r Academy of Management Journal 2016, Vol. 59, No. 6, 2021–2044. http://dx.doi.org/10.5465/amj.2015.0725

SCAFFOLDING: A PROCESS OF TRANSFORMING PATTERNS OF INEQUALITY IN SMALL-SCALE SOCIETIES

JOHANNA MAIR MIRIAM WOLF Hertie School of Governance

CHRISTIAN SEELOS Katholieke Universiteit Leuven

This study advances research on organizational efforts to tackle multidimensional, complex, and interlinked societal challenges. We examine how social inequality manifests in small-scale societies, and illustrate how it inheres in entrenched patterns of behavior and interaction. Asking how development programs can be organizing tools to transform these patterns of inequality, we use a program sponsored by an Indian non-governmental organization as our empirical window and leverage data that we collected over a decade. We identify “scaffolding” as a process that enables and or- ganizes the transformation of behavior and interaction patterns. Three interrelated mechanisms make the transformation processes adaptive and emerging alternative social orders robust: (1) mobilizing institutional, social organizational, and economic resources; (2) stabilizing new patterns of interaction that reflect an alternative social order; and (3) concealing goals that are neither anticipated nor desired by some groups. Through this analysis, we move beyond conventional thinking on unintended conse- quences proposed in classic studies on organizations, complement contemporary re- search about how organizations effect positive social change by pursuing multiple goals, and develop portable insights for organizational efforts in tackling inequality. This study provides a first link between the study of organizational efforts to alleviate social problems and the transformation of social systems.

Organizational scholars have (re)discovered the Sociologists consider inequality to be an organiz- ambition to study large unresolved societal chal- ing feature of societies (Tilly, 1998).1 It manifests in lenges. These challenges are seen as “grand”:they unequal access to opportunities and rewards for are important at a national and global scale; they different social positions or statuses within a group affect many people, including future generations; or society, and it is rooted in socially constructed and tackling them requires big, bold, and novel categories (such as gender, caste, or class) that de- ideas (Colquitt & George, 2011). Arguably one of termine boundaries for inclusion and exclusion and the most pressing grand challenges we face across demarcate positions of power and privilege. These and within societies is persistent inequality (United social categories shape and reflect a shared and Nations, 2015). taken-for-granted ordering of reality, which is closely tied to a complex institutional grammar of social norms, rules, and conventions that prescribe behavior We are grateful to Jennifer Howard-Grenville and and interactions and sanction deviance. three anonymous reviewers for their generous, con- For example, the caste system—widely consid- structive, and insightful comments. This paper results ered central to entrenched inequality in India— from a decade-long journey, and would not have been possible without the work that Gram Vikas has been involves a set of norms, rules, and beliefs that un- carrying out for more than three decades. We par- derpin categories of exclusion and reify social ticularly thank Joe Madiath and Chitra Choudhury for their patience and for sharing their experience. Kate 1 A review of extant literature on inequality goes beyond Ganly provided outstanding research assistance in the this paper. We follow a sociological tradition in analyzing early years and Marc Schneiberg and Woody Powell inequality that prioritizes inequalities across groups over offered valuable insights and comments throughout the inequality across individuals. Our approach is aligned journey. with the work of Tilly (1998) on “categorical inequalities.”

