The British Cotton GrowingAssociationandtheDemiseoftheLancashireTextileIndustry.AfricanEconomicHistory(28),89‐121.

Owusu, F. (2005). Livelihood strategies and performance of Ghana's health andeducation sectors: exploring the connections. Public Administration andDevelopment,25(2),157‐174.doi:http://dx.doi.org/10.1002/pad.337

Owusu, F. (2007). Conceptualizing Livelihood Strategies in African Cities. Journal ofPlanning Education and Research, 26(4), 450‐465. doi: http://dx.doi.org/‐10.1177/0739456x06298818

Painter‐Morland, M. (2011). Systemic leadership, gender, organization. In H. P.Werhane&M.Painter‐Morland(Eds.),Leadership,Gender,andOrganization(pp.139–165).LondonandNewYork:Springer.

Pamphilon,B.(1999).TheZoomModel:ADynamicFrameworkfortheAnalysisofLifeHistories. Qualitative Inquiry, 5(3), 393‐410. doi: http://dx.doi.org/‐10.1177/107780049900500306

Parsons, T. (2005). The Social System. In S. P.Hier (Ed.),ContemporarySociologicalThought:ThemesandTheories(pp.11‐21).Toronto,Ontario:CanadianScholars'PressInc.

Pesche, D., & Nubukpo, K. (2004). L'Afrique du Coton a Cancun: les Acteurs d'uneNegociation.PolitiqueAfricaine,3/2004(95),158‐168.Retrievedfrom:www.cairn.info/revue‐politique‐africaine‐2004‐3‐page‐158.htm.doi:http://dx.doi.org/10.3917/polaf.095.0158

Plickert, G., Côté, R. R., &Wellman, B. (2007). It's notwho you know, it's how youknowthem:Whoexchangeswhatwithwhom?SocialNetworks,29(3),405‐429.

Porta, R. L., Lopez‐de‐Silanes, F., Shleifer, A. & Vishny, R.W. (1997). Trust in largeorganizations.TheAmericanEconomicReview,87(2),333‐338.doi:http://dx.doi.org/10.2307/2950941

Poulsen,A.(2009).Cooperation:evidencefromexperiments.InG.T.Svendsen&G.L.H.Svendsen(Eds.),HandbookofSocialCapital:TheTroikaofSociology,PoliticalScience and Economics (pp. 36–56). Cheltenham and Northampton: Elgar,Edward.

REFERENCES

163

Putnam, R. (1993). The prosperous community: Social capital and public life. TheAmericanProspect,4(13),65‐78.

Putnam, R. (1995). Bowling alone: America's declining social capital.The JournalofDemocracy,6(1),65‐78.

Rai,A.S.,&Sjöström,T. (2004). IsGrameen lending efficient?Repayment incentivesandinsuranceinvillageeconomies.TheReviewofEconomicStudies,71(1),217‐234.

Ratcliffe, B.M. (1982). Cotton imperialism: Manchester merchants and cottoncultivation in West Africa in the mid‐nineteenth century. African EconomicHistory(11),87‐113.

Ridgeway, C.L., & Correll, S.J. (2004). Unpacking the gender system: A theoreticalperspective on gender beliefs and social relations.Gender and Society, 18(4),510‐531.doi:10.2307/4149448

Ritchie, J.,&Spencer,L.(1994).Qualitativedataanalysisforappliedpolicyresearch.InA.Bryman&G.R.Burgess (Eds.),AnalyzingQualitativeData. (2002ed., pp.173‐194).NewYork:Routledge.

Roberts,R.L. (1996).TwoWorldsofCotton:ColonialismandtheRegionalEconomy intheFrenchSoudan,1800‐1946.Stanford:StanfordUniversityPress.

Rogaly,B.,Coppard,D.,Safique,A.,Rana,K.,Sengupta,A.&Biswas,J.(2002).SeasonalmigrationandwelfareinEasternIndia:Asocialanalysis.JournalofDevelopmentStudies,38(5),89‐114.doi:http://dx.doi.org/10.1080/00220380412331322521

Rogers,E.M.(1995).DiffusionofInnovations(4thed.).NewYork:TheFreePress.Rosener,B.J.(2011).WaysWomenLead.InH.P.Werhane&M.Painter‐Morland(Eds.),

Leadership, Gender, and Organization (pp. 19–29). London and New York:Springer.

Roy,A.(2010).PeasantstrugglesinMali:fromdefendingcottonproducers'intereststo becoming part of the Malian power structures. Review of African PoliticalEconomy,37(125),299‐314.doi:http://dx.doi.org/10.1080/03056244.2010.510628

Rudie, I. (1995). The significance of eating: co‐operation, support and reputation inKelantanMalayhouseholds.InW.J.B.Karim(Ed.),MaleandFemaleinDevelopingSoutheastAsia(pp.227–245).Oxford,WashingtonDC:BergPublishers.

Saizonou, J.N. (2008). Gestion Interprofessionelle et Développement des FilièresAgricoles: Cas de la Filière Coton au (Master), Institut Universitaire du Bénin(IUB),Cotonou.

Schottman,W.(1998).ThejokingrelationshipaspractisedbytheBaatombuofBénin.In E. Boesen, C. Hardung & R. Kuba (Eds.), Regards sur le Borgou: Pouvoir etaltérité dans une région ouest‐africaine (pp. 155–171). Paris & Montréal:L'Harmattan.

Schottman,W.(2000).BaatonuPersonalNamesfromBirthtoDeath.Africa:JournaloftheInternationalAfricanInstitute,70(1),79‐106.

Schuller, T., Baron, S., & Field, J. (2000). Social Capital: A review and Critique. In S.Baron, J.Field&T.Schuller (Eds.),SocialCapital:Criticalperspectives. London:Oxford.

Scoones,I.(1998).SustainableRuralLivelihoods:AFrameworkforAnalysis(pp.21).IDS Working Paper No. 72: Institute of Development Studies, University ofSussex,Brighton.

Scrimshaw, S.C.M. (1990). Combining quantitative and qualitative methods in thestudyof intra‐household resource allocation. InB.L.Rogers&N.P. Schlossman

REFERENCES

164

(Eds.), Intra‐household resourceallocation: issuesandmethods for developmentpolicyandplanning(86‐98).Tokyo:UnitedNationsUniversityPress.

Sen, K.A. (1990). Gender and cooperative conflicts. In I. Tinker (Ed.), PersistentInequalities: Women and World Development (pp. 123–149). New York andOxford:OxfordUniversityPress.

Seuret, F. (2009). Mali: l'Or blanc transformé en plomb. Alternatives économiques,7/2009,No282,7,72‐77.

Shacklock,G.,&Thorp,L.(2005).LifeHistoryandNarrativeApproaches.InB.Somekh& C. Lewin (Eds.), Research Methods in the Social Sciences (pp. 156–163).London:SAGEPublications.

