building inclusive markets in rural bangladesh: how

HAL Id: hal-02276707https://hal.archives-ouvertes.fr/hal-02276707

Submitted on 3 Sep 2019

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Building Inclusive Markets in Rural Bangladesh : HowIntermediaries Work Institutional Voids

Johanna Mair, Ignasi Marti, Marc Ventresca

To cite this version:Johanna Mair, Ignasi Marti, Marc Ventresca. Building Inclusive Markets in Rural Bangladesh : HowIntermediaries Work Institutional Voids. Academy of Management Journal, 2012, pp.819-850 P. �hal-02276707�

https://hal.archives-ouvertes.fr/hal-02276707

https://hal.archives-ouvertes.fr

* Academy of Management Journal2012, Vol. 55, No. 4, 819-850.http://dx.doi.org/10.5465/amj.2010.0627

BUILDING INCLUSIVE MARKETS IN RURAL BANGLADESH:HOW INTERMEDIARIES WORK INSTITUTIONAL VOIDS

JOHANNA MAIRStanford University

IGNASI MARTÍEM LYON Business School

MARC J. VENTRESCAUniversity of Oxford

Much effort goes into building markets as a tool for economic and social development;those pursuing or promoting market huilding, however, often overlook that in too manyplaces social exclusion and poverty prevent many, especially women, from participatingin and accessing markets. Building on data from rural Bangladesh and analyzing the workof a prominent intermediary organization, we uncover institutional voids as the source ofmarket exclusion and identify two sets of activities—redefining market architecture andlegitimating new actors-^as critical for huilding inclusive markets. We expose voids asanalytical spaces and illustrate how they result from conflict and contradiction amonginstitutional bits and pieces from local political, community, and religious spheres. Ourfindings put forward a perspective on market huilding that highlights the on-the-grounddynamics and attends to the institutions at play, to their consequences, and to a morediverse set of inhabitants of institutions.

If someone who has no property rights under thelaw, who has had no formal education, who has nolegal right to divorce, who will very likely be beatenif sbe seeks employment outside the home, says thatshe endorses traditions of modesty, purity, and self-abnegation, it is not clear that we should considerthis the last word on the matter.

-Marta Nussbaum

This article is the outcome of a truly collaborative effort,and all three authors contributed equally. Although ourwork is about breaking conventions, we for once adhere toconventions in publishing and list authors in alphabeticalorder. We would like to tbank Tima Bansal for ber excep-tional guidance and tbree anon)rmous reviewers who en-couraged and helped us to find and refine the empirical andtheoretical nuggets in this project. This article would nothave been possible without BRAG and the people that makeBRAG. We particularly thank Fazle Abed for sharing in-sights and wisdom. We are also grateful for the followingpeople who graciously shared their concerns and sugges-tions: Yasser Bhatti, Dana Brown, David Gourpasson, Gré-goire Groidieu, Frédéric Delmar, Gatherine Dolan, IngridErickson, Fabrizio Ferraro, Royston Greenwood, GarolLeonard, Ray Loveridge, John Meyer, Philippe Monin, MariSako, Saras Sarasvathy, Marc Schneiberg, Dick Scott, andGhrisüan Seelos. This work also greatly benefited fromdiscussions at the 2006 Summer Institute "Trajectories ofCapitalism" at the Genter of Advanced Studies in the Be-havioral Sciences, an economic sociology panel at the 2006Annual Meetings of the American Sociological Association,the 7th Neo Institutionalism Workshop at EM LYON Busi-ness School, and the 2011 Strategy and Innovation confer-ence at the Said Business School.

Researchers and policy makers have long arguedthat markets are engines for economic growth andthat market-based activities constitute an importanitool for social progress, economic empowerment,and human development (Seelos & Mair, 2007;Stiglitz, 1989; UNDP, 2008). In this spirit, manycontemporary policy initiatives make poor wom-en's market access and participation central to theirexperiments and pilot programs—an aspect of pol-icy broadly referred to as "developing inclusivemarkets" (Mendoza & Thelen, 2008). Yet these well-intended interventions often overlook the role thatlocal institutions such as customs, religious credos,,and social norms play in compromising the potentialfor women's economic activify. One of our fieldworkinformants in Bangladesh said, "How can I go to themarket? I am a woman!" Her words exemplify howwomen in too many places are excluded from market-based activifies and expose how social conventionsprevent them from market access. The simple quotealso points to the limits of promoting the role ofmarkets without analyzing how local context andinstitutional arrangements shape markets and mar-ket-based activify (Banerjee & Duflo, 2011).

Institutions, metaphorically referred to as the"rules of the game" (North, 1990), matter for markets;they enable and support market activify (Campbell &Lindberg, 1990; De Soto, 2000; Greif, 2006; Sen,1999). Where such institutions are absent or weak,management and strategy scholars point to the pres-

819

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.

820 Academy of Management Journal August

ence of "Institutional voids," realities that can impactmarket formation, economic growth, and develop-ment (Khaima & Palepu, 1997; Webb, Tihanyi, Ire-land, & Sirmon, 2009). Researchers further note thatabsent and weak market institutions reinforce exist-ing social inequalities as market access and opportu-nity are governed by local institutional arrangements(Crow, 2001; Rodrik, 2007). Such local arrangementsconsist of complex interlocks of formal institutions,such as constitutions, laws, properfy rights, and gov-ernmental regulations, and informal institutions,such as customs, traditions, and religious beliefs(Fligstein, 2001; North, 1991), that not only enablebut also constrain market activity. They determinethe rules of the game and, importantly, who is al-lowed to play.

Local realities are the point of departure for thisstudy. Otir objective is to clarify market-buildingprocesses by interrogating the concept of institu-tional voids in institutionally complex contexts—particularly where markets and market-based activ-ities are seen as tools for economic development.Drawing on an in-depth qualitative study, we focusour empirical efforts on the case of Bangladesh andthe work of BRAC,̂ a prominent, pervasive localintermediary agency. Whereas many studies viewinstitutional voids as "empty" of specific institu-tions, our findings suggest that voids occinr amidstinstitutional plurality and are the intermediate out-come of conflict and contradiction among local po-litical, community, and religious spheres. Thisfresh perspective detects institutional voids as animportant driver of market exclusion and providesan analytical anchor for the study of market-build-ing processes. Our analysis of the varied work ofBRAC over several decades suggests that marketaccess and participation are negotiable and marketboundaries are potentially permeable for actorswho have been excluded.

The perspective on market building we put for-ward highlights "on-the-ground" dynamics and at-tends to the consequences of market building forthe people involved. Otn: findings complement andextend research on the institutional formation andinfrastructure of markets by highlighting marketbuilding as inhabited—that is, as replete with peo-ple and activity, albeit focused to varied purposes.

^ BRAC originally stood for "Bangladesh Rehabilita-tion Assistance Committee," in line with its focus onrelief work after natural disasters. In 1973, BRAC wasrenamed "Bangladesh Rural Advancement Committee"to reflect its focus on development. Almost three decadeslater, when BRAC was already working in the country'surban slums and had started to work internationally, itbecame "Building Resources Across Communities."

More specifically, we explain how markets can be-come inclusive places: legitimated arenas for inter-dependent social and economic activity where for-mal possibilities align with practical access acrossgender, race, religion, and social class.

MARKET BUILDING ANDINSTTTUTIONAL VOIDS

Markets, Institutions, and Context

Institutionalists throughout the social sciencesagree that markets are systems of economic ex-change and spaces for social interaction as well ascomplex bimdles of institutions (Geertz, 1978; LaPorta, Lopez-de-Silanes, Shleifer, & Vishny, 1998;McMillan, 2002; North, 1990). A context-specificand often complex assembly of institutions includ-ing both formal rules and informal norms generatesmarket microstructures and assembly of institu-tions that configure sociopolitical contexts and or-ganize the "terms of action" (Fligstein, 2001). Theseinstitutional architectures vary across polities,comprising rules for what issues are public andcollective and which actors are authorized to en-gage these issues (Jepperson & Meyer, 1991). Alongwith recognizing such cross-polity variation, schol-ars warn of the "need to maintain a healthy scepti-cism toward the idea that a specific type of institu-tion is the only type that is compatible with awell-functioning market economy" (Rodrik, 2007:162-163). This contention signals the need for in-quiries about market-building processes in institu-tionally complex contexts (Granovetter & McGuire,1998; Mair & Marti, 2009; Santos & Eisenhardt,2009; Spicer, McDermott, & Kogut, 2000).

Building markets is neither easy nor unproblem-atic. Regardless of how efficient a particular insti-tutional arrangement has proven itself in a specificcontext, the efficacy and impact of that same insti-tutional assembly will likely vary in a differentcontext. Institutional frameworks of meaning andauthority shape the conditions of market buildingand the particular tactics that actors use (Biggart &Guillen, 1999; Hamilton & Biggart, 1988). Recentwork in the economic sociology of markets (Flig-stein, 2001; Zelizer, 2005), the varieties of capital-ism (Hall & Soskice, 2001), and institutional eco-nomics (Greif, 2006; North, 1990; Rodrik, 2007)points to the varied intertwining of culture andpolitics in the social organization of markets andeconomies. Despite observing differences in fociand mechanisms, scholars argue that markets areconstructed rather than "natural" or spontaneousentities. This scholarship also challenges the dom-inant, simplified (and simplifying) view of markets

2012 Mair, Martí, and Ventresca 821

as cleanly distinguished from the infrastructure oftheir local contexts. Careful attention to the rules ofthe game in markets points to the relevance of thecontext and processes hy which they emerge (Flig-stein, 2001; Creif, 2006; North, 1990; Rodrik, 2007).The process of institutionalizing the rules that gov-ern exchange and market-based activity is ongoingand observable, and as such it provides a lens forohserving market huilding and the activities of di-verse institutional actors (Santos & Eisenhardt,2009; Spicer et al., 2000; Stark, 1996).

Institutional Voids

Recent work on institutional voids in strategy andeconomics recognizes the central role that insütu-fions—and their ahsence—play in developing marketeconomies and in shaping the behavior of a particularset of actors: firms and entrepreneurs. Standardclaims locate the source of key market institutionssuch as properfy rights and autonomy in state actionand rules (Camphell & Lindherg, 1990; La Porta et al.,1998; North, 1990). Where properly designed andimplemented, these institutions provide the rules ofthe game that support formaüon of ideal typical mar-kets to form. If these institutions are ahsent or weak,the argument goes, institufional voids occur, and acompensatory social structure is needed to spur mar-ket formation and operafion (Creif, 2006; IQiarma &Palepu, 1997). Building on new institutional econom-ics and agency theory, Khanna and Palepu (1997,2000) extended LefPs (1976) work on husiness groupsin developing economies, to contend that large busi-ness groups imitate and subsfitute for missing insü-tuüons to ensure market function in the event ofmarket failures due to the presence of insfitutionalvoids (Khanna, Palepu, & Sinha, 2005; Kharma &Rivkin, 2006). In this stream of research, institutionalvoids are typically presented as inhihitors of the es-tahlishment of Western-style markets. The proposedsolutions tj^ically favor the transfer of institutionaltechnology as a compensatory mechanism over localexperimentation and recombination.

Following Khanna and colleagues, and drawingfrom a broader spectrum of research on institutions(North, 1990; Powell & DiMaggio, 1991), recentscholarship in organizational theory and strategyhas focused on the consequences of institutionalvoids on business strategy. Meyer, Estrin, Bhaumik,and Peng (2009) showed that the relative strengthor weakness of various institutional ftameworksimpacts alternative modes of entry in India, Viet-nam, South Africa, and Egypt. Puffer, McCarthy,and Boisot (2009) revealed that weak and lackingformal institutions in Russia and China force entre-preneurs to rely on trust within networks as well as

on blat and guanxi, respectively (see also Ahlstrom& Bruton, 2006). Additionally, Chakrabarty (2009)showed that institutional voids influence familyownership patterns in 27 countries.

A second set of studies points to the impact thatinstituüonal voids have on entrepreneurial processesin "base of the pyramid" (BOP) markets and localinformal economies. Webb et al. (2009) suggested sev-eral ways that insfitutional incongruence and weakenforcement of formal insfitutions facilitate entrepre-neurial processes in an informal economy. Cross-sec-tor alliances between commercial companies and lo-cal social entrepreneurs have been shown to removethe "hurdles of implemenfing BOP models" (Seelos &Mair, 2007:49) by reducing the uncertainfy caused byweak market institufions (Webb, Kistruck, h-eland, &Ketchen, 2010). Together, these studies reinforce theimportance of understanding the institutional infra-structure of markets and foreground the impact ofinstitutional voids on effecüve market functioning(North, 1990; Peng, Sun, Pinkham, & Chen, 2009).They also showcase a productive conversation be-tween insfitutional economists and strategy research-ers, one that highlights a shared interpretation of in-stituüonal voids as spaces empfy of instituüons. Thisinterpretation stands in contrast to the interpretafionfavored by sociologists and anthropologists, who em-phasize the abundance and complexify of insfitutionspresent in similar contexts and situations and whofocus on the people parficipaüng (and not participat-ing) in markets (Banfield, 1958; Friedland & Alford,1991; Ceertz, 1978; Zelizer, 2010). Engaging the eco-nomic and strategy line of research with the sociolog-ical and anthropological line of research opens up asalient debate on the processes involved in buildingmarkets as inclusive arenas for social interaction andeconomic exchange. Although previous studies haveelaborated on a varied set of consequences resultingfrom the presence of institutional voids, much isstill unknown about how insfitutional voids are con-stituted, how they relate to existing institutional ar-rangements, and how they matter for localpopulations.

Market Building in and aroimd Institutional Voids

Institutional economists and their colleagues infinance have developed suhstantial empirical andtheoretical arguments about the institutional infra-structures and rules that support market formationand associated economic development (La Porta etal., 1998; Morrison, 2004; Morrison & White, 2009;North, 1990). This diverse work finds suhstantialpolicy expression in the legacy of North and in thework of neo-Hirschman scholars (Chan, 2002; Ro-drik, 2007). The gist of the argument is that specific


configurations of (micro and macro) institutionsconduce to market formation. In the absence ofthese institutional arrangements (i.e., in the pres-ence of institutional voids), there is much difficultyin establishing markets and the necessary outcomessuch as transparency and efficiency. Much of thepolicy translation of this work, however, reinforcesthe imagery that markets come into being in equi-librium. Moreover, although the attention to the"putting in place" of markets and market reforms isimportant, it neglects the complex work and les-sons for policy from attending to the early stagesand dynamics of market building.

