Studying Differences between Organizations: Comparative Approaches to Organizational Research: Volume 26

Comparative organizational analysis once dominated American organizational sociology, grounded in rich case studies about organizational processes and outcomes. The Columbia school's approach to organizational research was exemplary in this regard. Following the publication of Robert K. Merton's (1940) essay, “Bureaucratic Structure and Personality,” he attracted a group of talented doctoral students to his formal organizations seminar (Crothers, 1990), the core of whom would go on to write dissertations, books, and articles forming the substance of American organizational sociology in the decades to come. Among those students were Philip Selznick, Alvin Gouldner, Peter Blau, Seymour Martin Lipset, Rose Coser, and James Coleman. While their work varied greatly in substantive content, their studies shared a theoretical interest in explaining intra-organizational dynamics and the unexpected outcomes of bureaucratic administration. Organizations, they demonstrated, developed “lives of their own,” quite outside the intents of their founders (Haveman, 2009; refer, especially, Selznick, 1957). Organizations, in other words, were adaptive to the needs of their constituents, but adaptations did not always produce the intended results. One of the unintended consequences of organizational development was increasing variety in the kinds of organizations that emerged to meet particular societal goals or ends. Thus, an inherent focus of this early comparative research was the explanation of variety in organizational types, policies, and outcomes and an emphasis on the ways in which organizations diverged from ideal types.

In our 1983 paper, McKelvey and I (McKelvey & Aldrich, 1983) took the field of “organization science” to task for not paying sufficient attention to the scope conditions under which research findings are valid. (Today I would argue that the field also had not paid sufficient attention to matching theoretical ambitions with research designs.) We argued that the field fell short on three critical criteria: classifiability, generalizability, and predictability. We noted that samples of organizations were so poorly described that classifying them was impossible, that generalizations were being carelessly drawn, and that the predictive power of most theories was extremely weak.

If one examines the changing nature of organizations and the contexts in which they operate over the past half-century, a number of developments have worked to soften and erode early distinctions around which much comparative research was conducted. Three types of changes encourage convergence.

This chapter sets forth a form of comparative analysis that is explicitly organizational, in the sense that it uses cross-level, contextual or compositional, analysis to explain organizational-level observations, especially comparisons between organizations. Inter-organizational comparisons often surface paradoxical results, in the form of unexpected differences among similar kinds of organizations, or unexpected similarities among different kinds of organizations. The value of using comparative analysis in these cases is that the information required to unravel organizational-level puzzling results is often located at a higher or lower levels. The proposed form of comparative analysis thus extends the conventional top-down, unidirectional form by adding a bottom-up component – making it bi-directional. In addition to introducing an organization-centered form of comparative analysis, the chapter offers suggestions for its practice and speculates about the potential benefits of its broad application within organizational studies.

The contrast of multilevel and comparative research may seem counterintuitive at first. After all, one might argue that comparative research on organizations by necessity spans several levels of analysis (Rokkan, 1966). Yet, multilevel and comparative research on organizations present rather distinct traditions in organization studies, each with its own epistemological assumptions and associated methods. Accordingly, an approach that aims to incorporate both multilevel and comparative ideas needs to start with taking inventory of these prior literatures to situate itself. In the following, we thus turn to the literatures on multilevel and comparative research as different traditions with surprisingly little overlap.

The “first generation” (Lammers, 1978, p. 486) of comparative analysis of organizations in sociology (e.g., Blau, 1965; Stinchcombe, 1959) focused on the “nuts and bolts” of organizational structure as the key criterion with which to derive organizational typologies (Perrow, 1967; Pugh, Hickson, & Hinings, 1969). This initial cohort of analysts saw the intrinsic features – or “organizational attributes” (Blau, 1965, p. 326) – constitutive of the “technical core” of the organization, such as features related to the organization of the production process (Perrow, 1967) or the structure of allocation of discretion and authority (e.g., Etzioni, 1961), as the royal road to the development of a cogent approach to comparative analysis of organizations.

When we try to explain the dynamic relationship between actors and their environment, “unidirectional” paradigms clustered at either end of an agency–determinism continuum, theoretical absolutes, and a focus on final outcomes are of limited value. Comparative research is uniquely positioned to move beyond such limitations and toward accounts of organizing that incorporate variation, interests, and interaction. To guide comparative research toward this more relational approach, we (1) highlight a move toward the middle ground of the agency–determinism continuum, that is, varieties of interaction and mutual influence between actors and their environments across levels of analysis; and (2) propose to conceptualize actor–environment relations as a “negotiation.” We use this metaphorical lens to stimulate a focus on the variety of different “negotiation spaces” and “negotiation moves,” which actors may utilize in an organizational field to affect stability or change. We provide an exemplary application of the framework and conclude with some observation on the implications for future research.

Although extant theory has illuminated conditions under which organizations mimic each other in form and practice, little research examines how organizations seek to differentiate themselves from conventional forms. Our comparative ethnographic studies examine how the Burning Man and Open Source communities developed organizations to help coordinate the production of an annual temporary arts event and nonproprietary, freely distributed software. Both communities sought to differentiate their organizations from reference groups, but this was not a sufficient condition for sustaining organizational novelty. We found that the ability to pursue a differentiated strategy was moderated by environmental conditions. By exploring the organizing decisions that each community made at two critical boundaries: one defining individuals’ relationship with the organization; the second defining the organization's relationship with the market, we show how organizing practices were recombined from the for-profit and nonprofit sectors in unexpected, novel ways. This comparative research contributes a grounded theoretical explanation of organizational innovation that adjudicates between differentiation and environmental conditions.

Social movements are heterogeneous because they attract organizations from other movements and encourage activists to create organizations “indigenous” to the movement. This chapter examines the structural and technical differences between these kinds of organizations. Employing a contingency theory framework, it is shown that older “spill over” groups are much more likely to be multi-issue national organizations with particular organizational structures. Then, it is shown that these older groups have correlated environments and internal structures, but not their more contemporary counterparts. Finally, it is shown that the adoption of a new technology, the Facebook group, is mainly a path dependency outcome, and not correlated with contingency factors.

In this chapter, we describe a multilevel, longitudinal, comparative case study approach for investigating organizational heterogeneity based on our experience studying institutional change in the health care field (Scott, Ruef, Mendel, & Caronna, 2000). By examining the relationship between organizations, populations, and fields, differences between organizations can be captured in terms of their organizational identities as well as their (changing) relationships with the organizational field. We discuss the analytical strategies we used in our study of institutional change and describe our findings of organizational heterogeneity across levels and over time. We conclude with suggestions for future research that incorporate elements of our study design and lessons from our research process and outcomes.