Shaw, J., & Barrett‐Power, E. (1998). The Effects of Diversity on SmallWork GroupProcessesandPerformance.HumanRelations,51(10),1307‐1325.doi:http://dx.doi.org/10.1023/a:1016942231656

Siaens, C., &Wodon, Q. (2008). Cotton Production, Poverty, and Inequality in RuralBenin:Evidencefromthe1990s.InW.G.Moseley&L.Gray(Eds.),HangingbyaThread:Cotton,Globalization,andPovertyinAfrica(pp.159–176).Athens‐Ohio‐Uppsala:OhioUniversityPress;TheNordicAfricaInstitute.

Sinzogan, A.A.C. (2006). General Introduction. In Facilitating learning towardsustainablecottonpestmanagement inBenin:The interactivedesignofresearchfordevelopment (pp. 1‐13).Wageningen:WageningenUniversity andResearchCenter.

Sinzogan,A.A.C.,Jiggins,J.,Vodouhè,S.,Kossou,D.,Totin,E.,&vanHuis,A.(2007a).Ananalysis of the organizational linkages in the cotton industry in Benin.InternationalJournalofAgriculturalSustainability,5(2&3),213‐231.doi:http://dx.doi.org/10.1080/14735903.2007.9684823

Sinzogan,A.A.C.,Jiggins,J.,Vodouhè,S.,Kossou,D.K.,Totin,E.,&vanHuis,A.(2007b).An analysis of the organizational linkages in the cotton industry in Benin.InternationalJournalOfAgicultualSustainability,5(2&3),19.

Slangen,L.H.G.,vanKooten,G.C.,&Suchánek,P.(2004).Institutions,socialcapitalandagriculturalchangeincentralandeasternEurope.JournalofRuralStudies,20(2),245‐256.

Smith‐Sreen, P. (1996). Accountability to members in grassroots organizations:EvidencefromIndia.AgricultureandHumanValues,13(2),13‐23.doi:http://dx.doi.org/10.1007/bf01540689

Sprague, J.&Kobrynowicz,D. (1999).AFeministEpistemology. In J.S.Chafetz (Ed.),Handbook of the Sociology of Gender (Vol. 11, pp. 25–43). New York: KluwerAcademic/PlenumPublishers.

Sunseri, T. (2001). The Baumwollfrage: Cotton Colonialism in German East Africa.CentralEuropeanHistory,34(1),31‐51.

Svendsen, G.L.H., & Svendsen, G.T. (2004). The creation and destruction of socialcapital:entrepreneurship,co‐operativemovementsand institutions. Cheltenham:EdwardElgarPublishingLimited.

Tama‐Imorou, C., & Wenninck, B. (2007). ACooBéPA's role in pro‐poor serviceprovisioninBenin.InB.Wenninck,E.S.Nederlof&W.Heemskerk(Eds.),Accessof thepoortoagriculturalservices:Theroleof farmers'organizations insocialinclusion. In B. Shapland & J. Kampen (Series Eds.), Development Policy &Practice(Vol.376,pp.139‐153).Amsterdam:RoyalTropicalInstitute.Retrievedfromhttp://www.kit.nl/net/KIT_Publicaties_output/ShowFile2.aspx?e=1333.

Tama‐Imorou, C., Wenninck, B., & Nederlof, E.S. (2007). UCPC’s role in pro‐poorserviceprovisioninBenin.InB.Wenninck,E.S.Nederlof&W.Heemskerk(Eds.),

REFERENCES

165

Acces of the poor to agricultural services:The role of farmers' organizations insocial inclusion (Vol. Bulletin 376, pp. 105‐124). Amsterdam: Royal TropicalInstitute(KIT).

Thornton, J. (1990).PrecolonialAfrican IndustryandtheAtlanticTrade,1500‐1800.AfricanEconomicHistory(19),1‐19.

Tossou, R.C. (1993). Le Groupement Villageois: un cadre de participation commu‐nautaire au développement ou un instrument de réalisation d'intérêtsindividuelsetconflictuels.Bulletinde l'APAD[En ligne],5|1993.RetrievedfromUrl:http://apad.revues.org/3173

Tovignan, D.S., & Nuppenau, E.A. (2004). Adoption of organic cotton in benin: doesgenderplayarole?PaperpresentedatthemeetingonRuralPovertyReductionthroughResearchforDevelopmentandTransformation,Berlin.http://www.tropentag.de/2004/abstracts/full/173.pdf

Tschirly, D. (2009). Introduction. In D. Tschirley, C. Poulton & P. Labaste (Eds.),OrganizationandPerformanceofCottonSectorsinAfrica:LearningfromReformExperience. InWorldBank(SeriesEd.)AgricultureandRuralDevelopment (pp.3‐9). Washington DC: TheWorld Bank. doi: http://dx.doi.org/10.1596/978‐0‐8213‐7770‐3

Tuckman, B.W., & Jensen, M.A.C. (1977). Stages of Small‐Group DevelopmentRevisited.Group&OrganizationManagement,2(4),419‐427.doi:http://dx.doi.org/10.1177/105960117700200404

UNCTAD(2008,July1,2008).MarketInformationintheCommoditesAreaRetrievedJuly4,2008,fromhttp://www.unctad.org/infocomm/index.htm

UNDP(2005).InternationalCooperationataCrossroads:Aid,TradeandSecurityinanUnequal World Human Developement Report. New York: United NationsDevelopmentProgram.

UNDP(2011).SustainabilityandEquity:ABetterFutureforAll.HumanDevelopementReport.NewYork:UnitedNationsDevelopmentProgram.

Wallace,C. (2002).HouseholdStrategies:TheirConceptualRelevanceandAnalyticalScope in Social Research. Sociology, 36(2), 275‐292. doi: http://dx.doi.org/‐10.1177/0038038502036002003

Waller,R.(1997).Review:CottonIstheMotherofPoverty:Peasants,work,andruralstruggleincolonialMozambique,1938‐1961,byAllenIsaacman.AfricanAffairs,96(382),137‐141.

Weik,E.(2011).InstitutionalEntrepreneurshipandAgency.JournalfortheTheoryofSocial Behaviour, 41(4), 466‐481. doi: http://dx.doi.org/10.1111/j.1468‐5914.2011.00467.x

Wellman,B.,&Kenneth,A.F. (2001).NetworkCapital in aMultilevelWorld:GettingSupportfromPersonalCommunities.InN.Lin,K.Cook&S.R.Burt(Eds.),SocialCapital:Theoryandresearch(pp.233–273).NewYork:AldineDeGruyter.

Werhane,H.P.,&Painter‐Morland,M.(2011).Leadership,Gender,andOrganization.In H.P. Werhane & M. Painter‐Morland (Eds.), Leadership, Gender, andOrganization(pp.1–8).LondonandNewYork:Springer.