For this focus on early moments, we have re-course to economic and cultural sociologists whosework is reinserting activity into market formation.At the broader societal level, work on markets aspolitics views the institutional architecture ofmarkets as the outcome of social movement-likestruggles between incumbents and challengers(Bourdieu, 2005; Fligstein, 2001), highlighting thestruggles among actors to harness the cultrual, po-litical and institutional resources to organize mar-kets and economies (Anteby, 2010; Biggart & Guil-len, 1999; Hamilton & Biggart, 1988). This researchmakes explicit the link between social movements,corporations, the state apparatus, and markets(Lounsbury, Ventresca, & Hirsch, 2003; Schneiberg,2007; Tilly & Tilly, 1998). Such emerging imageryof market building draws on an understanding ofinstitutions as both obdurate and stable but also notwithout activity and contest.

A parallel stream of scholarship from the socialsciences prompts yet another useful reading on therole of institutions and institutional dynamics inthe process of market building. An important lineof work by sociologists shows that markets are of-ten built with, rather than on, the bit and pieces ofinstitutions (Stark, 1996). These insights repositionattention onto the plurality of incumbent institu-tional arrangements that support economies andmarkets (Hamilton & Biggart, 1988; Ostrom, 1990;Thelen, 2004). This scholarship also shows how, inmany cases, these incumbent institutional arrange-ments supplant institutions tbat support Khanna etal.'s model of the market economy or "market cap-italism." Moreover, whereas many studies on insti-tutional voids have, to a large extent, omitted peo-ple (for an exception see Webb, Tihanyi, Ireland, &Sirmon [2009]) and neglected the disenfranchised,an emerging body of literature is beginning to focuson tbe reality of activity present in "inhabited in-stitutions" (Hallett & Ventresca, 2006: 231). Thisnascent trajectory in the field is (re)infusing insti-tutionalism with a "lost" microsociology (Barley,

2008; Hallett, 2010; Kellogg, 2009), in turn enrich-ing the context of economic sociology.

We draw on this understanding of institutionalconfigurations and dynamics to investigate thesources of institutional voids. We treat voids asanalytical spaces at the interface of several institu-tional spheres, each with its own animating logic ofmeanings and social practices (Friedland & Alford,1991). This reading departs from the conventionalview that privileges modern/Western interpreta-tions of key market institutions and emphasizes thefunctioning of ideal markets. By accentuating thesituated and intermediate nature of institutionalvoids, we develop a view of voids that originates inthe presence of plural, often contending, institu-tional arrangements (Banfield, 1958; Fourcade,2007; Friedland & Alford, 1991; Kogut et al., 2000;Sako, 2009; Stark, 1996), rather than continuing theview of "empty" institutional space. The analyticpoint to be made here is that even the ideal typicalmarket promised by standard market institutionscan reflect broader inequalities in a society andthus result in exclusionary markets. Our approachlinks institutional voids to grounded sources forlimits on market participation and access. This ap-proach also generates conceptual and empiricalclaims about how institutional plurality, conflict,and contradiction contribute to tbeorizing on inclu-sive markets and toward tbeir potential implemen-tation. We also see this as an opportunity to inte-grate scholarship focused on market consequenceswith the standard work on markets and institu-tional voids. Studies of market consequences varyin focus, but are especially well developed in termsof policy issues such as the distributional impactsof markets (Easterly, 2002; Stiglitz, 2000), culturaleffects on endogenous preferences (Bowles, 1998),questions of social empowerment and the exclu-sion from labor markets (Banerjee & Duflo, 2011),and the emerging work on inclusive markets (Men-doza & Thelen, 2008).

To summarize, we extend the research on institu-tions and markets to put forward a perspective onmarket building that highlights tbe on-the-ground dy-namics in complex institutional contexts. To advancethis perspective, we pose two broad questions toguide our empirical analysis of building inclusivemarkets in nural Bangladesh: (1) How do institutionalvoids arise in institutionally complex settings, withwhat consequences for market access and participa-tion? (2) What organizational and other activitieswork these voids to build inclusive markets? To an-swer the first question, we focus on two standardmarket institutions, property rights and autonomy,engaging this standard view witb work on plural in-stitutional spheres. Drawing on this analysis, we ex-


plore the second question with a detailed case studyof a prominent local intermediary agency in Bangla-desh and its portfolio of initiatives and activities toaddress market inclusion.

RESEARCH DESIGN, DATA, AND METHODS

Research Setting

To probe the concept of institutional voids and toexamine market building processes in an institution-ally complex context, we focus on the case of BRAC(Mair & Marti, 2009; Sachs, 2005; Smillie, 2009). Ban-gladesh is a complex country setting in which toanalyze institutional voids and market building. De-spite substantial progress in poverfy alleviation andan overall economic growth rate of 6 percent in recentyears within Üie country, nearly half of the estimated156 million inhabitants of Bangladesh live below thepoverfy line (World Bank, 2010). BRAC, consideredto be the world's largest development organization inreach and staff scale {Economist, 2010; Smillie,2009),^ is present in all 64 districts of Bangladesh,operating in about 70,000 villages—a reach that af-fects the life of 80 percent of the entire population(BRAC, 2009).

Prevailing institutional configurations in Bangla-desh act to limit poor, rural inhabitants—particularlywomen—from accessing and participating in mar-kets. These arrangements have a disproportionate im-pact on access, despite formal constitutional and po-litical guarantees for the equal status of allBangladeshi citizens in all areas of public life (Crow,2001; Pereira, 2002). Market access and participationare further complicated by the amalgam of secularand religious dimensions tíiat define public and eco-nomic life in Bangladesh (Kabeer, 2000). Such struc-tural and institutional complexify also poses a seri-ous impediment to sustaining economic and socialdevelopment (Heritage Foundation, 2010).

Bangladesh's institutional arrangements make ita telling analytic case in which to explore the ex-perimental and "extreme" nature of rural market-building processes (Eisenhardt & Graebner, 2007;Flyvbjerg, 2011; Pettigrew, 1990; Yin, 1994). Ourhigh-relief observations also support a welcomeagenda of formulating generalizable insights in thisscholarly area. Finally, our case selection respondswell to recent calls for "unconventional" organiza-

^ The original and primary work of BRAG is in Ban-gladesh. Since 2002 BRAG has also expanded operationsinto other countries in Asia (Afghanistan, Pakistan, andSri Lanka), Africa (Liberia, Sierra Leone, Soutbern Su-dan, Tanzania, and Uganda), and Latin America (Haiti)with experiments to diffuse its core models.

tional research conducted to develop new knowl-edge about organizational phenomena (Bamberger& Pratt, 2010: 665): our emphasis on an organiza-tion-rich view of market-building processes di-rectly links organizational theory and studies ofmarket building representing other perspectives.

The Organizational Case of BRAC

Development researchers t)^ically distinguishamong a set of different development strategies andorientations among agencies (Körten, 1987). Körtenidentified three distinctive orientations: (a) relief andwelfare, (b) local self-reliance, and (c) sustainable sys-tems development. He further observed that thereexists an "underlying direction of movement thatmakes it appropriate to label these orientations asfirst, second, and third generation" (Körten, 1987:147). Although these generations of orientations si-multaneously coexist within the larger community ofnongovernmental organizations (NGOs), it is only inrare cases that they coexist in a single NGO (Körten,1987). BRAC is one of those rare cases.

Fazle Hasan Abed founded BRAC in 1972 as asmall-scale relief and rehabilitation project in Bangla-desh after the War of Liberation in 1971 (Chen, 1983;Smillie, 2009). Over the years, BRAC has shifted inmission and focus from being a relief operation fo-cused on strategic development and poverfy allevia-tion to being a social mobilization organizationrooted in a Marxist tradition, to, most recently, apoverfy alleviation agent characterized by systembuilding for and around markets for the least advan-taged. Currently, BRAC reaches about 80 percent ofthe total number of villages in the country via its coreorganizing vehicle, the village organization (VO)—adecentralized model of local activify andintervention.

BRAC has experimented with an array of differ-ent activities and programs ranging from micron-nance, health services, nonformal education, andhuman rights and legal aid support. In the 1990s,BRAC began to incorporate market mechanisms asa means of poverty alleviation into its primary ap-proach (Lovell, 1992). In parallel, BRAC shiftedfrom targeting village-level communities, as wascustomary in its 1970s and 1980s programs, to anear-exclusive focus today on women's economicempowerment and participation.

Over time, BRAC leadership has come to recog-nize that access to financial services is an impor-tant, but insufficient, means of involving poor andmarginalized people in market-based activities(Mair & Marti, 2009). This recognition led to thedecision to set up social enterprises that facilitateentrepreneurial efforts and sustainable livelihoods


in the late 1990s. These enterprises, which includelivestock and fisheries (e.g., dairy, poultry), health(e.g., iodized salt), and agriculture (e.g., cold stor-age, sericulture), provide access to assets, supportproduct marketing, and foster entrepreneurial andmarket activities that create local jobs.

BRAC and its contemporary portfolio of market-based programs provide a useful analj^c opporttmifyto examine market building in a complex institu-tional context in which market access and participa-tion are impeded for many. Thus, we focus onBRAC's initiatives to build inclusive markets and le-verage market-based activities as they simultaneouslyaddress the complex institutional context.

Data

The data we present come from multiple roundsof data collection and a variety of sources: partici-pant observation, retrieval of archival documents,and in-depth interviews by two of the authors in-termittently over six years, 2005-11. BetweenMarch 2005 and January 2006, two of the authorsconducted 58 semistructured interviews, primarilyat BRAC offices and in local villages in Bangladesh.The interviews increased in focus and depth overthe period because of the iterative and ctimulativenature of the fieldwork process.

We identified informants by sampling from var-ious programs across multiple hierarchical levelsin BRAC; we sampled within other organizations aswell. The bulk of our interviews were conducted inEnglish; interviews with Bangla-speaking infor-mants (primarily participants in rural areas) wereconducted with the assistance of a local interpreter.Each interview lasted between 20 minutes andthree hours, followed a standard protocol for cap-turing emerging themes in field research (Spradley,1979; Strauss & Corbin, 1998), and was audio-re-corded and later transcribed verbatim.

At the end of the first round of interviews, wedecided to focus our data collection on a few BRACprograms in greater depth. We applied theoreticalsampling (Denzin, 1989), a recommended approachfor analytical induction (Bansal & Roth, 2000), toidentify these programs. We sought to capture abroad set of activities and practices as well as dif-ferent periods in BRAC's strategic developmentwithin the sample. We used the cases to organizeand stimulate data analysis, rather than as a meansto expose variance.

In consultation with BRAC leadership, we se-lected four programs: Education; Social Develop-ment; Human Rights and Legal Education (HRLE);and Challenging the Frontiers of Poverty Reduc-tion/Targeting the Ultra Poor (CFRP/TUP). Table 1

displays a detailed list of the different data sourcesutilized to investigate each of the four target pro-grams. Each has a distinctive focus, target popula-tion, and inception date. They are also all ongoinginto the current period. Additionally, each programtook shape at a different point in the evolution ofBRAC's strategies for poverty alleviation. Becauseof this, we treat each program as exemplary of a keychallenge regarding market building (Rodrik,2007). With these selected areas of investigation inplace, we conducted 17 additional interviews be-tween 2008 and 2010 with internal and externalinformants. We used our direct and repeated inter-view access to the BRAC founder and chairman,Fazle Abed, to identify program directors for inter-views. Using these leads as a basis for snowballsampling (Lincoln & Guba, 1985), we identified asecond set of informants. Using a purposeful sam-pling strategy (Kumar, Stern, & Anderson, 1993),we sought to uncover key insights or informationabout the origins, development, or activities of thefour programs. Throughout the interview phase, weused a repeated comparison strategy to comparedata across both programs and informants andidentify substantive points of S5aiergy or juxtaposi-tion (Glaser & Strauss, 1967).

As we completed the second set of interviews,we started to refine emerging themes and askedrespondents to comment directly on specific as-pects of these nascent findings. We used otir con-versations with the chairman and the four programdirectors to check our analysis. This use of externalinformants mitigated the potential biases of anyindividual respondent (Miller, Cardinal, & Glick,1997) and enabled us to induce richer insights fromour aggregate data (Santos & Eisenhardt, 2009). Ap-pendix A summarizes the final roster of personsinterviewed from BRAC (n = 56) and from otheragencies and sectors [n = 19).

Interviews were supplemented with several pe-riods of participant observation. Two of the authorsobserved a total of 17 meetings, which includedboth on-the-ground meetings within the differentBRAC programs (e.g., popular theater performance,legal education classes) and also meetings at otherorganizations (e.g., microcredit meetings, garmentfactories). Meeting observation time totaled ap-proximately 40 hours over the course of the field-work. During observations, we took field notes onsite and wrote up detailed accounts after each visit.Field observation notes were not coded in detailbut were used to illuminate the complex nature ofthe situation in which BRAC works, particularlykey cultural and situational specificities thatemerged from direct contact with the women in


TABLE 1Overview of Focal Programs

Characteristics Education Social Development (SD)Human Rights and Legal

Education (HRLE)

Challenging theFrontiers of

Poverty Reduction/'Targeting the UltraPoor (CFPR/rUP)

Main objectives It aims to improve the qualityand delivery of services ineducation appropriate tothe needs of poor children,particularly girls, and toincrease their access tothese services. The purposeis to help fill the remaininggaps in coverage, retention,and quality of compulsoryprimary education inBangladesh.

Year of inceptionFocus on:

Property rightsAutonomy

Archival data

Internal sourcesExtemal sourcesNumber of

audio/videoNumber of

interviews

1985

Not directlyYes

1,700 pages1,900 pages3

10

It aims to enhance the humanand social capital of thepoor and marginalized,especially women, so thatthey are aware of theirrights and are empoweredto claim their entitlementsand resist exploitation.Also it attempts to helplocal government tobecome more transparentand responsive to the needsof Üie poorest.

1986

It aims to protect andpromote human rightsthrough legalempowerment and toensure access tojustice through bothformal and informalsystems, especially forthe poor andmarginalized.

1998

Not directly but increasingly Yesimportant YesYes1,800 pages i,300 pages1,600 pages i,200 pages4 3

It aims to assist theultra poorpopulationgraduate up frompoverty levelsand assist theultra poor getaccess to themainstreamdevelopmentprograms.

2002

YesYes

1,600 pages700 pages2

BRAC programs. Appendix A summarizes the par-ticipant observation details in full.

In addifion to interviews and observation, we alsocollected a wide variety of documents for analysis,including secondary historical, legal, and polificalstudies. We negoUated access to the extensive docu-mentafion generated hy BRAC's research departmentas well as newsletters and local news articles. Foreach of the four focal programs in BRAC, we ohtainedprocedural and organizational information, includingdescriptions of work tasks, project plans, trainingmaterials, and internal appraisal documents. Thesematerials provided us with a specific understand-ing of the institutional context in which BRAC'smarket building initiatives are situated. We alsoreviewed provisions of the Bangladesh legal frame-work and a set of specialized legal texts (Pereira,2002). This review proved especially important be-cause it permitted us to illuminate the gap hetweenthe "in text" and "in reality" embodiments of thetwo focal market institutions. Finally, we discussedlegal aspects of market participation with lawyers,BRAC staff members (some legally trained), andother NCO field staff.

