why do cooperatives emerge in a world dominated by ... · research into why and where cooperatives...

WHY DO COOPERATIVES EMERGE IN A WORLD DOMINATEDBY CORPORATIONS? THE DIFFUSION OF COOPERATIVES IN

THE U.S. BIO-ETHANOL INDUSTRY, 1978–2013

CHRISTOPHE BOONEUniversity of Antwerp, Belgium

SERDEN ÖZCANWHU – Otto Beisheim School of Management, Germany

Given the strong economic disincentives that exist, why do cooperatives continue toemerge? And why is it that, in some communities, these cooperatives are collocatedwith their commercial counterparts, yet, in others, territorial partitioning occurs? Inthis paper, we develop a community ecology approach that integrates economic andsociological accounts of cooperatives, in an attempt to reconcile these contradictoryobservations. Using a detailed panel data set for the county-level founding process ofcooperatives in the U.S. ethanol industry from 1978 to 2013, consistent with rationaleconomic arguments, we find that the founding rate of cooperatives decreases in thepresence of high, local, corporate ethanol production capacity. However, this negativecompetitive interdependence is attenuated in local communities where: (1) corpora-tions represent a potential threat to the autonomy of local farmers, (2) there is agenerally anti-corporate climate, and (3) there is a well-established organizationalinfrastructure supporting a cooperative ideology. Consistent with sociological theoriesthat emphasize the mobilizing force of ideology, these local conditions spur andfacilitate collective action among farmers to establish cooperatives in response to thelocal diffusion of corporations. We show further that the diffusion of plants owned bybig business (oil and agribusiness) in communities characterized by a general anti-corporate climate especially promotes greater ideological contestation and the mobi-lization of resources to form cooperatives.

The continuing attention to the cooperative as analternative organizational form over the past cen-tury stands in sharp contrast to the fact that theoverall number of cooperatives relative to the gen-eral firm population has always been small (Al-drich & Stern, 1983; Schneiberg, King, & Smith,2008). This somewhat contradictory situation pro-

vokes the question of whether it is true that “orga-nizing modes that are viable only among smallcommunities of highly motivated members arescarcely interesting for purposes of organizing eco-nomic activity in society at large” (Williamson,1980: 33). This question is especially relevantsince, time and time again, cooperatives seem toemerge in waves, within industries such as insur-ance, dairy, grains, and electricity (Schneiberg,2007; Schneiberg et al., 2008).

A similar trend has been observed in the U.S.bio-ethanol industry. This industry expanded rap-idly based, initially, on the entry of corporateplants in the late 1970s, which subsequently wascountered by the entry of cooperative ethanol pro-ducers (see Figure 1). Cooperative plants currentlyaccount for close to 40% of ethanol production,and are scattered over many U.S. counties andstates. This trend is especially remarkable giventhat—compared to corporations—cooperatives aresubject to strong economic disincentives, espe-

Christophe Boone gratefully acknowledges financialsupport from the Flemish Science Foundation’s (FWO)Odysseus program. We are very grateful to the editor andthree anonymous reviewers for their excellent commentsand suggestions. We appreciate insightful commentsfrom Glenn Carroll, Henrich Greve, Michael Hannan,Henry Hansmann, Klaus Heine, Özgecan Koçak, LimChin, Ivan Png, Jens Prüfer, Toke Reichstein, MilesShaver, Tal Simons, Bernard Yeung, seminar partici-pants at the National University of Singapore, and par-ticipants in the second European Organizational Ecologyworkshop and of the Tilburg Law and Economics Center(TILEC) workshop on Economic Governance and Organ-izations held at Tilburg University.

990

� Academy of Management Journal2014, Vol. 57, No. 4, 990–1012.http://dx.doi.org/10.5465/amj.2012.0194

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

cially in capital-intensive industries (Aldrich &Stern, 1983; Hansmann, 1996; Williamson, 1980,1985). While, in some states, cooperatives and cor-porate plants exist side by side, in others, there hasbeen a territorial partitioning that has resulted in cor-porate or cooperative plants dominating the scene.

This paradoxical situation calls for systematicresearch into why and where cooperatives emerge.These questions are important because they informus about the origins of institutional diversity(Greve, Pozner, & Rao, 2006; Ingram & Rao, 2004;Schneiberg et al., 2008). Systematic empirical re-search on this topic is scarce, and different strandsof the literature propose different explanations.Economists stress the relative inefficiency of thecooperative form, based on its lack of hierarchy andperformance-monitoring problems (Hansmann,1996; Williamson, 1985). Cooperatives are seen astransient compromises that emerge out of neces-sity, as a response to economic recessions, industryrestructurings, or market failures (Ben-Ner, 1984;Caves & Petersen, 1986). Sociological accounts em-phasize the collective entrepreneurship nature ofthe cooperative, and the importance of anti-corpo-rate social movements that contest mainstream cap-italist ideology (Schneiberg et al., 2008; Sine & Lee,2009). They suggest that cooperatives are deeplyrooted in the institutional repertoires of local com-munities and co-evolve with liberal market econo-mies (Greve & Rao, 2012; Schneiberg, 2007).

However, neither the economic nor the sociolog-ical arguments on their own are able to account forthe “mass of contradictions” (Aldrich & Stern,1983: 372) that characterize the literature on coop-eratives. The economic-based view that (especiallyproducer) cooperatives are “impossible organiza-tions” (Estrin & Jones, 1992: 328) is not easily rec-onciled with the rapid diffusion of the form and itstemporal stability in, for example, the U.S. ethanolindustry. It also does not explain why, evenacross similar resource environments, coopera-tive founding activity can be uneven. For in-stance, in corn-belt states such as Illinois andOhio, cooperative ethanol plants are virtually ab-sent, while, in Minnesota and South Dakota, theyoutnumber commercial plants. Economic motiva-tions matter for the collective creation of organi-zations, but so do non-pecuniary considerations.For instance, several farmers committed to grass-fed food and dairy operations despite the stronglikelihood of financial ruin; their commitmentthus became an expression of moral identity thatwent beyond economic concerns (Weber, Heinze,& DeSoucey, 2008).

On the other hand, sociological accounts gener-ally present a portrait of cooperative founderswhose decisions are affectively charged, guided notby economic rationale but by emotion, and tend tooverlook that the cooperative is costly to organize

Figure 1Density of Cooperative and Corporate Ethanol Plants in the US, 1978–2013

2014 991Boone and Özcan

and manage.1 Since there are several alternativesfor disciplining corporate behavior, ranging fromprotests and boycotting (Bell, 2009; Hiatt, Sine, &Tolbert, 2009; Ingram & Rao, 2004) to influencingstate power and regulation (Sine & Lee, 2009), thecooperative may not be an economically attractiveorganizational solution even for communities withanti-corporate sentiments. Indeed, Schneiberg et al.(2008) called for research on how social move-ments can moderate the effects of microeconomicconditions on the birth of cooperatives.

We argue that both the economic and sociologi-cal research traditions need to be integrated in or-der to understand the process through which coop-erative forms emerge. In this paper, we draw oncommunity ecology, which focuses on systematicstudy of the interdependency between different or-ganizational forms (Barnett & Carroll, 1987; Free-man & Audia, 2006). Since cooperatives competedirectly with corporations for resources, such anapproach is particularly well suited to understand-ing their respective diffusion processes as well astheir co-evolution. We propose that the nature ofthe interdependence between cooperatives and cor-porations depends heavily on the local communitycontext in which the actors are embedded, becausecooperative founding requires effective local, col-lective action (Greve & Rao, 2012).

We start from economic theory, which predictsthat rational agents will only establish cooperativesout of necessity when corporations are absent.However, we argue that this rational economiclogic (based on “cold cognition,” DiMaggio, 2002)will only dominate until specific local conditionsspur contestation between oppositional identities,fueling affectively laden “hot cognition” (DiMag-gio, 2002; Sine & Lee, 2009). Under such condi-tions, anti-corporate sentiments provide forcefulideological motives, which are essential for over-coming the economic disincentives for cooperativeentrepreneurship, and galvanizing local collectiveaction to marshal resources against the diffusion ofcorporations (Aldrich & Stern, 1983).

Based on social movement theory, we predictthat strong local anti-corporate sentiments will betriggered in communities where: (1) corporationsrepresent a potential threat to community auton-

omy and welfare, (2) there is a general anti-corpo-rate climate, or (3) there is a well-established or-ganizational infrastructure supporting cooperativeideology. In these contexts, the negative interde-pendence between corporations and cooperativespredicted by economic theory will be attenuated,and might even be reversed. Since not all corporatebusinesses are perceived equally as the “enemy,”we differentiate between the targets of antagonism,studying whether certain types of corporations (i.e.,big business vs. independent owners) are evenmore likely to transform local actors into a socialmovement that produces a cooperative.

We make important contributions to the litera-ture. By fully exploiting geographical variation incommunity characteristics and heterogeneity interms of entrepreneurial origins, we extend eco-nomic and sociological theories on cooperativeform emergence by delineating and refining theirboundary conditions. We contribute also to the de-velopment of community ecology by reintroducingthe “local community” into the theory of organiza-tional founding processes. Although early ecologi-cal work includes the spatial dimension in its con-cept of “community” (Barnett & Carroll, 1987), overtime, community has come to be seen as predomi-nantly a set of functional relations between organ-izational forms. Nonetheless, the social structures,ideologies, and social identities that characterize aspatial context are decisive in the birth and diffu-sion of organizational forms (Freeman & Audia,2006). Furthermore, much past research in this tra-dition focuses on the diffused and indirect rivalrybetween corporations and cooperatives, showing,for instance, that the number of commercial bankssuppresses the founding rate of worker coopera-tives with clear socialist identities (Ingram & Si-mons, 2000). The lack of research in settings wherecooperatives and corporations directly compete forresources within the same industry is a significantgap, given that ideological struggles, coalescing indirect competition for resources, are the main driv-ers of social, institutional, and political change(Barnett & Woywode, 2004; Haveman, Rao, & Paru-churi, 2007; Simons & Ingram, 2004; Weber etal., 2008).

We test our hypotheses on a detailed panel dataset for the county-level founding process of coop-erative plants in the U.S. ethanol industry fromJanuary 1978 to June 2013. This industry is an idealsetting to study form emergence because the ideo-logical rivalry between cooperative and corporateplants goes hand in hand with cutthroat competi-

1 This is partly because research that connects socialmovements to entrepreneurship primarily examines thecreation of firms (which are less complex to organize andmanage) as an outcome.

992 AugustAcademy of Management Journal

tion in local input (i.e., corn) and output (i.e., eth-anol) markets. Also, because the cooperative found-ing process is highly locally embedded, we canexploit community differences in order to under-stand the spread of cooperatives—not only overtime, but also in geographical space.