2021 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 express written permission. Users may print, download, or email articles for individual use only. 2022 Academy of Management Journal December divisions (Beteille,´ 1965; Dumont, 1980). These rules inequality in small-scale societies. Our main objec- and norms govern social practices in distinct spheres tive is to advance organizational studies of societal of social life: political, economic, and religious (Mair, challenges; we apply a multidisciplinary toolkit to Mart´ı, & Ventresca, 2012). As a result, inequality accomplish our goal (Howard-Grenville, Buckle, becomes explicit in everyday life; it is observable Hoskins, & George, 2014). in regularities of behavior and stable patterns of Understanding how organizations affect local so- interaction. These patterns perpetuate inequality, cial systems and are affected by them is central to because they enable opportunity hoarding by the organizational theory (for a review and a reminder, privileged. They restrict access to already scarce re- see Stern & Barley, 1996; also, Hinings & Greenwood, sources and opportunities for the underprivileged 2002). In a seminal study, General Gypsum Corpora- (Tilly, 1998). Therefore, inequality needs to be un- tion (a pseudonym), Gouldner (1954) showed how derstood as patterned: it inheres in patterns of in- one form of bureaucracy was replaced by another teraction and behavior. and how the change affected the community in un- The deeply relational and nested nature of in- expected ways. In TVA and the Grassroots, Selznick equality poses substantial challenges for purposive (1949) described how an organization navigated an organizational action in the form of programs environment that was full of power struggles and designed to tackle inequality. Organizational activity diverging interests. His analysis showed how the often suffers from a lack of legitimacy, motivation, and original goals and structures of the Tennessee Valley support from both the excluders and the excluded, Authority were transformed as a result of organi- because existing patterns of inequality are perceived zational commitment and formal and informal co- as “normal” and widely accepted. These aspects optation of powerful stakeholders in the local make asking how a single organization or a single community. Pressures and dynamics associated with program can tackle persistent patterns of inequality local social systems were also of theoretical interest, almost absurd. The nature of inequality also poses as they helped to uncover the sources of unintended analytical challenges for organizational scholarship, consequences in purposive action. Nevertheless, the because the complexity of social relations in a social primary focus rested on the transformation of orga- system makes determining an appropriate level for nizations and organizational structures. analysis difficult and requires a broader range of Contemporary organizational studies relate purpo- methods (Stern & Barley, 1996). sive action to societal problems. Research on organi- In this study, we analyze how purposive organi- zations that have an impact on society, create positive zational activity can tackle inequality locally—in social change, or address persistent or new social villages in rural India. We focus on locally bound problems has flourished over the last decade (see social systems, or small-scale societies (Douglas, Stephan, Patterson, Kelly, & Mair, 2016, for a recent 1986). Whereas challenges facing villages mani- review). Studies center on a diverse set of organiza- fest on a “small scale” and are confined to a physi- tions, including social enterprises, non-governmental cal place, challenges such as inequality remain organizations (NGOs), companies, and hybrid orga- “grand”—multidimensional, complex, and inter- nizational forms such as benefit corporations, that linked. Analyzing inequality in small-scale societies pursue goals related to societal problems ranging from allows us to be attentive to the social, cultural, and homelessness, unemployment, and poverty to HIV political realities on the ground, and makes organi- infection and disease (Lawrence & Dover, 2015; Pache zational research on grand challenges feasible as & Santos, 2013; Tracey & Phillips, 2016; Tracey, well as practically and theoretically meaningful. Phillips, & Jarvis, 2011). Centering on small-scale societies opens up pos- Some studies recognize the importance of local sibilities for organizations to engage deeply with institutional arrangements and place in affecting local realities and to experiment with multiple vil- the work and societal impact of these organiza- lages. It permits designing programs that not only tions (Lawrence & Dover, 2015; Mair et al., 2012; alleviate symptoms of inequality (such as unequal Venkataraman, Vermeulen, Raaijmakers, & Mair, access to social, political, and economic opportuni- 2016). Still, societal challenges are mostly treated ties), but also transform patterns of behavior and in- as setting or context (Battilana & Dorado, 2010; teraction and thus revises local social orders that McKague, Zietsma, & Oliver, 2015; Seelos & Mair, make inequality durable in the first place. In this 2007). This stream of work has paid relatively little paper, we ask how programs can be organizing tools attention to how organizations make progress on with which to transform entrenched patterns of a specific problem by affecting local social systems 2016 Mair, Wolf, and Seelos 2023 that give rise to the problem and perpetuate it. The with the organization’s purpose. Our findings can theoretical (and empirical) focus generally lies on or- be extended to a broader set of societal challenges ganizational form, internal structures, and processes that are based on interlocking normative and social (Besharov, 2014; Ebrahim, Battilana, & Mair, 2014; structures and require organizations not only to ad- Mair, Mayer, & Lutz, 2015; Tracey & Phillips, 2016), or dress social problems but to transform social systems. the relationship between the organization and rele- In the following sections, we summarize insights vant actors in the institutional field (Lawrence, Hardy, from research in social anthropology and develop- & Phillips, 2002; Mair & Hehenberger, 2014). ment studies to inform our empirical and theoretical Separating the study of social problems and social approach. Next, we introduce the context and focal systems might fit the existing paradigm in organiza- program, describe our research design and analytical tional theory. As Stern and Barley (1996: 149) con- strategy, and report our findings in the form of two tended, “The study of social systems and social narratives. In the discussion section, we elaborate on problems are different and should not be confused even how our findings contribute to the organizational though many social problems are systemic in nature.” study of societal challenges and efforts to alleviate But separation also limits the scope of relevance for inequality. We conclude with remarks for future or- organizational theory. Paying analytical and theoretical ganizational research on grand societal challenges. attention to the nested and interlinked relationship between social problems and social systems is critical BACKGROUND LITERATURE for making progress on grand societal challenges. In this paper, we report on the purposive organi- How to transform entrenched and durable patterns zational activity of Gram Vikas, an NGO best known of inequality poses puzzling questions for organiza- and widely recognized for its program that brings tional theory and practice. Finding answers and water and sanitation to rural villages in India. But this making progress on this grand challenge requires tak- aspect of its work does not reflect its members’ prin- ing seriously—building on and applying—knowledge cipal interest. They use the water and sanitation pro- generated in multiple disciplines. In this section, we gram as a means to transform patterns of inequality first draw from literature in social anthropology to within rural villages—a goal they conceal from the ground our analytical and theoretical approach and villagers they work with. The program is an adaptive help us unpack the interlinked and complex nature of structure between the organization and the local so- entrenched patterns of inequality in local social sys- cial system. It brings an array of social actors into play: tems. Second, we turn to research in development the organization sponsoring the program as well as studies to take stock of organized efforts to tackle in- groups across social divides. Using the program as our equality in the context of small-scale societies. empirical window, our analysis aims to unpack the micro-processes that transform patterns of inequality. The Patterned and Relational Nature of Inequality We identify scaffolding as a process that enables and organizes this transformation, and we specify Social anthropologists immerse themselves in “small- three generative mechanisms. First, scaffolding entails scale societies”—understood as locally bound sites— mobilizing institutional, social organizational, and to understand how these systems affect the lives of economic resources. Second, it hinges on stabilizing people on the ground. The work of Mysore Srinivas emerging patterns of interaction that reflect an alter- and his students Mary Douglas and AndreB´ eteille´ is native social order. And, third, scaffolding involves particularly helpful to us, as it is attentive to the in- concealing goals that are not desired by those affected terlock between normative and social structures on by the transformation. Most important, scaffolding the ground and pertains to the small-scale societies supports the institutionalization of a new local social we examined—villages in rural India (Beteille,´ 1965; order based on the transformation of normative social Douglas, 1986; Srinivas, 1976).2 Such villages do not structures rather than their replacement. Our findings complement classical and contem- 2 We are aware that anthropologists have used the term porary scholarship on organizations that seek to solve “small-scale societies” differently, such as in reference to societal challenges. We move beyond conventional primitive cultures in general or to isolated tribal villages thinking on hidden social forces as unintended con- specifically. Our emphasis is not on backwardness or the sequences of purposive action offered in classical unique features of a local order, but on denoting locally studies; we show how scaffolding supports pursuing bound sites that constitute small-scale societies. In this manifest and concealed organizational goals aligned paper, we use the term interchangeably with “village.” 2024 Academy of Management Journal December exemplify an ideal vision or sense of community Beteille´ ’s work in Sripuram—a rural village in the as Gemeinschaft (Tonnies,¨ 1887). Small size is not south of India documented in Caste, Class and Power necessarily associated with mutual trust or self- (Beteille,´ 1965)—exposes how class and caste shape stabilizing dynamics (Douglas, 1986). Instead, patterns of interaction and behavior. Inequality stability—regularity of behavior and patterns of inheres in such patterns. It is entrenched in local interaction—is based on a complex “grammar of orders, becomes visible in patterns of interaction and institutions,” a set of local norms and rules, shared behavior, and can be observed in three dimensions: perceptions of what is proper and improper, and the distribution of wealth, power structures, and shared commitments created and enforced by the status. Although empirically these three dimensions village (Crawford & Ostrom, 1995). are intertwined, they serve as a valuable analytical Institutions “do the thinking”; they shape bound- frame for analyzing inequality. They expose the aries between groups and reinforce local social or- morphology—form and structure—of inequality at der (Douglas, 1986). In Purity and Danger,Douglas the local level, and they also serve as markers to as- (1966) argued that norms that define and prescribe sess progress toward the transformation of behavior what is clean and what is dirty delimit judgment and interaction patterns. We build on these insights about objective reality, provide scripts for behavior in this paper to strengthen our analytical approach, and interaction, and act as evaluative devices that but also to expand current research on institutions govern daily life. Institutions safeguard access to and inequality in organizational scholarship. political, social, and economic opportunities. They The core insight of social anthropologists studying — determine who is eligible to participate in political small-scale societies that inequality not only be- decision making, market-based activities, or simply comes visible through patterns of interaction and recreational activities such as “hanging out” or behavior, but also gets reproduced through precisely — meeting in public spaces (Mair, et al., 2012). And, as these patterns is hardly surprising to organiza- Douglas’s (1966) work shows, they often make life- tional scholars interested in inequality. In a recent and-death decisions, defining who needs to be saved review, Amis, Munir, and Mair (2017) uncovered and summarized the ways in which inequality be- in situations of danger and who does not. comes reified in everyday actions and interactions This interlock between normative and social by and among individuals as members of organiza- structures characterizing small-scale societies is at tions or organizational fields. Research in social an- the center of anthropological research on the nature thropology allows us to expand the scope of inquiry and persistence of inequality exemplified in the from organizations to local systems—small-scale work of AndreB´ eteille.´ Grand designs to create more societies understood as normative and social struc- equal societies tend to concentrate on inequality tures in which patterns of inequality are perpetuated. among individuals, as acknowledged in the Indian In this study, we investigate how purposive organi- constitution and promoted in the work of Amartya zational activity can transform these entrenched Sen (1973;1992), but Beteille´ argued that inequality patterns. among social groups is more salient, especially in villages in rural India. Going beyond prominent work on caste and re- PRACTICAL DIFFICULTIES WITH TACKLING ligion (Dumont, 1980), Beteille´ showed that various ENTRENCHED PATTERNS OF INEQUALITY social categories such as class and caste are in- Programs that are established with the aim of de- terwoven, that this produces and reproduces pat- creasing inequality have a long tradition in devel- terns of inequality, and that inequality cannot be opment practice. Such programs are seen as tools reduced to one dimension. In his words (Beteille,´ with which to achieve a predefined and desired goal 2003: 2): or to empower individuals and communities to There are inequalities among individuals and dis- achieve their goals. The former perspective is asso- parities among groups; inequalities of income, esteem ciated with a social engineering logic and the latter and authority; and so on. These inequalities do not with a participatory logic. run along the same lines; they sometimes reinforce Programs following a social engineering logic rely and at other times cut across each other. Where simple on expert and scientific knowledge to define goals, judgements about increase or decrease of inequality design solutions, and assess impact. For example, are made without taking into account the multiplicity to increase access to economic opportunities for of dimensions, the judgements tend to be defective. marginalized women, program designs center on 2016 Mair, Wolf, and Seelos 2025 facilitating self-employment (entrepreneurship), train- anthropology, covering both social engineering and ing, and skill or cash transfers, and they integrate participatory programs, concluded that we actually randomized control trials to compare results with know very little about how local development in- a comparable group not benefitting from the program tervention may reshape local patterns of inequality (see Banerjee et al., 2015, for an example). Critics of (Viterna & Robertson, 2015). these programs argue that they typically focus on In this paper, we build on an understanding of a single dimension of inequality and pay little or no inequality in small-scale societies put forward by attention to the patterned nature of inequality (Viterna social anthropologists and take into consideration & Robertson, 2015). Social norms at the root of insights from development studies to understand entrenched patterns of inequality might be recognized, how programs can be tools for transforming deeply but they are ignored in planning and evaluating pro- entrenched patterns of inequality. The trans- grams (Cernea, 1985). As a result, the processes and formation requires (a) the institutionalization of an mechanisms underpinning the transformation of pat- alternative order, and (b) purposive organizational terns of inequality remain largely underspecified—a action that includes concealing undesired effects shortcoming that even those involved in planning and and preventing powerful elites from coopting the evaluating programs acknowledge (Banerjee et al., process. These two aspects make this study particu- 2015). larly relevant for organizational analysis. Programs following a participatory logic rely on the active involvement of villages or social groups in formulating program goals and the design and METHODS implementation of the program. For grassroots or- The question we asked in this study—How are ganizations, participation is an end in itself, and programs used as organizing tools to transform goals and strategies form during implementation. entrenched patterns of inequality?—is “problem For governments and international development driven,” and we apply a pragmatic approach to data organizations, participation is a means to ensure collection and analysis. A qualitative case study re- more equal access to the benefits of development. veals the difficult-to-observe mechanism and pro- For both, reversing power relations across social cesses that transform patterns of interaction and groups is crucial for tackling inequality in various behavior (Rojas, 2010; Yin, 2003). “Being there” domains, such as water supply (Cleaver, 1999; Katz & (Geertz, 1988), visiting multiple villages and gaining Sara, 1997) or education (Jimenez & Paqueo, 1996). exposure to local realities in small-scale societies we Evidence on whether and how these programs al- studied, helped us to refine the theoretically and ter patterns of exclusion or prevent powerful elites practically meaningful questions (Greenwood & from capturing program benefits is mixed at best (see Levin, 2005), to analyze data, and to derive impli- Mansuri & Rao, 2004, for a review). Defining features cations for old and new theories. such as public events to negotiate program goals make participatory programs inherently political and liable to relations of dependency, authority, and Inequality in Rural India gender. In fact, research by development anthropol- ogists has shown that most programs are dominated India is a textbook example of a hierarchical soci- by local elites. For example, in a review of water ety and an extreme case of social inequality. The projects, researchers found that benefits are largely most striking feature of inequality in India is its vis- appropriated by community leaders, with little at- ibility (Beteille,´ 2005). Placement in a social category tempt to include households let alone marginalized defines inclusion or exclusion that is evident in the groups. Even well-trained staff struggle to challenge way people live and how they dress, behave, and entrenched norms of exclusion (Katz & Sara, 1997). interact with others. An additional striking feature of Women are often systematically excluded (Agrawal, inequality in India is the subordination of the indi- 2001), and wealthier members of the community vidual to the group. Individuals have obligations to dominate decision making (Rao & Ibanez, 2003). As the group they adhere to by birth; groups have obli- a result, programs following a participatory logic gations to other groups according to their respective unintentionally but repeatedly reproduce estab- social ranks (Beteille,´ 2003). Socially constructed lished patterns of thought, behavior, and interaction. categories of caste, class, and gender define who is Not surprisingly, a recent review of studies from ranked higher and lower in society. In particular, the disciplines including economics, sociology, and caste system (Dumont, 1980; Srinivas, 1962), based 2026 Academy of Management Journal December on kinship and religion, prescribes distinct styles of tribes—are particularly pronounced (De Haan, 2004; life and regulates the social interactions within and Government of Odisha, 2011). Working with rural between groups. Caste is acquired by birth and villages in Odisha since 1979, Gram Vikas identified is believed to be changeable only through rein- poor health caused by waterborne diseases, resulting carnation, which is said to reward those who con- from widespread open defecation and the use of formed during life and punish those who did not. polluted water sources, as a concern of villagers Castes are grouped in hierarchical order, from Brah- across social and economic divides. Gram Vikas saw mins at the top down to Scheduled Castes,3 with many a water and sanitation program as an “entry-point subcastes in between, depending on the village. program,” as it allowed the organization to reach out Class is defined by ownership of land, property, to every household in a village and to communicate and means of production; it is another important direct benefits for them in the form of improved social structure and categorization scheme. Caste health. and class are similar in some aspects, but they differ In 1991, Gram Vikas initiated the Rural Health and in others. For instance, caste mobility is never Environment Program (RHEP), the manifest goal of downward, whereas class mobility may be. For which is to provide every household in a village— a more detailed account of the interrelationships regardless of the social category to which the between caste and class, see Beteille´ (1965; 2005). household belongs—with piped drinking water and Over the past decades, the caste system has be- a separate toilet and bathing room.4 The program is come closely interpenetrated with class structures unique among existing water and sanitation pro- (Beteille,´ 2003). Gender continues to be an important grams. It combines elements of social engineering category that prescribes exclusion for women from programs, focusing on the construction of water and many aspects of social, economic, and political life. sanitation facilities, with elements of participatory The Indian constitution prohibits discrimination on programs empowering rural villages to solve their grounds of social categories. Yet, particularly in rural own problems. areas, these social divides shape local realities. As Four features illustrate the RHEP’s uniqueness. research shows, there is little mobility between First, a “100% inclusion” rule prescribes working social groups—the boundaries between castes and together across religious, social, and economic di- classes appear almost insurmountable (Banerjee, vides to construct and use water and sanitation fa- Duflo, Ghatak, & Lafortune, 2013; Chaudhuri & cilities. The RHEP formally starts only after each Ravallion, 2006; De Haan, 2004). household in the village has agreed to the 100% rule. Second, a multilayered local governance structure formalizes equal representation of all social groups Gram Vikas and the Rural Health and Environment in decision making. It includes a Village Executive Program Committee, which is officially registered under In- Gram Vikas is an Indian NGO that confronts pat- dia’s Societies Registration Act of 1860. The com- terns of inequality in rural villages in Odisha, one of mittee is responsible for implementing the program India’s poorest states. Forty-seven percent of the and for maintaining and monitoring the use of the population lives below the poverty line, and dis- water and sanitation infrastructure built. Third, crimination and exclusion of social groups— a corpus fund, administered by the Village Executive particularly women, lower castes, and indigenous Committee, is collected to ensure that financial re- sources are available to build water and sanitation infrastructure for families entering the village after 3 “Scheduled Castes”—often referred to as “untouch- the completion of the RHEP. Each household is re- ables” or “Dalits”—refers to the lowest status groups in the quired to contribute to the fund according to its “ ”— ’ caste system. Together with Scheduled Tribes India s ability to pay. And, fourth, the water and sanitation “ ”— indigenous tribal people also known as Adivasis they infrastructure is collectively constructed. Each are among the poorest in the country. In areas where the household is required to contribute labor and mate- tribal population and Hindus are residentially mixed, rial resources, and Gram Vikas provides material Scheduled Tribes tend to have a low status within the caste hierarchy. Higher castes take extreme care to prevent physical and social contact through complicated pro- 4 In 2004, Gram Vikas changed the name of the program hibitions on the sharing of food and water and other forms from RHEP to Mantra (Movement and Network for the of interaction with these groups (Dumont, 1980; Srinivas, Transformation of Rural Areas). The change of name did 1955). not affect the objective and content of our analysis. 2016 Mair, Wolf, and Seelos 2027 support and know-how. According to Gram Vikas, as well as situational biases (Miller, Cardinal, & Glick, of January 2015, the program has been implemented 1997). For example, we interviewed the founder of in 1,140 villages across 28 districts in rural Odisha Gram Vikas, Joe Madiath, multiple times, with in- (Gram Vikas, 2015). terviews lasting from 15 minutes to 4 hours. We observed him in internal meetings in the Gram Vikas headquarters and while he was speaking to Data Sources different audiences at various events (see Table 1). The data for this study were collected and ana- We reduced interviewer biases by conducting in- lyzed over the course of ten years. Our study draws terviews involving different research team mem- on multiple sources of data and several rounds of bers and by comparing notes across interviews data collection. We conducted interviews, collected conducted by different team members; see Table 1 internal and external reports as well as video mate- for a list of interviews conducted. rial, engaged in participant observation and shad- Secondary data. We collected internal and ex- owing, and consulted with research teams that have ternal reports written by Gram Vikas staff, such as studied the focal program from a different theoretical annual reports, project reports, mission reports, and and empirical angle.5 Figure 1 provides an overview progress reports, as well as presentations. Reports of the data collection process. are an internal learning device for organizations Interviews. Our interview partners included (Seelos & Mair, 2016); they provided us with an un- members of the founding team of Gram Vikas, Gram derstanding of the historical development of the Vikas staff members coordinating the water and program and sense-making of the organization. We sanitation program with varied degrees of experi- consulted film material and documentaries created ence with the program, support and partner organi- by third parties. Film and documentaries allowed us zations, experts in development and water and to triangulate observations from our own fieldwork sanitation, village leaders, and villagers. The inter- and more systematically include voices of people views lasted from 15 minutes to several hours and affected by the program. Last, we ensured access were audio recorded and transcribed verbatim. For to program evaluation reports by independent interviews within villages, we relied on a local agencies. translator. The more than 100 documents we collected hel- Exploratory interviews in the early stages of data ped us to deepen our insights from interviews and collection helped us to generate an understanding of field trips, and to assess and compare internal and the program and the local context (Strauss & Corbin, external perceptions and representations of the 1990). During our second field trip, we conducted program. We relied on government reports, legal re- semi-structured interviews that allowed us to solid- ports, reports from development agencies (such as ify emerging conceptual and empirical patterns. We the World Bank’s World Development Reports, conducted a third round of interviews to triangu- World Health Organization reports, United Nations late and consolidate specific aspects of preliminary Children’s Fund reports, and reports commissioned findings. by the local government in Odisha), and local, In addition, we applied three measures to ensure national, and international newspapers (such as and enhance robustness in our data. First, we re- Odisha Diary,theHindu Times, the Economist)to peated interviews with the same interviewee in further develop our understanding of the local, re- different locales and different years. Second, we gional, and national context. We used these docu- corroborated the content of the interviews with ments and videos as an archival database for our multiple sources: reports, archival data, and public analysis. A full list is available from the authors. appearances. This method helped us to mitigate Participant observation and shadowing. We used potential biases of any individual respondent as participant observation to better understand local realities, and especially to overcome our biases as researchers trained and socialized in a different— 5 We exchanged views and discussed our findings with Western—context. Exposure to the daily life and an anthropologist who studied life in tribal villages that work with Gram Vikas, and with a research group at the struggles in villages in three field trips over the course Abdul Latif Jameel Poverty Action Lab at Massachusetts of seven years allowed us to appreciate particular Institute of Technology that studied the same program but local social environments and to explore the social focused on the effect of access to clean water on health context from within (Bechky, 2003; Van Maanen, outcomes. 1979). Besides conducting interviews and collecting 2028 Academy of Management Journal December