Data Analysis

Data analysis was conducted in two main stages, aprocess that allowed us to go back and forth betweenthe data and the emerging theoretical arguments(Miles & Huherman, 1994; Strauss & Corbin, 1998).Before either stage, we developed a narrative accountof our findings hy chronologically ordering the rawdata. The narrative account included quotes from in-terviews, documents, annual and committee reports,and field notes. To corrohorate our understanding ofhistorical events as reflected in the emerging narra-tive, we checked the accounts with a set of informantsthat included the BRAC founder and chairman, sev-eral BRAC managers, and independent historians, le-gal experts, and Islamic scholars. The production ofan historical narrative permitted us to better trace thehistory of BRAC's shift from a relief/assistanceorganization to a market-focused developmentorganization.

This historical overview revealed a way to organizethe data around the different BRAC programs. Weused the qualitative analysis software Nvivo 9.0 toaccomplish this analysis. Our primary data set in-


eluded data from four portfolio programs. We useddata collected on additional BRAC programs to com-plement and to corroborate these data. Our codingscheme built out a map and comparison of BRACprogram features, paying particular attention to thevariety of practices that the organization engaged in tointervene in market building. Our scheme also in-cluded a categorization of some institutional chal-lenges evident from theory and data (e.g., patriarchalsystem and kinship, political structures, socialnorms, religious beliefs) that BRAC addresses in itsefforts to promote women's market access.

Stage 1: Assessing the nature of the institu-tional voids. In our first stage of formal analysis, weidentified instances of on-the-ground market build-ing. We grouped these instances into relevant the-matic categories (performing open coding). In thiswork, we built on the Weberian conception of so-ciety as a multi-institutional space, and specificallyon the classic statement by Friedland and Alford(1991) that redescribes society not as an integratedwhole, but rather as system of interlinked institu-tional arenas. Per this view, society comprises sev-eral distinct spheres of activity, each one builtaround a central institutional logic that specifiesdistinct meaning systems and orderings of reality,along with social practices that support each ofthese "social worlds." Everyday activities oftentake form at the intersection of these spheres,through contradictions or reinforcement betweenlogics and practices (Friedland & Alford, 1991;Thornton & Ocasio, 1999). Our analysis in stage 1suggested that three central institutional spheres inrural Bangladesh have direct relevance for marketbuilding: commimity, politics, and religion. This isconsistent with findings from related studies thatfocus on the institutional constitution of society,specifically its economic activities (Campbell,2004; Greif, 2006; Hamilton & Biggart, 1988;Heimer, 1999; Thelen, 2004). Moving forward,again using Nvivo, our analysis identified relevantbits and pieces of institutions (Schneiberg, 2007)and assigned each of these instances to one of thethree identified institutional spheres. For instance,we allocated evidence related to "early marriage,""patriarchal system," or "kinship norms of behav-ior" to the community sphere category.

Once these institutional assignments were com-plete, we reviewed the data again to refine theinitial categorical assignments. Two BRAC mem-bers and two independent informants (a legal spe-cialist and the director of an indigenous NGO in thefield of education) were also asked to verify thecategorization. These outside reviewers agreed onall the assignments except one, the practice of pur-dah. One BRAC member suggested that this prac-

tice should be assigned to the community spherebecause it fit there more precisely than in the reli-gious sphere. Given the high degree of agreementamong the multiple reviewers, as well as corrobo-ration provided by the literatiure, we sustained oiurinitial decision to assign purdah practices to thereligious sphere.

Using our first research question as a lens (Eisen-hardt, 1989), we focused our analysis next on iden-tifying the interfaces between the three spheres andtwo focal market institutions. The community, po-litical, and religious institutional spheres are ana-lytically distinct and segregated, but in actual prac-tice, their boundaries blur and interface. Oinranalytic challenge, then, was to acknowledge thedistinct practices and systems of meaning thatcharacterize each sphere while closely examininghow their interfaces create possibilities for action.We intuited that conflicting and potentially contra-dictory accounts, demands, and solutions that oc-cur at these interfaces imprint the institutionalvoids that configure possible markets.

To develop our empirical analysis regarding thissupposition, we elected to focus on two specificand well-accepted institutions central to standardaccounts of markets and institutions: propertyrights and autonomy. In specific, we wanted todiscover how each the three societal spheres weidentified impacted these two market institutions.Market economies are understood to rely on thecreation and enforcement of property rights (DeSoto, 2000; La Porta et al., 1998). Property rightsexemplify a governing and stabilizing market insti-tution (Greif, 2006; Rodrik, 2007) because they de-termine "the social relationships between ownersand everyone else in society" (Fligstein, 2001: 33).Regarding autonomy, modem market dealings areunderstood to be made by—and only by—autono-mous actors (McMillan, 2002). As such, autonomyis an example of what development economists andexperts have called an enabling institution (Sen,1999): it influences whether individuals are able tooffer their goods and services or benefit firom theoffering. Moreover, both property rights and auton-omy are legal and institutional outcomes of author-ity and power relations (Campbell & Lindberg,1990; Carruthers & Ariovich, 2004; Friedland & Al-ford, 1991; Nussbaum, 2000; Sen, 2009).

These two market institutions provided focalpoints for exploration of the institutional interfacesand market-building challenges identified in theBRAC data. In our interviews, informants repeat-edly singled out autonomy as the primary focus ofNGO activity in Bangladesh. Many intervieweesalso mentioned that property rights have graduallygained preeminence as a focus for intervention be-


cause of the importance of property for women'sself-definition (Nussbaum, 2000). To understandhow the three societal spheres in Bangladesh affectproperty and autonomy, we developed a set of nar-ratives, or detailed memos, to describe each of theinterface possibilities. In total, 17 narratives (ap-proximately 5-30 single-spaced pages each) werecreated; each narrative contained direct quotes aswell as clarifying comments produced by the re-search team (Maitlis & Lawrence, 2007). By way ofexample, one of the narratives detailed the inter-face between autonomy and early marriage (associ-ated with the community sphere). We cross-checked each narrative with three informants (oneinternal to BRAC and two external) to validate theirfidelity with lived experiences.

Stage 2: Surfacing activities in and around thevoids. Our second formal stage of data analysisfocused on investigating BRAC's activities in rela-tion to the identified institutional interfaces. Wefocused specifically on four BRAC programs, asdescribed above. Although it would have been pos-sible to focus on a greater number of programs, itbecame evident during the initial phase of analysisthat few additional ideas and issues were emergingwhen we looked beyond the four portfolio pro-grams (Glaser & Strauss, 1967). We opted for anintensive analysis of these programs.

Our second stage of analysis comprised threesteps (Pratt, Rockmann, & Kaufmann, 2006). Figure1 provides a schematic overview of this analyticprocess. The first step involved the creation of pro-visional categories and first-order codes (VanMaanen, 1979). We used Nvivo to keep track of the

emerging categories and to view similarly codedtexts simultaneously, which helped to manage thelarge amount of data. Following the proceduressuggested by Miles and Huberman (1994), our firstcategorical codes provided descriptive labels forthe different sorts of activities that we observed atthe different interfaces. The codes were largelybuilt upon the vocabulary of the interviewees,which included, for instance, "giving voice," "rais-ing questions," and "building ties with the elites."Once codes were named and categories developed,we returned to the data to review categorical fidel-ity with the data. As Pratt et al. (2006) suggested,we either corrected a category or reconceptualizedit when the revisited data did not fit it well. Forinstance, after several iterations and discussionswe agreed that our initial category "embracing re-ligious arguments" inaccurately highlighted theuse of religion by BRAG, so we changed it to "de-mystifying [available religious arguments]."

The second step involved axial coding (Strauss &Gorbin, 1998), wherein we compared first-ordercodes with one another to clarify themes and tocreate second-order constructs. This was an induc-tive, recursive process through which we identifieda set of more abstract, theory-rich constructs. Toillustrate, when comparing the codes referring toBRAG'S facilitation of conscientization (Freiré,1970) with codes referring to building knowledgeof repertoires for poor women, we noticed thatBRAG'S work often helped actors develop their ca-pacity to make sense of their situational context. Tocapture this idea we created a second-order con-struct called "developing sensemaking capacity."

FIGURE 1Analytical Goding Process to Induce Theoretical Dimensions

First-Order codes Second-OrderConstructs

Aggregate TheoreticialDimensions

• Creating spaces for equals• Creating spaces for unequals

• Tying up with government systems• Teaming up with social service providers

• Building on local means of issue resolution• Making use of customary sources of support

• Facilitating conscientization• Knowledge of repertoires

• Demystifying• Adopting artistic traditional performances

k

»

»

>

»

Creating spaces forinteraction

Expanding resourcesystem

(Re)defining localarrangements

Developing sensemakingcapacity

(Re)combining norms andtraditions

Redefining marketarchitecture

Legitimating marketactors


The axial coding was done by individual research-ers as well as jointly by the research team. The teammet numerous times to create constructs and assessthe categorical fidelity of the emerging codes.These iterative discussions helped to refine thecode base and to delimit the emerging theory(Creed, Dejordy, & Lok, 2010; Pratt et al., 2006).Appendix B provides coding statistics for the keycodes from each of the four portfolio programs.

Finally, in the third step we identified importantdimensions from the sets of second-order con-structs. For example, some categories looked struc-tural (e.g., "creating spaces") whereas others ap-peared cognitive and cultviral (e.g., "recombiningnorms and traditions"). Next, we generated alterna-tive theoretical frameworks to make sense of howthese constructs related to one another and to theliterature on market building. Then we workedthrough the relevant insights each provided. Weconsolidated these available factors into two broadtheoretical dimensions: "redefining market archi-tecture" and "legitimating new actors and activi-ties." The theoretical dimensions resonated withthe data and provide further analytic guidance totmderstand BRAC activities to build inclusive mar-kets. Figure 1 provides a schematic overview of thisprocess, showing our first-order codes, second-or-der constructs, and derived theoretical dimensions.

ANALYSIS AND FINDINGS

Our analysis explores the resonance between theviews and experiences of people on the ground inBangladesh, the abstractions characteristic of theinstitutional voids literature, and the conceptualand practical work involved in market building.We present our analysis of market building in thissection to illustrate the issues and to develop aprovisional model of inclusive market building.Our initial findings reinforce the importance of in-stitutional interfaces, as exemplified when threeinstitutional spheres (community, politics, reli-gion) shaping life in rural Bangladesh meet formalmarket institutions of property and autonomy.

Our analysis of property rights and autonomyrelative to the institutional interfaces in ruralBangladesh points to a fresh understanding of in-stitutional voids as the intermediate outcomes ofconflicting institutional demands, norms, and rein-forcing mechanisms. Voids occur at an interfacebecause it is here that different bits and pieces ofinstitutions (Schneiberg, 2005) collide and recon-figure spaces for social (and economic) action. Ta-bles 2 and 3 summarize the three-way interfacesidentified by our analysis and provide illustrative,direct quotations from our fieldwork. Our investi-

gation of BRAC's initiatives also surfaced two keysets of market-building activities in and aroundinstitutional voids: (1) redefining market architec-ture and (2) legitimating new market actors. Tables4 and 5 present data that describe these two activ-ities in detail.

Institutional Interfaces: Conflict amongInstitutions as a Source of Voids

Property rights. In recent scholarship, propertyrights have emerged as a critical pillar of market econ-omies (De Soto, 2000; North, 1990). The legal corpusby which Bangladesh is governed includes extensivelegislation regarding property rights. According to theconstitution, citizens are equal with respect to theirentitlement to ownership and protection of property.Yet numerous studies have characterized Bangladeshas a country that has poor standards of property rights(Fernandez & Kraay, 2007; Islam & Asaduzzaman,2008) and a judiciary system that ineffectively en-forces these rights (Heritage Foundation, 2010; WorldBank, 2010). Although property rights are establishedby law, our analysis shows that they constitute veryweak pillars for markets to act as a means for povertyalleviation or a vehicle of economic and social prog-ress. The specifics of this situation are important inillustrating the effect of institutional interfaces.

Community sphere. Poor women's market- andnonmarket-based activities in Bangladesh follow therules of interaction established by the commimity. Inspecific, our data illustrate that institutions that relateto patriarchal and patrilineal systems, commimifynorms, and kinship norms are in conflict with andsometimes contradict constitutional properfy rights.This finding offers a potential explanation for whyformal properfy rights often go tmclaimed and unen-forced by many sectors of the population and there-fore govern market activities and transactions weakly(see Table 2 for examples).

Social organization in Bangladesh is consistentwith typical patriarchal and patrilineal rules andnorms (Kabeer, 2000). The patterns of behavior-andcognition associated with this social system struc-ture women's positions in society, impacting bothpublic and household economies. One informantassessed the situation of property rights as one inwhich women have little control over resourcesthey own:

Property rights are very unfavorable to women . . .even if tbey have something in tbeir name, legally, itis controlled by either husband or sons or whoever,family member . . . and because of their illiteracy,backwardness, they CEUinot claim ownersbip. (inter-view, man, social activist)


TABLE 2Institutional Interfaces: Property Rights in Rural Bangladesh"

Community and Societal Sphere Political Sphere Religious Sphere

Patriarchal and patrilineal system2.1 "The husband believes that hiswoman should not own anythingbecause that would make her powerful."(interview, man, microcredit officer)

Community norms of behavior2.2 "We are trying with this newAlternative Dispute Resolution toaddress some of these problems.However, often women have to face andare afraid of social sanctions for seekingjustice outside the community andbypassing the elites." (interview, man,SD manager)

Kinship norms of behavior2.3 "Kinship norms argue against thesale of land generally. Given an option,a landholder would prefer not to sellland, especially that belonging to hisfather or kin." [BRAG rural studiesseries 1: 161)

Patronage2.7 "In having justice, what reallymatters is whether or not you have thesupport of a patron . . . and if you havemoney to offer to officials . . . or yourpatron is willing to give you thatmoney." (interview, man, microcreditofficer)

Local powers corruption2.8 "We could not convince dewani(shalisker) unless we offered them atleast two hundred taka before thesession in village shalish." (man, HRLEinternal document)

Access to formal justice2.9 "Often we also found that some ofthe elite members believed that it wouldbe bad to involve women in formalcourts because it would reduce thesocial status of the victims and of theirfamily." (interview, man, SD officer)

Access to informal justice2.10 "Women are not welcome inshalish because they are said to createchaos." (interview, man, researchdepartment)

Purdah2.13 "It is mostly because of social andreligious norms that women do not go toshalish. They believe that women losepurdah if they participate in shalish."(interview, man, local journalist)

Local interpretations of Koran2.14 "Our observation identified lack ofrespect the existing laws of womenrights, specially from imam. It has alsobeen observed that whenever imamdissented on these issues, others got aninherent tendency to keep quiet." (HRLEinternal report)

Islamic pre-emption (shuf a) laws2.15 "According to the Muslim Shariahthere exists what is called the right ofpre-emption. Among other things, it isinterpreted as forbidding to sellinherited land outside the family or thekinship group provided a member of thefamily or some kin can offer the marketprice." (interview, man, Islamic lawscholar)

' The text and Table 1 give the spelled-out forms of abbreviaüons used here and in subsequent tables.