ETHANOL INDUSTRY: A HISTORY OFCONTESTATION

The history of the ethanol2 industry is character-ized by contestation and cycles of emergence anddeath. Each rebirth was fueled by social move-ments that mobilized economic and sociopoliticalresources. Ethanol activists pursued a reformiststrategy of market creation, as in the case of greenenergy (Sine, Haveman, & Tolbert, 2005) and grass-fed meat and dairy products (Weber et al., 2008).The values propagated (e.g., self-reliance, agrarian-ism) were progressive. The ethanol insurgencyquickly generated a countermovement from theincumbent oil industry. The pro- and anti-ethanolcoalitions pursued differing sets of institutionalchanges, and only since the late 1970s did thiscollective action lead to the entry of entrepreneursand plants dedicated to the production of ethanol,and, subsequently, to the birth of ethanol producercooperatives.

From the Origins of Ethanol Production to theModern U.S. Ethanol Industry

Ethanol appeared as a transportation fuel at theturn of the last century, but never caught on since itwas produced mainly at alcohol distilleries andProhibition squelched its production by shuttingdown the distilleries or capping production per-mits. The repeal of Prohibition in 1933 generated arebound, but efforts to promote ethanol failed. By1939, commercial ethanol production had ceasedand did not resume until 1978.

The first oil crisis of 1973 was a turning point forthis forgotten fuel. The rapid doubling in oil pricesturned the spotlight on ethanol. The state of Ne-braska took the lead, launching a multi-year etha-nol road test to mobilize the attention of the publicand policymakers. The test confirmed that mostvehicles could run on ethanol-blended gasoline.Subsequently, farm groups in the Midwest began to

lobby their state legislatures for subsidies to pro-duce this “home-grown fuel.” The breakthroughwas rapid: in 1978, Congress passed the NationalEnergy Tax Act to reduce the excise tax on ethanol-blended gasoline. This made ethanol competitivewith gasoline, triggering production and markingthe official beginning of the modern ethanol indus-try. The second oil price shock in 1979 increasedthe appeal of ethanol. The number of stations sell-ing it jumped from 100 in 1978 to 9,000 in 1980.Ethanol corporations multiplied and productioncapacity increased from 80 million gallons per year(mgy) in 1979 to 15 billion gallons in 2013. Today,ethanol is produced in 23 states, and 92% of allgasoline sold is blended with ethanol.

The Rise of Ethanol Cooperatives in an EmergentMarket Dominated by Corporations

Subsequently, the number of cooperatives alsobegan to rise. Farmer cooperatives espouse ideolog-ical values that run counter to the values of thedominant capitalist corporations (Schneiberg et al.,2008; Simons & Ingram, 2004). These cooperativesconstitute a distinct organizational form (Hans-mann, 1996; Schneiberg et al., 2008) the identity ofwhich is defined by three basic principles. They areorganized and owned by farmers, the members ex-ert democratic authority over primary policy deci-sions, and the members receive benefits on thebasis of use.

The cooperative form in U.S. agriculture is as oldas the nation itself. The widespread diffusion offarmer cooperatives that started at the end of the19th century was rooted in social movements suchas the Grange, which rejected corporate liberalideas of national markets and corporations. Theseanti-corporate movements stressed the values ofcommunity autonomy, social solidarity, and self-reliance, and were based on the principle of devel-opment through cooperation (Schneiberg, 2007;Schneiberg et al., 2008). More than a century later,these values still underpin the cooperative land-scape. Recent research shows that production co-operatives display a more collectivist identity thando their private counterparts (Brickson, 2005), andthat members prefer farmer cooperatives with alimited business focus (Foreman & Whetten, 2002).

These ideals were also the defining features ofthe ethanol movement. From its beginnings, therewas a close marriage between the promise of etha-nol and the ideological tenets of farmer coopera-tives. Soaring gas prices had produced a general

2 Ethanol is a fuel additive distilled predominantlyfrom corn.


feeling of helplessness among the public and strongresentment towards oil corporations. Farmer coop-eratives felt obliged to step in and spearhead amovement. Farming was not only about “feedingthe nation,” it was also about “growing towardsindependence,” and the cooperatives strove to cre-ate demand for ethanol. A similar ethos under-pinned the establishment of cooperative ethanolplants. These plants appealed to farmers and alsoresonated with nonfarmers in the same communi-ties, which boosted the social movement characterof their formation. The founders of ethanol cooper-atives not uncommonly found that almost all thelocal population was keen to invest in their plantsand to show their support for agriculture, energyindependence, and the creation of local jobs. Localmedia provided ongoing supportive coverage. Lo-cal businesses endorsed the cooperatives, and localcommunity colleges offered free training for coop-erative employees (Thompson, 2004).

THEORY AND HYPOTHESES

Community ecologists study how the prevalenceof one organizational population affects the vitalrates of another, as revealed by so-called “compe-tition coefficients” (Hannan & Freeman, 1977). Theinterdependence between two different forms issaid to be competitive if these coefficients are neg-ative, and mutualistic if they are positive (Barnett &Carroll, 1987; Freeman & Audia, 2006). Ecologicalwork differentiates between two types of mutual-ism. Direct mutualism occurs when organizationswork together concertedly to benefit from coopera-tion; for instance, to increase common resources orto enhance institutional standing (Barnett & Car-roll, 1987). Diffuse mutualism occurs without anycoordination of efforts. These benefits may accruefrom the adaptive behavior of organizations sharingthe same environment resulting in spillovers andopportunities to free-ride (Barnett & Carroll, 1987;Baum & Oliver, 1996). They can also arise whenorganizations provoke oppositional identities and,as a result, increase the carrying capacity of anotherform/subpopulation by unintentionally facilitatingcollective mobilization for it—by uniting oppo-nents into a social movement and serving as narra-tives for their frame. For instance, consumers whowant an alternative to for-profit day-care servicescan be stimulated by the prevalence of the latter tocreate their own non-profit day-care organizations(Baum & Oliver, 1996). It is to this form of diffusemutualism that we refer in the present study.

We argue that the nature of the cross-form effectbetween corporations and cooperatives is not con-stant but depends on specific local conditions (In-gram & Yue, 2008). Due to the strong disincentivesfor starting a cooperative, we expect the baselinecross-form effect, grounded in the “cold cognition”of rational economic arguments, will be negative.That is, cooperatives are generally born of neces-sity, and rational farmers only establish coopera-tives if ethanol corporations are absent. In contrast,in local communities with strong anti-corporatesentiments, the diffusion of corporations triggersideological rivalry, providing the necessary ideo-logical incentives grounded in affective-based “hotcognition,” to overcome the strong economic disin-centives for cooperative ownership (Aldrich &Stern, 1983). The negative cross-form effect will,accordingly, be attenuated and will become lessnegative, or even positive; that is, mutualistic.3

Necessity Entrepreneurship

There are major economic disincentives associ-ated with the formation of cooperatives. First,transaction cost economists claim that their inter-nal governance systems are less efficient than thoseof corporations. They are examples of the “peergroup” governance mode, while corporations’ gov-ernance systems exemplify the “hierarchy” mode.Transaction costs are assumed to be relativelyhigher in “peer groups,” because implementing hi-erarchical leadership and monitoring performanceare more difficult within groups of peers. In “hier-archies,” leadership and control are implementedby fiat; therefore, organizational problems associ-ated with opportunism and information asymmetryare more easily resolved (Williamson, 1980, 1985).The costs of collective decision-making in cooper-atives are particularly high if members’ interestsare heterogeneous; this variety undermines effi-ciency and spurs free-riding behavior among mem-bers (Hansmann, 1996; Staatz, 1987).

Second, common ownership in cooperatives to-gether with the limited possibility of cooperativeowners to sell their shares at a market price lead toa fragile financial system (Ben-Ner, 1984; Bonin,

3 As there are no clear theoretical arguments to predictthat local anti-corporate sentiments could be strongenough to completely invert this main effect from nega-tive to positive, we hypothesize monotonic interactioneffects in the remainder of the paper.


Jones, & Putterman, 1993; Estrin & Jones, 1992). Inparticular, it is difficult to justify to members thatthey should keep on investing in the cooperative.On the one hand, there will inevitably be differentinvestment horizons among the members. Thosewith shorter horizons will tend to refrain from com-mitting further, whereas those with longer horizonswill insist on not only making additional invest-ments, but also reinvesting the returns in the coop-erative. On the other hand, upon leaving a cooper-ative, the former member does no longer haveaccess to the accumulated assets. At the same time,new members have immediate access to all accu-mulated assets, creating a form of free-riding (Caves& Petersen, 1986; Cook, 1995).

Third, Aldrich and Stern (1983) link these disin-centives to the rise of modern monopoly capital-ism, which makes it easier for entrepreneurs toestablish corporations than cooperatives. The col-lective nature of the latter requires greater entrepre-neurial effort for their establishment, compared to atraditional firm. In addition, the quasi-rents fromthis entrepreneurial effort are more difficult to ap-propriate, which explains why cooperatives aregenerally assumed to suffer from underinvestment(Jones, 1979) and often “degenerate” and revert toconventional forms of ownership with the passageof time (Aldrich & Stern, 1983; Ben-Ner, 1984;Bonin et al., 1993).

Fourth, many economic actors prefer not to bepart of a cooperative simply because they valueautonomy, even when joining could reduce theirperformance hazard. Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson, and Moyano-Fuentes (2007), forinstance, find that family-owned olive oil mills insouthern Spain were reluctant to become membersof a cooperative due to a desire to preserve thefamily’s socio-emotional wealth (including its in-dependence), despite the bigger business risk asso-ciated with this choice.

The lack of material incentives would suggestthat rational economic actors establish coopera-tives only out of necessity. Several authors provideevidence that the collective entrepreneurial energyof a cooperative has survival–defensive roots andresponds to economic downturns, industry restruc-turings, or the negative economic impacts of marketfailure (Caves & Petersen, 1986; Cook, 1995; Sta-atz, 1987).

What do these arguments imply for our frame-work? Obviously, corn farmers benefit greatly fromthe fast-growing ethanol market in the UnitedStates, since the presence of local ethanol plants

guarantees steady demand for their produce. Etha-nol producers are the major buyers of local corn,and, in 2012, nearly half of the corn harvested inthe United States went to ethanol production.Thus, the presence of corporate plants producinglarge volumes of ethanol in the local communitymeans demand for local corn will be high, andthere will be no (economic) need for farmers toestablish a cooperative.4 In addition, given the al-leged efficiency advantages of corporations, farm-ers could not hope to gain much from integratingforward into ethanol production by founding a co-operative. In contrast, if there is an absence of cor-porate plants in the local community and/or etha-nol production is low relative to local cornproduction, establishing cooperative ethanol plantswill be necessary for farmers to secure a steadyoutlet for their production. Therefore:

Hypothesis 1. The lower the local corporateethanol production capacity, the greater thelikelihood that ethanol cooperatives will befounded.

Local Conditions that Spur CollectiveEntrepreneurship

We predict that the negative baseline cross-formeffect described in Hypothesis 1 will be attenuated incommunities where there is a strongly negative atti-tude towards corporations. In these circumstances,people commit to cooperation as an ideal and “mightbe willing to ignore the obvious material disincen-tives involved in creating cooperatives and insteadprovide the resources required” (Aldrich & Stern,1983: 387). We expect affective processes to fuel en-trepreneurial decision-making (“hot cognition”) ifcorporations represent a potential threat, or if localcommunities have an anti-corporate climate, or thereis an existing organizational infrastructure supportingcooperative ideology.