Woodward, A.R. (2007). The Impact of U.S. Subsidies on West African CottonProduction. Available from Cornell University Food Policy for DevelopingCountries:TheroleofGovernementintheFoodSystemRetrievedJuly8th,2008,fromCornellUniversityhttp://cip.cornell.edu/DPubS?service=Repository&version=1.0&verb=Disseminate&handle=dns.gfs/1200428204&view=body&content‐type=pdf_1#

REFERENCES

166

Woolcock,M.(2001).Theplaceofsocialcapitalinunderstandingsocialandeconomicoutcomes.CanadianJournalofPolicyResearch(2),12‐22.

World Bank (2000). World Developement Report 2000/2001: Attacking Poverty.WashingtonDC:TheWorldBank.

WorldBank(2002).BeninStrategicCountryGenderAssessment(BriefingNoteforthePRSPandtheCAS)(pp.35).WashingtonDC:TheWorldBank.

World Bank (2003). Benin Poverty Assessment Human Development II (pp. 186).WashingtonDC:TheWorldBank.

WorldBank(2007a).WorldDevelopmentReport2008:AgricultureforDevelopment.Retrievedfromhttp://siteresources.worldbank.org/INTWDR2008/Resources/WDR_00_book.pdf.doi:http://dx.doi.org/10.1596/978‐0‐8213‐7233‐3

WorldBank(2012).WorldDevelopmentReport2012:GenderEqualityandDevelop‐ment.Retrievedfromhttp://wdronline.worldbank.org//includes/imp_images/book_pdf/WDR_2012.pdf.doi:http://dx.doi.org/10.1596/978‐0‐8213‐8810‐5

WorldBank (Ed.). (2007b).Agriculture forDevelopment.WashingtonDC:TheWorldBank.

WTO(2005).Benin:DiagnosticTrade IntegrationStudy(DRAFTed.,pp.xxxiv+128).Cotonou:WorldTradeOrganization.

167

Appendix1:Focus‐groupdiscussionguideProvince: Date: District: DiscussionGroup(Men/Women):Village: Surveyor(s):

N° Topics Answers

I. Institutionalmapping(Socialnetworks/organisationsofalltypesoperatinginthevillage)

Networks/Organisations Hierarchy

1st2nd3rdEtc.

II. Meaningofassociationmembership

III. Cottonfarmersorganisationsoperatinginthevillage(rankthemostimportant)

Networks/Organisations Hierarchy

1st2nd3rdEtc.

IV. Possiblelinksbetweenfarmersorganisationsandotherssocialnetworks

V. Importanceofcottonproductioninthevillage

Economic Social Cultural Other

VI. Socialconstructs(songs,proverbs,etc.)aboutcottonandtheymeaning

VII.

Collectiveactionsinducedbycottonproduction(constructionsofcommunityinfrastructures,supplyofpublicservices,etc.)inthevillage

VIII.

Typesofconflictswithinandbetweenfarmersorganisations(leadership,redistributionofresources,etc.)

ConflictsHierarchy(decreasingorder)Numericalimportance

Criticalimportance

1st2nd3rd

Etc.

1st2nd3rdEtc.

IX. Consequencesofconflictsfortheorganisationsandforcottonproduction

APPENDICES

168

X. Causesofandmotivationbehindthecreationofneworganisations(rank)

Causes Motivations

1st2nd3rdEtc.

1st2nd3rdEtc.

XI. Profileofneworganisations’leaders

XII. Factorsdeterminingthedecisiontoadheretoagiven(new)organisation

Factors Hierarchy

1st2nd3rdEtc.

XIII. Reasonsofcottonproductionabandonment

Reasons Hierarchy

1st2nd3rdEtc.

XIV. Consequencesofthemultiplicityoffarmers’organisationsforcollectiveactionatvillagelevel

XV. Income‐generatingactivitiesthatsubstituteforcotton

Activities

HierarchyNumberofpeopleinvolved

Amountofincomegenerated

XVI. Suggestionsofreformactionsfortherevivalofcottonsector

APPENDICES

169

Appendix2:Semi‐structuredquestionnairePreamble(tothesurveyors)

1. Presentyourselfascompleteaspossible;2. Presentthecontextandobjectivesoftheresearchandensuretheintervieweethattheinformationcollectedwillbekeptconfidentialandthattheresearchhasno‐commercialpurpose;

3. Solicit thepermissiontorecordthe interview(try toconvince therespondent incase there isanyresistance);

4. Solicittherespondent’spatiencefortheinterviewtobeconductedtoitsend;5. Invitetherespondenttointroducehim(her)self;6. Thenleadtheinterview.I. SOCIO‐ECONOMICANDDEMOGRAPHICCHARACTERISTICSOFTHERESPONDENTN° Characteristics Code Answers1.1 Name 1.2 Gender 0=Female;1=Male 1.3 Age 1.4 Religion 1.5 Ethnicgroup 1.6 Maritalstatus 1=Married;2=Single

1.7 Levelofinstruction

0=Withoutanyformalinstruction1=Primaryschool[1‐6years]2=Secondaryschool[6‐13years]3=University[+13years]4=Informaleducation(specify)

1.8 Householdsize

1.9

Numberofhouseholdmembersatworkingage

8‐14yearsMale Female

15‐60yearsMale Female

>60yearsMale Female

1.10 Residencestatus1=Borninthevillage2=Migrant

1.11 Ifmigrant,sincehowmanyyears 1.12 Experienceincottonproduction 1.13 Experienceinagriculture11 1.14 Mainincomessource

1.15

Placeofagricultureinincomessources1205‐06 06‐07 07‐08 08‐09 09‐10R % R % R % % R % R

1.16

Placeofcottoninagriculturalincomes05‐06 06‐07 07‐08 08‐09 09‐10R % R % R % % R % R

1.17

Cottonproductionoverthepast5years1305‐06 06‐07 07‐08 08‐09 09‐10A P A P A P A P A P

1.18 Membershipstatusinanetwork1=Excluded;2=Dismissed;3=Stillmember

1.19 Statusofcottonproduction 0=Abandon;1=Stillproducing;2=On

11Consider the time from when the person has his own farm.12

R = Rank; % = contribution. 13

A = Area in ha; P = production in tons.

