Abstract
Organizations across sectors appear to be shifting their ambitions from solving social problems to changing entire social systems. This phenomenon offers a timely opportunity to revisit what came to be known as the third mandate of organizational theory. In this paper we interrogate how organizational scholarship can productively explore and theorize the relationship between organizations and social systems in organized system change – an effort by organizations to alter the conditions that generate the characteristics of social problems and their dynamics of change. As a basis for theorizing organized system change, we develop an analytical scaffold that helps researchers to attend to fundamental aspects of the phenomenon and to achieve parsimony without blanking out complexity. Grounded in realist metatheory and principles, the scaffold reduces ambiguity, provides a backbone for empirical analysis, and favours mechanism-based explanation. We suggest that generating theoretically interesting and practically adequate knowledge on organized system change requires attention to three system realms: First, the subjectively constructed problem realm of systems concerned with processes of evaluating and problematizing situations. Second, the objectively constituted situational realm that attends to factual characteristics of situations and their dynamics of change. And third, the realm of causality understood as the mechanisms that generate both the objective characteristics of situations and the subjective criteria by which situations are evaluated as problems. In concluding, we reflect on the topics of boundaries and power as two promising areas for theorizing organized system change.
Introduction
Recent studies published in management journals have expanded a traditional focus on organizational problems by a new focus on social problems such as sex trafficking (Ruebottom & Toubiana, 2021), child marriage (Claus & Tracey, 2021) or homelessness (Lawrence & Dover, 2015). Sociologists refer to social problems as conditions or circumstances that are perceived and interpreted as troubling by a collective (Hilgartner & Bosk, 1988). This new focus provides an important opportunity to ground organizational research in the observation of real-life phenomena relevant to society at large (Walsh, Weber, & Margolis, 2003) instead of finding holes in the management literature (Hambrick, 2007). In this paper we focus on a recently emerging phenomenon at the nexus of organizations, social problems and social systems. Organizations across sectors appear to be shifting their ambitions from solving social problems to changing entire social systems (Seelos &Mair, 2018). An example of this shift is a group of philanthropic organizations with a combined asset base of US$50 billion that joined forces to promote and implement a system change approach. Managed and coordinated by the organization Co-Impact, their collective effort aims ‘to ensure systems that provide the most fundamental services of health, education, and economic opportunity result in improved outcomes for millions of women, men, and children.’ 1 Policy makers and social enterprises have also endorsed system change as the new mantra. Ursula von der Leyen, President of the European Commission, claimed in her foreword to A System Change Compass that the report ‘provides guidance for a systemic realisation of the European Green Deal.’ 2 Ashoka, an organization that supports social entrepreneurs globally, celebrates ‘system-changing social entrepreneurs’ whose approaches involve ‘changing the participants, rules, or goals of the game in a lasting way.’ 3 The multi-business organization Sekem won the Right Livelihood Award, also referred to as the ‘Alternative Nobel Prize’, for implementing a mission of transforming Egypt’s cultural, economic, social and environmental systems (Seelos & Mair, 2007a). The adoption of the term system that conveys ambition, broad scope and grand scale is an intriguing change in organizations’ rhetorical vocabulary. But beyond the rhetoric, many organizations are spending millions of dollars and are basing core strategic and operational decisions to move from solving problems to changing systems. We perceive this ambition for organized system change as a timely phenomenon for rejuvenating what came to be known as the third mandate of organizational theory.
The inaugural issue of Administrative Science Quarterly published in 1956 constitutes an important imprint for organizational theory. At the occasion of the journal’s 40th anniversary, Stern and Barley (1996) refer to this imprint as the three mandates for organizational theory: First, to understand internal organization structure and process; second, to examine relations between organizations and environmental actors; and third, to analyse the impact of organizations on the broad social systems in which they were embedded. Stern and Barley (1996) noted that the third mandate of organizational scholarship remained incomplete and, even worse, had ‘faded from the research agenda’ (p. 146). Twenty-five years later, we rarely integrate social systems – the central focus of the third mandate – into the empirical and conceptual repertoire of the organizational theories in use (Bansal & Song, 2017). Reflecting on the recent ambitions of organizations to enact system change, we ask: How can we productively explore and theorize the relationship between organizations and social systems in the form of organized system change—an effort by organizations to alter the conditions that generate the characteristics of social problems and their dynamics of change?