Threat of vertical integration by ethanol corpo-rations into farming. Extensive involvement ofcorporate ideology in agriculture is more likely tomobilize farmers to organize ethanol cooperativesas a reaction against commercial ethanol plants.

4 McNew and Griffith (2005) estimated the impact onlocal grain prices of ethanol plants that opened between2001 and 2002. They found that corn prices increased by12.5 cents per bushel at the plant site, and that positiveprice responses were observed up to 68 miles fromthe plant.


This is because corporate farming poses a compet-itive threat, and violates a number of the ideologi-cal beliefs of independent farmers. An invasion onfarmers’ primary activity challenges their funda-mental identity, which is built upon a family-basedproduction system and the values of empower-ment, engagement, and autonomy (Thompson,2004). The corporate mentality clashes with thefarmers’ view that agriculture is more than a purelyeconomic activity, and is likely to unsettle deep-seated values in their communities such as com-monwealth and equality (Schneiberg et al., 2008).Farmers have long feared that corporate farmingwill alter the agrarian community structure by cre-ating a segment of farmers who are structurallyequivalent to factory production workers, eventu-ally polarizing families (Lobao, 1990).

Ideological resentment will also be triggered bythe erosion of farmers’ control over production de-cisions, and the fear that corporate domination ofagriculture will result in their exploitation. More-over, when corporations own the farmland, as ab-sentee owners, they contribute very little to thelocal economy and cohesion. In contrast, the familyfarm earns and spends locally, and is often a morecommitted steward of land, water, and other re-sources that enrich the farm environment (Lyson &Welsh, 2005). Indeed, research shows that the qual-ity of life, standard of living, level of income equal-ity, and social participation in counties dominatedby corporate farming are significantly lower than infamily-farm-dominated counties (Lobao, 1990).

To avoid corporate involvement in farming,many communities adopted anti-corporate farmingregulations. Although these laws reflect local, pre-existing, anti-corporate sentiments, their existencesettles the ideological tension between corn farm-ers and ethanol corporations by preventing the lat-ter from integrating backwards into corn produc-tion. As a result, corn farmers in counties with suchprotective laws benefit maximally from growth inlocal ethanol production by corporations sincetheir position vis-à-vis those corporations is secure.The law provides immunity, and reduces the re-source overlap between corporations and farmerssince the former can only be buyers and not (po-tential) producers of corn (Lyson & Welsh, 2005). Incounties where there are no prohibitions in place,corporations have ample opportunity to engage infarming, which represents a serious threat to farm-ers. The absence of laws generally depresses cornprices and spurs absentee ownership, leavingfarmer communities with very few economic ben-

efits (Lobao, 1990; Lyson & Welsh, 2005). This, inturn, can trigger anti-corporate sentiments and ide-ological opposition to the proliferation of corporateethanol plants (Jonsson, Greve, & Fujiwara-Greve,2009). Thus, we hypothesize that farmers are morelikely to become mobilized to establish ethanol co-operatives in response to commercial ethanol en-terprises if the latter have the potential to enter intofarming (i.e., in communities that have no anti-corporate farming laws). This will moderate thenegative cross-form main effect of Hypothesis 1.

Hypothesis 2. The negative relationship be-tween local corporate ethanol production ca-pacity and cooperative founding will beweaker in counties where corporate ethanolproducers can vertically integrate into farming.

General anti-corporate climate. Since the birthof the United States, ideological clashes betweentwo approaches—the community logic of gover-nance, stressing autonomy, and the nationallydominant, corporate logic of governance focusingon centralization and mass-market consolidation—have affected the dynamics of several industries(Marquis & Huang, 2009: 1241). Anti-corporate sen-timents embedded in deep-seated distrust of thelatter mindset have spurred social movements totarget corporations (Rao, Yue, & Ingram, 2011). Forinstance, local communities have mobilized re-sources to prevent large retail chains such as Wal-Mart from opening stores, in an effort to protectlocal economic activity (Ingram & Rao, 2004). Cul-tural resistance to large-scale organizations has alsobeen forceful among communities with respect tothe commercial banking industry (Marquis &Huang, 2009; Marquis & Lounsbury, 2007), the U.S.brewing industry (Carroll & Swaminathan, 2000),and U.S. radio broadcasting industry (Greve etal., 2006).

These studies reveal great variety among commu-nities in the prevalence of pro- versus anti-corpo-rate sentiments. For instance, Marquis and Huang(2009) found that local agrarianism suppressed thepositive impact of state legislation allowing state-wide bank branching on the growth of U.S. com-mercial banks during the 20th century. Bell (2009)documents that, even within the same state, corpo-rate entry through the buying-out of small indepen-dent mines was unproblematic in some communi-ties, while, in others, the community did not givein without a fight. Rao et al. (2011) showed that, instates without right-to-work laws (which serve as asignal of the “pro-business climate in a state”),


there was a higher risk of protest activism targetingcorporations.

The geographical variety that characterizes ananti-corporate climate in communities generallytraditionally applies also to farmer communities.The Grange—the leading anti-corporate move-ment—did not enjoy the same level of supportacross U.S. farming regions (Schneiberg et al.,2008). Ingram and Rao (2004), in their study of theanti-chain legislation, describe how rural Americais divided in the sense that, in some communities,farmers were sympathetic to the “Northeastern cap-italistic values” that chains represented, and effec-tively rallied for pro-chain legislation, while, inothers, farmers were outraged by the advances ofcorporate America and allied themselves with theanti-chain movements to protect rural interests.

We expect that, in communities with a stronganti-corporate climate, the diffusion of corporateethanol facilities could generate a backlash in theform of cooperative founding. In such communi-ties, ideological rivalry between forms that are di-rectly competing for resources makes commoncauses and sentiments sharper, distinct, and salient(Boone, Divarci, & van Witteloostuijn, 2011; Carroll& Swaminathan, 2000; Haveman et al., 2007; Si-mons & Ingram, 2004). It galvanizes opposinggroups, and creates a strong local “we feeling”among farmers characterized by loyalty and devo-tion to the ideals of autonomy and developmentbased on cooperative commonwealth (Schneiberget al., 2008). This spurs the formation of coopera-tives that carry forward these oppositional identi-ties and aim to protect the local social order (Greveet al., 2006; Marquis & Lounsbury, 2007; Pozner &Rao, 2006). Therefore:

Hypothesis 3. The negative relationship be-tween local corporate ethanol production ca-pacity and cooperative founding will beweaker in counties with an anti-corporateclimate.

Organizational infrastructure in support of co-operative ideology. Since alternative organizationalforms such as cooperatives challenge existing inter-ests, their birth is a contentious process that re-quires collective action, ideological activism, andresource mobilization (Carroll & Swaminathan,2000; Greve et al., 2006; Greve & Rao, 2012; King &Pearce, 2010). As “organization-generating organi-zations” (Stinchcombe, 1965), anti-corporate socialmovements “foster cooperatives in their efforts toestablish alternative economic orders” against cor-

porate capitalism (Schneiberg et al., 2008: 638).They provide the necessary supportive social struc-tures that facilitate collective action and resourcemobilization, boosting entrepreneurship aroundnew forms in several ways (Schneiberg et al., 2008).

First, alternative organizational forms are institu-tional projects that must be assembled, theorized,politically defended, legitimated, and made avail-able as an organizational solution to a problem aswell as becoming integrated with the prevalent in-stitutional order (Rao, 1998; Schneiberg et al., 2008;Sine & Lee, 2009). Movements are important cul-tural forces that operate as agents of theorizationand framing (Rao, 1998; Schneiberg et al., 2008). Byconstructing and propagating theories about unre-solved problems, contesting the value of existingsolutions and showing how cooperatives can solvethem more effectively, anti-corporate social move-ment organizations can disrupt existing market ar-rangements, direct market actors’ attention to alter-native forms, and embed their norms and visions inregulation (Greve & Rao, 2012; Sine & Lee, 2009). Inso doing, cooperatives gain legitimacy as a taken-for-granted rational alternative to corporate capital-ism, boosting cooperative entrepreneurship (Sch-neiberg et al., 2008).

Second, by creating cultural codes and a support-ive organizational infrastructure, anti-corporatemovements provide cooperators with organizationaltemplates that enhance the diffusion of coopera-tives (Schneiberg et al., 2008). This increases thelikelihood that entrepreneurs will recognize newopportunities, acquire new information aboutthem, and marshal the resources necessary to ex-ploit them (Schneiberg & Lounsbury, 2008; Sine &Lee, 2009). Moreover, by morally sanctioning par-ticular activities and linking them to deeply heldvalues, movements can inspire a class of entrepre-neurs to engage in those activities even if they areextremely risky (Schneiberg, 2007; Sine et al., 2005;Weber et al., 2008).

Third, anti-corporate movements provide solu-tions to the technical, organizational, and market-related challenges that entrepreneurs face when de-veloping their businesses (Weber et al., 2008). Theycreate favorable environments for supplying actorswith resources, knowledge, and social networksneeded to establish cooperatives (Schneiberg etal., 2008).

Finally, they decrease the costs of collective ac-tion by cultivating social ties within local commu-nities. This social fabric unites people aroundcommon values and ideologies, which increases


loyalty, trust, and cohesion in the community andcurtails free-riding (Ingram & Simons, 2000; Sch-neiberg et al., 2008). This is particularly importantfor the successful emergence of cooperatives, giventhe collective nature of cooperative entrepreneur-ship. For all the reasons mentioned, we expect thatlocal actors will be more likely to mobilize againstthe diffusion of corporations if anti-corporate sen-timents are embedded in a well-developed, activistorganizational infrastructure that supports thevalue of the alternative form.

Hypothesis 4. The negative relationship be-tween local corporate ethanol production ca-pacity and cooperative founding will beweaker in counties with more established or-ganizational infrastructures supporting coop-erative ideology.

Corporate Plant Heterogeneity: Who is theCommon Foe?

The presence of a salient and more powerfulenemy, which triggers the development of a strongin-group identity, is a major prerequisite for effec-tive collective action and the emergence of neworganizational forms (Pozner & Rao, 2006). Themore salient the enemy, the more in-group identi-fication is triggered and the more forceful the reac-tion against the out-group (Pozner & Rao, 2006).According to this logic, not all corporate ethanolplants will fuel anti-corporate sentiments to thesame extent. Some corporations will be more likelythan others to be perceived as a salient threat andthe common foe, with ethanol plants owned by oilcompanies and agribusiness companies more likelyto produce rifts. The New York Times summed upwhy ethanol plants owned by oil firms are viewedas adversaries:

For decades, the big oil companies and the farmlobby have been fighting about ethanol, with thefarmers pushing to produce more of it and the refin-ers arguing it was a boondoggle that would do littleto solve the country’s energy problems.

(Krauss, 2009)

In most cases, oil companies not only stalled theethanol industry’s advancement; they brought itdown. For instance, by pushing for mandatorypump labeling requiring service stations to postnotices telling consumers whether their fuel con-tained ethanol, oil producers brought the ethanolindustry to the brink of extinction in several states.