FIGURE 1 Data Collection Timeline

Field trip

Interview, participant observation Crosscheck, expert / team discussion

Desk research

Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 archival data, we used these visits to shadow the window and analyzed how the program affected founder and senior and junior staff members in the process of transforming entrenched patterns of team meetings and in their interactions with inequality. First, we coded instances referring to villagers. program dynamics, and how villagers across social divisions participated in the program. We identified three sets of resources mobilized in the process Data Analysis of implementing the program in villages: (1) in- We analyzed the data by continually going back stitutional resources, such as local beliefs about and forth between data and emerging theoretical ar- purity and pollution; (2) social organizational re- guments (Glaser & Strauss, 1967; Strauss & Corbin, sources, such as village committees or self-help 1990). Several rounds of coding and involving dif- groups; and (3) economic resources, such as finan- ferent analysts in the process helped to spot flaws cial means, labor, or material. We then coded why, and/or personal biases. We discussed emerging cat- when, and how different types of resources were egories and theoretical themes within our research mobilized; who mobilized the resources (Gram Vikas team, with other research groups and Gram Vikas. or the villagers), and how and why behavior and in- Within our research team, we did not always agree teractions between social groups started to change. on categories and themes. These instances of di- Second, we coded instances of resistance and con- vergence helped us to sharpen our analytical focus flict emerging as a result of implementing the pro- and—most importantly—our theoretical contribu- gram, and of how conflict was resolved. This allowed tion. For example, discussions and a lack of consen- us to trace patterns of cooptation and different forms sus on the characteristics of stabilizing as a defining of sanctioning when groups or individuals chal- mechanism of scaffolding triggered an insight that lenged the implementation of the program. And, encouraged us to revisit the work of Selznick (1949) third, we coded for instances of variance in how the and resulted in the proposition of concealing as an program was implemented across villages. Although additional constitutive mechanism of scaffolding. this last aspect of our analysis was not directly Our analysis and coding strategy included two aligned with our theoretical interest—specifying major steps. In a first step, we surfaced local patterns the mechanisms underpinning the process of trans- of inequality before and after the RHEP, our focal forming entrenched patterns of inequality—it en- program. Building on the work of Beteille´ (1965, hanced the robustness of our findings. 2003, 2005), we used data across villages to code patterns of access to economic assets and in- FINDINGS frastructure, distribution of power, and attribution of status before the start of the program. Next, we coded We report our findings in the form of a narrative how such patterns changed as a result of the pro- and in two parts. In the first part, we take stock of how gram, to show that transformation is possible. In inequality manifests in patterns of behavior and in- a second step, we used RHEP as our empirical teraction in small-scale societies and showcase the 2016 TABLE 1 Summary of Our Interviews and Fieldwork

Date Location Data

2004 (Jan.) Geneva, Switzerland c first contact and interview with Gram Vikas founder Joe Madiath (1 hour) c Schwab Foundation Social Entrepreneurship Summit 2005 (Jan.) Campinas, Brazil c participant observation of founder with peers and interview c Schwab Foundation Social Entrepreneurship Summit 2006 (Jan.) Davos, Switzerland c participant observation of founder with peers and with corporate executives c Schwab Foundation Social Entrepreneurship Summit c World Economic Forum 2006 (Mar.) Oxford, UK c participant observation of founder with peers and with funders c Skoll Foundation World Forum c interview with founder (30 mins) 2007 (Jan.) Zurich, Switzerland c participant observation of founder with peers c Schwab Foundation Social Entrepreneurship Summit c interview with founder (30 mins) 2007 (Mar.) Oxford, UK c participant observation of founder with peers and with funders c Skoll Foundation World Forum c interviews with founder and staff member (30 mins each) 2007 (Mar.) Barcelona, Amsterdam, London, New York, Mexico City c interview with 3 senior partners from PricewaterhouseCoopers who had recently c telephone conference completed a two-month strategy project for Gram Vikas on site in India (1 hour) c 2007 (May) Gram Vikas Headquarters, Mohuda, India interviews with founder (4 hours, 1 hour) Seelos and Wolf, Mair, c field trip / site visit c shadowing of founder (1 day) c contact with communities implementing or having implemented water and sanitation program c interviews with 3 senior staff members (average 1 hour each) c interviews with 2 coordinator level staff members (average 1 hour each) c group interview with senior management team (1 hour) c shadowing of senior manager (2 days) c participant observation among all levels of staff (3 days) 2008 (Jan.) Zurich, Switzerland c participant observation of founder with peers c Schwab Foundation Social Entrepreneurship Summit c interview with founder (30 mins) 2008 (Mar.) Gram Vikas Headquarters, Mohuda, India c interviews with water and sanitation program senior managers (4 hours over 2 days) c field trip / site visit c shadowing of water and sanitation program senior manager (2 days) c contact with communities implementing or having implemented water and sanitation program c interviews with 3 senior staff members (different staff members to previous site visit: 30 mins, 1 hour, 1.5 hours) c interviews with 3 coordinator level staff members (conducted over 3 days; ranging from 15 mins to 1 hour) c interview with independent partner organization implementing a renewable energy project on-site c interviews with founder (over 2 days; 3 hours) c participant observation among all levels of staff including volunteers (4 days) 2008 (Mar.) Oxford, UK c participant observation of founder with peers and with funders c Skoll Foundation World Forum c interviews with founder and staff member (20 mins) 2013 (Jan.) Mohuda, India c interviews with founder (2 hours over 3 days) c field trip / site visit c interviews with several senior members of staff (5 hours over 3 days) c interviews with water and sanitation program across ranks (3 hours over 2 days) c shadowing of water and sanitation program senior manager (2 days) 2029 2030 Academy of Management Journal December

transformation of such patterns. In the second part, we document and explain the process of trans- forming entrenched patterns of inequality.

