The Nature of Institutional Change

Hargrave and Van de Ven (2006) note that part of the difficulty in understanding institutions is their diversity and pervasiveness. Prior work on institutional change certainly supports their view. This work varies substantially in that it focuses on explaining changes in a wide range of institutions (Jepperson, 1991). These include templates or blueprints for organizing (Powell, 1991; Greenwood and Hinings, 1996; D'Aunno, Succi, and Alexander, 2000; Johnson, Smith, and Codling, 2000); organizational form (Fligstein, 1990, 1991; Davis, Diekmann, and Tinsley, 1994); organizational fields (Hoffman, 1999; Leblebici et al., 1991; Alexander and D'Aunno, 1990); institutional logics (Thornton and Ocasio, 1999, 2008; Thornton, 2002; Scott, Ruef, Mendel, and Caronna, 2000; Lounsbury, 2002); organizational programs (Kraatz and Zajac, 1996; Kraatz and Moore, 2002); and management or organizational practices (Townley, 2002; Zilber, 2002; Sherer and Lee, 2002; Smets, Morris, and Greenwood, 2012; Lawrence et al., 2009).

Previous work varies not only in its substantive focus on different institutions, but also in its focus on institutions at varying levels of analysis, ranging from individual-level phenomena to field-level ones. For example, some papers examine change at the field level of analysis (Hoffman, 1999), while others are concerned with radical changes within organizations (Greenwood and Hinings, 1996). Explanations for institutional change also vary by level of analysis; in other words, analysis levels differ for both independent and dependent variables.

Some observers critique earlier work on two related grounds (Haveman and David, 2008; Mizruchi and Fein, 1999). First, and more generally, critics argue that key concepts in institutional theory, including the concept of institution itself, often are not clearly defined or measured. Similarly, they claim that studies of institutional change may be examining changes in practices that are not institutionalized: not all change in organizational fields, organizations, or practices is institutional change (Tolbert and Zucker, 1996). The importance of these critiques is obvious, but they merit highlighting. Without an accepted definition and measure of a focal institution, it is not possible to examine institutional change, let alone make comparisons and draw conclusions across studies of such change. I am sympathetic with these critiques and aim to be clear and consistent in distinguishing institutional change in health care from less fundamental changes that are occurring in the field.

Explanations for Institutional Change

Smets et al. (2012) provide a useful summary of explanations for institutional change. They argue that earlier work can be categorized into three approaches, each with its view of the origins, mechanisms, and processes involved in change: exogenous shocks, endogenous field-level contradictions, and intraorganizational dynamics.

The exogenous shocks view argues that institutional change occurs when current arrangements collide with new social values and technologies, demographic shifts, or different political or economic models and conditions that originate either outside an organizational field (Oliver, 1992; Clemens and Cook, 1999; Rao, Monin, and Durand, 2003) or from its periphery (Battilana, 2011; Maguire, Hardy, and Lawrence, 2004). To the extent that actors encounter jolts from external sources, they are forced to rethink fundamental assumptions about practices that they had taken for granted as the right way to do things. In this view, new ideas and ways of doing things typically originate from outside a field or from actors who are marginal to the field because they are neither socialized in widely accepted views and practices nor as likely to see their value. Thornton (2004), for example, showed how actors from outside the field of publishing in higher education carried new ideas about business strategy and operations into the world of publishing and resulted in substantial changes in the field.

Furthermore, once shocks occur, even well-entrenched actors can see the need for change and can become so-called institutional entrepreneurs (DiMaggio, 1988; Hardy and Maguire, 2008; Battilana et al., 2009; Tracey, Phillips, and Jarvis, 2011). These are actors, either organizations or individuals (or combinations of these), who leverage resources, including social and organizational networks, to create new institutions or transform existing ones. In the past few years, researchers have given much attention to the role of such entrepreneurs in institutional change. Battilana and colleagues (2009) recently reviewed sixty published papers and chapters on this topic, and they propose a model of the process of institutional entrepreneurship (reviewed below).

At the same time, controversy remains about assumptions underlying the study of institutional entrepreneurs. Holm (1995) stated the problem quite well: “How can actors change institutions if their actions, intentions, and rationality are all conditioned by the very institutions they wish to change?” (p. 401). Some analysts argue that the idea of institutional entrepreneurs is not consistent with the basic tenets of institutional theory, which hold that social and cultural environments, with their taken-for-granted norms and rules, shape organizational and individual behavior. In this view, studies of “hypermuscular heroes” who somehow have the power to change society (e.g., Steve Jobs or Bill Gates) are outside the bounds of institutional theory (Delmestri, 2006; Meyer, 2006). In contrast, other analysts argue that institutional theory has an oversocialized view of human behavior, rendering actors as passive “cultural dopes” whose social environments imprison them. In this view, institutional theory needs to develop a more compelling account for agency in human activity.