APPENDICES

170

break1.20 Membershipofapoliticalparty/movement 0=No;1=Yes 1.21 Numberofyearsofpoliticalactivism II. EXPLORINGTHECURRENTNETWORKOFMEMBERSHIP

2.1 Emergenceandgrowth.

N° Characteristics Answers2.1 Nameofthenetwork2.2 Originalnetworkfromwhichitderived2.3 Dateofcreation2.4 Placeofregistration2.5 Registrationnumber2.6 Numberofpromoters2.7 Sizeofthefirstboard2.8 Numberofboardofficialrenewalssinceitscreation2.9 NumberofPresidentssincethecreation

2.10 FarmergroupscompositingthenetworkAtcreationNow

2.11 NumberofnetworkmembersAtcreationNow

2.12 Maximumandminimumsizesreachedbythenetwork

Farmers’groups

Members

Maximum Minimum

2.2 Relationsbetweentheproponenets.

A B C D E FA X B X C X D X E X F X

2.3 Howwouldyounowqualifythecreationofyourcurrentnetwork,spontaneousormotivatedfromoutside?Justify.

2.4 Relationshipsbetweenthefirstboardmembers.

President Secretary Treasurer CoordinatorPresident X Secretary XTreasurer XCoordinator X

APPENDICES

171

III. NETWORKINGDYNAMICS

3.1 Membershipofcottonnetworksandboardpositionsoccupied.

N° CottonNetworkPeriod

(Year/month) PositionsPeriod

(Year/month)N1 N2 N3

3.2 Mappingfarmer’spersonalnetwork(withregardtocottonproduction).

N° Organisations

Membershipwithregardtocottonproduction(1=Yes;0=No)

ResponsibilityBeforetheprime

Duringtheprime

Afterthedecline

A1 A2 A3

3.3 Interactionsbetweentheseassociations/organisationsandthecottonnetworks(0=Nocommunication/norelation;1=Exchangeofinputs;2=Supportfromcottonnetwork;3=Consultationforcollectiveaction).

CottonNetworksOrganisations

Network1 NetworkN2 NetworkN3

A1 A2 A3

3.4 Rankthefollowingcausesofbreakingawayaccordingtotheirimportance.

N° Factors Rank Remarks1 Lackofpayment2 Manipulation(andcorruption)3 Mismanagementofresources4 Lackofcommunication5 Extensionofvillages6 Intra‐groupconflicts7 Imitation

3.5 Rankthefollowingmotives/objectivesinthecreationofbreakawaynetworks.

N° Motives/Objectives Rank Remarks1 Leadership 2 Altruism 3 Selfishness

3.6 HowdidyoubecomePresident/Secretary/Treasurer/Coordinator…?(Forleaders)

3.7 Rankthefollowingpersonalabilities/assetsaccordingtotheirimportanceinhelpingleaderstogetpositionintonetworkboards?

N° Abilities/Assets Rank Remarks1 Levelofinstruction 2 Experienceinapreviousboard 3 Extendofcottonproduction 4 Goodmorality(deservingtrustworthiness&faith) 5 Levelofprosperity

APPENDICES

172

6 Masteryofcottonsector 7 Relationswithoutsiders(inputssuppliers,etc.) 8 Beinganopinionleader(fame) 9 Youthfulness 10 Commitmentandavailability(forgroup) 11 Courage 12 Abilitytonegotiate

3.8 Rankthefollowingsourcesofsupportintheprocessofcreatingbreakawaynetwork.

N° Typesofsupport Rank Remarks1 FamilyandRelatives2 Friends3 Religiousgroups4 Ethnicsolidarity(sharingofculturalvalues)5 Politicalmembership

6 Professionalgroups(Producers,inputssuppliers,ginners,etc.)

7 Personnelfromadministration

3.9 Messages/discoursesofferedtothe“supporters”bytheleaderstoconvincethemofsupportinginthecreationofthenetwork.

3.10 Messages/discoursesdeliveredtocottonfarmerstodecidethemtojointhenetwork.

3.11 Rankthefollowingdeterminantsinfarmers’decisiontoadheretoanetwork.

N°

Determinantfactors Rank Remarks

1 Paymentontime2 Expectationofprofits3 Tastefornovelty4 Trustinnetworkboardmembers5 Kinship6 Sizeofnetworks7 Facilityofgettinginputs8 Hopeofoccupyingaboardposition

3.12 Rankthefollowingreasonsaccordingtotheirimportanceinfarmers’decisiontogiveupcottonproduction.

N°

Reasonsofabandonment Rank Remarks

1 Absenceofpayment2 Declineofcottonprofitability

3 Generalizationofthe“cautionsolidaire”(jointliabilitymechanism)

4 Latesupplyofbadqualityinputs5 Inter‐andintra‐networksconflicts6 Imprisonmentoffarmers

7 Declineofsoilfertilityanditscorollaryoflowyield

8 Incapacitytowork(oldage)

APPENDICES

173

IV. NETWORKCONFLICTS

4.1 Rankthefollowingcausesofconflictswithinandbetweennetworksaccordingtotheirfrequencyandtheacuityoftheirrelated‐conflicts.

N° Typesofconflict Frequency Acuity1 Resourcesdistribution 2 Leadership 3 Mismanagementofresources 4 Cottoncommercialization 5 Absenceofpayment 6 Extensionofthe“cautionsolidaire” 7 Lackofcommitmentatwork 8 Paymentofmembershipfees 9 Outlets(Expressionofotherrivalries)

4.2 Factorsthatmaketheconflictscritical(internal,external,etc.).4.3 Ways/strategiesofsettlementandtheirrelationwiththetypesofconflicts.4.4 Effectsoftheconflictsontherespondent(membershipofnetworks,cottonproduction,etc.).4.5 Effectoftheconflictsoncollectiveaction(communitylevel).

V. INTERACTIONS/RELATIONSBETWEENCOTTONNETWORKS

ANPC FENAPRAFENAPROC

FENAG‐ROP

AGROPBENIN

AGROP‐NOUVELLEVISION

AGROPDEDE

UNAPRO‐B

UPROCOB

ANPC X

FENAPRA X

FENAPROC X

FENAGROP X

AGROPBENIN X

AGROP‐NOUVELLEVISION

X

AGROPDEDE X

UNAPRO‐B X

UPROCOB X

VI. COLLECTIVEACTION

6.1 Typesofactionsrealisedinthevillagefromcottonresources.

N° ActionsRealisation

(0=No;1=Yes)Effects

1Constructionofcommunityinfrastructures

Agriculturalinputsstores SchoolsWaterdigsBridgesandruralroadsHealthcentresLeisurecentresFootballfieldMunicipalbuildingsPubliclighteningKerosenesellingpoints

APPENDICES

174

Ruralradio

2 Payments

Compensationforinfrastructuresconstruction

SalariesofcommunityteachersSupporttostudentsSupporttomunicipalcounsel

3 Maintenanceofinfrastructure 6.2 Waythecollectiveactionaffectedthehouseholdlivelihood.6.3 Continuityofsuchactionsafterthedecline/abandonmentofcottonproduction.6.4 Effectsonthehouseholds’dailylife.6.5 Adaptationstrategiestofacethefreezeofcollectiveaction.