The patriarchal system in rural Bangladesh rein-forces norms that confer control of women's prop-erty, income, and lahor to men. Because of the riskof social sanction, women often avoid rightfulclaims of ownership or inheritance and do notmake use of existing [formal, legal] means to con-vey and protect property rights. A male BRAC So-cial Development Program manager explained:

Wives know they must listen to and obey their hus-bands, and this includes the in-laws. If the womanwould ask for shalish^ or other means of disputeresolution, immediately community people passbad comments towards the woman and also thewhole family, and that damages the social statuswithin the community.

Kinship norms also constrain women's opportu-nities to sell and rent assets or property becausethey foreclose the possihility of transacting withmembers outside kinship groups.

^ Shalish is a community-based, largely informal pro-cess through which small panels of influential local fig-ures help resolve community members' disputes and/orimpose sanctions on them.

Political sphere. Village life in Bangladesh isshaped in important ways by the decisions of localvillage councils. Our examination of the interactionbetween property rights and the local patterns of in-fluence (i.e., patronage, corrupfion, and justice pro-cesses) suggest that rural Bangladesh is characterizedhy strong power asymmetries hetween different seg-ments of the village. For example, having a patron isa key factor in favorably resolving asset and propertyconflicts. An informant illustrates this point:

Many poor women were exposed to a high risk thatthe assets they received would be stolen or dam-aged. Because of their lack of connections with morepowerful actors . . . these women cannot get thesupport of elites as patrons in either formal or infor-mal—shalish—courts to enforce their own rightsover their assets, (interview, woman, GFPR/TUPdirector]

Corruption and hrihery also affect the legal andpractical implementation of property rights (Trans-parency International of Bangladesh, 2009). Severalof the people we interviewed in rural villages ex-plained that the police were quick to accept "infor-mal payments." Women, however, are excluded


from participating in these practices. As one SocialDevelopment program officer, put it, "Husbandscan bribe the police or lawyers to prevent going tocourt," but not women.

Finally, we found that the legal protection ofwomen's property rights was further impeded bytaken-for-granted beliefs and practices regardingparticipation in courts of informal law. As a SocialDevelopment Program coordinator explained, "Wehave been long trying to engage our members intraditional [informal] forms of dealing with con-flicts [shalish] . . . but it is very difficult since theyare dominated by the male elite, which makeswomen's participation rather limited" (see alsoquotation 2.9, in Table 2).

Religious sphere. Practices and beliefs associatedwith religion also impact the scope and content ofBangladeshi women's social and economic activities(Kabeer, 2000). Very often these practices and beliefsare at odds with modem conceptions of propertyrights. Purdah is an exemplary instance of this ten-sion. Purdah, which literally means "cinlain," refersto the obligation that Muslim women have to stayclose to their family relations, limit contact with un-related men, and avoid being visible in public venuessuch as the village market or a court (Chen, 1983). Inthis way, purdah directly limits women's ability toclaim or protect their property rights.

Local interpretations by rural clergy—i.e., mul-lahs—often promote a version of Islam that reinforcesnorms about virtuous women as docile and submis-sive. These norms create a set of social expectationsthat reinforce women's seclusion and foster the invis-ibility of women in the public sphere. These religiousinterpretations also constrain women's use and en-forcement of property rights and conflict direcflywith women's constitutional rights.

When women seek to control assets and re-sources or participate in formal or informal courts,these claims are often interpreted as challenges toreligious norms and laws. A male Social Develop-ment Program manager explained:

In disputes where some people or groups intend toappropriate some resources that have heen stolenfrom a family or a group, there are always excusesfoimd and in many cases fabricated, in stating that awomem from that family or group has broken thenorms of Islam.

Our data reveal that the interaction between prop-erty rights and local institutional spheres is both com-plex and multifaceted. One of the reasons for this isthat the boundaries that separate these societalspheres are highly permeable (Heimer, 1999). For ex-ample, traditional kinship norms in Bangladesh, a setof behaviors that we attributed to the community

sphere, are also reinforced by Islamic pre-emptionlaws (shufa), which pertain to the religious sphere.This interaction of elements across multiple institu-tional spheres is particularly important to explainingthe weak enforcement of property rights in Bangla-desh. By way of illustration, Islamic law contains aprovision for the pronouncement of fatwa (religiousopinion), which many clerics and village patriarchsgrossly misuse. Although the High Court made thepractice of fatwa illegal in Bangladesh in 2001, reli-gious leaders at the local level continue to exercise itwidely (Pereira, 2002). An imam quoted an Intemalreport from the Hinnan Rights and Legal Education(HRLE) program in an interview that described thechallenges of using alternative dispute resolutionmechanisms when religious-legal practices are socommon:

We cannot give the legitimacy of practicing suchlegal rights, which is controversial with Islamiclaws. If we find such activities, we will protest first,and we will go against whoever goes for so.

Autonomy. As mentioned, scholars fiom differ-ent research traditions have emphasized the centralrole that autonomy plays in market building andthe distributional effects of markets (McMillan,2002; North, 1990; Sen, 2009). Bangladesh's consti-tution places women on an equal footing with menin all dimensions of public life. For example. Arti-cle 28(1) states that the state shall not discriminateagainst any citizen on grounds only of religion,race, caste, sex, or place of birth (Pereira, 2002).However, our data reveal that local community,political, and religious spheres act to limit wom-en's autonomy and erode the ability of poor womento participate in markets.

Community sphere. Women in rural Bangladeshare socialized to be dependent. Our examination ofthe interaction between autonomy and the conunu-nity sphere illustrates that patriarchal and patrilinealnorms, as well as customs such as early marriage,stand at odds with constitutional provisions andmodem conceptions of women's autonomy (see Ta-ble 3 for illustrations). The patriarchal and patrilinealsystem is omnipresent in Bangladesh. Women are"spoken for" first by their fathers, later by their hus-bands, and finally by their sons, brothers, or othermale relatives. Women are also considered subordi-nate in most situations. When there is not enoughfood to feed an entire family, for example, girls aregiven less than their brothers. The husband of a mi-crocredit borrower articulated, "Girls must be beatento maintain strict control." A female manager inBRAC's Education Program explained further:

Once she is married, a girl has to sacrifice her life.She has to give service to her husband and family


TABLE 3Institutional Interfaces: Autonomy in Rtu-al Bangladesh

Community and Societal Sphere Political Sphere Religious Sphere

Patriarchal and patrilineal system2.4 "You know, what we are taughtis that a father always wants whatis best for his son and his daughter.This means he can oblige hisdaughter to get married without herconsent." (interview, woman, localjournalist)

Early marriage2.5 "Parents give much moreattention to their sons' schoolingand education than to theirdaughters' education. Parents feelthat the best they can do for theirdaughters is to find a good groomfor them. Hence, they startsearching for brides for theirdaughters, not for a good school."(interview, woman, educationprogram manager)

Community norms of behavior2.6 "The social stigma attached todivorce is so pernicious for womenas opposed to men, that a womanin many cases would remain withher husband, even though thiswould mean an intolerable andsometimes life-threateningexistence. Frequently a woman hasno choice but to remain in anunpleasant marriage because in thealternative she would be destitute,without any financial or socialmeans of survival." (interview,woman, HRLE program, director)

Patronage2.11 "For instance, to get a position as ateacher in a public school, while"education" merits are important, whatreally matters is having the support of apatron." (interview, man, SDcoordinator)

Civil laws2.12 "Sections 64 and 65 of theCriminal Procedure Code deal withthe service of summons of personwho cannot be found. . . . Here wesee a law that intentionally bypasseswomen, indicating all women lackagency, all being variations of someform of pardianishln. In most otherlaws the assumption is the threadinto the procedure, sometimesmanaging to obscure the normativeassumption." (Pereira, 2002: 8-9)

Purdah2.16 "Women are imder constant surveillance.It is necessary to ensure that they do nothingthat brings sharam (shame) to their kin."(interview, woman, local journalist)

Local interpretations of Koran2.17 "Well, there exist some progressivelaws in recent years. However, they areoften ineffective because of extraneousfactors. One of the main such factors in thecase of Bangladesh is that of religiouspolitical pressure. Particular religiousattitudes are so entrenched in Bangladeshisociety that they strike at the foundation ofprogressive laws and deprive them of muchof their effectiveness." (interview, woman,HRLE director)

members. She is supposed to follow her husbandwhatever he orders to do.

Gustoms related to family and marriage forcewomen to defer to men, both consciously and un-consciously, in ways that constrain their auton-omy. In the case of marriage, fathers make all of thedecisions. In the case of early marriage,* which isillegal, giving a dowry^ is perceived as a moralobligation. Since the amount of dowry goes up withthe age of the bride, early marriages are typically

* The legal marriage age in Bangladesh is 21 for boysand 18 for girls.

^ Dowry refers to the money, goods, or estate that awoman brings to her husband in marriage. In the last fewdecades, the practice of dowry has become widespread,though supported neither by state law nor personal law.It is a very common source of violence against women.

favored. Early marriage further reduces women'sautonomy by limiting their access to education.Gommimity norms also stigmatize divorce, whichheavily restricts a woman's autonomous capacity toterminate her marriage. Elements from the religioussphere reinforce marriage norms. In particular, Is-lamic law accords impose severe conditions onwomen seeking divorce by requiring them to un-dergo a process that is "time-consuming, expensiveand in most of cases socially humiliating for thewoman and her family" (Pereira, 2002: 25).

Political sphere. The public and private life ofpoor women in rural Bangladesh is also sharplyinfluenced by rules and norms associated with pa-tronage and civil laws. In the countryside, the poorsecure protection from patrons by providing themservices such as proxy voting in elections. In ex-change, the poor receive remunerated positions or


opportunities, such as "a position as a teacher in apublic school" (quotation 2.11), access or voice inshalish processes, social security, or links to publicresources such as food cards or food-for-work pro-grams (see Matin, 2002). Well-entrenched patron-client relationships and their resulting asymmetricexchanges conflict with the quest for women's au-tonomy in the country. As the director of the CFPR/TUP program explained.

It comes at the cost of dependency obligationswhich may endure over generations and becomedemeaning or arduous.

Our data also suggest that different elements ofBangladesh's legal framework, as constituted byconstitutional law and general law, contradictWestern conceptualizations of autonomy both incontent and spirit. As we pointed out above, theconstitution, especially part III ("FundamentalRights"), embodies multiple provisions that grantrights and reinforce women's autonomy. However,our examination of institutional interfaces suggeststhat constitutional law may be only one of severalforces in play. Several legal scholars have revealedthat different pieces of legislation within the gen-eral law do contribute to hinder women's auton-omy. For example.

There are several examples of what I term ultra-protective laws, impinging on fundamental rights ofwomen witb impunity, calculatedly taking on only aselective notion of women's capacity. All of tbeselaws gravely restrict women's rigbt to movement orchoice of emplojrment. For instance, these laws pro-hibit employment of women and children betweentbe hours of 8 pm and 6 a.m. or other than between7 a.m. and 8 p.m. (Pereira, 2002: 9)

Religious sphere. Finally, the autonomy of poorwomen is severely restricted by many local reli-gious beliefs and practices. In particular, normsassociated with purdah reinforce the existing gen-dered division of labor that is prevalent throughoutsociety. Because women are strictly confined to theprivate sphere, they are prevented from becominginvolved in market transactions and income gener-ation. Even marketing products in public is prohib-ited for women. As one of our informants, a micro-credit borrower, explained:

I cannot go to the market. My busband and my soncan go . . . I am a woman.

Women's economic contributions are tradition-ally restricted to activities that can be performed onfamily property, such as rearing poultry or posthar-vest activities. Selectively, women do disobey therestrictions of purdah, yet our analysis suggeststhat restrictive institutional rules often prevail. One

woman we met in a primary health care officeconfessed, "I used to work in the fields at night orwhen it was difficult to be seen." However, anotheryoung woman told us that her parents pressuredher to stop working the fields because they wereafraid that she "might fall in love with someone."

According to a female BRAC volunteer healthworker, rural clergy also reinforce norms associatedwith the practice of purdah, solidifying the com-monly held belief that "Allah made womenweaker." Although this interpretation can be un-derstood as the enforcement of a religious tenet, itis also appears as a clear cultural exemplar of mentrying to consolidate their power and reinforce pa-triarchal structures.

Up to this point in our analysis we have looked athow the two focal institutions of properfy rights andautonomy are often contradict and conflict with ex-isting beliefs and practices associated with the com-munify, political, and religious spheres in Bangla-desh. As a result of our investigation, we suggest thatcontexts such as Bangladesh can be seen as an areawhere multiple institutions exist simultaneously toshape market dynamics rather than as a locale devoidof institutions. One informant, a male social activist,articulated this thesis as follows:

Of course we bave laws on property rights; of coursewomen are, in theory, equal before the law in ourconstitution. The issue is not that we have few in-stitutions, but tbat we bave way too many! Andoften, well almost always, the ones tbat matter inour communities go against women baving a moreactive role in society, in markets, at bome, etc.

The lack of primacy regarding market institu-tions is due to the plurality of institutions thatsupport local action. As evident from our infor-mants, local practices and beliefs interact withWestern conceptions of market logics; coding thisinteraction as simply "weakness" or "absence" ofmodern market institutions misspecifies the situa-tion and underrecognizes the significance of theinstitutional plurality.

The results of our analysis provide evidence for atheory that institutional voids are situated, inter-mediate outcomes of contestation at institutionalinterfaces. Rather than empty spaces, institutionalvoids are, we suggest, dynamic spaces reconfiguredby conflicting and contradictory institution flux.We emphasize the situated and intermediate fea-tures of voids as a way of better understanding whyand how market exclusion occurs. This insight alsoserves as a starting point for both practical andpolicy-related efforts to build inclusive markets.

Next, we report findings from our analysis of theon-the-ground activities by BRAC as its members


have endeavored to build inclusive markets in re-cent years in a context of institutional plurality andcomplexity.

Building Inclusive Markets

Building on data from our four identified BRACprograms, we found two broad repertoires of inter-ventions that address the indeterminacy of the keymarket institutions, private property and autono-my: (1) activities that redefine market architectinreand (2) activities that legitimate new actors. "Rede-fining market architecture" refers to the renegotia-tion of existing institutional arrangements to definewho can access and participate in markets andunder which conditions. "Legitimating activities"include building awareness and identity as wellconstructing social narratives that support and au-thorize women's roles in and access to markets.