Local farmers and activists can also be expectedto find the involvement of agribusiness in ethanolproduction especially threatening. These compa-nies are often accused of exploiting local farmers,thereby reducing rural welfare (Lyson & Welsh,2005). In 1990, the four largest agribusiness firmsalready controlled over 60% of the U.S. grain busi-ness. By 2008, these firms controlled 95% of theU.S. corn market. Because they deal in high vol-umes, these vertically and horizontally integratedcompanies have significant leverage in terms ofsetting the purchase price, and exert a tight grip onthe farming economy (Murphy, Burch, & Clapp,2012). Some farmers even contend that these agri-business corporations “should be required to do-nate capital and resources to the development ofthe bio-economy as reparations for a history of fun-neling farm profits away from the farmer” (Meyer,2008: 115).

We thus expect the local diffusion of “big busi-ness” ethanol plants to trigger collective entrepre-neurship among farmers, and thus cooperative en-try in communities with latent anti-corporatesentiments. Specifically, the negative cross-form ef-fect of corporate prevalence on cooperative found-ing will be attenuated if the corporate enemy isparticularly salient (i.e., big business corporations)and if the local communities share anti-corporatesentiments. Formally, this implies:

Hypothesis 5. (a) The attenuating effects of thethreat of vertical integration (Hypothesis 2), (b)anti-corporate climate (Hypothesis 3), and (c)supportive infrastructure (Hypothesis 4) will bestronger the more that corporate ethanol pro-duction is controlled by oil and agribusinesscompanies.

METHODS

Data

Our population consists of 365 plants across 36states that entered the industry between January1978 (the year of entry of the first ethanol plant)and May 2013. Of these, 85 plants, spread across 17states, were cooperative-owned; 78 were incorpo-rated according to state incorporation statutes gov-erning traditional cooperatives (the 1922 StandardAct provides the basis for these statutes), and 7were chartered under the “limited cooperative as-sociation acts,” which allowed them to raise fundsfrom non-farmer local community residents. Mostcooperative entries occurred in Iowa (22) and Min-


nesota (16), with Arkansas, California, Kentucky,Michigan, Ohio, Pennsylvania, and Virginia host-ing only one plant each.

We assembled the data based on archivalsources, including Renewable Fuels Association re-ports, industry analyses by Information ResourcesInc., the U.S. Department of Energy and Depart-ment of Agriculture, magazines such as OilgramNews, Oxy-Fuel News, and Ethanol Producer Mag-azine, newspaper archives, and press releases.

Dependent Variable and Estimation

We analyze the rate of cooperative plant found-ing in a given U.S. county. Founding can be under-stood as an arrival process and therefore can bemodeled as a semi-Markov process. Accordingly,we estimate event history models, and, followingHaveman et al. (2007), use Cox proportional haz-ards regressions (see also Sine et al., 2005; Sine &Lee, 2009). Unlike parametric event history mod-els, the Cox model makes no assumption regardingthe shape of the baseline hazard, which gives itgreater flexibility. The hazard rate is given by:

hi(t) � h0(t)exp(�*z)

where h0(t) is the baseline hazard function, � is thevector of parameters associated with the explana-tory variables included in the model, and z is thevector of the explanatory variables. We estimatecoefficients and confidence intervals using maxi-mum likelihood techniques, and report robust Hu-ber–White standard errors to account for clusteringof observations within states.

Our county-level study includes all U.S. countiesexcept those in Hawaii and the District of Columbia(D.C.), giving a sample of 3,127 counties and a totalof 444,034 spells. Hawaii and D.C. are excludedbecause of lack of data on key variables; the ex-cluded counties have not hosted a plant. The re-maining counties are included in the risk set for theperiod January 1978 to June 2013. We define timeof entry as the date when the plant officially beganproduction. We obtained precise entry dates for332 plants, and precise failure dates for all but twofailed plants. For the rest, we can identify themonth when the plant came online or was closedpermanently. Given the large number of counties,we constructed quarterly spells. No two coopera-tive plants were founded in the same quarter/county.

Our dependent variable is a dummy, which iscoded “1” if, in a given quarter/county, a coop-

erative plant officially commenced production.The clock variable in each county starts in Janu-ary 1978 and ends when a founding event takesplace in that county, which then resets thecounty clock. Our archival search showed thatthe process of founding—from the first equitydrive to start of production—takes 18 months onaverage. We therefore lagged our explanatoryvariables by six quarters.5 We also tested thesensitivity of our results by lagging all explana-tory variables by eight instead of six quarters. Theestimates obtained differed very little.

Focal Independent Variables

For a given county/quarter, local corporate etha-nol production capacity is the logged sum of theproduction capacity of all corporate plants operat-ing in that county and the counties adjacent to it.We included adjacent counties since researchshows that the impact of a mid-sized cooperativeethanol plant on corn prices and supplies is felt upto 70 miles’ distance from the plant (McNew &Griffith, 2005). We chose to examine productioncapacity, rather than density, because the formermatters more in terms of its effect on the process ofcompetition, and much can be learned about theecological dynamics of organizational populationsby studying the evolution of size distributions(Hannan, 1988). Following this operationalization,we constructed the variable local cooperative eth-anol production capacity.

We measure vertical integration threat with adummy that indicates whether the county permitscorporate farming (“1” � yes) or not (“0” � other-wise). Currently, all counties in South Dakota,Minnesota, Wisconsin, Nebraska, North Dakota,Oklahoma, Iowa, and Kansas; five counties inPennsylvania; and all but three counties (Putnam,

5 A few cooperatives bought idle plants and retooledthem, which accelerated the process of founding. Amongthose who built their own plants is KAAPA. This coop-erative started an equity drive in January 2002, and, inNovember 2003, its plant processed its first corn. Groundwas broken for the Show Me Ethanol plant in March2007, and the plant began production in May 2008. Con-struction of the CVEC plant began in June 1995, andfinished in April 1996. Big River Resources held its firstcapital meeting in February 2002, and the plant groundcorn in April 2004. Corn Plus raised its equity in 3 weeks,and constructed its plant in less than a year.


Sullivan, and Mercer) in Missouri have laws pro-hibiting corporate farming.

Following Holmes (1998) and Rao et al. (2011),we use the absence of right-to-work (RTW) laws—which ban compulsory union membership as a req-uisite for employment, among other provisions—toserve as a general indicator of anti-corporate cli-mate (dummy � “1” if RTW law absent). RTW lawsare a good proxy, because, to opponents of corpo-rations, they signify not just a management–laborissue but a general threat posed by those corpora-tions to local autonomy, commonwealth, and self-sustainment, as well as the ability to defend againsttheir excesses.6 This expanded view is reinforcedby the fact that the RTW laws were accompanied bythe adoption of other policies (e.g., low capitaltaxes, lax environmental regulations) especially fa-vorable to corporations, and the same actors wholobbied for RTW laws later worked to realize manycorporate-friendly policies (Holmes, 1998; Rao etal., 2011). By June 2013, 24 states were RTW states.

We operationalize the supportive organizationalstructure by Farm Bureau tenure, which indicatesthe history of the Farm Bureau in a given county,measured by the logged number of quarters. Nearly2,600 counties had active bureaus in 1978. By June2013, this number had risen to 2,772. We obtainedthe founding dates of Farm Bureaus from states’business entity filings.

We chose Farm Bureau for four reasons. First, itis the largest agrarian social movement organiza-tion in the United States. Second, it represents theideological extension of the Grange, which was aleading anti-corporate social movement (Sch-neiberg et al., 2008). Many county Farm Bureauswere founded by local Grange leaders and mem-bers, rallying around the same principles of equal-ity and independence (Kile, 1948). As the Grangedeclined in influence, the Farm Bureau absorbedits members and ideology, and, by 1960, Bureaumembership totaled more than all other farmer in-terest organizations combined. Third, the FarmBureau is pro-cooperative. In fact, it successfully

revived the ailing cooperative movement in theUnited States, which had first surged in thelate 19th century but then rapidly declined. Un-like Grange offices, county bureaus quickly sawthe need for the organization of economic activ-ities as a means of providing services to farmersthat public agencies were unable to offer. Butthere was more: for members, economic cooper-ation seemed a way of gaining some measure ofcontrol over market processes and preserving therural way of life (Berlage, 2001). Throughout thelast century, county bureaus created marketing,supply, and insurance and electricity coopera-tives. Fourth, from its beginnings, the Farm Bu-reau Federation has been a fervent proponent ofethanol. Farm Bureaus lobbied in legislativecourts, sponsored education campaigns, distrib-uted ethanol through their cooperative servicestations, petitioned the state to use ethanol instate vehicles, and threatened oil and large foodcompanies to stop their campaign againstethanol.

We believe that tenure in a particular county is abetter construct for our theory than a simpledummy variable. Long tenure implies organizationalstrength and entrenchment within the social andeconomic fabric of the local community. The longerthe Farm Bureau’s presence, the stronger should beits capacity to inspire collective entrepreneurshipand mobilize the necessary resources (Sine & Lee,2009). Further, imprinting of the Grange’s anti-cor-porate and commonwealth ideologies should bestronger for earlier county bureaus than later ones(Marquis & Lounsbury, 2007). In similar vein,Greve and Rao (2012) show that Norwegian com-munities that were the earliest to establish collec-tive social movement organizations were also morelikely to experience founding of cooperative stores.The earlier was the founding of a formal nonprofitorganization, the greater was the civic capacity of acommunity.

To test Hypothesis 5, we split local corporateethanol production capacity into two componentsbased on ownership. One component representsthe production capacity of 46 commercial plantsowned by agribusiness (ADM, Cargill, LD, Bunge,Kraft, Tate & Lyle, Heinz, MGP, GPC, Penford, Ma-nildra, JR Simplot) and oil companies (Williams,Marathon, Valero, Sunoco, Murphy, Ashland,Texaco, Flint Hills). The other component capturesthe production capacity of the remaining commer-cial plants not owned by agribusiness or oil com-panies. These other plants were created mainly by

6 For instance, Wall-Mart’s expansion met greater op-position in communities with no RTW laws than in com-munities with such laws, but the mobilizing codes hadlittle to do with the chain’s specific relationship to labor(Rao et al., 2011). More broadly, this opposition was notjust from the labor unions. From environmental organi-zations (i.e., Sierra Club) to civil rights organizations (i.e.,NAACP), a broad-based coalition mounted fierce resis-tance to these laws and mobilized supporters.


owners of businesses within the farm economy,plant engineering, or renewable energy. During ourperiod, plants owned by agribusiness and oil com-panies entered with an average capacity of 70 mgy,whereas other plants began production with 40mgy capacity on average. In 1990, agribusiness- andoil-owned plants accounted for 59% of total U.S.production capacity. By June 2013, their share haddeclined to 21%.