Entrenched Patterns of Inequality in Small-Scale Societies In the following narrative, we illustrate how in- equality manifests in small-scale societies before and after the implementation of the RHEP. We followed the work of Beteille´ (1965) to analyze our data according to (a) how economic assets and infra- structures are accessed by different social groups, (b) Data how power is distributed between social groups, and (c) how status and privilege are attributed to mem- bers of different social groups. Access to economic assets and infrastructure. Before the program, access to and use of infrastruc- ture related to water, sanitation, and other services was largely controlled by localelites and higher castes. A local woman described how higher castes restricted access to local water sources for lower castes: We are Dalits. That’s why they [those belonging to

contact with communities implementing or havingsanitation implemented program water and interview with independent consultant to Gramover Vikas 2 and days) volunteers (45 mins interviews with founder (2 hours overparticipant 3 observation days) among all levels ofinterview staff with including manager volunteers of (3 water days) and sanitation program interview with a member of theobservation top management (2 hours) andparticipant participant observation of founder with peersinterview and with with founder funders (30 mins) higher castes] denied us water. If we went to get water, c c c c c c c c there would be a fight. They would tell us we are untouchable. (Lower-caste woman, Video) TABLE 1 (Continued) Lack of clean water and safe sanitation was a major cause of illness, and families had to take out loans to pay for medicines and treatments. Borrowing had severe consequences for poor villagers and led to an upsurge in economic inequality at the local level (Keirns, 2007). After the program—a local woman explains—not only members of privileged groups but every house- hold in the village gained access to a piped water sup- ply and safe sanitation: We have [...] piped water supplies to all families, without exception. Every family has its own toilet and bathing room as well. (Local woman, Interview) A study of 100 villages that had implemented the program shows that the availability of clean field trip workshop organized by the Rockefeller Foundation Skoll Foundation World Forum c c c water and safe sanitation in the villages reduced incidences of waterborne diseases by up to 50%— whichalsoresultedinasignificantdecreasein illness-related indebtedness (Duflo, Greenstone, Guiteras, & Clasen, 2015). Our findings also point to changes in access to fi-

Date Location nancial resources and relationships of dependency 2014 (Nov.)2015 (Apr.) Bellagio, Italy Oxford, UK 2014 (Mar.) Dhaka, Bangladesh between social groups. Before the program, it was 2016 Mair, Wolf, and Seelos 2031 common practice for members of Scheduled Castes alter the status quo. Low levels of education among and Tribes to mortgage trees, small pieces of land, or members of Scheduled Castes and lower classes per- other private possessions to local elites in order to petuated the pattern. Certainly, women in leadership pay for medicine, funerals, or wedding dowries. positions were unimaginable. Despite government Another common social practice—inherited from programs and international development schemes, it one generation to the next and perpetuating relations was extremely difficult to bring Scheduled Castes or of dependence—was bonded labor. Because caste women into public offices. A Gram Vikas staff member, and class membership are inherited, dependent re- suggested why: lationships were reproduced, and escaping the ... vicious circle became extremely difficult. Women [Although the] government may insist that [ ] Dalits, were confined to a household with very limited or no women, and minority community members should have appropriate representation in these local bodies, access to the household wealth or rights to property. the problem [is] [...] the Tribes or the Dalits or the In the villages where the program was fully imple- women have always been considered as incapable, as mented, existing dependencies such as bonded people not required to give their personal opinion. So, labor and mortgaging diminished. Formal courts of- the control of administration was always with [...] the ten validated the changes, an important signal for upper castes, and women per se were always thought villagers who may now rely on laws to ensure their to be people of domestic involvement [...]. So, these rights. Women started to generate extra income for people are [...] not able to [...] perform their respon- the household as they were trained as masons and sibilities as anticipated. learned to operate fisheries in the local pond. A local (Senior staff member, Interview) woman attributed her improved finances to the availability of running water: After the RHEP program was implemented, access to positions of power and leadership were no longer We can finish our chores quickly and have more time exclusively tied to kinship and gender. In a report to earn extra income. Gram Vikas conducted in 2002 to document the (Local woman, Video) program’s progress, a local woman explained how Women also formed savings groups in which they patterns of interactions with men have changed: pooled their income, improving their economic We never thought we could sit on the same mat as the status. men. But things have changed now. We no longer Access to education changed too. Before the pro- draw the veil over our faces, and we can talk with men gram, a minority of boys from Scheduled Castes as equals. attended school, and few girls were literate. After (Local woman, Archival data from 2002) the program’s implementation, almost all children gained access to education, regardless of caste or Our own observations, corroborated by interview gender, according to Joe Madiath, a member of Gram data from Gram Vikas staff members, revealed that Vikas’s founding team. Increased education gave women and members of Scheduled Castes and tribes women and lower castes important skills to help had gained an active voice in village-level meetings, them access economic opportunities. These groups and their presence in meetings was considered ap- grew more independent and were able to access new propriate. Archival documents record that, in one of opportunities, such as working with government the villages that completed the RHEP program, agencies or attaining positions of power at the village a woman was voted into the regional government. level and beyond. The changes in the distribution of power between Distribution of power. Before the program, posi- social groups were closely intertwined with a change tions of power were inherited and available only to in the attribution of status to social groups—a marker men belonging to a certain caste and class. A senior of transformation we explore in the next section. staff member of Gram Vikas explained how local Status assigned to members of social groups. leaders came into office: Before the program, restricted access to economic assets and power for women and members of By birth, such fellows became the leaders. And, once Scheduled Castes or lower classes was not only taken that fellow dies, his son used to become the leader. for granted, but perceived as appropriate by all social (Senior staff member, Interview) groups—privileged and disadvantaged alike. Mem- Members of higher castes and classes who benefit- bers of Scheduled Castes were considered impure ted from these informal rules had little motivation to and polluted, and therefore physical distance was 2032 Academy of Management Journal December required. Changing caste in this life was considered To summarize, our findings show that, although an offense to a god’s will, because Scheduled Castes categories of caste, class, and gender are still in place, are believed to be in penitence for misbehavior in they are less salient in determining patterns of be- a previous life. Conformity to one’s fate is considered havior and interactions between social groups at the the only option to escape divine punishment in the local level. After the program, access to economic next life: assets and infrastructures, the distribution of power, and the attribution of status were determined not What can I do? I am born Dalit, so, in this birth, I will only by membership in a social group. Our findings be excluded. In this birth, this will be my fate. Maybe resonate with the work of Beteille´ (1965) and Douglas [in the] next birth, I’ll be different. (1986): the interlock between normative and social (Member of the Gram Vikas founding structures provides a complex system of rules, be- team recasting the voice of a member of the Scheduled Castes, Interview) liefs, and norms that make patterns of inequality durable—unquestioned, uncontested, and hard to Higher-caste members feel entitled to exclude change. Whereas anthropologists have focused their them from most aspects of social, political, and analytical attention on the local reproduction and economic life: reification of inequality, we found that patterns of inequality can be purposefully transformed—by We have a right to exclude them [...] we are the privileged. means of a development program. (Member of the Gram Vikas founding team recasting the voice of a member Scaffolding: A Process of Transforming of the village elite, Interview) Entrenched Patterns of Inequality Before the program, for the local elites, conforming In the previous section, we illustrated the outcome to norms associated with purity and pollution en- of the transformation process. Here, we show how sured privilege and signaled superiority, and, for the it unfolds. Scaffolding underpins and shapes the Scheduled Castes, it allowed for expiation and hope transformation of durable local orders that entrench for redemption in the next life. Both groups perceived patterns of behavior and interaction. Specifically, it the order as god-given fate, and the resulting exclu- is a process of transforming deeply entrenched pat- sion was neither questioned nor problematized. In- terns of interaction and behavior that account for stead, much effort went into safeguarding boundaries durable social inequality in small-scale societies. between social groups to protect the purity of the Our findings suggest three generative mechanisms of higher castes and confine pollution to the Scheduled scaffolding: mobilizing, stabilizing, and concealing. Castes or Tribes. First, mobilizing resourcesinvolvescreating,un- After the program, the status attributed to members locking, and repurposing new and existing resources of various social categories changed. Video data from from both the organization and the villagers. Three sets our archival database corroborates this: a local man, of resources were mobilized in the process we analyzed: asked if a member of a Scheduled Caste or Tribe was (1) institutional resources—formal and informal rules working with him on the community hall, replied: directly related to the water and sanitation program, as well as local norms and beliefs about purity and pol- I don’t know, we don’t discuss these matters anymore. lution; (2) social organizational resources—formal or (Local man, Video) informal governance structures organizing the in- An elderly woman explained that the importance teractions within or between groups in the villages; and of caste membership in daily life has changed, par- (3) economic resources—material or economic means ticularly in interactions among younger people: accessed from within the villages or from outside. Second, scaffolding hinges on stabilizing new For ages, we’ve practiced untouchability. [...] Now- patterns of interaction that reflect an alternative so- adays, the boys say “what untouchability”! cial order. Stabilizing enables the institutionaliza- (Local woman, Video) tion of new ways of thinking, acting, and relating. During our fieldwork, we observed members of It includes formal and informal enforcement and different castes sitting next to each other, sharing sanctioning mechanisms that are built to protect the meals, working together doing the same type of work, new shared goal. Stabilizing is critical in a trans- conversing together at village meetings, and working formation that revises a social order rather than together to build community libraries. replacing it. 2016 Mair, Wolf, and Seelos 2033