Redefining market architecture. Within thelarger category of activities that redefine marketarchitecture, our analysis exposed three sets of ac-tivities (second-order constructs) that BRAC en-gages in that provide structural interventionswithin a complex institutional context. Table 4maps the relevant first-order codes to direct quota-tions from BRAC workers, beneficiaries, experts,and policy observers that illuminate these activi-ties. Appendix B displays descriptive statistics forthe complete set of codes related to the BRAC pro-grams in our analysis.

Create (social) spaces for interaction. The firstset of activities within the frame of market archi-tecture redefinition correspond to the construct"creating spaces for interaction." These activitiesconsist of building platforms for interaction anddialogue. This second-order construct aggregatestwo first-order codes: creating "spaces of equals"and creating "spaces of unequals."

From its beginning, BRAC's social mobilizationapproach has been an attempt to build equal ratherthan hierarchical communities. The objective hasbeen to create "free places" (Goffman, 1961) inwhich women feel they belong—places that con-tribute to the development of women's sense of selfand break women's dominant relations of depen-dence. For instance, in referring to BRAC's villageorganizations (VOs), a VO member told us that"[we] had learned how to get together and helpeach other." VOs, which consist of 35 to 50 womenfrom a single village, are seen by BRAC as "the key"and "the base" of their activities (quotation 3.2)because of their ability to redefine women's tradi-tionally passive roles. Building these VO spacesrequires a small, but continuous, set of interactions.According to Social Development Program organizer:

Then you have a small group in another communitythen you ultimately expand the small group into thevillage level and then you meet the VO each andevery day, well formally once a week in a meetingsetup. But informally, everyday day and night,morning and evening you're visiting and you arehaving contact with them. That is how you're be-coming very close to them and part of them. That ishow it begins.

These spaces for equals isolate and "de-inte-grate" (Touraine, 1995) women from their tradi-tional positions and encourage them to take socialaction.

BRAC also builds spaces for "unequals" to re-integrate women into arenas where traditionalrelations, roles, and practices can be understoodas elements of a negotiated and negotiable order(Strauss, 1978). The Village Poverty ReductionCommittees [Gram Daridro Bimochon Commit-tees, or GDBCs) are examples of these spaces forunequals. These include members of the localelite along with members of BRAC, VO women,,and the ultra poor. These spaces are contingent,,tentative, and carefully managed by BRAC be-cause they often create conflicts at political, cog-nitive, and emotional levels by their very nature.BRAC facilitates consensual solutions by helpingelites to reinterpret their support for women ratherthan withdraw it. The CFPR/TUP program directorexplained further:

Tbe language we use in motivating them [the elites],is the most important aspect... we tell them, listen,we are a stranger here in your village, but yon are thepeople wbo have been supporting them for hun-dreds and hundreds of years. Otherwise it would bevery, very difficult to mobilize them and have themsitting with tbe ultra poor women.

Outreach to existing systems of services provid-ers. The second category of activities that emergedfrom our analysis highlights the way that BRACreaches out to existing service provision systems.This second-order construct aggregates the first-order codes "tying up with government systems"and "teaming up with social service providers." Asour prior analysis shows, women are often impededfrom accessing many of the structures, services, andorganizations that exist to support them (i.e., judi-ciary, education, health services). To address andmodify such restricted access, BRAC has worked to"build partnerships and referral linkages with theGovernment" (quotation 3.12). One example of thisactivity is the primary education school BRAC de-veloped when it recognized that interrupted edu-cation was a major issue impacting the develop-


u0)

I

eo

I

a

O i

QI

cSCA

Scda.

t/3

Í

O•3a,

,í3ü en *^t S T)

oT3O•S 00 B Ö ^ >£

S" '5 fei tí en o "3

* S g g O aa-ïî

l̂ípí II"* 9 «. 5- 2

o

I

^ tíQJ O g cî

ftOjcngcBP&cl

. t ; CO en> g o

MCO

O S

>O

s .« ^ •£

,2 0)f 1 T^ T ^

... sa 5

> 2 SS « .S^ r

tí '-a

•S a

COS I 2 .s g -g I ;§ .

o

CJ

o3" oi)

I. (B ' S g ^ . . e n

" _ '»H 0 0 s

g s ° tí|

9 s tí -^ -̂^ s S | ^ §>

g ^ « g ^ •§ ^ "-̂C U O ^ ^ U ' d t í c o c o

<duS

2012 Mair, Martí, and Ventresca 83E

ao

••Scd

Iuaa.en

I

13g.II

II

3

•a

uo

CA


Se

II

0} 0)

'S ^ -B

- " S o P< U2 w M <B 0)

i^-s-M^ B a s

I

cdb

73

ac

•I

13

I£

2012 Mair, Marti, and Ventresca 837

¡I

^ 1 1 ^ 2 tí =.c o K í u g j j S c o


ment of girls' autonomy. A female EducationProgram manager told us:

Well, this was for children (70 percent of girls and30 percent of boys] who had dropped out from thegovernment school or who do not have access togovernment school. And these children are defini-tively poor from rural areas. They will be in schoolfor one year, and after the conclusion of the programthey will go to the government public schools, not toBRAG schools.

BRAC also teams up with service providers out-side the government, such as NCOs. These alli-ances serve to increase women's "exit options"(Nussbaum, 2000). According to research, a womanis far more likely to stand up to ahuse if she is ableto read and access alternative means of issue reso-lution, such as legal aid clinics, employment coop-eratives, and traditional healers. BRAC's partner-ship with Ain O Salish Kendra (ASK), a legal aidand human rights organization, exemplifies thisform of outreach. An internal document from theHRLE program elaborates on the nature and ohjec-tives of such collaboration:

BRAG-ASK joint legal aid program is designed as apartnership. ASK provides orientation and trainingto BRAG staff . . . to familiarize them with existinglaws, court procedures, eind filing of cases at thepolice station.

Purposeful integration of BRAC initiatives withexisting local support structures. The third activ-ity we observed was the purposeful way that BRACintegrated local support structures into its variousprograms. The second-order construct labeled"(Re)defining local arrangements" aggregates twofirst-order codes: "Building on local means of issueresolution" and "making use of customary sourcesof social support." According to our analysis, mak-ing market access and participation for women pos-sible seems to require learning about, making useof, and adapting to prevailing institutional arrange-ments—even if they are considered to be dysfunc-tional or if they reinforce patterns of exclusion.

Access to formal justice is tedious, costly, andfrequently unavailable to women, as describedabove. As an alternative, BRAC attempts to engageits members in shalish, a "traditional form of deal-ing with conflicts" (quotation 3.18) when issuessuch as "land, divorce, illicit relationships, orfights hetween individuals" (quotation 3.19) arise.A BRAC Social Development Program man-ager said:

Shalish serves the needs of the villagers . . . well, weknow this is not the whole picture. Many times theyare not fair and can be degrading, but we also know

that it is what the poor people prefer. We know alsothat they see the memhers of the shalish court ascloser to them and that they speak in a language they(the poor people] can imderstand.

In promoting the use of shalish under certainconditions, BRAC acknowledges that the needs ofpoor women might well he served by alternative,local sets of arrangements (Unger, 1998). Similarly,BRAC recognizes and incorporates the value of ex-isting "older, village-hased practices of assistanceto the poor" (quotation 3.23) into its programs. Thisis a radical departure from BRAC's traditional ap-proach to poverty alleviation, since it makes use ofpractices that are thought to perpetuate relation-ships of dependence. As the director of the CFPR/TUP program explained:

We thought that undermining older, village-hasedpractices of assistance to the poor would . . . reducetheir already rather limited sources of support. Butit's good that we learned from our field and re-sponded immediately . . . we asked (the villagecommittee] to stand up for the women in shalishprocesses, in ensuring the education of the childrenof our ultra-poor members . . . to ensure . . . thatwhatever support they are providing to the commu-nity . . . in a more organized, in an even morepopular manner, will... produce a more long-term,sustainable result.

In sum, we have identified a set of activities thatredefine market architecture in ways that hegin toallow poor women to engage in markets and market-based activities. Our analysis reveals how BRAC hasshifted the boundaries of spaces and activities to bemore accessible and available to women. BRAC ac-complishes this redefinition by extending resourceand support systems and building various platforms.In so doing, BRAC brings poor women into posifionsthat have the potential for increased discretion andautonomy at the local level.

However, development scholars and practitionershave long warned about the limits of this narrow typeof liberation and empowerment (Nussbaum, 2000;Sen, 1999). Although these arrangements may tem-porarily ameliorate local difficulties, deeper, moretaken-for-granted institutions that "identify catego-ries of social actors and their appropriate activitiesor relationships" (Barley & Tolhert, 1997: 97) canremain unchanged. Our findings highlight a secondset of activities that deepen and embed market-supporting institutional infrastructures into localinstitutional arrangements. This set of market-building activities focuses directly on legitimatingwomen as market actors.

Legitimating new market actors. We found thatBRAC engaged in two specific types of activities


(second-order constructs] that legitimate women'saccess and participation in markets. Tahle 5 pro-vides quotations from our data that illustrate eachof these activities along with the correspondingfirst-order codes. Appendix B displays statistics for

the key codes used to categorize these instanceswithin the BRAC program data.

Develop "sensemaking" capacity. The first setof legitimating activities we foiuid, "developingsensemaking capacity," aggregates the two first-or-

TABLE 5Data Supporting the Theme of Legitimating New Market Actors

Developing Sensemaking Capacity

ProgramFacilitating

Conscientization Knowledge of Repertories

(Re)combining Norms and Traditions

Adopting ArtisticDemystifying Traditional Performances

Education 4.1 "In our educationprograms, one of ourmain objectives is to helpthem to collect and reflecton their experiences,what they know, the workthey do, that their parentsdid. This is important toincrease their dignity.And of course, creatingthe VO is the key. Theyhave to learn that it istheir organization, thatthey manage it."(interview, woman.Education manager)

SocialDevelopment

4.2 "IWe always] start withthe VO. It is the tool thatpermits the women tostart thinking ahout theirprohlems and their lives.Also to start thinkingabout what they can do ifthe work together. This ishow we have alwaysunderstood our work, likeFreiré. It has to contributeto raising awareness andstart breaking unequalrelationships." (interview,man, SD manager)

4.5 "These are peoplewhere maybe they're notable to see that they arenot ignorant. They canthink, and they can buildup their own, you know,you can facilitate it. So, ateacher, she or he is afacilitator in a commonthinking process toimprove understandingamong people, and so westill believe that it's actionand refraction whichprovides the knowledge,the source of knowledgeand not from books orfrom other people; thesource of knowledge isintemal. You act andrefract and yourknowledge, sort of, comesfrom a process of actionand refraction rather thanfrom books." (interview,man, chairman)

4.6 "So what the womenmembers of the PalliShamajs have realized isthat they can includethemselves as members ofthe local committees, themosque committee,madrasa committee,school committee. Andmany have succeeded . . .so they then become morepowerful." (interview,man, SD program officer)

4.9 "With the localcommunity leadersworkshops what we try todo is basically toconvince the otherstructvires, apart for thepower structure, thatwe're trying to developthe villages. . . . We inviteall of them to come andto help our organization:what should we do,basically to have, to try toconvince them to work asa support group for thesepoor people, as a kind ofsupport group for thesepoor people. We also talkabout the laws we teachand that they are notcontradictory to religiousIslamic laws." (interview,man, SD manager)

4.10 "When POs visithouseholds, they areusually offered withchairs or stools to sit.However, as a matter ofstrategy, POs instead ofsitting on stool/chair, siton the ground. Thismakes the peopleembarrassed, but happy!They are embarrassedbecause they are not usedto seeing an educatedoutsider sitting on theground with them. Butthey are happy becausethe PO sits with them inan informal way as anearer one which creates afellow feeling among themand the gap becomesnarrower." (SocialDevelopment report)

4.13 "Popular theatre alsoattempts to engenderbuilding a differentsystem of educationwithin the community—nonformal educationaccessible to all. Implyingthat it does not rely onliteracy—that would atonce exclude the vastmajority of the poor. It isa process centered roundthe people, starting andending with theexpressive potential ofthe human body. Thus,learning over here is morethan entertainment, as theprocess of learning isuser-friendly. Populartheatre is inexpensive toorganize as it does notrequire an expensiveoutlay in equipment orinfrastructure."(education report)

4.14 "It comes fi-om ourfolk, popular culture. Werealized we can use it asa communication tool,communication network.As in the early days therewas no electricity, noradio, no television,people used to do thiskind of drama in the ruralareas. So we actuallywent back to borrow thisidea. People have nothingelse to do, so they comeand see the drama. Peoplelove to see, and throughthis people can becomeeducated, more awareabout their situation, theirrights, their problems,issues." (interview, man,SD manager)

Continued


TABLE 5(Continued)

Program

Developing Sensemaking Capacity [Re)combining Norms and Traditions

FacilitatingConscientization Knowledge of Repertories Demystifying

Adopting ArtisticTraditional Performances

HRLE

CFPR/TUP

4.3 "It is very difficult tostop early marriage, sonow we organizecommunity workshopswith imams, religiouspeople, with those whohold power in thecommunity. We arecreating awareness amongstakeholders, likereligious leaders, etc."(interview, woman, HRLEdirector)

4.4 "The objective isawareness about gendernorms and relations;better understanding ofrationality of jointownership by men andwomen of familyresources and women'sparticipation in decision-making in the family."(interview, woman,CFPR/TUP director)

4.7 "In our trainingprograms one of ourobjectives is to help themto collect and reflect ontheir experiences, whatthey know, the work theydo, that their parents did.This is important toincrease their dignity.And of course, creatingthe VO is the key. Theyhave to learn that it istheir organization, thatthey manage it."(interview, woman, HRLEdirector)

4.8 "Understanding thatknowing the law andrights is not enough forseeking redress—one hasto know where to go, whocan help, what do. Tohelp women gain basiclegal skills." (interview,man, CFPR/TUP seniorprogram officer)

4.11 "Our objective withthe HRLE Program is,mostly, to let the poorknow their own rights,that they have the lawand human rights. Andalso to convince themthat they can ask for theserights, and fight for them.In a way, we want todemystify the law."(interview, man. SocialDevelopment manager)

4.12 "To convince elites toparticipate in thecommittees (GBDCs), weexplain the tasks andactivities in terms that arevery close to commonunderstandings of thetraditional and religiousobligation they have tohelp the poor."(interview, man, CFPRsenior program officer)

4.15 "But popular theatrealso has problems. Webelieve that it is veryimportant to have actorsthat are women. Butsometimes religiousleaders or the villageleaders are against it,because there are womenacting and also becausethey think that the topicsare not correct. However,actors and also programorganizers resist thatopposition. They talkabout how important it isto talk about problemsthat the poor have, butalso problems ofsanitation, violence."(interview, woman, HRLEand Advocacycoordinator)

4.16 'The theaters not onlyview society as adichotomy of good andbad where the latteroppresses the former, butby following the trend ofpopular theater highlightsthe cause of theoppressed in the society.The theatres not onlyhighlighted theimplication of theproblems, but also thecauses behind theproblems, and mostimportantly hinted onhow these problemscould be resolved."(CFPR/TUP report)

der codes "facilitating conscientization" and"knowledge of repertoires." "Gonscientization" re-fers to efforts to build awareness by provoking in-dividual and collective self-reflection (Freiré, 1970).Gonfronted with the strong inner sense of nonentitle-ment that women have, BRAG starts by leadingwomen to think about their lives. In the words of amale Social Development Program manager:

This is a . . . two stage process. First they wouldcome to see themselves as in a bad situation, op-pressed; and then would come to see themselves ascitizens who had a right to a better situation.