Control Variables

We include several key economic controls in ourmodels. Fortunately, many of these controls are atcounty level, providing even richer variety. Cornprice denotes the state’s quarterly average of themonthly corn price per bushel. Corn accounts fornearly 80% of ethanol’s cost per gallon. Corn pro-duction denotes the focal county’s logged annualproduction of corn in bushels, and determines thelocal carrying capacity for ethanol plants. Ethanolplants are not the only destination for corn. Farm-ers can choose to sell their corn to plants that grindgrain for food. We thus control for the number ofcorn processors in the county. The presence ofmany processors increases the number of alterna-tive outlets for corn farmers, but also reduces therelative market power of ethanol corporations. Thissuppresses transaction costs between corn farmersand corporations, and, thus, the incentive for farm-ers to establish cooperatives (Caves & Petersen,1986). We obtained data from Census County Busi-ness Patterns and include flour mills, wet cornmills, and cereal mills. County livestock is thelogged number of cattle on feed in a given county ina given quarter. We include this variable becausedistillers’ grain is an important co-product of etha-nol production, and is fed to livestock. We obtainedthese data from the National Agricultural StatisticsService (NASS). Average farm size (in loggedacres), obtained from the NASS, proxies for con-centration of farm ownership within the countyover time. We expect a negative relationship sincefarmers with large fields can negotiate better termswith buyers, leaving fewer incentives for creatingcooperatives. Racial diversity has been found toaffect collective mobilization (Ingram & Rao, 2004),thus we also computed a county racial diversityindex at the county level, which is 1 minus theHerfindahl index of three racial groups (Whites,Blacks, and Others) for a given year. Retail gasolineprice is the quarterly average of the monthly priceof gasoline per gallon in the state. As expected,

gasoline prices are highly correlated with the rackethanol price. We thus opted to exclude the latterfrom our models. Demand is measured by Stateethanol consumption, which represents a state’squarterly ethanol consumption (in British ThermalUnits, scaled by trillions) as reported by the U.S.Energy Information Administration. State ethanolexemption controls for excise tax exemption (in ¢)in a given quarter/state. These consumer creditshave varied between and within the states overtime. We coded this variable based on HighwayStatistics reports. If the state governor is a memberof the Governors’ Ethanol Coalition, then the vari-able, Coalition member, is coded “1” (which im-plies the likely presence of more pro-ethanol insti-tutional behavior in these states), and “0”otherwise. Minimum efficiency scale is the loggedmean of the production capacities of all plants thatexisted in a given quarter. As the industry matured,the average size of ethanol plants dramatically in-creased. By including this variable, we control forchanges in scale economies. We include the loggednumber of farmer cooperatives in a given state (alltypes), as reported by the U.S. Rural DevelopmentAgency. Finally, we include region dummies basedon the U.S. Census Bureau’s definition of regions.

FINDINGS

Table 1 reports the summary statistics and corre-lations. Correlations between the variables are lowto moderate. Table 2 presents the results (coeffi-cients) of our estimates. Model 1 excludes localcorporate ethanol plant capacity; Model 2 is the fullbaseline model; Models 3–5 include the interac-tions sequentially; in Model 6, all interactions areincluded simultaneously.7

Looking across the models, several control vari-ables are significant as expected. Counties withgreater corn production are more conducive to co-

7 In Models 1–4, all variance inflation factors (VIFs)are below 3, indicating low multicollinearity. The cross-product of local corporate ethanol production capacityand Farm Bureau tenure are highly correlated with thetenure component in Models 5 and 6, creating potentialmulticollinearity problems. To check this, we re-ranModels 5 and 6 after mean-centering both variables (nei-ther is a dummy) and then generating the interactionterm, which reduces the maximum VIF to below 3. Asexpected, these results were nearly identical to the mod-els without mean-centering. We report the latter, as in-terpretation of these models is more straightforward.


operative founding. The higher the minimum effi-ciency scale, the lower the likelihood of coopera-tive founding, because larger plant size makes itmore challenging to mobilize the community. Re-tail gas prices in the state are negatively related tocooperative founding because the higher the gasprice, the higher the demand for ethanol—and, inturn, grain, and the lesser the need for cooperativefounding. Membership by the state’s governor in apro-ethanol lobbying organization facilitates therise of cooperatives, albeit weakly.

In Model 2, the coefficient of local corporateethanol production capacity is significantly nega-tive, which confirms Hypothesis 1. Since we con-trol for the number of corn-processing plants re-lated to the food industry, county’s supply of corn,and average farm size, this is clearly a necessityissue. In this industry, cooperatives have been

founded as a response to a lack of corporate buyersin the immediate environment.

Models 3 to 6 add interaction terms to Model 2.Consistent with Hypothesis 2, Hypothesis 3, andHypothesis 4, we find that all three interactionterms are significantly positive at less than p �0.01. Model 6 includes all three interactions. Ourestimations are cluster-corrected at state level mak-ing them robust, using the Huber–White Sandwichtechnique. For this reason, the log-likelihood val-ues become pseudo values, and hence the standardlog-likelihood ratio tests for evaluating relativemodel fitness are not valid (StataCorp, 2013). Wetherefore apply a Wald test to check whether thecoefficients of these three interactions are simulta-neously equal to 0. From the Wald test, we obtaineda p-value of 0.0039 associated with a chi-square of13.35 with 3 degrees of freedom. Based on this

TABLE 1Descriptive Statistics for Observations in Baseline Model

Variables Mean SD 1 2 3 4 5 6 7 8 9

1. Cooperative founding 0.00 0.01 1.002. Local ethanol production capacity—Corp. 0.56 1.38 0.01 1.003. Local ethanol production capacity—Coop. 0.23 0.95 0.03 0.39 1.004. Vertical integration threat (VIT) 0.76 0.43 �0.02 �0.23 �0.33 1.005. Anti-corporate climate (AC) 0.50 0.50 �0.00 �0.01 �0.02 0.06 1.006. Supportive organizational structure (SOS) 4.38 1.78 0.01 0.13 0.10 �0.11 �0.07 1.007. Corn price 2.82 0.94 �0.01 0.15 0.11 0.12 �0.03 0.03 1.008. County corn production 5.51 2.62 0.02 0.35 0.31 �0.36 0.05 0.38 �0.07 1.009. Mean farm size in the county 5.75 1.06 0.00 0.11 0.06 �0.21 �0.11 �0.08 �0.04 0.04 1.00

10. Livestock in the county 0.04 0.09 0.00 0.03 0.01 �0.04 �0.02 0.01 0.01 0.05 0.0711. County racial diversity 0.17 0.17 �0.01 �0.16 �0.14 0.29 �0.32 �0.02 0.09 �0.19 �0.0912. Corn processors in the county 0.14 0.48 �0.00 0.05 0.01 0.00 0.06 0.05 0.00 0.08 �0.1113. State ethanol consumption 5.78 12.52 0.00 0.19 0.11 0.08 0.05 0.11 0.46 0.09 �0.0414. State ethanol exemption 0.87 2.09 0.00 0.01 0.01 �0.08 �0.09 �0.10 �0.02 �0.00 0.1215. Coalition member 0.36 0.48 0.01 0.29 0.29 �0.22 �0.01 0.18 0.17 0.24 0.1716. Retail gas price 1.13 0.63 0.00 0.25 0.26 0.00 �0.04 0.09 0.67 0.03 �0.0217. Min. efficiency scale 3.28 0.92 0.01 0.18 0.17 �0.02 �0.07 0.13 0.26 0.04 �0.0118. No. of farmer coops in the state 4.37 1.04 0.01 0.10 0.10 �0.41 0.09 0.08 �0.20 0.32 0.22

TABLE 1(continued)

10 11 12 13 14 15 16 17

11. County racial diversity 0.04 1.0012. Corn processors in the county 0.03 0.04 1.0013. State ethanol consumption �0.00 �0.00 0.02 1.0014. State ethanol exemption �0.00 �0.08 �0.02 �0.14 1.0015. Coalition member �0.00 �0.10 �0.02 0.31 �0.20 1.0016. Retail gas price �0.00 0.02 �0.01 0.56 �0.08 0.42 1.0017. Min. efficiency scale �0.00 0.07 �0.02 0.33 �0.18 0.44 0.47 1.0018. No. of farmer coops in the state 0.03 �0.26 0.04 0.03 0.13 0.06 �0.23 �0.22

Note. Correlation coefficients below �0.02 and above 0.02 are significant at a 5% level.


TA

BL

E2

Cox

Pro

por

tion

alH

azar

dE

stim

ates

ofC

oop

erat

ive

Pla

nt

Fou

nd

ing

Mod

el1

Mod

el2

Mod

el3

Mod

el4

Mod

el5

Mod

el6

Cor

np

rice

�0.

69(0

.47)

�0.

64(0

.46)

�0.

63(0

.45)

�0.

64(0

.45)

�0.

65(0

.46)

�0.

64(0

.45)

Cou

nty

corn

pro

du

ctio

n0.

59**

*(0

.17)

0.65

***

(0.1

8)0.

65**

*(0

.18)

0.66

***

(0.1

9)0.

66**

*(0

.18)

0.66

***

(0.1

9)M

ean

farm

size

inth

eco

un

ty�

0.07

(0.1

7)�

0.09

(0.1

6)�

0.14

(0.1

7)�

0.12

(0.1

6)�

0.09

(0.1

6)�

0.15

(0.1

6)L

ives

tock

inth

eco

un

ty1.

04(2

.02)

1.11

(1.9

6)1.

07(1

.93)

1.11

(1.9

3)1.

15(1

.96)

1.13

(1.9

2)C

oun

tyra

cial

div

ersi

ty�

1.15

(1.7

2)�

1.13

(1.7

3)�

1.15

(1.7

3)�

1.16

(1.7

4)�

1.19

(1.7

0)�

1.21

(1.7

2)C

orn

pro

cess

ors

inth

eco

un

ty�

0.19

(0.2

0)�

0.19

(0.2

1)�

0.19

(0.2

1)�

0.19

(0.2

1)�

0.19

(0.2

1)�

0.19

(0.2

1)S

tate

eth

anol

con

sum

pti

on�

0.02

(0.0

3)�

0.02

(0.0

3)�

0.03

(0.0

3)�

0.03

(0.0

3)�

0.02

(0.0

3)�

0.04

(0.0

3)S

tate

eth

anol

exem

pti

on0.

08(0

.06)

0.08

(0.0

6)0.

08(0

.06)

0.08

(0.0

6)0.

08(0

.06)

0.08

(0.0

6)C

oali

tion

mem

ber

1.52

(0.9

2)1.

50*

(0.8

9)1.

55*

(0.8

8)1.

51*

(0.9

0)1.

53*

(0.8

8)1.

56*

(0.8

9)R

etai

lga

sp

rice

�1.

50**

*(0

.32)

�1.

43**

*(0

.33)

�1.

42**

*(0

.33)

�1.

43**

*(0

.33)

�1.

43**

*(0

.33)

�1.

42**

*(0

.33)

Min

.ef

fici

ency

scal

e�

4.33

***

(1.6

4)�

4.18

***

(1.6

1)�

4.14

***

(1.5

9)�

4.15

**(1

.62)

�4.

18**

*(1

.61)

�4.