Battilana and I (2009) propose a middle ground and argue that enabling conditions at the field, organizational, and individual levels can promote or inhibit human agency. For example, institutional entrepreneurs are more likely to emerge in organizational fields that are less mature and hold more heterogeneous norms and values (D'Aunno et al., 2000; Dorado, 2005; Kraatz and Block, 2008). In these conditions, people have more freedom to challenge the status quo in organizational arrangements and practices.

Furthermore, following Emirbayer and Mische (1998), Battilana and I (2009) argue for a more complex and nuanced view of human agency. In this view, individuals have a range of levels of self-consciousness as well as a wide range of temporal orientations. At one extreme, self-consciousness is limited, and individuals are oriented mainly to the present or past; they reproduce habits and practices without much question. At the other extreme, self-consciousness is heightened and people project themselves into the future. In doing so, they can imagine alternative ways of behaving and thus have the ability to make choices that challenge the status quo.

Our approach poses research questions about the conditions under which different dimensions of agency and temporal orientations predominate and predispose individuals to behave more or less as cultural dopes than as muscular heroes. Debate about the concept of institutional entrepreneurs and their role in institutional change clearly is not settled, and thus I argue that research should focus on the conditions that promote or inhibit agency.

Two final points are important to consider about the exogenous shocks view. One is that this view emphasizes conflict and power struggles as actors challenge an institutionalized practice or organizational form and incumbents defend the status quo (Hargrave and Van de Ven, 2006; Van de Ven and Hargrave, 2004). For example, Rao, Morrill, and Zald (2000) draw on social movement theory to discuss “power plays” among actors involved in such struggles, and Hoffman (1999) uses the term institutional war to refer to open conflict among actors. Second, as illustrated in works by Fligstein and McAdam (2012) and Davis and colleagues (2005), the exogenous shocks perspective, as well as the endogenous contradictions view discussed below, often includes at least some analyses at the field and societal levels to account for changes in institutions.

A second approach to explain institutional change emphasizes endogenous contradictions at the field level of analysis. In this view, the potential for change increases to the extent that tensions exist within fields, which hold logics that are inconsistent or contradict each other (Friedland and Alford, 1991; D'Aunno, Sutton, and Price, 1991; D'Aunno et al., 2000; Seo and Creed, 2002). Organizations and individuals facing inconsistent or conflicting demands often must choose among them, thus creating the possibility for choice, which opens the door for these actors to create changes in their practices or models of organizing (Greenwood et al., 2011; Pache and Santos, 2010; Kraatz and Block, 2008). Furthermore, this view holds that tensions among inconsistent logics are likely to increase as organizational fields mature, making it less likely that actors can buffer themselves from conflicting or inconsistent demands (Oliver, 1991).

Smets and colleagues (2012) note that research that focuses on institutional contradictions within fields has emphasized the role that rhetoric and political struggles play in resolving conflicts. Studies show how discourse (Maguire and Hardy, 2009) and political skills are used to mobilize allies (Lounsbury, 2007; Weber, Heinze, and DeSoucey, 2008). Similarly, studies show tactics that actors use to promote collective movements and the ways in which these movements help actors gain advantage in conflicts (Hargrave and Van de Ven, 2006).

Kellogg (2009) provides an excellent example of this approach in her careful ethnographic study of the implementation of a mandate from the American Council of Graduate Medical Education that limited residents to an eighty-hour workweek. She examined responses to the mandate over two and one-half years in three similar teaching hospitals. Based on extensive data from participant observation and interviews, Kellogg argues that although field-level forces, such as mandates from credentialing or regulatory bodies, often may be necessary to prompt change in institutionalized practices, it is microlevel processes that make such changes happen within organizations. She terms these microlevel processes collective combat, in which defenders of the status quo and reformers engaged in a relatively protracted struggle to promote their causes. Actors tried to advance their goals by building coalitions, and they used a variety of tactics to mobilize resources, including labeling and denigrating opponents and interfering with their daily activities. Kellogg quickly discovered that changing daily work practices proved difficult—and indeed was unsuccessful in two of the three hospitals—because it required challenging long-standing beliefs, roles, and authority relations. She concludes that to understand institutional change, we need to analyze power and politics within organizations.

In short, in the endogenous contradictions view, institutional change often originates in fields that hold heterogeneous, conflicting, or contradictory beliefs, values, and logics that give actors within such fields opportunities, and indeed license, to challenge the status quo. Change then unfolds in conflict-laden processes that are similar to those reported by researchers who have examined how exogenous shocks prompt institutional change.