Third, scaffolding involves concealing goals that Gram Vikas’s tactics included demonstrating how are neither anticipated nor desired by the members feces dilute in water used for drinking and washing. of some social groups. Concealing requires focusing Staff members organized “walks of shame”—asking on the desired and uncontested manifest goal rather local elites to show the defecation spots around the — than the covert goal of revising the social order. village to persuade local leaders to include marginal- Concealing helps generate consensus across social ized social groups in the water and sanitation program. groups, enables collective purposive action, and Asaresultofthissortofargument,localelitesac- tively promote and support the process: they leverage prevents cooptation efforts by specific groups. their position of power and privilege to influence, We document the process in a narrative that further motivate, and coerce other villagers to commit to the illustrates the rationalization, routinization, and so- program. As Chitra Choudhury, a member of the lidifying of new patterns of interaction and behavior. management team, told us, established hierarchies Coopting power structures, norms, and beliefs: and traditional enforcement mechanisms start to work Rationalizing interaction. When approaching vil- in favor of implementing the program because the lages to join the RHEP, Gram Vikas adheres to local voice of the elites is heard and accepted by the village. rules by first establishing contact with village leaders Gram Vikas continued to insist that all social groups and local elites. Although the elites perceive sanita- in the villages—not only elites—must actively endorse — tion and running water as desirable for themselves and participate in the program. Members of Gram the pure, clean, and superior people—they do not Vikas repeatedly pointed out that starting the pro- support involving the Scheduled Castes and Tribes— gram without full commitment from all social groups the dirty, polluted, and inferior ones. From this in the community backfires later on: perspective, using the same water source with the ’ Wherever we have not got a consensus and we have Scheduled Caste people would put the higher castes tried to cut short those methodologies, we have suf- superiority and purity at risk, as relayed here: fered for it. (Senior staff member, Interview) [The higher-caste women] get water by the turn of the tap, and the Scheduled Caste women also by the turn Reaching consensus requires intensive effort and of the tap will get water. That means there is no dif- time. From various archival documents and our in- ference between us, and we are supposed to be higher terviews, we learned that, in the early years of the ’ and they re supposed to be lower. program, Gram Vikas tried to implement the program (Senior staff member, Interview) in a village close to its headquarters. To speed up the In an effort to gain the local leaders’ support, Gram process, Gram Vikas did not insist on the 100% in- Vikas coopts institutional resources—local norms, clusion rule. The RHEP could not be completed in values, and beliefs about purity and pollution—to that village. More importantly, interactions between promote the program. Joe Madiath, a co-founder, members of different social groups could not be recalled how he approached village leaders: consolidated, and existing patterns of entrenched inequality could not be transformed. As Sojan You want to drink Dalit shit? Now, you are drinking Thomas, a senior member of the Gram Vikas team, everybody’s shit so, by not [including them in the observed, once 100% commitment is reached, it is rare program], you are only drinking their waste. for villages to halt implementation or discontinue the (Member of Gram Vikas founding program, and thus halt the transformation process. team, Interview) Even if only a single household does not commit to Madiath recalled a pitch to persuade local elites the program, Gram Vikas will not officially start the that involving 100% of village households, in- water and sanitation program in that village: cluding the Scheduled Castes, was instrumental for We do not take a “no” from anyone [the privileged and retaining higher caste purity and thus for safe- the underprivileged]. We do not accept positions like guarding the existing social order: “we do not want them included” or “we do not want to be included.” Even if 99% is there, but one family shits all over the (Senior staff member, Interview) place, that family can pollute the water, the sur- rounding, everything of the village. Once consensus to implement the program has (Member of Gram Vikas founding been reached among all social groups in the village, team, Interview) Gram Vikas launches a formal contract that binds 2034 Academy of Management Journal December each household to the prescribed rules of the health and children’s diseases. As women gain con- program. The contract entails multiple rules to en- fidence, Gram Vikas encourages them to form “Self- force a collective effort and thus is an additional Help Groups” that meet in public. Meetings of these mechanism to rationalize new ways of interacting. groups are safe spaces (Mair et al., 2012) where women The contract establishes the villagers’ commitment can raise and discuss issues or problems of a contested at a stage during which they do not recognize nor nature, such as saving money, alcoholism, and do- anticipate the full extent of the transformation. mestic violence. Meanwhile, their husbands are fo- In sum, Gram Vikas coopts local norms and be- cusing on the requirements for implementing the liefs about purity and pollution and local power water and sanitation program. Self-Help Groups are structures to turn water and sanitation into an un- a valuable social organizational resource, as they sup- contested but confined space for breaking taken- port women in learning to be more vocal about their for-granted patterns of interaction. Clean water and concerns and gain self-confidence (Sanyal, 2009). sanitation as manifest goals of the program unite all As women begin to express their needs and con- members of the village, and interaction between so- cerns outside the confines of the household, Gram cial groups is perceived as critical to protect the pu- Vikas encourages them to create a new social orga- rity of powerful elites. Elite pressure and the fear of nizational resource in the form of an all-female Vil- being punished in the next life are enforcement and lage Body, which formally and officially represents sanctioning mechanisms in service of the program. A the interests of women in the village. It is composed formal contract between Gram Vikas and the village of the female heads of all households independent of is an additional stabilizing mechanism to enforce changes in patterns of interaction and behavior. As caste or class. The Village Body gives women a plat- a result, the local population rationalizes interac- form to raise and discuss issues that concern the tions across social divides on the basis of the same entire village. Examples include village-level access local normative order that created and reproduced to health services to reduce child mortality and re- categories of exclusion in the first place. strictions of local alcohol sales to combat alcoholism Redesigning local governance structures: Rou- in the village. In parallel to the all-female Village tinizing interaction. Contractually enforced pro- Body, Gram Vikas initiates another Village Body gram rules require that the villages establish formal consisting of the male heads of all households. Al- governance structures involving all social groups, though the men—both lower and higher caste—are notwithstanding gender, class, or caste. Our findings used to voicing their concerns within their own so- point toward a number of challenges and also re- cial groups, for many, it is the first time they interact sistance to this rule. For example, women have never and consider opinions across groups. Gram Vikas been permitted to engage in decision making at the actively moderates the interchange, nudging mem- village level, and their exclusion has become taken bers of Scheduled Tribes and Scheduled Castes to for granted. A female member of Gram Vikas’s found- voice their concerns, and higher castes and members ing team recalled the response of a local woman: of local elites to listen. Men willingly comply with All the centuries our men decided everything—so this requirement in light of the prospect of obtaining why should we decide anything? access to running water and sanitation. (Member of the Gram Vikas founding As the two committees begin to develop routines team, Interview) for meeting and decision making, Gram Vikas insists that the gender-segregated groups merge into one Husbands often oppose the idea of women leaving Village General Body. In most villages, the merger the house to join village meetings. We observed in our idea meets with skepticism and resistance. Never- fieldwork that, even if women do join, they often ad- theless, men and local elites agree to compromise here to traditional ways of interaction: men and women within the uncontested domain of water and sanita- won’t sit together in a meeting. Several members of the tion and the confines of the program. The following Gram Vikas team corroborated our observation. quote, retrieved from archival data in the form of an To facilitate change in how women behave and interact with others, Gram Vikas initially adheres to internal report, illustrates this point: local norms and hierarchies. As women are not tra- Gram Vikas says we must involve women to make ditionally allowed to move freely, team members decisions, but this is just for [the] RHEP, not for other approach women within the confines of their homes activities. to discuss “appropriate” topics, such as maternal (Local man, Archival data from 2003) 2016 Mair, Wolf, and Seelos 2035

Gram Vikas does not correct these views. In fact, the Committee collects a corpus fund to which every organization conceals the extent of the transformation household contributes, regardless of class and caste. unfolding. Once the Village General Body operates Through the course of implementing the RHEP, with all social groups represented and equally in- members of privileged groups become particularly volved in discussions and decision making, Gram interested in gaining access to piped water and safe Vikas prompts and oversees the election of a Vil- sanitation to avoid being polluted by Scheduled lage Executive Committee. The Village Executive Caste feces. Therefore, to prevent delays caused by Committee—four women and four men—is elected the limited financial means of the marginalized, by the Village General Body, and it stays in power elites frequently subsidize poorer families, covering for three years. It gets registered as a legal entity, their contributions. assumes formal decision-making power over all topics related to water and sanitation, and takes Because there is nobody in that village who can be responsibility for adherence to the program rules excluded, the poorer villagers are now subsidized. within the village. The Executive Committee also (Senior staff member, Interview) interacts with outside agencies, such as govern- In addition, Gram Vikas motivates and supports ment agencies or public administration. the Village Executive Committee to unlock addi- As interaction between members of different tional financial resources, such as subsidies, avail- castes, classes, and genders became a daily routine, able in the form of development schemes at the state ways of dealing with group boundaries in daily level. Mobilizing material and economic resources is routines also started to change. For instance, before a distributed effort: village households across social — — the program, a purification ritual bathing was groups raise the corpus fund, Gram Vikas provides prescribed each time a higher-caste member inter- a financial top-up, village households provide basic acted with a Scheduled Caste member. In the archival materials for construction, Gram Vikas provides ’ video material, we analyzed a higher-caste woman s masonry training, village households build basic explanation of how this ritual was altered. Wearing brickwork, Gram Vikas provides the material for a good saree is now enough to prevent being polluted roofing, and so on. This process continues until the by the lower castes: water tank is constructed. The tank is visible from We wear a good saree—a synthetic saree. [...]We afar and symbolizes the villagers’ collective effort. don’t bathe on return if we wear a good saree. The collective mobilization of economic resources (Higher-caste woman, Video) forces all social groups within the village to work together, to gain confidence and solidify the routine Both the villagers and Gram Vikas mobilize and create social organizational resources. As all members of interacting across social groups. Our study, how- of the village—both privileged and marginalized— ever, also exposed a high risk to program success: if keep their eyes on the goal of gaining access to water one group or a single household resists participation, and sanitation, taken-for-granted patterns of behavior both construction and the covert transformation of and interaction are compromised. The behavior of the social order could stagnate or collapse. In many social groups and interaction between them start to villages, the self-esteem of members of Scheduled change. The local governance structures that are cre- Castes has grown by this time, and power games and ated, including formal rules established by the con- conflicts begin to surface between groups. Madiath tractual agreement and informal rules such as the told us that formerly marginalized groups, in par- need to safeguard purity of the elites, further enhance ticular, start abusing their new power to block higher a shared commitment to the program. These efforts castes’ access to infrastructure: stabilize the ongoing transformation of patterns of They agree [...] and then, as the work begins, they say, behavior and interaction between men and women “We will not even contribute physical labor.” So the and members of different castes and classes, and thus rest of the village has to bring the materials for those allow the villagers to build routines that involve people and even provide the labor for them in order to interacting across social divides. complete. So this [is] a revenge mentality—[the mar- “Constructing together”: Solidifying interaction. ginalized groups realize] that this is one time when Once cross-group divides are bridged by new in- they can blackmail these people: “These people have teraction routines, Gram Vikas initiates the mobili- blackmailed us all throughout or exploited us all zation of economic resources to build the water and through life.” sanitation infrastructure. The Village Executive (Member of founding team, Interview) 2036 Academy of Management Journal December