13**

(1.6

0)N

o.fa

rm.

coop

sin

the

stat

e0.

73*

(0.4

3)0.

74*

(0.4

0)0.

75*

(0.3

8)0.

70*

(0.4

2)0.

72*

(0.4

0)0.

71*

(0.3

9)L

ocal

eth

anol

pro

du

ctio

nca

pac

ity

(LE

PC

)—C

oop

.0.

13(0

.09)

0.13

(0.0

9)0.

14(0

.09)

0.13

(0.0

9)0.

13(0

.09)

0.14

(0.0

9)V

erti

cal

inte

grat

ion

thre

at(V

IT)

�0.

87*

(0.4

9)�

0.81

*(0

.47)

�1.

21**

(0.4

9)�

0.85

*(0

.47)

�0.

83*

(0.4

6)�

1.16

**(0

.47)

An

ti-c

orp

orat

ecl

imat

e(A

C)

0.34

(0.4

2)0.

21(0

.40)

0.22

(0.4

1)0.

01(0

.45)

0.23

(0.3

9)0.

13(0

.42)

Su

pp

orti

veor

g.st

ruct

ure

(SO

S)

0.01

(0.0

7)�

0.01

(0.0

7)�

0.02

(0.0

7)�

0.01

(0.0

7)�

0.11

(0.0

8)�

0.09

(0.0

7)L

EP

C—

Cor

p.

�0.

13**

(0.0

5)�

0.17

***

(0.0

5)�

0.19

***

(0.0

4)�

0.41

***

(0.1

1)�

0.42

***

(0.1

1)L

EP

C—

Cor

p.�

VIT

0.26

***

(0.1

0)0.

19*

(0.1

0)L

EP

C—

Cor

p.�

AC

0.18

***

(0.0

7)0.

10**

(0.0

5)L

EP

C—

Cor

p.�

SO

S0.

05**

*(0

.02)

0.04

**(0

.02)

Reg

ion

du

mm

ies

Yes

Yes

Yes

Yes

Yes

Yes

Nu

mbe

rof

spel

ls44

4034

4440

3444

4034

4440

3444

4034

4440

34L

og-l

ikel

ihoo

d�

788.

52�

786.

26�

784.

54�

784.

87�

785.

71�

783.

81

Not

e:R

obu

stst

and

ard

erro

rsca

lcu

late

dby

clu

ster

ing

arou

nd

stat

es.

*p

�0.

10**

p�

0.05

***

p�

0.01

;C

oeff

icie

nt

sign

ific

ance

test

sar

etw

o-ta

iled

p-value, we are able to reject the null hypothesis,indicating that including these interactions createsa statistically significant improvement in model fit.

Given the correlation between these interactionterms, their effect size and significance levels be-come weaker, as expected. Two (anti-corporate cli-mate and supportive organizational structure) re-main significant at p � 0.05, respectively, withtwo-sided tests. The third interaction term, whichexamines the moderating effect of a vertical inte-gration threat, is significant at p � 0.055 with atwo-sided test. To illustrate the unique effect ofeach interaction, we present figures that show themarginal impact of local corporate production ca-pacity on the multiplier of the cooperative found-ing rate, based on the estimates from Model 6.

The figures show a strong negative effect of local,corporate, ethanol production capacity on the mul-tiplier of the founding rate (supporting Hypothesis1). Hypothesis 2, Hypothesis 3, and Hypothesis 4are also clearly confirmed, as the moderators allattenuate this negative interdependence. When cor-porate ethanol production capacity reaches 100mgy (log capacity � 4.61), the multiplier of the rateis 2.5-times larger in counties with no anti-corpo-rate farming laws (high threat) compared to coun-ties where vertical integration is not allowed (seeFigure 2A). The multiplier is also twice as large incounties with a high (mean � SD) compared to lowFarm Bureau tenure (mean � SD) at the same levelof local corporate production capacity (Figure 2C).The effect of the anti-corporate climate interactionis somewhat smaller: the multiplier of the rate isabout 1.6-times larger in counties with an anti-business compared to a pro-business climate (at100 mgy; see Figure 2B).

Table 3 presents the estimates for Hypothesis 5.The first column provides estimates of local cor-porate production capacity split into two compo-nents: owned by oil and agribusiness, and ownedby others. Both estimates are negative, but onlythe coefficient of oil and agribusiness is signifi-cant, and is much more negative than the coeffi-cient of “other” capacity. A Wald test reveals thatboth coefficients are significantly different fromeach other at p � 0.06. In Model 2, we add theinteraction between both production capacityvariables and the threat of vertical integration.Hypothesis 5a is not supported since both corpo-rate production capacity variables interact posi-tively with the threat of vertical integration, andin a similar way. A Wald test shows that neitherinteraction is significantly different from the

other. So, the attenuating effect of vertical inte-gration threat (VIT) does not depend on the cor-porate identity of the producer. The same conclu-sion applies to Hypothesis 5c, which is tested inModel 4. The attenuating effect of a supportiveorganizational structure (SOS) is not affected bycorporate identity. Both interaction coefficientsinvolving SOS are not statistically different, asrevealed by a Wald test.

However, Hypothesis 5b is supported. A generalanti-corporate climate (AC) attenuates the negativeimpact of local corporate ethanol production ca-pacity, but only if the owners are oil and agribusi-ness corporations. Wald tests show that both coef-ficients are significantly different from each other,although only at the 10% level and are not simul-taneously equal to 0, meaning that including thesevariables creates a statistically significant improve-ment in the model fit. We also tested Hypothesis 5in a different but equivalent way by analyzingthree-way interactions creating product terms be-tween overall local corporate production capacity(as used in Table 2), the context moderators (i.e.,VIT, AC, or SOS), and the proportion of total localproduction capacity produced by oil and agribusi-ness corporations. The results are the same as thosereported here. That is, only the three-way interac-tion with AC is significantly positive (at p � 0.05,two-tailed test). We report the findings as in Table 3because they are easier to interpret.

Figure 2D illustrates that in communities with astrong anti-corporate climate (based on the interac-tion estimate in Model 3 of Table 3), the effect of oiland agribusiness corporate production capacity iseven reversed—showing a boost in cooperativefounding.

ROBUSTNESS

We assessed the robustness of our results in sev-eral ways. First, we experimented with additionalcontrols, including state per capita income, countypopulation size, whether the state has an ethanolproducers’ association, whether the state requirespump labeling, period dummies to capturethe presence of a production income tax credit forsmall ethanol producers, federal (excise) tax ex-emptions, whether the county is a border county,and number of ethanol plant failures in the state inthe previous four years. The results were substan-tively the same. Since 2001, 11 states in the Mid-west have gradually introduced limited coopera-tive association acts (as a complement to, not a


substitute for, traditional cooperative statutes). Aperiod dummy capturing these new regulationsdoes not affect our main findings, and is not in-cluded in the models due to its degree of correla-tion with our region dummies. These analyses areavailable upon request.

DISCUSSION

Social movement organizations have producedsome of the most significant institutional and social

changes in the United States (Hiatt et al., 2009;Schneiberg & Lounsbury, 2008). Over the last fourdecades, however, few have been as drastic intransforming rural America as the ethanol move-ment. The collective push for ethanol has createdlocal wealth, accelerated technological advances infarming, and promoted physical and institutionalinfrastructure within communities, all of whichhave contributed to a rural renaissance. The found-ing of cooperative plants has strengthened commu-nity bonds, increased collective optimism, passion,

Figure 2(a). Vertical Integration Threat and the Hazard of Cooperative Plant Founding. (b). Anti-corporate Climateand the Hazard of Cooperative Plant Founding. (c). Supportive Organizational Structure and the Hazard

of Cooperative Plant Founding. (d). Anti-corporate Climate and the Hazard of Cooperative Plant Founding


and confidence, and paved the way to further col-lective action to improve communities (Thomp-son, 2004).

Such revivals emerge and fade, depending onchanges in the community orientation of inhab-itants. Simons and Ingram (2003) and Schneiberget al. (2008), for instance, link the decline of earlyfarmer cooperatives in Israel and the UnitedStates, respectively, to immigration and residen-

tial instability. Other accounts from alternativeenergy fields document that environmentalgroups foster entrepreneurial activity by promot-ing an agenda that includes increased use of re-newable energy (Sine & Lee, 2009). Yet, none ofthese explanations fully describes the emergenceof cooperative ethanol plants where local endog-enous processes were the driving force. We ad-opted a community ecology approach precisely

Figure 2(continued)


because it allows analysis of the endogenous dy-namics of form emergence while simultaneouslycapturing the heterogeneity of community char-acteristics (Freeman & Audia, 2006; Marquis &Lounsbury, 2007). We hope we have shown thatthis focus on direct strategic and/or contentiousinteractions between different organizationalforms, in different geographical settings, pro-vides a more comprehensive understanding ofthe diffusion of the cooperative form.

Our key findings can be summarized as follows.The diffusion of ethanol corporations in local com-munities generally serves to decrease the foundingrate of cooperatives as predicted by economic ac-counts. However, in communities motivated by“hot cognition,” corporate prevalence mobilizesfarmers to establish cooperatives. This applies es-pecially when corporations can vertically integrate,or when communities have anti-corporate senti-ments or a well-established activist organizationalsupport structure. These contextual features atten-

uate the negative, competitive interdependence be-tween these forms.

We found also that the identity of the corporationmatters a great deal in communities with a generalanti-corporate climate. When the common foe has aclear oppositional identity (agribusiness and oil),the competitive interdependence between the or-ganizational forms may even be reversed, and be-come mutualistic in such communities. Studiesshow that large businesses with controversial iden-tities generally avoid expansion into communitiescharacterized by an anti-corporate climate (Bell,2009; Holmes, 1998; Rao et al., 2011), and, if theydo enter, are met with community protests andpickets (Bell, 2009; Rao et al., 2011). We identifieda more dramatic response of the community orga-nizing as a cooperative to challenge these largebusinesses directly, in both input and outputmarkets.