BRAG'S use of the term "conscientization" stemsfrom the deep influence that the Brazilian educatorFreire's (1970) work had on Fazle Abed. Activitiesthat trigger conscientization aim to provoke womeninto seeing and questioning their condition so thatthey can begin "de-naturalizing" it (Douglas, 1986).Gonscientization processes involve wonien's be-coming aware of their own sense of self and visu-alizing the possibilities for expanding the boundar-ies of permissible behavior. To develop women'ssensemaking in this direction, BRAG works to gen-erate a repertoire of resources and inner capabili-


ties for women to mobilize (Nussbaum, 2000). Mar-tha Chen, an early BRAC staffer, wrote in a note in1976 (published later in 1983):

These village women possess many skills. Theymust be made to feel these skills are valuable. Thenthe women must decide for themselves what theycan and want to do. (Ghen, 1983)

BRAC also provides tools, skills, and resourcesfor women. These tools and resources range frombasic math and accounting training to more sub-stantial knowledge lessons about citizen rights. Forinstance, referring to a new initiative by the HRLEProgram to empower poor women through propertyrights, the program director explained:

Before property rights can empower people, peoplemust be empowered to claim their property rights.Our legal literacy courses combine legal aid witbrigbts articulation, taking legal empowerment be-yond courts and to hands-on application at thegrassroots.

(Re)combine norms and traditions. Developingwomen's sensemaking capacity is inherently polit-ical. BRAC's efforts to build self-awareness andfoster women's mobilization are not uncontrover-sial and often generate opposition. In response,BRAC works to provide alternative definitions totraditional notions such as "household property."This second type of legitimating work is about"(re)combining norms and traditions" and aggre-gates the two first-order codes, "demystifying" and"adopting artistic traditional performances." Ourepigraph at the beginning of this paper, from thework of Marta Nussbaum on human capabilities,provides a direct statement of the rationale forthese activities.

Efforts to change prevailing institutionalizedpractices, customs, and beliefs challenge manyconventions and sometimes breed opposition. Avery graphic example of this occurred in the 1990s,when 110 BRAC schools were set on fire by radicalswho used religious arguments to claim that mixingboys and girls in class violated the values andnorms of Bangladeshi society (Riaz, 2005). Opposi-tion can be understood as an expression againstdominant actors' produced meanings—meaningsthat articulate implicit hierarchies, reproduce theiradvantages, and stabilize a particular "local world"(Fligstein, 2001; Tilly, 1998).

The evidence from our informants suggests thatin traditional arrangements, women are often so-cialized to occupy and treated as holding depen-dent positions. Our analysis suggests that BRAC'sengagement with poor women via participation inVOs denaturalizes and demystifies the traditional

relations of male-female subordination and patron-age and helps women feel at ease in the company ofnonfamily "strangers." As the following quote froman internal report from the Social DevelopmentProgram illustrates, this démystification ispurposeful:

When Program Organizers visit households, tbey areusually offered with chairs or stools to sit. However,as a matter of strategy, POs instead of sitting onstool/chair, sit on the ground. This makes the peopleembarrassed, but bappy! They are embarrassed be-cause they are not used to see an educated outsidersitting on the ground witb tbem. But tbey are bappybecause the PO sits with tbem in an informal way.Then tbe gap between them becomes narrower.

BRAC also works with women to demystify theirview of the law as opaque, out of reach, and inap-plicable to themselves. In HRLE training in partic-ular, women discuss and learn that the law is notnecessarily "contradictory to Islamic religious law"(quotation 4.9).

BRAC also works with local elites—e.g., villagechiefs, religious leaders, teachers, policemen—todemystify BRAC's own interventions. For example,when CFPR program organizers approach elites toprovide support to women (e.g., granting fair jus-tice in traditional shalish), they emphasize thatsuch support is nothing more than a more formal-ized version of what the local elites have alwaysbeen responsible for in the community. The CFPRdirector told us:

First of all we acknowledged their contribution totheir community, so they come to see what we askthem to do as notbing extraordineiry, but . . . wbattheir father did.

In addition, BRAC carefully uses religious argu-ments to support its own initiatives, such as fram-ing elite support for the poor in "terms very close tocommon understandings of the [elites'] traditionaland religious obligation." A local journalistexplained:

It directly resonates with one of the Five Pillars ofIslam, which states that the rich have a moral obli-gation to belp tbe needy.

The issue of public framing and presentation isimportant when dealing with institutions such aswomen's autonomy and property rights. We ob-served that BRAC purposively draws upon and in-tegrates traditional artistic and cultural perfor-mances in its activities. These actions facilitatesensemaking and help to legitimize discourse thatfavors inclusion. This integration can also be inter-preted as an experiment with old rules and prac-tices that attempts to make sense of new situations


(Fligstein, 2001). For instance, BRAC draws onBangladesh's rich popular folk culture by incorpo-rating popular theater performances into its advo-cacy efforts and educational projects. Theater pro-vides a medium for discussing controversial issuesthat all villagers understand. A Social Developmentmanager referred to theater's ability to

ease the public discussion about issues such asdomination and exploitation, land grabbing, womendiscrimination, village arbitration, or harassmentof women.

Elaborating on the plot of a performance one of theauthors attended that gathered 300 people, includ-ing local government members and a religiousleader, the manager further recalled:

Someone is beating his wife, they know there is thisproblem in the village, because they are from there.They take story, and they represent the drama andthey ask to the audience: Do you think this shouldbe done? Should he beat bis wife? And at the end oftbe drama tbey ask to tbe audience, what can belearned from this drama?

Thus, via the highly legitimized institution of pop-ular theater, BRAC introduces a new image of wom-en's autonomy into a community and provides avenue that safely fosters debate about communitynorms, appropriate behaviors, and the evolution ofold practices. Furthermore, by employing the me-dium of theater, BRAC not only permits viewers tovisualize this new realify and its possible implica-tions, but also creates a reason to discuss the causesand potential responses for the situations on display.

DISCUSSION AND CONCLUSION

This paper centers on market building in institu-tionally complex contexts. Our analysis of the in-stitutional context in Bangladesh and the activitiesof BRAC allow us to develop a grounded theory ofinstitutional voids built on a rich empirical case inwhich market building is being undertaken as atool for economic and social development. We nowelaborate on how our findings contribute to andextend existing institutional accounts of the forma-tion and functioning of markets.

From Institutional Voids to InstitutionalInterfaces

There may well be some instances and arenasthat are "empty" of institutional arrangements;however, they are uncommon. Although the ex-treme nature of such instances makes them poten-tially useful for analysis, our study focuses not on

the absence and weakness of modern market insti-tutions but rather on their "situatedness" within amulti-institutional context—constituting what wehave called an "institutionally complex context"(Greenwood, Magán, Li, & Céspedes, 2010; Greif,2006). The unitary view of voids makes it difficultto imagine how markets can be built or operatewithin anything other than a very narrow set ofinstitutional contexts. This view also reinforces acompensatory view of institutional arrangements,rather than recognizing how indigenous institu-tions do support varied complex market activitiesand governance (Ostrom, 1990; Pierson & Skocpol,2002). Our research enables a move away fromabstraction toward a grounded knowledge of insti-tutional voids shaped by investigating how theytake form and are managed in the process of marketbuilding. We arrived at this insight by acknowledg-ing the existence of multiple institutional logicsand analytically identifying interfaces as the pointsat which these logics come together. This perspec-tive made it possible to conduct a fine-grainedexamination of two core institutions related tomarket creation and stability: property rights andautonomy.

Our data reveal that property rights and autono-my—taken for granted as pillars of markets in mod-ern societies and economies—often contradict andthus stand in conflict with existing rules of thegame in local community (e.g., patriarchy, earlyforced marriage), political (e.g., corruption and pa-tronage), and religious (e.g., interpretations of reli-gious credos, purdah) spheres in Bangladesh.These institutional interfaces configure exclusion-ary markets, a point we note is secondary in theliterature on institutional voids and merely notedin passing by institutional economists concernedwith distributional outcomes of markets. The ap-proach introduced in this study focuses on thosewho are denied the chance to "play the game" andexplains how many poor women in Bangladesh areexcluded from market-based activities.

Attention to the interfaces between market insti-tutions and local institutional arrangements has atleast two important implications for the study ofinstitutional voids. First, our analysis illustratesthe importance of looking at a full array of inter-linked institutional spheres (Friedland & Alford,1991). Understanding what prevents women fiomaccessing and participating in markets permits usto see how various institutional logics shape exist-ing rules of the game (North, 1990). Our findingsthat patriarchal systems, religious beliefs, and localconceptions of "proper" behavior limit women'saccess to and participation in markets suggests thatexisting institutional arrangements, and the result-


ing institutional voids, can be seen as outcomes ofcultural and political contention among actors withdifferential power and competing frames (Bartley,2007; Camphell, 2004; Rao, 1998). We show thatthe protection of property rights is variahly grantedby the powerful to some but not to others. As aresult, the institutional arrangements that shapeinstitutional voids can be understood as relativelydurable, but contestable, compromises based onspecific coalitional dynamics that are potentiallyvulnerable to shifts.

Second, our focus and mapping of both marketand nonmarket institutions specifies more fully theinstitutional arrangements relevant for understand-ing market-based activities in developing coun-tries. In referring to bottom of the pyramid markets,Wehh et al. pointed out that there is "little to noproperty rights protection available in the event ofviolations" (2010: 506). Our analysis suggests thatalternatives such as traditional means of issue res-olution have consequences for market building andaccess. Not only do these alternative mechanismsexist, but they are often preferred by local actorsbecause of their cultural and cognitive proximityand ease of use. Of course, such mechanisms arenot ideal; as we show, they are often captured hydominant actors and serve to reproduce existingpatterns of subordination and exclusion. However,our investigation of BRAC's interventions showsthat it is not always necessary to create replicas ofWestern institutions when they are ahsent or weak.Rather, adding to one of the main lessons derivedfrom developing countries (Rodrik, 2007) and"marketization" processes in Eastern Europe(Kogut et al., 2000; Stark, 1996), we echo Dixit, whowrote "It may be possible to work with such alter-native institutions as are available, and build onthem" (2004: 4).

In sum, we propose that apparent institutionalvoids can be seen as useful problem-sensingtools. They can help to diagnose conditions thatneed to be addressed for inclusive market initia-tives to develop. They are analytical spaces (Ma-honey & Thelen, 2010) in which elements from avariety of institutional spheres, each built aroundcentral systems of meanings and social practices(Friedland & Alford, 1991), come together andaffect the interpretation, enforcement, or embodi-ment of certain focal institutions—such as, in ourcase, property rights and autonomy (Carruthers &Ariovich, 2004).

On the Work of Building Inclusive Markets

This article builds on existing scholarship onmarkets and institutions but refocuses on actors

and activities on the ground. In contrast to previouswork that highlights the role of the state, firms,social movements, or entrepreneurs in market-building processes, we begin with a focus on a keyintermediary actor, BRAC, hut we also provide amore detailed analysis of a range of individual ac-tors and communities, along with various inhahit-ants of existing social structures and institutionallogics. We focus on market building, not simplymarket formation, to emphasize the activity andprocesses involved. And we also focus on marketbuilding that has the explicit purpose of includingthe formerly excluded.

Our reading of autonomy and property rights asvoid for many hecomes the starting point for exam-ining BRAC's efforts to huild inclusive markets.Like other liminal spaces, the interfaces we exam-ine represent spaces that illuminate the conflictthat occurs over and within institutions (Mahoney& Thelen, 2010; Morrill, 2006). These interfaces,though seemingly fixed in many ways, are alsounstable and contested, and therefore represent op-portunities for actors to create and transform therelations, boundaries, or rules of the game withinthem (Fligstein, 2001; Lawrence & Suddaby, 2006;Santos & Eisenhardt, 2009). Our approach reflectsthe centrality of such microsocial processes, inter-actions, and (re)negotiations for understandinghow institutions "act and play" (Barley, 2008).

We find that on-the-ground market-huilding ac-tivities situated in institutionally complex contextsenable market access and participation through therenegotiation of existing social orders. An interme-diary organization initiates this process of renego-tiation, hut it unfolds on behalf of and with thepeople affected, the marginalized (poor women)and the ones who marginalize (elites, religiousleaders, etc.). It includes two distinct categories ofactivities: redefining market architecture and legit-imating new actors.

Our findings support a perspective on markethuilding in which markets are viewed as builtalong with rather than on top of existing local in-stitutions and that allows "markets to hecome"rather than pushes for "markets to exist." In con-trast to previous work, we integrate market conse-quences into our argument. This move underscoresthe value of attending to plural institutionalspheres and provides a more complete understand-ing of how markets form, how intermediaries im-pact markets, and why it is important to take intoaccount dimensions of inclusiveness. This perspec-tive suggests that the inclusive market that BRAC isstriving to build amalgamates with the existing lo-cal institutional arrangements rather than consti-tuting an isolated institutional sphere.


Our findings illustrate how market-huilding ac-tivities are located at the interfaces of institutionalspheres and how they often modify existing norms,heliefs, and practices to alter the imderlying socialorder (Strauss, 1982). Our study sees market accessand participation as negotiable and market hound-aries as potentially permeable for actors who havebeen excluded. These findings and interpretationsreinforce ongoing conversations ahout inhabitedinstitutions (Barley, 2008; Hallett, 2010; Scully &Creed, 1999).

Our findings related to redefining market archi-tecture extend the current view that markets canadopt different architectiu-es. According to the so-ciology of markets literature, markets and their sup-porting institutions are built through rules, conven-tions, and the codification of meaning in the formof standards (Beckert, 2007; Biggart & Beamish,2003; Fourcade, 2007). Once a particular architec-ttu-al configuration takes form, it defines who cando what and who has access to what, and it be-comes difficult to stray from the configuration(Fligstein, 2001).