A third set of explanations for institutional change focuses on intraorganizational dynamics (Greenwood and Hinings, 1996). In this view, organizations, similar to fields, can hold inconsistent, uncoordinated, and sometimes conflicting and contradictory logics (Meyer and Rowan, 1977). In turn, actors within such organizations hold values and interests that are similarly fragmented, and this situation creates tensions that prompt challenges to established ways of doing things and potentially result in organizational change. How senior managers (Edelman, 1990) and powerful occupational groups (Dobbin, 2009) respond to these internal organizational conflicts may amplify or resolve them. Meyer and Rowan (1977) observe that loose coupling among actors and practices within organizations is an antidote to such conflicts, but there are limits to loose coupling, especially to the extent that organizations face external pressures for efficiency.

In sum, previous accounts for institutional change vary in some important ways, especially by focusing on either the processes in change or the causes of change, the mechanisms proposed to explain change, the role of agency, and levels of analysis for causal mechanisms. More recent work focuses on the processes in institutional change (Greenwood and Suddaby, 2006), as compared to earlier work that focused on understanding the antecedents of such change. Earlier work asked, “Why, or under what conditions, do institutions change?” as opposed to asking, “How does institutional change occur?” (Scott, 1994, p. 81; Scott, 2001). Accompanying the shift to an increased focus on the processes involved in institutional change has been a shift in research methods, moving from quantitative approaches that test hypotheses about why change occurs to qualitative methods that examine processes involved in change (Lawrence and Suddaby, 2006). This movement is producing a rich set of explanations for how change occurs that I discuss in more detail in the following sections.

A Field-Level Model

Hinings et al. (2004) and Greenwood et al. (2002) propose a process model for understanding institutional change at the field level of analysis. The proposed model consists of five stages that unfold in a sequence.

In stage 1, a range of powerful forces creates pressure for change. These forces include changes in political (changes in elected officials), technical (invention of new technologies), social (changes in demography), and market (entry of new competitors) conditions, and, importantly, the poor performance of established institutions (Oliver, 1992). For example, actors both internal (the Institute of Medicine) and external (employers) to the field of US health care have called for change in its institutional arrangements (how providers are paid) due to widely held views that our system is underperforming.

In stage 2, institutional entrepreneurs introduce new practices to replace old ones. Dorado (1999) argues that these entrepreneurs may be of different types: innovators who operate within a field and see opportunities for improvements in established practices; engineers who are gatekeepers within fields, controlling access to key resources needed to promote change; and catalysts who are outsiders to a field (such as business leaders who want improvements in health care). If these entrepreneurs are successful, their efforts lead to the next stage.

In stage 3, processes unfold, primarily at the field level of analysis, to legitimate new practices and delegitimate old ones. One key process is that of theorization, in which conceptual rationales are elaborated that both specify why old approaches fail to address current problems and justify why a new approach will solve these problems (Maguire and Hardy, 2009). These ideas and arguments to support new practices develop as they are tested by trial and error in local settings (Tolbert and Zucker, 1996; Strang and Meyer, 1993).

Furthermore, once new practices have been theorized and begin to receive some legitimacy, they can be disseminated across a field. Westphal, Gulati, and Shortell (1997) provide a useful example of how such dissemination occurs in their study of the adoption of total quality management (TQM) among US hospitals. They showed that early adopters customized practices to improve their performance, while later adopters were more likely to use TQM practices as a standard set, even if these practices did not improve their performance. Later adopters seem to be more interested in gaining the legitimacy of using practices that are widely accepted than in actually improving their performance.

In stage 4, as new practices are disseminated across a field, dynamics of change occur within organizations that may (or may not) adopt the new practices, as actors work to promote or oppose them (Greenwood and Hinings, 1996). Kellogg's (2009) study of hospital responses to changes in residency training hours, summarized above, nicely illustrates the dynamics of values, interests, power, and conflict that Hinings and colleagues (2004) discuss in their model.

Finally, in stage 5, powerful field-level actors promote institutionalization of the new practices. Hinings and colleagues (2004) emphasize that even at this stage, there is no guarantee that new practices will gain enough support to dominate a field. Rather, it is often the case that newly institutionalized practices and the logics that support them (Thornton, Ocasio, and Lounsbury, 2012) sit side by side in fields with other practices that may be uncoordinated or even conflicting (Meyer and Rowan, 1977). Heterogeneity of practices and logics at the field level is often common (Kraatz and Block, 2008).

The proposed model is relatively comprehensive, addressing the fundamental causes of institutional change (the pressures identified in stage 1) as well as the origins of new practices (institutional entrepreneurs of different types) and how they replace established practices at the organizational and field levels. Indeed, one could argue that the model of field-level institutional change by Hinings and colleagues (2004) incorporates all of the key elements and arguments from the models reviewed above. Yet the virtues of the proposed model also entail a limitation: it operates at a level of generality that suggests the need for more precise models of the processes involved in institutional change. I now turn to discuss three such models.