Contrary to expectations, the moderating role of avertical integration threat was found not to depend

TABLE 3Cox Proportional Hazard Estimates of Cooperative Plant Founding

Model 1 Model 2 Model 3 Model 4

Corn price �0.67 (0.46) �0.66 (0.45) �0.67 (0.45) �0.68 (0.46)County corn production 0.65*** (0.19) 0.64*** (0.19) 0.65*** (0.19) 0.66*** (0.19)Mean farm size in the county �0.13 (0.18) �0.19 (0.19) �0.22 (0.17) �0.14 (0.18)Livestock in the county 0.85 (2.04) 0.77 (2.04) 0.91 (2.01) 0.88 (2.07)County racial diversity �1.05 (1.75) �1.09 (1.74) �1.09 (1.75) �1.12 (1.73)Corn processors in the county �0.20 (0.21) �0.21 (0.21) �0.18 (0.21) �0.20 (0.21)State ethanol consumption �0.02 (0.03) �0.03 (0.03) �0.03 (0.03) �0.02 (0.03)State ethanol exemption 0.08 (0.06) 0.08 (0.06) 0.08 (0.06) 0.08 (0.06)Coalition member 1.51* (0.89) 1.58* (0.88) 1.55* (0.88) 1.55* (0.88)Retail gas price �1.46*** (0.33) �1.45*** (0.33) �1.47*** (0.33) �1.47*** (0.33)Minimum efficiency scale �4.21*** (1.62) �4.16*** (1.60) �4.19*** (1.62) �4.23*** (1.63)No. farm. coops in the state 0.75* (0.43) 0.79* (0.41) 0.70 (0.44) 0.72* (0.42)Local ethanol production capacity (LEPC)—Coop. 0.12 (0.09) 0.14* (0.08) 0.12 (0.09) 0.12 (0.09)Vertical integration threat (VIT) �0.80* (0.48) �1.22** (0.50) �0.94* (0.51) �0.83* (0.47)Anti-corporate climate (AC) 0.18 (0.43) 0.21 (0.44) 0.02 (0.48) 0.19 (0.42)Supportive org. structure (SOS) 0.01 (0.07) �0.00 (0.07) 0.00 (0.07) �0.11 (0.08)LEPC—Corp. (Agro � Oil) �0.13*** (0.05) �0.17*** (0.05) �0.22*** (0.05) �0.59 (0.42)LEPC—Corp. (Others) �0.02 (0.05) �0.07 (0.05) �0.05 (0.05) �0.31*** (0.11)LEPC—Corp. (Agro � Oil) � VIT 0.19* (0.10)LEPC—Corp. (Others) � VIT 0.28*** (0.10)LEPC—Corp. (Agro � Oil) � AC 0.24** (0.11)LEPC—Corp. (Others) � AC 0.04 (0.07)LEPC—Corp. (Agro � Oil) � SOS 0.08 (0.07)LEPC—Corp. (Others) � SOS 0.05*** (0.02)Region dummies Yes Yes Yes YesNumber of spells 444034 444034 444034 444034Log-likelihood �786.29 �783.75 �784.67 �785.43

Note. Robust standard errors calculated by clustering around states.* p � 0.10

** p � 0.05*** p � 0.01; Coefficient significance tests are two-tailed


on the identity of the local corporations. This isperhaps because the potential backward integrationof corporate plants into grain production threatensthe core identity of the farmer, regardless of whichactors are involved. Similarly, the presence of asupportive organizational infrastructure attenuates thenegative effect of corporate production capacity on co-operative founding, irrespective of corporate identity.Interestingly, this overturns the alternative view that alongstanding social movement organization, such as theFarm Bureau, might work as a mobilizing structureagainst, rather than in favor of, a cooperative; either dueto the former’s fear of loss of status and function as a keycommunity organization, or by assuming a brokeragerole between corporate buyers and farmers, hence sup-pressing the emergence of a cooperative while strength-ening its own economic role and social status.

These “non-findings” do not necessarily under-mine our claim that “hot” cognitions and ideolog-ical mobilization are important prerequisites forthe emergence of the cooperative form. Rather, theysuggest that complementary mobilization processesare also at work, which act as substitutes for the sa-liency of the particular corporate identity. We spec-ulate that a local organizational infrastructure thatsupports the cooperative form as an alternative tocapitalism facilitates mobilization, irrespective ofthe extent to which the corporate identity of the ri-val is controversial. Similarly, when all types ofcorporations represent a threat to community au-tonomy, corporate identity does not matter. In con-trast, we find that corporate identity becomes sa-lient only when antagonistic ideologies clash most;that is, when the mass-market logic of big businessinvades communities with an anti-corporate cli-mate. Since anti-corporate campaigning is broaderbased, more vibrant, and more inclusive (generalist)in its corporate targets in these communities thanthose in which farmers live under corporate entrythreat, community members are also more likely tohear about and eventually generalize from big busi-nesses in other industries, which will make the iden-tities of big oil and agribusinesses even more salient(Jonsson et al., 2009). Unraveling the subtleties ofhow different features of the local context interactwith corporate identity in affecting mobilization is animportant avenue for further research.

Our study contributes to organization theory byinforming us about the origins of form diversity andthe emergence of cooperatives in a novel way. Bymodeling the dynamic interplay between corpora-tions and cooperatives, we unraveled the relativeimpact of instrumental, cold decision factors

(which have been underemphasized in socialmovement explanations of cooperatives) and ideo-logical, hot decision factors (which have relativelybeen ignored in economic accounts) in the cooper-ative form emergence process. These explanationswould appear to be complementary. In this study,we showed that the trade-off between these twoexplanations depends on a local community’s ide-ological blueprint and, to a lesser extent, the iden-tity of the enemy. Hansmann (1996) argues that amajor problem related to cooperatives is their highcost of ownership, especially when member prefer-ences are heterogeneous. Our study suggests thatthe local context is important precisely because itunites local actors behind common ideologies, re-ducing such heterogeneity and the cost of collec-tive decision-making.

These insights allow the observation that coop-eratives are organizational anomalies due to thepresence of economic disincentives; nevertheless,there are waves of cooperative formation. For in-stance, the peak of cooperative formation in theUnited States at the turn of the 19th century corre-sponds to major changes in the U.S. economicstructure, which triggered collective action to de-fend the values of autonomy and craftsmanshipagainst the rise of impersonal administrative hier-archies (Aldrich & Stern, 1983). As a result, ideo-logical incentives contributed strongly to the diffu-sion of cooperatives in this period (Schneiberg etal., 2008). In periods marked by low ideologicalcontestation, cooperative founding was intermit-tent. Similarly, our approach explains why territo-rial partitioning between cooperatives and corpora-tions is juxtaposed to geographical areas whereboth forms exist side by side. In the absence of localanti-corporate sentiments, corporations will domi-nate provided there are sufficient economic incen-tives; otherwise, cooperatives might fill the gap.

Our study contributes also to the so-called “neworganizational synthesis” in social movement re-search, which aims to study how social movementsshape organizational fields, and vice versa (Sch-neiberg et al., 2008). Recently, scholars have stud-ied the impact of social movements on the diffu-sion of organizational forms. Schneiberg et al.(2008) analyze how the Grange affected the birth ofcooperatives at the beginning of the 20th century;Sine and Lee (2009) examine the impact of envi-ronmental movements on entrepreneurial activityin the wind energy sector; and Hiatt et al. (2009)link the American Temperance movement to thefounding of firms that produced new kinds of bever-


ages. Haveman et al. (2007) show that social move-ments also have diffuse effects across various do-mains of social life, demonstrating how theProgressive movement strongly affected the spread ofbureaucracy in the early Californian thrift industry.All these studies underscore that collective actionand associated resource mobilization play a majorrole in the evolution of organizational fields, and thatmovements have mobilized against bureaucracies,markets, and corporations (Schneiberg et al., 2008).

Although collective action is generally targetedtowards a common foe (Pozner & Rao, 2006; Rao etal., 2011), direct rivalry—with respect to both re-sources and ideology—is seldom explicitly takeninto account. The present study shows that a com-munity ecology approach, with a focus on directrivalry between different populations, can fill thisgap. Our study shows that the emergence of coop-eratives critically depends on the joint presence ofanti-corporate movements and direct rivals. It isinteresting that neither of these factors in isolationappears to spur cooperative formation. For in-stance, we do not find a direct effect of Farm Bu-reau tenure on the emergence of cooperatives. Sim-ilarly, although Schneiberg et al. (2008) find thatGrange membership has a strong positive effect onthe emergence of the cooperative form at the begin-ning of the 20th century, its effect on cooperativediffusion in the subsequent three decades was neg-ative. Apparently, direct rivalry is an importantprerequisite for the emergence of mobilization and,ultimately, diffuse mutualism; this finding stressesthe added value of a community ecology approachto social movement research.

Finally, there is a growing body of research thatshows that the legal environment affects popula-tion dynamics. These studies typically analyze thedirect effects of regulation on the vital rates oforganizational populations (Marquis & Huang,2009). We contribute to this literature by showingthat regulation strongly affects the interdependen-cies between organizational forms, and, as a result,shapes the dynamics of organizational diversity.Specifically, we have provided evidence that farm-ers in communities with no RTW laws (implying ageneral anti-corporate climate) react against corpo-rations by forming cooperatives. We found also thatcorporations are more likely to be regarded as athreat, spurring the emergence of cooperatives incommunities with no anti-corporate farming laws.Interestingly, the presence of a legal settlement inresponse to anti-corporate sentiment over the rightto own farms tempers future ideological rivalry in

related domains of economic activity; that is, be-tween ethanol corporations and corn farmers.

This study has some limitations. First, the mea-sures we use to assess the likelihood of “hot” cogni-tion in a local community have the advantage ofallowing us to analyze cooperative founding over along period of time. However, our proxy for generalanti-corporate climate is based on the presence orabsence of RTW laws at state level while we modelcooperative establishment at county level (Rao et al.,2011). This ignores potential within-state heterogene-ity in anti-corporate sentiments among counties. Fu-ture work could explore the possibility of developingmore fine-grained measures; for instance, by collect-ing data on local anti-corporate activist events orcounty referendums targeting corporations (see e.g.,Ingram, Yue, & Rao, 2010). Similarly, although FarmBureau tenure is a better proxy for supportive organ-izational infrastructure than a simple dummy vari-able that captures the presence or absence of localFarm Bureau, the cumulative number of local coop-eratives established by the Farm Bureau might be abetter alternative. However, this information is notavailable in conventional data sources.

Second, we model cooperative entry as a functionof corporate prevalence, ignoring the impact of in-cumbent cooperatives on corporate vital rates. As afirst step, this is a sensible approach, because coop-eratives are often assumed to emerge in response tothe presence of the dominant corporate form. How-ever, future work should develop theories of howcorporations react to the emergence of cooperatives.

Third, we study the birth of cooperatives in a sin-gle, albeit important, industry. Future research mightexplore the generalizability of our findings. Neverthe-less, we believe that the community ecology ap-proach developed in this paper should prove a valu-able tool for future work on the emergence ofcooperatives, and institutional diversity more generally.

Finally, a complete understanding of the dynam-ics of the cooperative form would require a focuson the processes of both cooperative founding andcooperative survival. Their relative scarcity can berelated to high failure rate and/or the fact that manyeventually revert to the corporate form (Aldrich &Stern, 1983; Ben-Ner, 1984). Whyte and Blasi(1982) claim that the weakness of many coopera-tives is the result of an absence of strong guidingideologies. Thus, it could be particularly fruitful toinvestigate whether the type of incentives, i.e., ma-terial or ideological, that guide the founding of acooperative also affect its longevity. Such an inves-tigation would help to explain the scarcity of coop-


eratives, and the simultaneous success, in certaincircumstances, of this alternative organizationalform in a world dominated by corporations.

REFERENCES

Aldrich, H. E., & Stern, R. N. 1983. Resource mobilizationand the creation of US producer’s cooperatives,1835–1935. Economic and Industrial Democracy,4: 371–406.

Barnett, W. P., & Carroll, G. R. 1987. Competition andmutualism among early telephone companies. Ad-ministrative Science Quarterly, 32: 400–421.