This stream of work emphasizes market struc-tures, however, in ways that are sometimes overlystylized or abstract. Instead, we suggest that atten-tion to negotiation activities surrounding existinginstitutional patterns can show how inclusive mar-kets can be renegotiated even from initially restric-tive structures (e.g., to formal justice). Moreover,focusing on market architecture highlights the piv-otal role that different types of social spaces playrelative to market building. Some purposively de-signed spaces may be privileged settings in whichindividuals from disparate groups can (re)negotiateexisting social orders and seek microinstitutionalchange. Recent work by social movement scholars(PoUeta, 1999) and organizational theorists (Kel-logg, 2009; Zietsma & Lawrence, 2010) suggests theimportance of different social spaces for these ends.Our findings speak to and provide an empiricalpath to continue these conversations.

Our analysis also reveals how purposeful effortto legitimate new market actors complements ini-tiatives to define and shape market architecture.Our findings echo perspectives in entrepreneur-ship that combine legitimacy and cognitive-basedstrategies and suggest that persuasion and infiu-ence can be used to overcome the skepticism andresistance of those who guard the status quo (Al-drich & Fiol, 1994). In addition, various studieshave shown how actors bestow legitimacy on prac-tices, products, and services (Suchman, 1995;Vaara & Tienari, 2008) by using speaking and writ-ing (Lawrence & Suddaby, 2006: 56), rhetoric(Creen, 2004; Suddaby & Creenwood, 2005), or

analogies and metaphors (Etzion & Ferraro, 2010;Hargadon & Douglas, 2001).

In line with this tradition, our study illustrateshow rhetorical strategies and culture can be used asa tool kit (Swidler, 1986) (e.g., popular theater per-formances; local means of issue resolution) and canbe actively and strategically deployed by an inter-mediary actor to convey legitimacy. Moreover, ourstudy allows for an in-depth look at how legitimacyis constructed and negotiated (Barley, 2008;Strauss, 1982). "As institutional analysis takes itsinterpretive turn, it is well worth remembering thatwriting, reading and rhetoric are important for ne-gotiating legitimacy, but words break no bones"(Barley, 2008: 507). Our study represents an at-tempt to do so by applying a more diverse reper-toire of methods (see also Dover & Lawrence, 2010)to study institutions inhabited and at play in lessthan comfortable or conventional contexts (Bam-berger & Pratt, 2010).

Limitations and Conclusions

This study reports on the activities of BRAC tobuild inclusive markets in rural Bangladesh. We donot intend to claim that our findings represent theonly way in which inclusive markets might bebuilt. In fact, we do not believe that there is onlyone way to build markets (Rodrik, 2007). Our in-tention was to illustrate how exclusion from marketactivities can be traced back to institutional voidsand to surface the microprocesses involved inbuilding inclusive markets by examining the activ-ities and role of an intermediary.

Markets are not constructed de novo (Stark,1996). The choice of setting and the local condi-tions exemplified in this case restrict the general-izability and transferability of its findings, particu-larly with regard to modern societies in whichprocesses of secularization have diminished thecentrality of religious institutions and to societiesin which what we call the emporium of the lawgrants inclusion effectively. However, our analyti-cal approach of spotting possible voids and theo-retical insights on the negotiated order of marketaccess and permeability of market boundariesmight provide a useful lens for researchers anddecision makers studying a variety of phenomena.For example, the recent demographic shifts and theconsequent rise of "Islamic banking" in Europe rep-resent an interesting setting for analyzing how tomake [financial] markets more accessible in mod-ern societies.

Despite these boimdary conditions, an extremecase in complex institutional context such as theone presented in this paper offers an opportunity to


study a familiar set of processes and phenomena onfresh terms and to focus in on key elements thatexisting work has neglected (Bamberger & Pratt,2010). Moreover, too much research on institutionshas been "embarrassingly . . . constructed by U.S.scholars based on data collected from U.S. organi-zations" (Scott, 2005: 478). Our case stands apart inshowcasing an "unusual" (albeit, in reality, verycustomary and ubiquitous) and a largely unex-plored setting. This case of market building in ruralBangladesh is uncommon and anal3^cally extreme inthe literature, but common in the world and hence,relevant and timely for our research community

REFERENGES

Ahlstrom, D., & Bruton, G. D. 2006. Venture capital inemerging economies: Networks and institutionalchange. Entrepreneurship Theory and Practice, 30:299-320.

Aldrich, H. E., & Fiol, C. M. 1994. Fools rush in? Theinstitutional context of industry creation. Academyof Management Review, 19: 645-667.

Anteby, M. 2010. Markets, morals, and practices of trade:Jurisdictional disputes in the U.S. commerce in ca-davers. Administrative Science Quarterly, 55: 606-638.

Bamberger, P. A., & Pratt, M. G. 2010. Moving forwardby looking back: Reclaiming unconventional re-search contexts and samples in organizationalscholarship. Academy of Management Journal, 53:665-671.

Banerjee, A. V., & Duflo, E. 2011. Poor economics: Aradical rethinking of the way to flght global pov-erty. New York: Public Affairs.

Banfield, E. 1958. The moral basis of a backward soci-ety. Glencoe, IL: Free Press.

Bansal, P., & Roth, K. 2000. Why companies go green: Amodel of ecological responsiveness. Academy ofManagement Journal, 43: 717-736.

Barley, S. 2008. Goalface institutionalism. In R. Green-wood, C. Oliver, R. Suddaby, & K. Sahlin-Andersson(Eds.), The Sage handbook of organizational insti-tutionalism: 490-516. Newbury Park, CA: Sage.

Barley, S. R., & Tolbert, P. S. 1997. Institutionalizationand structuration: Studying the links between actionand institution. Organization Studies, 18: 93-117.

Bartley, T. 2007. Institutional emergence in an era ofglobalization: the rise of transnational private regu-lation of labor and environmental conditions. Amer-ican Journal of Sociology, 113: 297-351.

Beckert, J. 2007. The great transformation of embed-dedness: Karl Polanyi and the new economic soci-

ology. Discussion paper. Max Planck Institute for theStudy of Societies, Colonia.

Biggart N. W., & Guillen, M. 1999. Developing difference:Social organization and the rise of the auto indus-tries in South Korea, Taiwan, Spain, and Argentine.American Sociological Review, 64: 711-7M.

Biggart, N. W., & Beamish, T. D. 2003. The economicsociology of conventions: Habit, custom, practiceand routine in market order. In K. Cook & J. Hagan(Eds.), Annual review of sociology, vol. 29: 443-464. Palo Alto, CA: Annual Reviews.

Bourdieu, P. 2005. The social structures of the economy.Cambridge, U.K.: Polity.

Bowles, S. 1998. Endogenous preferences: the culturalconsequences of markets and other economic insti-tutions. Journal of Economic Literature, 36: 75-111.

BRAC. 1979. Who gets what and why: Resource alloca-tion in a Bangladesh village (RED Rural Study Se-ries 1). Dhaka, Bangladesh: BRAC.

BRAC. 2009. Annual report, http://wvirw.brac.net/oldsite/useruploads/files/brac-ar-2009.pdf. AccessedFebruary 11, 2011.

Campbell, J. L. 2004. Institutional change and global-ization. Princeton, NJ: Princeton University Press.

Campbell, J. L., & Lindberg, L. N. 1990. Property rightsand the organization of economic activity by thestate. American Sociological Review, 55: 634-647.

Carruthers, B. G., & Ariovich, L. 2004. The sociology ofproperty rights. In K. Cook & J. Hagan (Eds.), Annualreview of sociology, vol. 30: 23-46. Palo Alto, CA:Annual Reviews.

Chakrabarty, S. 2009. The influence of national cultureand institutional voids on family ownership of largefirms: a country level empirical study. Journal ofInternational Management, 15: 32-45.

Chan, H.-J. 2002. Kicking away the ladder: Develop-ment strategy in historical perspective. London:Anthem.

Chen, M. A. 1983. A quiet revolution: Women in tran-sition in rural Bangladesh. Rochester, VT: Shenk-man.

Creed, D., Dejordy, R., & Lok, J. 2010. Being the change:resolving institutional contradiction through iden-tity work. Academy of Management Journal, 531336-1364.

Crow, B. 2001. Markets, class and social change: Trad-ing networks and poverty in rural South Asia. NewYork: Palgrave.

De Soto, H. 2000. The mystery of capital: Why capital-ism triumphs in the West and fails everywhereelse. New York: Basic Books.


Denzin, N. 1989. The research act: A theoretical intro-duction to sociological methods (3rd ed.). Engle-wood Gliffs, NJ: Prentice Hall.

Dixit, A. K. 2004. Lawlessness and economics. Prince-ton, NJ: Princeton University Press.

Douglas, M. 1986. How institutions think. Syracuse, NY:Syracuse University Press.

Dover, G., & Lawrence, T. 2010. A gap year for institu-tional tbeory: Integrating institutional work and par-ticipatory action research. Journal of ManagementInquiry, 19: 305-316.

Easterly, W. 2002. The elusive quest for growth. Gam-bridge, MA: MIT Press.

Economist 2010. Face value: BRAG in business, http://www.economist.com/node/15546464. Accessed Feb-ruary 17, 2011.

Eisenhardt, K. M. 1989. Building tbeories from casestudy researcb. Academy of Management Review,14: 532-550.

Eisenbardt, K. M., & Graebner, M. E. 2007. Tbeory build-ing from cases: Opportunities and challenges. Acad-emy of Management Journal, 50: 25-32.

Etzion, D., & Ferraro, F. 2010. The role of analogy in theinstitutionalization of sustainability reporting. Or-ganization Science, 21: 1092-1107.

Fernandez, A. M., & Kraay, A. G. 2007. Property rightsinstitutions, contracting institutions and growth inSoutb Asia: Macro and micro evidence. In S. Ahmed& E. Ghani (Eds.), South Asia: Growth and regionalintegration: 80-109. New Delhi: Macmillan.

Fligstein, N. 2001. The architecture of markets: an eco-nomic sociology of capitalist societies. Princeton,NJ: Princeton University Press.

Flyvbjerg, B. 2011. Gase study. In N. K. Denzin & Y. S.Lincoln (Eds.), The Sage handbook of qualitativeresearch (4th ed.): 301-316. Thousand Oaks, GA:Sage.

Fourcade, M. 2007. Theories of markets, tbeories of society.American Behavioral Scientist, 50: 1015-1034.

Freiré, P. 1970. Pedagogy of the oppressed. New York:Herder and Herder.

Friedland, R., & Alford, R. R. 1991. Bringing society backin: Symbols, practices, and institutional contradic-tions. In W. W. Powell, & P. DiMaggio (Eds.), Thenew institutionalism in organizational analysis:232-263. Ghicago: University of Ghicago Press.

Geertz, G. 1978. Tbe bazaar economy: Information andsearcb in peasant marketing. American EconomicReview, 68: 28-32.

Glaser, B., & Strauss, Y. A. 1967. The discovery ofgrounded theory: Strategies for qualitative re-search. New York: Aldine.

Goffman, E. 1961. Asylums: essays on the social situa-tion of mental patients and other inmates. NewYork: Ancbor Books.

Granovetter, M., & McGuire, P. 1998. The making of anindustry: Electricity in the United States. In M. Gal-lon (Ed.), The ¡aw of markets: 147-173. Oxford,U.K.: Blackwell.

Green, S. 2004. A rhetorical theory of diffusion. Acad-emy of Management Review, 29: 653-699.

Greenwood, R., Magán, A., Li, S. X., & Géspedes, J. 2010.The multiplicity of institutional logics and the het-erogeneity of organizational responses. Organiza-tion Science, 21: 521-539.

Greif, A. 2006. Institutions and the path to the modemeconomy: Lessons from medieval trade. Gam-bridge, U.K.: Cambridge University Press.

Hall, P. A., & Soskice, D. W., 2001. Varieties of capital-ism: The institutional foundations of comparativeadvantage. Oxford/New York: Oxford UniversityPress.

Hallett, T., & Ventresca, M. J. 2006. Inhabited institu-tions: social interactions and organizational forms inGouldner's "patterns of industrial bureaucracy."Theory and Society, 35: 213-236.

Hamilton, G. G., & Biggart, N. W. 1988. Market, culture,and authority: A comparative analysis of manage-ment and organization in the Far East. AmericanJournal of Sociology, 94: S52-S94.

Hargadon, A. B., & Douglas, Y. 2001. When innovationsmeet institutions: Edison and tbe design of the elec-tric light. Administrative Science Quarterly, 46:476-501.

Heritage Foundation. 2010. Economic opportunity S-prosperity: 2010 Index of economic freedom.http://heritage.org/index/Gountry/Bangladesh. Ac-cessed July 26.

Islam, N., & Asaduzzaman, M. (Eds.). 2008. A ship adrift:governance and development in Bangladesh (Ban-gladesh Institute for Development Studies), http://www.bids.org.bd/bids-bd/images/a_ship_adrift.pdfAccessed March 10, 2011.

Jepperson, R. L., & Meyer, J. W. 1991. The public orderaind tbe construction of formal organizations. InW. W. Powell & P. J. DiMaggio (Eds.), The newinstitutionalism in organizational analysis: 204-231. Ghicago and London: University of GhicagoPress.

Kabeer, N. 2000. The power to choose: Bangladeshiwomen and labour market decisions in Londonand Dhaka. London: Verso.

Kellogg, K. G. 2009. Operating room: Relational spacesemd microinstitutional cbange in surgery. AmericanJournal of Sociology, 115: 657-711.

2012

Khanna, T., & Palepu, K. 1997. Why focused strategiesmay he wrong for emerging markets. Harvard Rusi-ness Review, 75(4): 41-51.

Khanna, T., Palepu, K. G., & Sinha, J. 2005. Strategies thatfit emerging markets. Harvard Rusiness Review,83(6): 63-76.

Khanna, T., & Rivkin, J. W. 2006. Interorganizational tiesand business group boundaries: Evidence from an emerg-ing economy. Organization Science, 17: 333-352.

Körten, D. (Ed.). 1987. Community management. WestHartford, GT: Kumarian.

Kumar, N., Stern, L., & Anderson, J. G. 1993. Gonductinginterorganizational research using key informants.Academy of Management Jaumal, 36: 1633-1651.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny,R. 1998. Law and finance. Joumal of Political Econ-omy, 106: 1113-1155.

Lawrence, T. B., & Suddaby, R. 2006. Institutions andinstitutional work. In S. Glegg, W. R. Nord, & T. B.Lawrence (Eds.), Handbook of organization stud-ies: 215-254. London: Sage.

Leff, N. 1976. Gapital markets in the less developed coun-tries: The group principal. In R. McKinnon (Ed.),Money and finance in economic growth and devel-opment: 97-122. New York: Dekker.