Process Models of Institutional Entrepreneurship

Greenwood and Suddaby (2006) and Battilana and colleagues (2009) have proposed overarching models of processes involved in institutional entrepreneurship. Greenwood and Suddaby's (2006) model has several key components that are linked in a causal stream. To begin, a fundamental cause for institutional change is field stratification, that is, the existence of hierarchies of status and power within and between communities of organizations. By definition, stratification means that some organizations are more elite and central to a field than others. This distinction matters because elite organizations have access to more information than lower-status organizations do and thus elite organizations, more than others, are aware of contradictions that may exist at the field level.

Organizations also vary in the extent to which their forms and practices are aligned with the regulatory and market rules of their field; misalignment exposes organizations to contradictory criteria for their performance (D'Aunno et al., 2000). Furthermore, organizations whose work involves spanning the boundaries of their fields learn about logics that differ from their own. In turn, misalignment and boundary spanning prompt organizational awareness, motivation, and openness to alternative logics. Finally, institutional entrepreneurship is more likely to occur under these conditions.

In brief, stratification and network positions expose organizations differentially to field-level contradictions in expectations for their behavior and performance. Once exposed to such contradictions, organizations may experience negative feedback on their performance (e.g., they cannot meet conflicting expectations and thus disappoint some actors). Furthermore, exposure to contradictions makes organizations aware of alternative models and logics. In turn, negative feedback and awareness of alternatives make organizations open to change and enable institutional entrepreneurs to launch change efforts, which are likely to be successful to the extent that powerful, centrally located organizations are involved in them.

The model by Battilana and colleagues (2009) has some important similarities with that of Greenwood and Suddaby (2006). These include an emphasis on field characteristics and actors' positions within them as factors that enable change, as well as similar observations about the power and politics of change once entrepreneurs initiate their efforts. Yet some different points of emphasis in the two approaches make them more complementary than redundant.

First, Battilana and colleagues (2009) emphasize that institutional change and entrepreneurship are more likely to occur in fields that are newer and less settled (Tolbert and Zucker, 1996). In contrast, Greenwood and Suddaby (2006) are explicitly concerned with why and how change occurs in mature organizational fields. Second, Battilana and colleagues (2009) argue that actors on the periphery of organizational fields (those with lower status) are more likely to challenge the status quo because they are less invested in it (see Leblebici et al., 1991). In contrast, Greenwood and Suddaby (2006) argue that actors who are more elite in a field have more possibilities to learn about alternative logics and adequate power to change arrangements in their fields. Greenwood and Suddaby also argue that even organizations with central positions in a field can become institutional entrepreneurs when they are exposed to field-level contradictions as a result of either boundary-spanning roles or misalignment that they experience. In short, these models differ in their views about the roles that central versus peripheral positions in organizational fields play in motivating and enabling institutional change.

As was the case for the field-level model of institutional change (Hinings et al., 2004) discussed above, both models of institutional entrepreneurship operate at a level of abstraction that leaves open the possibility for complementary accounts. Tracey and colleagues (2011) recently provided such an account. They examined processes involved in the creation of a new organizational form in England: a social enterprise to help homeless people in which they would operate a household catalogue business. Profits from this business not only pay employees a living wage, but also are used to hire staff members whose responsibility is to link employees to needed social services and other job opportunities. To launch this new organization, two entrepreneurs, Paul Harrod and Mark Richardson, engaged in six types of institutional work that clustered around three levels of analysis: individual, organizational, and societal.

At the individual level, creating a new organizational form required Harrod and Richardson to frame the problem in a novel way and engage in counterfactual thinking to imagine alternative approaches to homelessness (e.g., homeless people need jobs and careers rather than homes). Harrod and Richardson also needed to build a new organizational template for their work and did so by borrowing ideas from multiple sources about how to organize and operate a profitable business. Finally, to institutionalize their approach, these entrepreneurs changed societal discourse about the homeless and raised awareness of their new model by using the media and aligning their approach to highly legitimate actors, including, for example, the British royal family.

The approach to institutional entrepreneurship by Tracey and colleagues (2011) combines elements of models by Battilana and colleagues (2009) and Greenwood and Suddaby (2006). In particular, much of what Tracey and colleagues (2011) discuss concerns processes involved in what Battilana and colleagues (2009) term “implementing divergent change.” Tracey and colleagues (2011) nicely theorize the processes in building, launching, and institutionalizing the new organizational model and illustrate what Lawrence et al. (2009) term institutional work. At the same time, Tracey and colleagues (2011) illustrate the importance of Greenwood and Suddaby's (2006) concept of network position, especially the role that boundary spanners such as Harrod and Richardson can play in importing new organizational models, or parts of new models, into a focal field (in this case, homelessness) from other fields (business).

In sum, in the past few years, researchers have made much progress in developing models of key processes involved in institutional entrepreneurship and in articulating its role in institutional change. I turn next to a final set of three models that, unlike those discussed above, focus on changes in institutionalized practices within organizations.