VII. HOUSEHOLDS’LIVELIHOODANDGENDERRELATIONS

7.1 Labourdivisionincottonproduction.

N° TasksExecutedby(Men=1;Women=2;Children=3)

Timeconsumption(Low–Medium–High)

Observations(Otherinterveningworker)

1 Clearance 2 Ploughing 3 Sowing

4 Expandingfertilisers

5 Weeding

6Pesticidetreatment

7 Harvesting 8 Transport 9 Other

7.2 Evolutionhouseholds’incomesources.

N°Sourcesofincomes(Activities)

Performedby

Rank Contribution(%)

Duringtheprimeofcotton

After thedecline/abando

nment

Duringtheprimeofcotton

Afterthedecline/abando

nment1 2 3 4 5

7.3 Households’livelihoodassets.

N° AssetsEvolutionofassetswithregardtocottonproduction

BeforetheboomDuringtheprimeof

cottonAfterthe

decline/abandonment1 Typeofhouse 2 Agriculturalequipment 3 Household’sbasicequipment 4 Luxuriousequipment 5 Transportationmeans

APPENDICES

175

7.4 Evolutionofspouses’contributiontohouseholdexpenditures.

N° Householdneeds

Spouses’Contribution(Low– Medium–High)Duringtheprimeofcotton Afterthedecline/abandonment

Husband Wife(ves) Other Husband Wife(ves) Other1 Education 2 Clothing 3 Healthcare 4 Food 5 Accesstopublicservices

6 Ceremonies(wedding,baptisms,etc.)

7.5 Effectofthedecline/abandonmentofcottonproductiononthehouseholdlivelihood.7.6 Evolutionofgenderrelationswithregardtocottonproduction.

N° DomainsNature/intensityofrelations(Low–Medium– High)

Duringtheprimeofcotton

Afterthedecline/abandonment

1 Intensityofwomen’sextra‐domesticactivities(Timedevoted)

2 Women’saccesstofarmingland3 Autonomyindecisionmaking

7.7 Whichofthefollowingculturesareusedtosubstituteforcottonasproviderofincomes.

N° Cultures Rank Contributiontoincomes(%)

1 Maize 2 Soybean 3 Groundnuts 4 Rice 5 cowpea 6 Sorghum 7 Yam 8 Cassava 9 Sheabutter 10 Cashewnut

7.8 Areruralexodusormigrationanincomeearningmeansthatsubstituteforcotton?7.9 Membersofhouseholdwhomigratedorwentonexodus?7.10 Placeofexodus/migrationandtheprocessofdecisionmaking(whereandwhen?)

APPENDICES

176

VIII. PERSPECTIVESONCOTTONPRODUCTIONANDRELATED‐NETWORKING8.1 Isitstillpossibletomakethecottonsectorrevive?How?8.2 Rankthefollowingreformactionsaccordingtotheirlikelihoodtomakethecottonsector

revive.N° Suggestedreformsactions Rank Remarks1 Paymentontime2 Improvingtheinputsqualityandsupply3 Loweringinputsprices4 Raisingcottonprice5 Re‐nationalisationofthecottonsector6 Restoringtherefunds7 Mergingthenetworksinone8 Suppressingallcottondebts9 Eliminatingthe“cautionsolidaire”(jointliability

mechanism)10 Sharingtherisksbetweenfarmers&inputssuppliers

IX. CHARACTERISTICSOFTHEVILLAGE(Forleaders)N° Characteristics Answer4.1 Province4.2 District4.3 Village4.4 Mainethnicgroupsinthevillage4.5 Agro‐ecologicalzone

4.6 Evolutionofcottonproducersthelast5years

05‐06 06‐07 07‐08 08‐09 09‐10Men Women

4.7Evolutionofcultivatedlandincottonthelast5years

05‐06 06‐07 07‐08 08‐09 09‐10

4.8Evolutionofcottonproductionthelast5years

05‐06 06‐07 07‐08 08‐09 09‐10

4.9 Distancefromthevillagetothemaincity4.10 Accessibilityofthevillage 1=Pavedroad ;2=Ruraltrack ;3=Smalltrack

4.11Numberofcottonnetworksoperatinginthevillage

4.12Numberoffarmers’associations/organisations

4.13 NumberofNGOsoperatinginthevillageDuringcottonprimeAfterthedecline/abandonment

X. WHATDOYOUTHINKABOUTTHEON‐GOINGREFORMTOTRANSFORMCOTTONFARMERS

ORGANISATIONSINTOCOOPERATIVES?

177

SUMMARY

Cotton production in Benin, West Africa, is intertwined with colonialism, whichcontributedtothetransformationofthecrop’sproductionsystemfromtraditionaltomodern.Throughouttheyears,theimportanceofthecropforthestakeholdersvaried.Thelastdecadeshavewitnessedagrowinginterestincottonoffarmers,businessmen,andtheState.Fromhavingamarginalstatusduringtheseventiesandthefirsthalfofthe eighties, cotton grew in importance during the nineties, both in terms of areacoveredandincomegenerated,averaging37percentofthetotalcultivatedareainthecountry. Thus, cotton has a critical cash function and plays a key role in Benin’seconomicgrowth,accountingforanimportantshareintheState’srevenuesandfarmhouseholdsincomes.Indeed,theshareofcottonexportsrepresented75percentofthecountry’stotalagriculturalexportsduringthe2000s,andthecropprovidedupto80percentofruralhouseholdsincomesintheNorth.Thoughcottonisgrownthroughoutthe country, its production was always concentrated in the North, where it isembedded in a farming system formerly dominated by food crops. Hence, cottontransformedsubsistencefarmingintosemi‐subsistencefarming.

Thecentralpositionofthecropinthecountry’seconomy,whichloomedlargeatthebeginning of the 1990s, led to agricultural and economic policies being greatlyinfluencedby the crop fordecades.TheStructuralAdjustmentProgramof theearly1990sprescribedtheliberalisationofthecottonsector,whichhadhugeeffectsonthesector.This resulted inan increased importanceofcotton farmerorganisations thatelapsedintothefirsteverhierarchicalnetworkinthecountry,andthecropbeingputattheforefrontofagriculturaldevelopmentprograms.Enduringbenefitsforfarmers,farmingcommunities,privateactors,andtheStatewerederivedfromthatevolution.Thisgainedcottonthestatusof‘whitegold’.However,theinstitutionaldynamicsthatfollowed in the wake of liberalisation and their corollary of actors’ interactionsgenerated never‐ending conflicts of various kinds, particularly within the cottonfarmers’ networks. These resulted in atomised networks. As a consequence, thebenefits attached to cotton then started to wane and cotton production became adilemmaforfarmers,asreflectedinasteepdeclineofcottonproduction.