Lincoln, Y., & Guba, E. 1985. Naturalistic inquiry. NewYork: Sage.

Lounsbury, M., Ventresca, M. J., & Hirsch, P. 2003. Socialmovements, field frames, and industry emergence.Socio-Economic Review, 1: 71-104.

Lovell, G. H. 1992. Rreaking the cycle of poverty: TheRRAC strategy. West Hartford, GT: Kumarian.

Mahoney, J., & Thelen, K. A. 2010. A theory of gradualinstitutional change. In J. Mahoney & K. A. Thelen(Eds.), Explaining institutional change: 1-37. Gam-bridge, U.K.: Gambridge University Press.

Mair, J., & Marti, I. 2009. Entrepreneurship in and aroundinstitutional voids: A case study from Bangladesh.Joumal of Rusiness Venturing, 24: 419-435.

Maitlis, S., & Lawrence, T. 2007. Triggers and enablers ofsensegiving in organizations. Academy of Manage-ment Joumal, 50: 57-84.

Matin, I. 2002. Targeted development programmes forthe extreme poor: Experiences fiom RRAC exper-iments. Working paper no 20, GPRG. http://www.chronicpoverty.org/uploads/publication_files/PRGPB_WP_2.pdf. Accessed March 11, 2011.

McMillan, J. 2002. Reinventing the bazaar: A naturalhistory of markets. New York: Norton.

Mendoza, R. U., & Thelen, N. 2008. Innovations to makemarkets more inclusive for the poor. DevelopmentPolicy Review, 26: 427-458.

Mair, Marti, and Ventresca 847

Meyer, K. E., Estrin, S., Bhaumik, S. K., & Peng, M. W.2009. Institutions, resources, and entry strategies inemerging economies. Strategic Management Jour-nal, 30: 61-80.

Miles, M., & Huberman, A. M. 1994. Qualitative dataanalysis. Beverly Hills, GA: Sage.

Miller, G. G., Gardinal, L. B., & Glick, W. H. 1997. Retro-spective reports in organizational research: Reexam-ination of recent evidence. Academy of Manage-ment Joumal, 40: 189-204.

Morrill, G. 2006. Institutional change and interstitialemergence: The growth of alternative dispute reso-lution in American law, 1965-1995. In W. W. Powell& D. Jones (Eds.), Haw institutions change. Ghicago:University of Ghicago Press.

Morrison, A. D. 2004. Gompetition and information pro-duction in market maker models. Joumal of Rusi-ness Einance and Accounting, 31: 1171-1190.

Morrison, A. D., & White, L. 2009. Level playing fields ininternational capital regulation. Joumal of Einance,64: 1099-1142.

North, D. G. 1990. Institutions, institutional change andeconomic performance. Gambridge, U.K.: Gam-bridge University Press.

North, D. G. 1991. Institutions. Joumal of Economic Per-spectives, 5: 97-112.

Nussbaum, M. 2000. Women and human development:The capabilities approach. Gambridge, U.K.: Gam-bridge University Press.

Ostrom, E. 1990. Governing the conmions: The evolu-tion of institutions for collective action. Gambridge,U.K.: Gambridge University Press.

Peng, M. W., Sun, S. L., Pinkham, B., & Ghen, H. 2009.The institution-based view as a third leg for a strat-egy tripod. Academy of Management Perspectives,23(4): 63-81.

Pereira, F. 2002. The fractured scales: The search for auniform personal code. Galcutta: Stree.

Pettigrew, A. M. 1990. Longitudinal field research onchange: Theory and practice. Organization Science,1: 267-292.

Pierson, P., & Skocpol, T. 2002. Historical institutional-ism in contemporary political science. In 1. Katznel-son & H. Milner (Eds.), Political science: State ofthe discipline: 693-721. New York: Norton.

Polletta, F. 1991. "Free spaces" in collective action. The-ory and Society, 28: 1-38.

Powell, W. W., & DiMaggio, P. J. 1991. The new institu-tionalism in organizational analysis. Ghicago: Uni-versity of Ghicago Press.

Pratt, M. G., Rockmann, K. W., & Kaufmann, J. B. 2006.Gonstnicting professional identity: The role of work


and identity learning cycles in the customization ofidentity among medical residents. Academy ofManagement Journal, 49: 235-262.

Puffer, S. M., McCarthy, D. J., & Boisot, M. 2009. Entre-prenevu-ship in Russia and China: The impact offormal institutional voids. Entrepreneurship The-ory and Practice, 34: 441-467.

Rao, H. 1998. Caveat emptor: The construction of non-profit watchdog organizations. American Joumal ofSociology, 103: 912-961.

Riaz, A. 2005. Traditional institutions as tools of politicalIslam in Bangladesh. Joumal of Asian and AfricanStudies, 40: 171-196.

Rodrik, D. 2007. One economics, many recipes. Prince-ton, NJ: Princeton University Press.

Sachs, J. D. 2005. The end of poverty: Economic possi-bilities for our time. New York: Penguin.

Sako, M. 2009. Global strategies in the legalservices marketplace: Institutional impacts onvalue chain dynamics. Working paper. Universityof Oxford.

Santos, F., & Eisenhardt, K. M. 2009. Constructing mar-kets and shaping boundaries: Entrepreneurial powerin nascent fields. Academy of Management Jour-nal, 52: 643-671.

Schneiberg, M. 2005. Combining new institutionalisms:explaining institutional change in American prop-erty insurance. Sociological Forum, 1: 93-137.

Schneiberg, M. 2007. What's on the path? Path depen-dence, organizational diversity and the problem ofinstitutional change in the US economy, 1900-1950. Socio-Economic Review, 5: 47-80.

Scott, W. R. 2005. Institutional theory: Contributing to atheoretical research program. In K. G. Smith & M. A.Hitt (Eds.), Great minds in management: 460-484.Oxford, U.K.: Oxford University Press.

Scully, M., & Creed, W. E. D. 1999. Restructured families:Issues of equality and need. Annals of the AmericanAcademy of Political and Social Science, 562: 47-65.

Seelos, C, & Mair, J. 2007. Profitable husiness modelsand mcirket creation in the context of deep poverty:A strategic view. Academy of Management Per-spectives, 21(4): 49-63.

Sen, A. 1999. Development as freedom. Oxford, U.K.:Oxford University Press.

Sen, A. 2009. The idea of justice. Cambridge, MA: Har-vard University Press.

Smillie, I. 2009. Freedom from want: The remarkablesuccess story of BRAC, the global grassroots organ-ization that's winning the fight against poverty.Sterling, VA: Kumarian.

Spicer, A., McDermott, G. A., & Kogut, B. 2000. Entre-preneurship and privatization in Central Europe:The tenuous balance between destruction and cre-ation. Academy of Management Review, 25: 630-649.

Spradley, J. P. 1979. The ethnographic interview. NewYork: Wadsworth.

Stark, D. 1996. Recombinant property in East Europeancapitalism. American Joumal of Sociology, 101:993-1027.

Stiglitz, J. E. 1989. Markets, market failures, and devel-opment. American Economic Review, 79: 197-203.

SUglitz, J. E. 2000. Capital market liberalization, eco-nomic growth, and instability. World Development,28: 1075-1086.

Strauss, A. 1978. Negotiations: Varieties, contexts,processes, and social order. San Francisco: Jossey-Bass.

Strauss, A. 1982. Social worlds and legitimation processes.Studies in Symbolic Interactions, 4: 170-190.

Strauss, A., & Corbin, J. 1998. Basics of qualitative re-search: Techniques and procedures for developinggrounded theory (2nd ed.). Thousand Oaks, CA:Sage.

Suchman, M. C. 1995. Managing legitimacy: Strategicand institutional approaches. Academy of Manage-ment Joumal, 20: 571-610.

Suddahy, R., & Greenwood, R. 2005. Rhetorical strategiesof legitimacy. Administrative Science Quarterly,50: 35-67.

Swidler, A. 1986. Culture in action: symbols andstrategies. American Sociological Review, 51: 273-286.

Thelen, K. 2004. How institutions evolve: The politicaleconomy of skills in Germany, Britain, the UnitedStates, and Japan. New York: Cambridge UniversityPress.

Thornton, P., & Ocasio, W. 1999. Institutional logics andthe historical contingency of power in organizations:Executive succession in the higher education pub-lishing industry, 1958-1990. American Joumal ofSociology, 105: 801-843.

Tilly, C. 1998. Durable inequality. Berkeley: Universityof California Press.

Tilly, C, & Tilly, C. 1998. Work under capitalism. Boul-der, CO: Westview Press.

Touraine, A. 1995. Critique of modernity. Oxford, U.K.:Blackwell.

Transparency International of Bangladesh. 2009. AnnualReport: Survey on corruption in Bangladesh. Dha-ka: Uttara.


UNDP. 2008. Annual report 2008. http://www.undp.org/publications/annualreport2008/pdf/IAR2008_ENG_low.pdf. Accessed March 10, 2011.

Unger, R. 1998. Resisting gender. Thousand Oaks, CA:Sage.

Vaara, E., & Tienari, J. 2008. A discursive perspective onlegitimation strategies in MNCs. Academy of Moui-agement Review, 33: 985-993.

Van Maanen, J. 1979. The fact of fiction in organizationalethnography. Administrative Science Quarterly,24: 539-550.

Webb, J., Kistruck, G. M., Ireland, R. D., & Ketchen,D. K. 2010. The entrepreneurial process in bottom-of-the-pyramid markets: The case of multinationalcorporation/nongovernment organization alli-ances. Entrepreneurship Theory and Practice, 34:555-581.

Webb, J., Tihanyi, L., Ireland, R. D., & Sirmon, D. G. 2009.

You say illegal, I say legitimate: Entrepreneurship inthe informal economy. Academy of ManagementReview, 34: 492-510.

World Bank. 2010. World development report: Globaleconomic prospects. http://siteresources.worldbank.org/INTGEP2010/Resources/FullReport-GEPSummer2010.pdf. Accessed on January 26, 2011.

Yin, R. K. 1994. Case study research: Design and meth-ods (3rd ed.). Thousand Oaks, CA: Sage.

Zelizer, V. A. 2005. The purchase of intimacy. Prince-ton, NJ: Princeton University Press.

Zelizer, V. A. 2010. Economic lives: How culture shapesthe economy. Princeton, NJ: Princeton University Press.

Zietsma, C, & Lawrence, T. B. 2010. Institutional workin the transformation of an organizational field:The interplay of boundary work and practicework. Administrative Science Quarterly, 55: 189-221.

APPENDIX A

Interview and Participant Observation Data

Informant Type of Interviewee (Number of Interviews)Type of Participant Observation(Total Number of Observations)

InternalsBRAC managementBRAC universityBRAC programsEducation Program

Social DevelopmentProgram

HRLE Program

CFPR/TUPEconomic DevelopmentHealth Program

Research and EvaluationDepartment

Social enterprisesAarong (retail)Dairy & Food ProjectBRAC Agriculture &

Livestock EnterprisesExternals

Other NCOs and socialactivists

Microfinance institution 1Microfinance institution 2NCO in education field 1NGO in education field 2Other

Chairman (4), advisory board member (2)Pro-vice chancellor (1)

Director (1), senior manager (1), manager (2), programorganizer (2), teacher (2), students (2)

Senior manager (3), program officer (2), village organizationleader (1), village organization member (2)

Director (2), manager (1), program officer (1), beneficiary (2)

Director (2), manager (2), jimior manager (1), Beneficiary (2)Senior manager (2), program officer (2), borrowers (6)Senior manager (1), program officer (1), volunteer health

worker (1)Senior researcher (1), junior researcher (1)

Senior manager (1)Senior manager (1)Senior manager (1)

General manager (1), program organizer (1), branch manager(1), borrowers (2)

Chairman (1), program organizer (1)General managers (1)Teachers (2)

Local journalist (2), Islamic law specialist (2), missionaries (2),lawyer and member of Supreme Court of Bangladesh (1),expatriate manager (2)

Attended courses in primaryschool (2)

Attended popular theatreperfonnance (1)

Attended legal education classes(1)

Attended microcredit meeting (2)Attended health meeting (1)

Visited (2)Visited (1)

Attended microcredit meeting (2)Visited local branch office (1)Attended microcredit meeting (1)

Visited garment factories (3)


APPENDIX B

Statistics for Key Codes from the Programs Studied"

Sproncl-OrderConstructs

Creating spaces forinteraction

Expanding resourcesystems

(Re) defining localarrangements

Developing sensemakingcapacity

(Re)combining normsand traditions

Code Name F

For equalsFor unequalsTying up witb government

systemsTeaming up witb social

service providersBuilding on local means of

issue resolutionMaking use of customary

sources of social supportFacilitating

conscientizationKnowledge of repertoiresDemystifyingAdopting artistic

traditional performances

Education

'assage!

201115

19

5

16

44

241214

i Sources

1168

14

3

9

26

1788

SocialDevelopment

Passages

341221

14

15

15

34

1916

7

Sources

178

15

8

10

10

24

1213

4

HRLE

Passages

311919

26

23

23

25

183121

Sources

181514

18

16

16

19

132014

CFPR

Passages

113712

24

26

26

9

191418

Sources

5

237

14

18

18

6

127

10

" "Passages" presents the number of passages witb tbe code indicated. "Sources" presents tbe number of sources witb passages bavingtbis code.

- / s û ^ -

Johanna Mair ([email protected]) is the Hewlett Foun-dation Visiting Scholar at the Stanford Genter on Philan-thropy and Givil Society and a professor of management,organization, and leadership at the Hertie School of Gov-ernance in Berlin. Her research focuses on how novelorganizational and institutional arrangements generateeconomic and social development and the role of entre-preneurship and innovation in tbis process. Sbe receivedher Ph.D. in strategy from INSEAD.

Ignasi Marti ([email protected]) is an associate profes-sor of organization theory and entrepreneurship at theEM LYON Business School, where he is member of theOGE Research Genter. He received his Ph.D. from theEESE Business School at tbe University of Navarra. Hisresearch focuses on exclusion, dignity, resistance, entre-

preneurship, power and politics, and other institutionalprocesses.

Marc J. Ventresca ([email protected]) is onthe Strategy, Innovation and Marketing Faculty at theUniversity of Oxford and is a Faculty Fellow at tbe Insti-tute for Science, Innovation and Society. His researcbfocus is the economic sociology of strategy and innova-tion, with empirical studies of institutionally complexcontexts and market building. Gurrent projects investi-gate governance innovations in knowledge-intensive or-ganizations and industries, including global financialmarkets, "ancient" universities, and Amazon ecosystemsservices markets. He received his Ph.D. in organizationalsociology from Stanford University.

Copyright of Academy of Management Journal is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

building inclusive markets in rural bangladesh: how

Documents