Our findings indicate that, in such critical sit- infrastructure. Our interviews and archival data in- uations, village members sanction deviance and dicated various stabilizing mechanisms that enable enforce compliance to the process, with the result scaffolding and a revised social order to persist after that emergent changes in patterns of behavior and Gram Vikas withdraws. A multilayered net of gover- interaction are not reversed. According to archival nance structures establishes legal and formal mecha- film material provided by Gram Vikas, in one village, nisms that ensure transparency, accountability, and for instance: the representation and participation of all village households. The interest earned from the corpus fund [O]ne person [a man] tried to stall the work. He re- ensures financial viability and substitutes for the fused to participate. The village collectively made subsidies that Gram Vikas has provided or made ac- a decision to excommunicate him from village life. He cessible as part of implementing the RHEP. was banned from meetings and feasts, denied all ser- As a result, the “100% inclusion” rule is main- vices, could not even arrange the marriage of his ... tained even after the manifest goal—access to water children. Finally, [ ] he came around, requested — a compromise. and sanitation has been achieved. Newly formed (Video) households and families moving to the village are automatically represented in the Village General The Village Executive Committees frequently es- Body, and members of disadvantaged groups are tablish their own sanctions, not prescribed by the actively integrated into various committees and Self- formal contract, to ensure each household’scom- Help Groups. Although new households are required pliance. Such sanctions include cutting attendance to build their own water and sanitation infrastruc- at weddings or other festivities, public shaming, and ture, they can use the interest earned from the corpus denial of access to water. In a report Gram Vikas fund to partly finance this effort. Savings and newly developed in the early days of the program, a local acquired skills enable women, men, and lower-caste woman elaborated as follows on options to ensure members to engage productively in a broader set of compliance from men: economic activities. Economic opportunities are more abundant because of the legal status of the Village If the men do not co-operate in the implementation of [the] RHEP in our village, we will employ laborers Executive Committee, which facilitates access to from outside for the construction of [the] RHEP in- government funds and development programs for frastructure and will go for an indefinite kitchen strike infrastructure projects. Such projects include build- to ensure their co-operation. ing local fishponds, asphalt roads, community cen- (Local woman, Archival data from 2003) ters, schools, health centers, and improved houses. These projects are collective efforts, and the infra- Women assume a particularly important role in structure is made accessible to all villagers regardless policing the implementation of the program and, of gender, caste, and class. although it is not the center of their attention, the Summary of scaffolding. Through our analysis of transformative process. One village we visited was the RHEP program, we identified the process of ’ particularly affected by alcoholism. Men s liquor scaffolding, which transforms deeply entrenched consumption dried up the already-scarce financial patterns that account for durable social inequality in resources of families and hampered efforts to raise small-scale societies. We identified three mecha- the corpus fund. In this village, women—organized nisms of scaffolding: (1) mobilizing institutional, in Self-Help Groups—collectively decided to take social organizational, and economic resources; (2) action. They expelled the liquor vendor from the stabilizing emerging patterns of interactions that re- village and called a village meeting to declare that flect an alternative social order; and (3) concealing liquor consumption was prohibited in the village goals that are neither anticipated nor desired by and penal action would be taken against those who some groups. These interrelated mechanisms make sell or consume alcohol. As this and other similar the transformation processes adaptive and emerging instances show, informal and formal mechanisms to alternative social orders robust. enforce rules and impose sanctions are important We found that scaffolding requires active planning anchors to implement the program, and also—more and cannot rely on self-organizing. First, the program important for our theoretical interest—to stabilize we analyzed here is the result of many years of the emerging new social order. experimenting, failing, and learning (for more de- The RHEP program officially ends with the con- tails, see Seelos & Mair, 2016). It builds on the ac- struction of a functioning water and sanitation cumulated knowledge of an external actor—an 2016 Mair, Wolf, and Seelos 2037 organization the purpose of which is to alter deeply establish functional governance structures in which entrenched patterns of inequality. Second, whereas all social groups were represented. In Kholo Saman- scaffolding is instigated and guarded by the organi- tapur, one of the caste villages, for instance, it took 162 zation, villages need to build the first platform of the village meetings to reach initial consensus about ad- scaffold and help lay the foundation of the new social hering to the explicit rules of the program. In Bimala order—forming a consensus on the 100% inclusion Jani, another caste village, it took ten years until the rule formalized in a contract. And, third, the involve- process was completed. ment of the external actor is temporary; its withdrawal Tribal villages—predominantly populated by needs to be carefully planned so that the new norma- members of Scheduled Tribes with less heterogene- tive and social structure cannot be detached. Rather, ity in social classes and castes—struggled much the process builds a strong foundation to support more to raise the corpus fund. Two factors might ongoing transformation and development efforts. account for their difficulty. First, local economic Figure 2 provides a visual overview of the process elites do not typically reside in the tribal villages and we uncovered. therefore did not subsidize or support financing for Scaffolding: Robustness and adaptation across tribal villagers; and, second, because of their geo- villages. The main objective of our analysis was to graphic isolation, villagers are less accustomed to specify the process of transformation. Scaffolding engaging with outside actors such as government was remarkably robust across villages in our study. agencies. In such cases, Gram Vikas put in extra effort But villages did not “undergo” scaffolding at the to teach the villagers how to access external funds same pace. Although it was not the main focus of our and mobilize government support. analysis, our results point to heterogeneity of social Specific social organizational resources such as groups within villages as a main driver of variation Self-Help Groups played an important role in en- across villages. suring a robust scaffolding process. Our preliminary For example, caste villages and tribal villages dif- analysis showed that scaffolding is adjusted to ac- fered in the time required for changing patterns of count for variance in social structures and available behavior and interaction. Caste villages—deeply resource types within small-scale societies. For segregated into multiple castes and classes—took example, the specific types of norms or persuasion significantly longer to reach consensus and to tactics that were used varied according to the

FIGURE 2 Scaffolding: A Process of Transforming Patterns of Inequality

Villagers

Persistent New patterns of patterns of Rationalizing new Routinizing new Solidifying new interaction and interaction and interactions interactions interactions behavior behavior Coopting local Redesigning local Constructing breaking with reinforcing norms and power governance together social divides social divides structures structures

Gram Vikas Mobilizing Unlocking, creating, and repurposing institutional, social organizational, and economic resources

Stabilizing Enforcing and sanctioning through formal and informal governance structures