Barnett, W. P., & Woywode, M. 2004. From red Vienna tothe Anschluss: Ideological competition among Vien-nese newspapers during the rise of national socialism.American Journal of Sociology, 109: 1452–1499.

Baum, J. A. C., & Oliver, C. 1996. Toward an institutionalecology of organizational founding. Academy ofManagement Journal, 39: 1378–1427.

Bell, S. E. 2009. “There ain’t no bond in town like there usedto be”: The destruction of social capital in West Virginiacoalfields. Sociological Forum, 24: 631–657.

Ben-Ner, A. 1984. On the stability of the cooperative typeof organization. Journal of Comparative Econom-ics, 8: 247–260.

Berlage, N. K. 2001. Organizing the farm bureau: Family,community and professionals: 1914–1928. Agricul-tural History, 75: 458–469.

Bonin, J. P., Jones, D. R., & Putterman, L. 1993. Theoret-ical and empirical studies of producer cooperatives:Will ever the Twain meet? Journal of EconomicLiterature, 31: 1290–1320.

Boone, C., Divarci, A., & van Witteloostuijn, A. 2011.deological polarization and organizational formevolution: A study of Islamic–secular rivalry andhigh schools in Turkey, 1971–1998. University ofAntwerp, unpublished manuscript.

Brickson, S. L. 2005. Organizational identity orientation:Forging a link between organizational identity andorganizations’ relations with stakeholders. Adminis-trative Science Quarterly, 50: 576–609.

Carroll, G. R., & Swaminathan, A. 2000. Why the micro-brewery movement? Organizational dynamics of re-source partitioning in the U.S. brewing industry.American Journal of Sociology, 106: 715–762.

Caves, R. E., & Petersen, B. S. 1986. Cooperatives’ sharesin farm industries: Organizational and policy factors.Agribusiness, 2: 1–19.

Cook, M. L. 1995. The future of U.S. agricultural cooper-atives: A neo-institutional approach. AmericanJournal of Agricultural Economics, 77: 1153–1159.

DiMaggio, P. 2002. Why cognitive (and cultural) sociol-ogy needs cognitive psychology. In K. A. Cerulo(Ed.), Culture in mind: Toward sociology of cultureand cognition: 274–282. New York, NY; London,England: Routledge.

Estrin, S., & Jones, D. 1992. The viability of employee-owned firms: Evidence from France. Industrial andLabor Relations Review, 45: 323–339.

Foreman, P., & Whetten, D. A. 2002. Members’ identifi-cation with multiple-identity organizations. Organ-ization Science, 13: 618–635.

Freeman, J., & Audia, P. G. 2006. Community ecology andthe sociology of organizations. Annual Review ofSociology, 32: 145–169.

Gómez-Mejía, L. R., Haynes, K. T., Núñez-Nickel, M.,Jacobson, K. J. L., & Moyano-Fuentes, J. 2007. Socio-emotional wealth and business risks in family-con-trolled firms: Evidence from Spanish olive oil mills.Administrative Science Quarterly, 52: 106–137.

Greve, H. R., Pozner, J.-E., & Rao, H. 2006. Vox populi:Resource partitioning, organizational proliferation,and the cultural impact of the insurgent microradiomovement. American Journal of Sociology, 112:802–837.

Greve, H. R., & Rao, H. 2012. Echoes of the past: Organ-izational foundings as sources of an institutionallegacy of mutualism. American Journal of Sociol-ogy, 118: 635–675.

Hannan, M. T. 1988. Organizational population dynam-ics and social change. European Sociological Re-view, 4: 95–109.

Hannan, M. T., & Freeman, J. 1977. The population ecol-ogy of organizations. American Journal of Sociol-ogy, 82: 929–964.

Hansmann, H. 1996. The ownership of enterprise. Cam-bridge, MA: Harvard University Press.

Haveman, H. A., Rao, H., & Paruchuri, S. 2007. Thewinds of change: The progressive movement and thebureaucratization of thrift. American SociologicalReview, 72: 117–142.

Hiatt, S. R., Sine, W., & Tolbert, P. 2009. From Pabst toPepsi: The deinstitutionalization of social practicesand the creation of entrepreneurial opportunities.Administrative Science Quarterly, 54: 635–667.

Holmes, T. J. 1998. The effect of state policies on thelocation of industry: Evidence from state borders.Journal of Political Economy, 106: 667–705.

Ingram, P., & Rao, H. 2004. Store wars: The enactmentand repeal of anti-chain legislation in America.American Journal of Sociology, 100: 446–487.

Ingram, P., & Simons, T. 2000. State formation, ideolog-ical competition, and the ecology of Israeli workers’


cooperatives, 1920–1992. Administrative ScienceQuarterly, 45: 25–53.

Ingram, P., & Yue, L. Q. 2008. Structure, affect and iden-tity as bases of organizational competition. Acad-emy of Management Annals, 2: 275–303.

Ingram, P., Yue, L. Q., & Rao, H. 2010. Trouble in store:Probes, protests, and store openings by Wal-Mart,1998–2007. American Journal of Sociology, 116:53–92.

Jones, D. 1979. U.S. producer cooperatives: The record todate. Industrial Relations, 18: 342–357.

Jonsson, S., Greve, H. R., & Fujiwara-Greve, T. 2009.Undeserved loss: The spread of legitimacy loss toinnocent organizations in response to reported cor-porate deviance. Administrative Science Quar-terly, 54: 195–228.

Kile, O. M. 1948. The Farm Bureau through three de-cades. Baltimore, MD: Waverly Press.

King, B. G., & Pearce, N. A. 2010. The contentiousness ofmarkets: Politics, social movements, and institu-tional change in markets. Annual Review of Sociol-ogy, 36: 249–267.

Krauss, C. 2009. Big oil warms to ethanol and biofuelcompanies. New York Times, 26/5.

Lobao, L. M. 1990. Locality and inequality: Farm andindustry structure and socioeconomic condition.Albany, NY: State University of New York Press.

Lyson, T. A., & Welsh, R. 2005. Agricultural industrial-ization, anti-corporate farming laws, and rural com-munity welfare. Environment and Planning, 37:1479–1491.

Marquis, C., & Huang, Z. 2009. The contingent nature ofpublic policy and the growth of U.S. commercialbanking. Academy of Management Journal, 52:1222–1246.

Marquis, C., & Lounsbury, M. 2007. Vive la résistance:Competing logics and the consolidation of U.S. com-munity banking. Academy of Management Journal,50: 799–820.

McNew, K., & Griffith, D. 2005. Measuring the impact ofethanol plants on local grain prices. Review of Ag-ricultural Economics, 27: 164–180.

Meyer, A. 2008. Farming fuels: Searching for revitaliza-tion in agricultural bioeconomy. Doctoral disserta-tion, The Pennsylvania State University. (UMI No.3346347)

Murphy, S., Burch, D., & Clapp, J. 2012. Cereal secrets:The world’s largest grain traders and global agri-culture. Oxfam research reports. Cowley, Oxford,England: Oxfam GB for Oxfam International.

Pozner, J-E., & Rao, H. 2006. Fighting a common foe:

Enmity, identity and collective strategy. Advancesin Strategic Management, 23: 445–479.

Rao, H. 1998. Caveat emptor: The construction of non-profit watchdog organizations. American Journal ofSociology, 104: 912–951.

Rao, H., Yue, L. Q., & Ingram, P. 2011. Laws of attraction:Regulatory arbitrage in the face of activism in right-to-work states. American Sociological Review, 76:365–385.

Schneiberg, M. 2007. What’s on the path? Path depen-dence, organizational diversity and the problem ofinstitutional change in the U.S. economy. Socio-Eco-nomic Review, 5: 47–80.

Schneiberg, M., King, M., & Smith, T. 2008. Social move-ments and organizational form: Cooperative alterna-tives to corporations in the American insurance,dairy and grain industries. American SociologicalReview, 73: 635–667.

Schneiberg, M., & Lounsbury, M. 2008. Social move-ments and institutional analysis. In R. Greenwood,C. Oliver, K. Sahlin-Andersson & R. Suddaby (Eds.),The handbook of organizational institutionalism:648–670. London, England: Sage Publications.

Simons, T., & Ingram, P. 2003. Enemies of the state:Interdependence between institutional forms andthe ecology of the kibbutz, 1910–1977. Administra-tive Science Quarterly, 44: 592–621.

Simons, T., & Ingram, P. 2004. An ecology of ideology:Theory and evidence from four populations. Indus-trial and Corporate Change, 13: 33–59.

Sine, W. D., Haveman, H. A., & Tolbert, P. S. 2005. Riskybusiness? Entrepreneurship in the new indepen-dent-power sector. Administrative Science Quar-terly, 50: 200–232.

Sine, W. D., & Lee, B. H. 2009. Tilting at windmills? Theenvironmental movement and emergence of the U.S.wind energy sector. Administrative Science Quar-terly, 54: 123–114.

Staatz, J. M. 1987. Farmers’ incentives to take collectiveaction via cooperatives: A transaction-cost approach.USDA ACS Service Report, 18: 87–107. Washing-ton, D.C.: U.S. Dept. of Agriculture, Agricultural Co-operative Service.

StataCorp. 2013. Stata 13 Base reference manual. Col-lege Station, TX: Stata Press.

Stinchcombe, A. L. 1965. Social structure and organiza-tions. In J. G. March (Ed.), Handbook of organiza-tions: 142–193. Chicago, IL: Rand McNally.

Thompson, S. 2004. Community investments helpedlaunch plant. Rural Cooperatives, 71: 4–10.

Weber, K., Heinze, K. L., & DeSoucey, M. 2008. Forage forthought: Mobilizing codes in the movement for


grass-fed meat and dairy products. AdministrativeScience Quarterly, 53: 529–567.

Whyte, W. F., & Blasi, J. R. 1982. Worker ownership,participation and control: Toward a theoreticalmodel. Policy Sciences, 14: 137–163.

Williamson, O. E. 1980. The organization of work: Acomparative institutional assessment. Journal ofEconomic Behavior and Organization, 1: 5–38.

Williamson, O. E. 1985. The economic institutions ofcapitalism. New York, NY: Free Press.

Christophe Boone ([email protected]) isprofessor of Organization Theory and Behavior at the

Faculty of Applied Economic Sciences at the Universityof Antwerp (Belgium). He received his PhD from thesame university. His research interests focus on the an-tecedents and consequences of team and organizationaldiversity, and on the neuroeconomics of decisionmaking.

Serden Özcan ([email protected]) is a professorand chair of Innovation and Organization at WHU–OttoBeisheim School of Management (Germany). He receivedhis PhD from Copenhagen Business School. His researchinterests focus on the demography of organizations andproducts, the evolution of industries, entrepreneurial fi-nance, social entrepreneurship, and the life course mod-els of entrepreneurs.


Copyright of Academy of Management Journal is the property of Academy of Managementand its content may not be copied or emailed to multiple sites or posted to a listserv withoutthe copyright holder's express written permission. However, users may print, download, oremail articles for individual use.

why do cooperatives emerge in a world dominated by ... · research into why and where cooperatives...

Documents