This thesis aims at understanding the dynamic interactions between the economicactivityofcottonproductionandthestructureofsocialrelationsfromcommunitytohousehold and individual level. It addresses the question of how farmers’ agencyaffectedtheirorganisations,thecottonsystem,andthecollectiveactionthatevolvedaroundthecrop.Theresearchwasalignedalong threemainaxes: theemergenceofbreakaway networks, the decline of social cohesion and the squeeze of collectiveaction,andthelivelihoodsreconstructionafterthedemiseofcottonproduction.Themain theoretical perspectives underlying the conceptual frameworkwere an actor‐orientedapproach,actor‐networktheory,livelihoodtheory,andagenderperspective.

SUMMARY

178

TheresearchisbasedonfieldworkcarriedoutinfourprovincesintheNorthofBeninfromJanuary2009toApril2011.Beninisacountrywhoseemploymentcapacityandeconomicgrowthheavilyrelyontheagriculturalsector,inwhichcottonisadominantfactor.ThisisstillthecaseforruralareasintheNorth,whereruralhouseholdshavebeen heavily dependent on cotton as a critical cash crop for poverty alleviation.NorthernBeninsuppliesmore than75percentof thecottonyearlyproduced in thecountry, thanks to the favourable agro‐ecological conditions prevailing there, andbecausethereis lesspopulationpressurethaninthesouthernpart.Theexploratoryphase of the research covered four provinces: Borgou, Alibori, Atacora and Donga.SincetheprovincesofBorgouandAliborihosttheheartofthecottonbelt,subsequentdatacollectionprogressivelyfocussedonthesetwoprovinces.

Theresearchadoptedamixed‐methodsdesign,applyingquantitativeandqualitativemethodsofdatacollection.Asurveywascombinedwithfocus‐groupdiscussions,in‐depth interviews and the life history method, to unveil the dynamic interactionsbetweensocialactorsandtheirinteractionswiththematerialandtechnicalelementsofthecottonsystem.Thelifehistorymethodwasusedtodocumenttheexperienceofwomen leaders that had made them exceptions to the rule among women cottonfarmers. Apart from cotton farmers and their leaders, other targets groups of theresearch,likeinputssuppliersandexecutivesofcottonbodies,oftenhadtobefoundbeyond the twoprovinces in other parts of the country. The research covered ninecottonnetworksintenvillagesinthefourprovinces.Surveyinterviewsandin‐depthinterviewswereconductedwith148headsofcottonfarminghouseholds,menaswellaswomen.

About80percentofthefarmersinthesamplewereintheir40sor50s,andmorethanhalfofthemhadnoformaleducation.Educatedwomenrepresentedonly17percentoftheircategory,suggestingthatmalecottonfarmersaresignificantlymoreeducatedthan their female counterparts. The average household sizewas 16, with about 11workersinmaleadultequivalents.Whileagricultureisthemainoccupationandoftentheonlysourceof incomeinthearea,womenturnedouttorely lessonagriculturalincomesthanmen.

Withregardtonetworking,theprocessofatomisationresultedinabout20percentofstayersinremnantnetworks,51percentofjoinersofoperatingnetworks,andabout28percentofcreatorsofnewnetworks.Itwasfoundthatmorethanthreequartersofcotton farmers broke away from their original network at least once during theircotton cropping career, and that creators of new networks weremore likely to beleadersthanstayersor joiners.Theresultsfurthertellusthatmorethanoneintwocotton farmers (ever)hada leadershipposition.A significantassociationwas foundbetween these three categories of farmers and leadership status. Finally, a greaterstockofsocialcapitalwascorrelatedwiththeabilityofleadingcottonnetworks.

The research indicates that the liberalisation of an agricultural value‐chain can beharmfulratherthanbeneficialwhentheStatefailstoplayacoherentroleduringthe

SUMMARY

179

shift from State monopoly to private interest. Cotton proved to be the lifeline forfarmerorganisations, anddrove collective action in rural areas from the importantresourcesitgenerated.However,thedeclineoftrustwithinnetworksinconjunctionwith poor management of cotton resources led to a reversed dynamic that torenetworks apart, which resulted in their atomisation. Social relations deterioratedwhen the financial stakes became higher. As attested by the way the process ofnetwork atomisation evolved, cooperation within large groups requires legalsanctions to be sustainable. The qualitative results showed that the process ofatomisation was nurtured by ties of friendship, kinship, and ethnicity at the start,afterwhichnetworksextendedtoincludeotherareasandmoregeneralmembership.From the survey results it can be inferred that push and pull factors interacted toinfluence the process of cotton network atomisation. Themost influential of thesefactors were, on the one hand, mismanagement of network resources andmanipulation of farmers by outsiders, and, on the other hand, trust in boardmembers,hopeforboardpositions,theexpectationofprofit,andsupportfrompublicofficialsandethnicorreligiousconnections. The research further demonstrates that gender myths and stereotypes obstructwomen'sactiveinvolvementinmanagingorganisations,inspiteoftheirkeypositioninthecottonproductionsystemathouseholdlevel.Womenwerefound21timeslesslikelytobea leaderthanmenincottonorganisations,andtheirpresenceonboardshardlyempoweredthembecausetheyspendtheirenergystrugglingtomeetpracticalneeds.Women’sadmissiontocottonboardsappearstobeinstrumentalformenandhidesmen'srealmotives,judgingbythewaymaleboardmemberstendtorestrictthepoweroftheirfemalecolleagues.However,menareinclinedtogivemorefreedomtowomenwhentheyfindtheiractivitiesbenefittingthemselves,aswasrevealedbythedataonlivelihoodadaptationstrategies.

Theresearchclearlyascertainsthatfarmersaremorerationalthanoftenassumedandthattheygrowacropaslongasitisasourceoflivelihoodandfoodsecurity.Despiteits current low to negative returns, cotton remains part of the livelihooddiversification strategies of households because cotton production gives access toresources that can then be used for food crops. However, relying on one source ofincomeputsthe livelihoodsystemofruralhouseholdsatrisk.Facedwiththecottonproblems,householdsdiversifiedtheirsourcesofincome,firstandprimarilyon‐farmwithfoodcropsincreasinglygainingacashfunction.Additionally,theywoulddeploybeyond‐farm alternative strategies, includingmigration of youth. It was also foundthat thedecline of cottonproductionproved to result inmore freedom forwomen.Because of theirmultiple extra‐domestic activities,women are less vulnerable thanmen when it comes to coping with livelihood shortages. Their contribution to theprovisionforhouseholdneedsincreasedduringthedeclineofcottonproductionandthe ensuing income shortages compared to that of men. The livelihood adaptationstrategiesshowedthedecisionmakingaboutincomediversificationtomovefromthecentreofthehouseholdtoitsperiphery.