Concealing Drawing and focusing attention on a desired and uncontested goal 2038 Academy of Management Journal December sensitivities of local leaders and the marginalized Contribution to the Organizational Study of members of the village. The issues that Self-Help Societal Challenges Groups dealt with were adapted to the specific needs The main theoretical objective of this paper was to and problems in the communities, and economic advance organizational study of societal challenges. resources were mobilized in accordance with avail- Our findings on scaffolding establish a first but ability and intra-village dynamics. Additional re- important link between the study of organizational search is needed to examine variance across villages efforts to alleviate social problems and the trans- and identify sources of variation as well as enabling formation of social systems. The concept and the and disabling conditions more systematically. three mechanisms of scaffolding—concealing, sta- bilizing, and mobilizing—build on classical and DISCUSSION contemporary work on purposive organizational activity that affect society and extend it in impor- In this paper, we studied how an organization tant ways. confronts what mathematician David Hilbert dubbed First, we revive a theoretical and analytical tradi- a “grand challenge.” We examined how a grand tion that deeply engages with organizations and ex- challenge such as social inequality manifests in amines how their activities affect the lives of people small-scale societies and inheres in patterns of be- and communities (Gouldner, 1954; Selznick, 1949). havior and social interaction. Local patterns of in- The main objective of these studies was to reveal the equality constitute local realities—they are visible hidden social forces in a system—the unintended and provide comfort to those who marginalize and consequences of organizational activity (Selznick, those marginalized. Alternative realities or social 1949). Instead, our findings on concealing show that orders are unthinkable and not wanted; therefore, a program sponsored by an organization can be used programs confronting inequality are vulnerable, as a tool to organize these hidden forces. We show contested, and likely to fail. Yet, transforming dura- that purposive action can entail pursuing both ble patterns of inequality is possible—in particular, manifest and concealed goals. Scaffolding allows our findings show that transformation of such pat- organizations to hide consequences in pursuit of terns is possible through scaffolding. a concealed goal—transforming social orders and As a metaphor for building temporary support entrenched patterns of inequality—and concur- structures, scaffolding has been used by cognitive rently reveal consequences aligned with a manifest scientists to study individual transformation pro- goal: access to clean water and safe sanitation. cesses (Vygotski, 1964) and by evolutionary bi- Our findings allow us to turn attention to the “the ologists to understand the evolution of complex limiting function of end-in-view”—aconcept adaptive systems (Davidson & Erwin, 2006). In this Selznick (1949) identified as significant, but dis- study, we demonstrate the relevance of scaffolding regarded as sociologically irrelevant in explaining for research on organizations and institutions. We go purposive organizational action.6 According to beyond using scaffolding as a metaphor (Orlikowski, Selznick (1949: 255), “The very necessity to keep 2006) and specify it as a process that enables and your eye on the ball—which demands the con- organizes institutional transformation. Scaffolding struction of a rational system explicitly relating involves mobilizing, stabilizing, and concealing; means and ends—will restrain the actor from three generative mechanisms to unlock entrenched taking account of those consequences which in- normative and social structures and, in parallel, to directly shape the means and ends of policy.” institutionalize emerging alternative social orders. In our case, scaffolding enabled Gram Vikas to Although it is temporary, scaffolding blends in and proactively manage villages to keep their eyes on the becomes integral to new normative and social water-and-sanitation “ball.” Scaffolding helped to structures. In other words, scaffolding is an under- provide scope for what people can and will do, and recognized process for building and strengthening to hide the emergence of a new social order that alternative local social orders. was neither desired nor anticipated by some social In the next sections, we first elaborate how our groups. Concealing was important because members findings contribute to existing organizational schol- arship on societal challenges and to practical en- deavors to tackle entrenched patterns of inequality. 6 Selznick (1949) identified two sources of unintended We then conclude with ideas for future research on consequences and considered this one as sociologically grand challenges. insignificant compared to commitment. 2016 Mair, Wolf, and Seelos 2039 of specific social groups would resist and potentially transformation efforts and those who inhabit the compromise the process of transformation. In the institutional arrangements to be transformed (Hallett & case reported here, the formulation of an end-in- Ventresca, 2006; Mair et al., 2012). view—probably surprising to Selznick—was a com- In addition, our findings reemphasize the impor- mitment structure in itself, and it was an important tance of mobilizing resources for organizational ef- organizing device for the organization sponsoring forts to affect institutional and societal change the program. (Battilana & Leca, 2009; Mair & Mart´ı, 2007; Ganz, Second, our findings complement contemporary 2000; Seelos & Mair, 2007). Our study complements research that has identified pursuing multiple goals this tradition and specifies that mobilizing resources as a defining feature of organizations that effect is intertwined with mechanisms such as stabilizing positive social change.7 Standing on the shoulders of and concealing in the process of transformation. Our Selznick, this literature has emphasized conflict, findings also show how collective mobilizing and the tension, and challenges involved in pursuing mul- interplay between organization and village distributes tiple organizational goals. Such goals are typically agency in the transformation process, a critical feature associated with distinct and competing institutional of scaffolding. logics—a commercial logic and a development logic As we mentioned before, our analysis focused on (Battilana & Dorado, 2010; Ebrahim et al., 2014). the process of transformation. Future research based Similar to Selznick’s, these authors’ main interest on designs that allow for specific attention to vari- lies in understanding how conflict affects or trans- ance across small-scale societies will help to sys- forms organizations (Besharov & Smith, 2014; Mair tematically unpack how resources are combined and et al., 2015) or the organization’s environment un- configured in the process of scaffolding. This will derstood as organizational fields (Lawrence et al., connect to yet another set of literature, social move- 2002; Mair & Hehenberger, 2014). Gram Vikas also ment theory, which argues that access to re- pursues multiple goals. But these goals do not reflect sources determines the likelihood of collective “institutional conflicts.” Instead, they refer to action, and hence dynamics and outcomes of a specified (uncontested) social problem nested in/ transformation processes within societies (Edwards & related to deeper relational and institutional (con- McCarthy, 2004; Schneiberg & Lounsbury, 2017). tested) challenges. The concept of scaffolding put forward in this Contribution to Efforts to Alleviate Inequality paper is a first attempt to connect the study of goals related to a social problem and the study of goals In line with showcasing the relevance of organi- related to transforming a local social system (Stern zational studies, the second main objective of this & Barley, 1996). Doing so allows us to expand paper was to develop portable insights for orga- Selznick’s ideas on institutionalization as a process nizational efforts to tackle inequality. Scaffolding of strengthening, stabilizing, and infusing organiza- has the potential to transform prevailing condi- tions with value (Selznick, 1957). We show how tions of the social, political, and economic lives of scaffolding plays an important role in transforming people and thereby alter patterns of thought and existing social orders and strengthening and stabi- behavior. lizing alternative ones at the level of small-scale so- Well-intended programs, especially in develop- cieties. Our insights complement existing research ing countries, often target specific dimensions of on institutional change that predominantly studies inequality—for example, seeking to decrease the attempts to directly challenge or replace existing number of households below the poverty line by institutional orders (Greenwood, Suddaby, & Hinings, fostering entrepreneurship, or working to decrease 2002; Hargrave & Van de Ven, 2006). Instead, scaf- the number of children not going to school by building folding refers to an ongoing and gradual process of a school system parallel to the public school system. transforming local orderings of reality. Our findings These efforts might enable access to opportunities on scaffolding provide a more granular view and for marginalized groups. Yet, the effect of such analysis of the interplay between those who prompt programs is often elusive, temporary, and revers- ible. Programs are not designed to monitor the 7 More recently, Markman, Russo, Lumpkin, Jennings, emergence of hidden social forces leading to un- and Mair (2016) have extended this perspective to the en- intended consequences or to deal effectively with trepreneurship literature and have portrayed entrepre- them. For example, elites might not respect new neurship as a platform for pursuing multiple goals. rules, or, as research on microfinance has shown, 2040 Academy of Management Journal December benefits for women of a community may be appro- “patterns [of inequality] can never be observed by priated by their husbands (Goetz & Gupta, 1996). We measurement” and “change from one pattern to an- argue that assisted processes of transformation such other is even less observable” (Barth, 1967: 662). Our as scaffolding create more robust and more enduring objective was not to capture change in the patterns of results. Scaffolding creates a guarded and supported inequality in real time and in situ. Instead, our con- space where manifest and concealed goals can be tinuous engagement with the organization and expo- pursued in parallel until renewed or altered local sure to villages over ten years allowed us to capture orders emerge. instances of transformation across villages, and Our findings also have implications for evaluating allowed us to better understand how an organi- the success of organizational efforts to alleviate in- zation can intervene and confront entrenched equality. The progress each village makes in trans- patterns of inequality in institutionally complex forming local realities so that marginalized groups settings. are included in many aspects of social, economic, We studied local experimentation without losing and political life is an important marker of success. sight of the grand challenge. The characteristics of Yet, evaluating ongoing progress is not included in grand challenges are often thought of as being scale the standard repertoire of development practice. dependent (Ferraro, Etzion, & Gehman, 2015). In this Evaluations of programs such as the one we report on study, we draw attention to characteristics of grand in this study typically assess effectiveness in allevi- challenges as they manifest locally. We propose that ating a specific problem that is perceived as impor- the important characteristics of inequality are scale tant by those affected by it and those seeking to independent. Inequality remains multidimensional, solve it. A study on the RHEP—the program we complex, and interlinked even in locally bound set- studied—by economists from the Abdul Latif tings. Yet, scale-independent grand challenges at the Jameel Poverty Action Lab at Massachusetts Institute level of small-scale societies—local villages—enable of Technology, for example, assessed how the pro- experimentation and learning. gram affects levels of diarrhea and malaria in villages This study centered on a specific and locally (Duflo et al. 2015). No doubt, positive health outcomes bound societal challenge—inequality entrenched are an important indicator for progress. But, as our in durable patterns of interaction and behavior in study shows, health problems are deeply rooted in rural villages in India. Which other problems could entrenched patterns of interaction and behavior pre- scaffolding shed light on? Societal challenges scribed by a local system of rules, norms, and beliefs. rooted in normative and structural interlocks Our study shows that social problems and change abound in the developing and developed worlds. in social systems can be addressed conjointly. Or- Scaffolding provides a theoretical and practical ganizational research can complement efforts by lens with which to reflect on important societal economists in important ways. Confronting and al- challenges and the ways we address them. The leviating stubborn social problems requires taking following two examples illustrate the potential of local realities seriously to diagnose the root cause of scaffolding to make progress on established as well the problem but also to include local factors such as as new social problems. place and local governance structures for under- First, a well-documented example represents standing processes that are conducive to trans- microfinance, in which a specified problem—women formation and positive social change (Lawrence & or marginalized groups not able to access finance—is Dover, 2015; Mair et al., 2012; York, Hargrave, & addressed by a technical solution: microloans. Such Pacheco, 2016). solutions often overlook the relational aspects of the problem rooted in normative and social structures prevalent in both small-scale and large-scale societies Extending Research on Organizations and (Mart´ı &Mair,2009). Grand Challenges More recent and unexpected societal challenges, Research on how to confront grand challenges al- such as the influx of refugees in Europe, can be lows us to recover the ambition to combine rigor and assessed through a scaffolding lens too. Most well- relevance in organizational research (Lawrence, 1992) intended efforts prioritize the alleviation of a specific and, at the same time, to engage with knowledge de- problem; for instance, providing language courses or veloped in other disciplines (Howard-Grenville et al., job training to support the integration of refugees. 2014). In this study, we built on literature from social Our insights on scaffolding suggest that programs anthropology. Social anthropologists recognize that might also involve various groups within the host 2016 Mair, Wolf, and Seelos 2041 societies in programs. They should identify mani- Besharov, M. L. 2014. The relational ecology of identifi- fest goals to keep the eye on the ball while adapting cation: How organizational identification emerges normative and social structures necessary for in- when individuals hold divergent values. Academy of – tegration to work. Maintaining focus on a technical Management Journal, 57: 1485 1512. problem is far easier, but ultimately success will Besharov, M. L., & Smith, W. K. 2014. Multiple in- depend on facing relational challenges defined by stitutional logics in organizations: Explaining their normative and power struggles. Exploring scaf- varied nature and implications. Academy of Man- – folding across a diverse set of societal challenges agement Review, 39: 364 381. will help to further clarify the boundary conditions Beteille,´ A. 1965. Caste, class and power: Changing pat- of scaffolding. terns of stratification in a Tanjore village. Berkeley, We hope that our detailed specification of scaf- CA: University of California Press. folding as a productive tool with which to organize Beteille,´ A. 2003. Equality and universality: Essays in transformation will serve as a template for re- social and political theory. Oxford, England: Oxford searchers and practitioners to expand our under- University Press. standing of this process and the potential and limits Beteille,´ A. 2005. Anti-Utopia: Essential writings of of its generalizability. AndreB´ eteille´ . New Delhi, India: Oxford India Paperbacks. Cernea, M. M. 1985. Putting people first: Sociological REFERENCES variables in rural development. New York, NY: Agrawal, A. 2001. Common property institutions and Oxford University Press. sustainable governance of resources. World Devel- Chaudhuri, S., & Ravallion, M. 2006. Partially awakened opment, 29: 1649–1672. giants: Uneven growth in China and India. Wash- Amis, J. M., Munir, K. A., & Mair, J. 2017. Institutions and ington, D.C.: World Bank. economic inequality. In R. Greenwood, C. Oliver, Cleaver, F. 1999. Paradoxes of participation: Questioning T. Lawrence & R. Meyer (Eds.), The SAGE handbook participatory approaches to development. Journal of of organizational institutionalism (2nd ed.). Thou- International Development, 11: 597–612. sand Oaks, CA: Sage. Colquitt, J. A., & George, G. 2011. Publishing in AMJ: Topic Banerjee, A., Duflo, E., Ghatak, M., & Lafortune, J. 2013. choice. Academy of Management Journal,54: Marry for what? Caste and mate selection in modern 432–435. India. American Economic Journal. Microeconom- Crawford, S. E. S., & Ostrom, E. 1995. A grammar of in- – ics,5:33 72. stitutions. In E. Ostrom (Ed.), Understanding institu- Banerjee, A., Duflo, E., Goldberg, N., Karlan, D., Osei, R., tional diversity: 137–174. Princeton, NJ: Princeton Pariente,́ W., Shapiro, J., Thuysbaert, B., & Udry, C. University Press. 2015. A multifaceted program causes lasting progress Davidson, E. H., & Erwin, D. H. 2006. Gene regulatory for the very poor: Evidence from six countries. Sci- networks and the evolution of animal body plans. ence, 348: 772–788. Science, 311: 796–800. Barth, F. 1967. On the study of social change. American De Haan, A. 2004. Disparities within India’s poorest re- Anthropologist, 69: 661–669. gions: Why do the same institutions work differently Battilana, J., & Dorado, S. 2010. Building sustainable hy- in different places? Washington, D.C.: World Bank. brid organizations: The case of commercial micro- Douglas, M. 1966. Purity and danger: An analysis of the finance organizations. Academy of Management concept of pollution and taboo. New York, NY: Journal, 53: 1419–1440. Routledge. Battilana, J., & Leca, B. 2009. The role of resources in in- Douglas, M. 1986. How institutions think. New York, NY: stitutional entrepreneurship: Insights for an approach Syracuse University Press. to strategic management that combines agency and Duflo, E., Greenstone, M., Guiteras, R., & Clasen, T. 2015. institution. In L. A. Costanzo & R. B. MacKay (Eds.), Toilets can work: Short and medium run health Handbook of research on strategy and foresight: impacts of addressing complementarities and ex- 260–274. Cheltenham, England: Edward Kluwer. ternalities in water and sanitation (NBER working Bechky, B. A. 2003. Sharing meaning across occupational paper series, no. 21521). Cambridge, MA: National communities: The transformation of understanding Bureau of Economic Research. on a production floor. Organization Science, 14: Dumont, L. 1980. Homo hierarchicus. New Delhi, India: 312–330. Oxford University Press. 2042 Academy of Management Journal December

Ebrahim, A., Battilana, J., & Mair, J. 2014. The governance of Jimenez, E., & Paqueo, V. 1996. Do local contributions af- social enterprises: Mission drift and accountability chal- fect the efficiency of public primary schools? Eco- lenges in hybrid organizations. Research in Organiza- nomics of Education Review, 15: 377–386. – tional Behavior, 34: 81 100. Katz, T., & Sara, J. 1997. Making rural water supply sus- Edwards, B., & McCarthy, J. D. 2004. Resources and social tainable: Recommendations from a global study. movement mobilization. In D. A. Snow, S. A. Soule & Washington, D.C.: World Bank. H. Kriesi (Eds.), The Blackwell companion to social Keirns, P. 2007. Water supply and sanitation services for – movements: 116 152. Oxford, England: Blackwell the rural poor: The Gram Vikas experience. Rugby, Publishing. England: Practical Action Publishing. Ferraro, F., Etzion, D., & Gehman, J. 2015. Tackling grand Lawrence, P. R. 1992. The challenge of problem-oriented challenges pragmatically: Robust action revisited. research. Journal of Management Inquiry,1:139–142. Organization Studies, 36: 363–390. Lawrence, T. B., & Dover, G. 2015. Place and institutional Ganz, M. 2000. Resources and resourcefulness: Strategic work creating housing for the hard-to-house. Admin- capacity in the unionization of California agriculture, istrative Science Quarterly, 60: 371–410. 1959-1966. American Journal of Sociology, 105: 1003–1062. Lawrence, T. B., Hardy, C., & Phillips, N. 2002. Institu- tional effects of interorganizational collaboration: The Geertz, C. 1988. Works and lives: The anthropologist as emergence of proto-institutions. Academy of Man- author. Stanford, CA: Stanford University Press. agement Journal, 45: 281–290. Glaser, B. G., & Strauss, A. L. 1967. The discovery of Mair, J., & Hehenberger, L. 2014. Front-stage and back- grounded theory. New York, NY: Aldine. stage convening: The transition from opposition to Goetz, A. M., & Gupta, R. S. 1996. Who takes the credit? mutualistic coexistence in organizational philan- Gender, power, and control over loan use in rural thropy. Academy of Management Journal,57: credit programmes in Bangladesh. World Develop- 1174–1200. – ment, 24: 45 63. Mair, J., & Mart´ı, I. 2007. Entrepreneurship for social im- Gouldner, A. W. 1954. Patterns of industrial bureau- pact: Encouraging market access in rural Bangladesh. cracy. Glencoe, IL: Free Press. Corporate Governance, 7: 493–501. Government of Odisha. 2011. Annual report 2010–2011. Mair, J., Marti, I., & Ventresca, M. J. 2012. Building in- Orissa, India: ST & SC Development, Minorities & clusive markets in rural Bangladesh: How intermedi- Backward Classes Welfare Department. aries work institutional voids. Academy of Management Journal, 554: 819–850. Gram Vikas. 2015. Annual report (2014–2015). Bhuba- neswar, Odisha, India: Gram Vikas. Mair, J., Mayer, J., & Lutz, E. 2015. Navigating in- stitutional plurality: Organizational governance in Greenwood, D. J., & Levin, M. 2005. Reform of the social sci- hybrid organizations. Organization Studies,36: ences and of universities through action research. In 713–739. N. K. Denzin & Y. S. Lincoln (Eds.), The Sage handbook of qualitative research:43–64. Thousand Oaks, CA: Sage. Mansuri, G., & Rao, V. 2004. Community based (and driven) development: A critical review. The World Greenwood, R., Suddaby, R., & Hinings, C. R. 2002. Theo- Bank Research Observer, 19: 1–39. rizing change: the role of professional associations in the transformation of institutionalized fields. Acad- Markman, G. D., Russo, M., Lumpkin, G. T., Jennings, P. D., & emy of Management Journal, 451: 58–80. Mair, J. 2016. Entrepreneurship as a platform for pur- suing multiple goals: A special issue on sustainability, Hallett, T., & Ventresca, M. J. 2006. Inhabited institutions: ethics, and entrepreneurship. Journal of Manage- Social interactions and organizational forms in Gould- ment Studies, 53: 673–694. ner’s Patterns of Industrial Bureaucracy. Theory and ı Society, 352: 213–236. Mart´, I., & Mair, J. 2009. Bringing change into the lives of the poor: Entrepreneurship outside traditional Hargrave, T. J., & Van de Ven, A. H. 2006. A collective boundaries. In T. Lawrence, R. Suddaby & B. Leca action model of institutional innovation. Academy of – (Eds.), Institutional work: Actors and agency in Management Review, 314: 864 888. institutional studies of organizations: 172–223. Hinings, C. R., & Greenwood, R. 2002. Disconnects and Cambridge, England: Cambridge University Press. consequences in organization theory? Administrative McAdam, D., McCarthy, J. D., & Zald, M. N. (Eds.) 1996. – Science Quarterly, 47: 411 421. Comparative perspectives on social movements: Howard-Grenville, J., Buckle, S. J., Hoskins, B. J., & George, G. Political opportunities, mobilizing structures, and 2014. From the editors: Climate change and management. cultural framings. New York, NY: Cambridge Uni- Academy of Management Journal, 57: 615–623. versity Press. 2016 Mair, Wolf, and Seelos 2043

McKague, K., Zietsma, C., & Oliver, C. 2015. Building the Srinivas, M. N. 1962. Caste in modern India and other social structure of a market. Organization Studies, 36: essays. Bombay, India: Asia Publishing House. – 1063 1093. Srinivas, M. N. 1976. The remembered village. Berkeley, Miller, C. C., Cardinal, L. B., & Glick, W. H. 1997. Ret- CA: University of California Press. rospective reports in organizational research: A Stephan, U., Patterson, M., Kelly, C., & Mair, J. 2016. Or- reexamination of recent evidence. Academy of ganizations driving positive social change: a review – Management Journal, 40: 189 204. and an integrative framework of change processes. Orlikowski, W. J. 2006. Material knowing: the scaffolding Journal of Management, 42: 1250–1281. of human knowledgeability. European Journal of Stern, R., & Barley, S. 1996. Organizations and social – Information Systems, 15: 460 466. systems: Organization theory’s neglected man- Pache, A.-C., & Santos, F. 2013. Inside the hybrid organi- date. Administrative Science Quarterly, 41: zation: Selective coupling as a response to competing 146–162. institutional logics. Academy of Management Jour- Strauss, A. L., & Corbin, J. M. 1990. Basics of qualitative – nal, 56: 972 1001. research: Techniques and procedures for de- Rao, V., & Ibanez, A. M. 2003. Thesocialimpactofsocial veloping grounded theory. Thousand Oaks, CA: Sage. funds in Jamaica: A mixed methods analysis of par- Tilly, C. 1998. Durable inequality. Berkeley, CA: Uni- ticipation, targeting and collective action in commu- versity of California Press. nity driven development (World Bank Policy Research, Tonnies,¨ F. 1887. Gemeinschaft und Gesellschaft: working paper 2970). Washington, D.C.: World Bank, Grundbegriffe der reinen Soziologie [Community Development Research Group, Poverty Team. and society: Fundamental concepts of sociology] Rojas, F. 2010. Power through institutional work: Acquiring (in German). Darmstadt, Germany: Wissenschaftliche academic authority in the 1968 third world strike. Buchgesellschaft. Academy of Management Journal, 53: 1263–1280. Tracey, P., & Phillips, N. 2016. Managing the consequences Sanyal, P. 2009. From credit to collective action: The role of organizational stigmatization: Identity work in of microfinance in promoting women’s social capital a social enterprise. Academy of Management Jour- and normative influence. American Sociological nal, 59: 740–765. Review, 74: 529–550. Tracey, P., Phillips, N., & Jarvis, O. 2011. Bridging in- Schneiberg, M., & Lounsbury, M. 2017. Social Movements stitutional entrepreneurship and the creation of new and the Dynamics of Institutions and Organizations. In organizational forms: A multilevel model. Organiza- R. Greenwood, C. Oliver, T. Lawrence & R. Meyer tion Science, 22: 60–80. (Eds.), The SAGE handbook of organizational insti- United Nations 2015. Transforming our world: The tutionalism (2nd ed.). Thousand Oaks, CA: Sage. 2030 agenda for sustainable development.New Seelos, C., & Mair, J. 2007. Profitable business models and York, NY: United Nations Division for Sustainable market creation in the context of deep poverty: A Development. strategic view. The Academy of Management Per- Van Maanen, J. 1979. The fact of fiction in organizational spectives, 21: 49–63. ethnography. Administrative Science Quarterly, 24: Seelos, C., & Mair, J. 2016. Innovation and scaling for 539–550. impact: How effective social enterprises do it. Stan- Venkataraman, H., Vermeulen, P., Raaijmakers, A., & Mair, ford, CA: Stanford University Press. J. 2016. Market meets community: Institutional logics Selznick, P. 1949. TVA and the grass roots: A study in the as strategic resources for development work. Organi- sociology of formal organization. Berkeley, CA: zation Studies, 37: 709–733. University of California Press. Viterna, J., & Robertson, C. 2015. New directions in the Selznick, P. 1957. Leadership in Administration. New sociology of development. Annual Review of Soci- York: Harper & Row. ology, 41: 243–269. Sen, A. 1973. On economic inequality. Oxford, England: Vygotski, L. S. 1964. Denken und Sprechen. Berlin, Clarendon Press. Germany: Akademie-Verlag. Sen, A. 1992. Inequality Reexamined. Oxford, England: Yin, R. 2003. Case study research, design and methods. Clarendon Press. Thousand Oaks, CA: Sage. Srinivas, M. N. 1955. The social system of a Mysore village. York, J., Hargrave, T., & Pacheco, D. 2016. Converging In A. R. Beals & M. Marriott (Eds.), Village India: winds: Logic hybridization in Colorado wind energy Studies in the little community:1–35. Chicago, IL: field. Academy of Management Journal,59: University of Chicago Press. 579–610. 2044 Academy of Management Journal December

Entrepreneurship as a Force for more Inclusive and Innovative Societies). Her research interest lies in or- ganizing processes around social or environmental is- Johanna Mair ([email protected]) is a professor of sues. She received her PhD from Leeds University organization, strategy, and leadership at the Hertie School Business School. of Governance in Berlin. She is also a distinguished fellow at the Stanford Center on Philanthropy and Civil Society. Christian Seelos ([email protected]) is the Leo Tinde- Her research focuses on how organizational and in- mans Chair of Business Model Innovation at KU Leuven stitutional arrangements affect social and economic de- and a research fellow at the Stanford University Center on velopment, and the role of innovation in this process. She Philanthropy and Civil Society. His research focuses on received her PhD from INSEAD. how social sector organizations enact innovation processes and on the link between innovation, scaling, and impact. Miriam Wolf ([email protected]) is a post-doctoral researcher at the Hertie School of Governance in Berlin and a project manager for the SEFORIS Project (Social