181

SAMENVATTING

In Benin, West Afrika, is de katoenproductie verweven met het kolonialisme, datbijdroegaandetransformatievaneentraditionelenaareenmoderneproductiewijzevan katoen. Door de jaren heen veranderde het belang van het gewas voor de be‐trokkenen.Gedurendede laatstedecenniawaser sprakevaneengroeiendebelang‐stelling voor katoenproductie onder boeren, zakenmensen, en nationale overheid.Terwijlkatoeneenmarginalestatushadindejaren70tothalverwegdejaren80,namhet belang van katoen sterk toe in de jaren 90, zowel in termen van areaal als intermen inkomen. Gemiddeldwerd in die tijd op 37 procent van de landbouwgrondkatoenverbouwd.Daarmeespeeldekatoeneencrucialerolinhetgenererenvangeldeneconomischegroei,enhadheteenbelangrijkaandeelinhetnationaalinkomenende inkomens van boeren. In de jaren 2000 bestond 75 procent van de agrarischeexport uit katoen en kwamen75 tot 80 procent vande inkomsten van rurale huis‐houdens in noord Benin uit katoen. Ofschoon katoen door het hele land wordtverbouwd is de productie geconcentreerd in het noorden van Benin. Daar wordtkatoenverbouwdnaastvoedselgewassenenheeftkatoenookdeelsvoedselgewassenverdrongen.

Decentrelepositievanhetgewasindeeconomievanhetlandvanafdejaren90leiddeer toedat landbouwbeleideneconomischbeleiddecennia langdoorkatoenwerdengedomineerd. In het kader van het Structurele Aanpassingsprogramma van beginjaren 90 werd de katoen sector geliberaliseerd. Dit resulteerde in een toenemendbelangvanboerenorganisatiesdieopgingen in eennationaalhierarchischnetwerk ,waarbijkatoenhetspeerpuntwerdvanagrarischeontwikkelinsprogramma’s.Hiervanprofiteerden de boerengemeenschappen, particuliere ondernemers en de Staat.Katoen kreeg de status van ‘wit goud’. De institutionele dynamiek die dit teweegbracht endebelangendie ophet spel stonden, leiddenechter tot talloze conflicten,vooral in de katoenorganisaties. Daarmee viel het hierarchische netwerk uiteen enontstond een versplintering van netwerken. Als gevolg daarvan leverde katoenminder op, liep de productie terug, en werd katoen verbouwen voor boeren eendilemma.

In dit proefschrift wordt de dynamische interactie onderzocht tussen enerzijds deproductieactiviteiten enhet economischbelang van katoen en anderzijds de socialerelaties van de betrokken actoren, op gemeenschaps‐, huishoud‐ en individueelniveau.Devraagdiewordtgesteldishoedekeuzesenactiesvanboerenvaninvloedwarenophunorganisaties,hetkatoenproductiesysteemendecollectieveactiesdiedaaruit voortkwamen. Het onderzoek werd gestructureerd langs drie assen: hetopkomenvan splinternetwerken,deafnamevan sociale cohesieen collectieveactie,en de aanpassingen in de wijze waarop de huishoudens van katoenboeren in hunlevensonderhoudgingenvoorziennadatdekatoenproductie steedsproblematischerwerd. De belangrijkste theoretische perspectieven die aan het onderzoek ten

SAMENVATTING

182

grondslag lagen zijn een actor‐georiënteerde benadering, actor‐netwerk theorie,livelihoodtheorieeneengenderperspectief.

Het onderzoek is gebaseerd op veldwerk datwerd uitgevoerd in vier provincies inNoordBeninindeperiodejanuari2009‐april2011.InBeninzijnwerkgelegenheideneconomischegroeizeerafhankelijkvandeagrarischesectorendaariniskatoeneendominantefactor.DitiszekerhetgevalinhetNoorden,waarruralehuishoudensvankatoenafhankelijkzijnvoorhungeldinkomstenenomarmoedeopafstandtehouden.Dankzijdevoorkatoengunstigeagro‐ecologischeomstandighedenwordt75procentvan de katoen in dit deel van het land geproduceerd. In de exploratieve fasewerdonderzoekgedaanindevierprovinciesBorgou,Alibori,AtacoraenDonga.Ineenlaterstadiumwerd het onderzoek geconcentreerd inBorgou enAlibora.Deze provinciesvormenhethartvanhetkatoengebied.

In het onderzoekwerd eenmix van kwantitatieve zowel als kwalitatievemethodentoegepast. De survey werd gecombineerd met focus groep discussies, diepteinterviewsenhetdocumenterenvanlevensgeschiedenissen,omdedynamiekvandeinteractie tussen de actoren onderling en die van hun omgangmet demateriële entechnische aspecten van het katoensysteem te kunnen blootleggen. Delevensgeschiedenisvaneenaantalvrouwelijkeleidersindekatoenorganisatieswerdopgetekend om te begrijpoen hoe zij hun leiderschapspositie hadden verworven enweerkwijtraakten.Naastkatoenboerenenhun leiderswerdenandereactorenzoalsopkopers en staf van katoenbedrijven ondervraagd voorwat betreft hun rol in hetkatoensysteem.Dezemensenwonenvaakvervanhetgebiedwaardekatoenvandaankomt en werden daar thuis bezocht. Acht netwerken in tien dorpen in de vierprovincies werden geanalyseerd. De survey interviews werden gedaan met 148hoofdenvanhuishoudensvankatoenboeren,zowelmannenalsvrouwen.

Ongeveer80procentvanderespondentenwastussende40en60jaaroud.Ietsmeerdandehelfthadgeenformelescholinggehad.Voordevrouwenindesteekproefwasdit percentage 83, waaruit blijkt datmannelijke katoenboeren beter geschoold zijndanhun vrouwelijke collega’s.De gemiddeldehuishoudgroottewas16,waarvan11volwassen, werkende personen. Terwijl landbouw het belangrijkste middel vanbestaanis,blekenvrouwenvakerinkomstenuitniet‐agrarischeactiviteitentehebbendanmannen.

Uit de analyse van het proces van de versplintering van netwerken bleek dat eenvijfde van de ondervraagde boeren waren gebleven bij de resten van eenuiteengevallen netwerk, 51 procent hadden zich aangesloten bij opererendenetwerken,enderestbestonduitboerendieeennieuwnetwerkhaddengecreëerd.Het bleek dat meer dan driekwart van de katoenboeren tenminste een keer hetnetwerk waar ze oorspronkelijk toe behoorden, had verlaten. Degenen die nieuwenetwerken hadden opgericht waren vooral boeren met leidersposities., waarbijaangetekendmoetwordendatmeerdanéénoptweeboereneenleiderspositiehadof

SAMENVATTING

183

ooit had gehad. Leiders bleken te beschikken over meer sociaal kapitaal dan niet‐leiders.

Het onderzoek wijst uit dat liberalisering in de landbouw eerder schadelijk dangunstiguitpaktalsdeStaatverzuimteencoherentesturenderoltespelentijdensdetransitie van staat monoplie naar vrije markt. Katoen was de drijvende kracht inboerenorganisatiesendemotorachtercollectieveactievooralgemeennutindeuralegebieden.Hetverlorengaanvanhetvertrouwentussendeledenvandenetwerkenenhetslechtebeheervandekatoenopbrengstenleiddenechtertotuiteenvallenvandeorganisaties. Dit laat zien dat samenwerking binnen grote groepen waarbij grotebelangenophetspelstaannietzonderwettelijkesanctiesenreguleringkan.Uithetkwalitatieve onderzoek bleek dat factoren zoals vriendschaps‐ en verwantschaps‐relaties, woonplaats en etniciteit een belangrijke rol speelden in het proces vanversplintering en hergroepering van netwerken. Van de factoren die boeren dedenbesluiteneennetwerkteverlatenblekenslecht leiderschap,slechtgeldbeheerendemanipulatievanboerendoorbuitenstaandershetmeestbepalend.Indeaantrekkings‐kracht van andere netwerken bleken het vertrouwen in de leiders, de hoop op eenleiderschapspositie,deverwachtingvanwinsten,desteunvanambtenarenenpolitici,enetnischeenreligieuzeconnectiesdedominantefactoren.

Het onderzoek laat ook zien dat gender mythen en stereotypen een active rol vanvrouwenindekatoenorganisatiesindewegstaan,ondankshunbelanrijkerolinhetproductieproces.Vrouwenhadden81keerminderkansopeenbestuursfunctiedanmannen en als ze een bestuursfunctie hadden was het een probleem om die teverenigen met hun andere verplichtingen. Het lijkt erop dat vrouwen met eenbestuursfunctie worden gebruikt door hun mannelijke collega’s die er tevens voorzorgendatzenietteveelmachtkrijgen.Oppersoonlijkniveauzijnmannenechterwelgeneigdhunvrouwruimtevooreconomischeactiviteitenbuitenshuistegevenalszedaardevoordelenvaninzien.

Uithetonderzoekblijktduidelijkdatboerenrationelerzijndanveelalwordtaange‐nomen; ze verbouwen een gewas zo lang het een goede bron van bestaans‐ envoedselzekerheid is. Ondanks de crisis in de katoen, wordt het gewas nog veelverbouwd,vooralomdatboerenviakatoentoegangtotkunstmesthebben.Diekunnenzedangebruikenvooranderegewassenalsdiemeeropleveren.Hieraanzijnechterrisico’sverbonden.Daarnaastvindenandereaanpassingenplaats,zoalsmigratievanjongerenenmeereconomischeactiviteitenvanvrouwen.Hunbijdrageaanhethuis‐houdinkomenbleekgestegen invergelijkingtotdie tijdensdehoogconjunctuurvande katoen. De besluitvorming over aanpassingsstrategieën in huishoudens lijkt alsgevolgvanderecenteontwikkelingenminderbijhethoofdenmeerbijandereledenvandehuishoudingteliggen.Methetverdwijnenvandecentralerolvankatoenlijkterophetmicroniveauvanhethuishoudenooksprake tezijnvaneenverschuivingvandemachtvanhetcentrumnaardeperiferie.

185

AbouttheAuthorGuirguissouMaboudou Alidouwas born in Perere, Benin. In 2001, he obtained hisMaster of Arts degree in Sociology and Anthropologywith amajor in OrganisationStudies from the National University of Benin (UNB), now University of Abomey‐Calavi(UAC).In2003,heobtainedaMasterofScienceinSociologyattheUniversityofLomé, Togo, for which he had received a Research Graduate Fellowship from theConsultativeGrouponInternationalAgriculturalResearch(CGIAR).Heconductedhisresearch for this master degree on technology adoption and diffusion at theInternationalInstituteofTropicalAgriculture(IITA)inaprojectfundedbytheDanishInternational Development Agency (DANIDA). In 2008, Guirguissouwas awarded afellowship by the Netherlands University Foundation for International Cooperation(NUFFIC) to do a PhD about cotton networking in Benin at the Social SciencesDepartment ofWageningenUniversity in theNetherlands. The research focused ontheinteractionsbetweentheeconomicactivityofcottonproductionandthestructureofsocialrelationsinvolvedinthecottonsystem,withgenderasacross‐cuttingissue.

Guirguissou Maboudou Alidou has been successively research assistant andresearcher at the Agricultural Policy Analysis Unit of the National Institute ofAgricultural Research of Benin (PAPA/INRAB), where he works on technologyadoptionanddiffusion,and impactassessment.His research interests includesocialcapital,organisationstudies,participatoryagriculturalinnovationsdevelopment,andgenderstudies.

187

GuirguissouMaboudouAlidouWageningenSchoolofSocialSciences(WASS)CompletedTrainingandSupervisionPlan

Nameofthecourse Department/Institute

Year ECTS(=28hrs)

GeneralpartFieldResearchMethods:MethodsandToolsforQualitativeData MG3S 2008 2.3

InformationLiteracy,includingIntroductionEndnote WGS 2008 0.6

TheoryandToolsforNarrativeInquiry MG3S/CERES 2008 1.4

EffectiveBehaviorinyourProfessionalSurroundings WGS 2009 0.7

ResearchMethodology:Fromtopictoproposal MG3S 2009 4

ProjectandTimeManagement WGS 2010 1.5DoingInterpretativeAnalysis MG3S 2010 3

TechniquesforWritingandPresentingScientificPapers WGS 2011 1.2

Reviewingascientificpaper WGS 2011 0.1Mansholt‐specificpartMansholtIntroductioncourse MG3S 2008 1.5

TheoreticalandMethodologicalApproachestotheStudyofLocalPoliticsinDevelopingCountries(PhDWorkshop)

Roskildeuniversity(Denmark)

2010 4

ExecutiveCourseonSustainableDevelopmentDiplomacy(SDD)/InternationalProgrammeonthemanagementofSustainability(IPMS)

WUR/SDF/TUFTS 2011 2

‘CottonProductioninBenin:Fromwhitegoldtopest’ PosterattheNVAS,ASC

2011 1

‘GenderRoleinCottonProductioninBenin’ WASSPhDDay 2012 1

‘GeneratingIncomeandConflict:ProfitabilityandSocialCohesioninBenin'sCottonproduction’

7thInt.Conf.onInterdisciplinarySocialSciences,Barcelona

2012 1

Discipline‐specificpartSocialNetworksAnalysis ECPRSummer

SchoolLjubljana2008 5

RuralGenderStudiesTutorialPhD SCH 2008 2

SociologicalTheoriesofRuralTransformation(RDS30306) RDS 2008–2009

6

Socialcapital(SCH51306) SCH 2009 6

Total 44.3

Printedby:WöhrmannPrintService,Zutphen,theNetherlands