System perspectives became fashionable sixty years ago and affected debates among organization theorists (Waldo, 1961). Thus, perspectives on social systems were never absent from first- and second-mandate research. For our proposed third-mandate perspective, we depart from first-mandate perspectives that view organizations as social systems (Burns, 1961; Harris & Sutton, 1986; Siggelkow & Levinthal, 2003). We also depart from second-mandate perspectives that centre on the interdependencies of an organization and social systems understood as the context that affects organizational outcomes (Anderson, 1999; Gulati, Sytch, & Tatarynowicz, 2012). In first- and second-mandate research, the main explanandum in terms of competitive success such as profits, market share, or organizational survival remains largely uncontested (Margolis & Walsh, 2003). Our objective in this paper is to reinvigorate the tradition of third-mandate research by focusing on the phenomenon of organized system change. This focus complements existing perspectives on how organizations influence the socio-economic system they are embedded in and the roles and responsibilities of organizations in a societal context (Barley, 2007; Matten & Crane, 2005). Research on organized system change is concerned with the efficacy of organized efforts to change social systems and the legitimacy and desirability of the nature and outcomes of such efforts. In Table 1 we provide a stylized overview of these perspectives.
System perspectives in organizational research.
What explains the hesitation to integrate social system perspectives in our theoretical and analytical repertoires? Stern and Barley (1996) cited analytical challenges and the increasing complexity of social relations as factors that might contribute to this hesitation. Social complexity obfuscates the determination of appropriate levels of analysis and the boundaries of social systems and exposes the limits of well-vetted methodological tools such as regression analysis (March & Sutton, 1997). These difficulties are also apparent in recent work by management scholars interested in sustainability. Adjacent to our effort here, Bansal and Song (2017) have suggested to apply system perspectives to explore topics of organizational responsibility and sustainability. Focusing on the role of firms in the social-ecological systems in which they are embedded, sustainability scholars have turned to older perspectives of systems thinking (Meadows & Wright, 2008) and systems dynamics (Forrester, 1979) as a source of inspiration (see Williams, Kennedy, Philipp, & Whiteman, 2017, for a review). However, to advance knowledge on the phenomenon we characterize as organized system change, we advise against importing and adopting concepts from system perspectives developed in adjacent fields too hastily as we might unconsciously accelerate a trend observed by Clegg (2002) as ‘organization theory [becoming] increasingly its own series of language games’ (p. 435).
The construct system was subjected to its own language games, which generated several system perspectives across disciplines whose ontological positions and epistemological assumptions are inconsistent (Jackson, 2000; Rousseau, Wilby, Billingham, & Blachfellner, 2016). Specifying the empirical referents of the construct system is far from trivial. In the natural world, distinguishing settings by whether they constitute a system or a non-system provides hypotheses that are analytically important and that can be confirmed or rejected. For example, the absence of any emerging properties of an ensemble of parts may categorize these parts as a set rather than a system. A set of rocks has no inherent structure that would make it a whole with properties distinct from those of its parts. Water molecules, in contrast, interact to form water, a system that has properties such as wetness that are not properties of its parts. Can this distinction between systems and non-systems also be applied to social systems to help specify the empirical referents of third-mandate research? We are not aware of any clear criteria that would enable this distinction. Relational complexity arises from a vast behavioral repertoire enabled in human beings. Unlike in a water molecule, where relations between its atoms are stable and thus causally deterministic for the parts and the whole, human behaviour is situationally triggered by a myriad of intrinsic and extrinsic factors (Harré & Madden, 1975). This propensity generates dynamically changing systems with ambiguous boundaries that display only temporal regularities (Archer, 1995; Lawson, 1995). 4
Unfortunately, this perspective on social systems does not provide much analytical guidance for third-mandate research; in fact, it raises the question: How do we avoid getting lost in complexity, idiosyncrasy and ambiguity about the scale, scope and boundaries of social systems? Parsimony, which we define as a minimal set of empirical lenses necessary for adequate explanation, should also be a principle for third-mandate research. Without an analytical scaffold of systems that specifies parsimony, we may disengage with the complexity inherent in system perspectives in favour of narrow, established perspectives within the comfort zone of editors and reviewers. However, applying mainstream perspectives might run the risk of developing theory that remains socially underspecified. At the other extreme, the ambiguity of the construct system might risk turning it into a ‘dead metaphor’ (Tsoukas, 1991). Hirsch and Levin (1999) pointed out that ambiguous constructs risk eventual collapse: ‘When a scholarly idea becomes dangerously close to meaning all things to all people, that idea’s validity cannot be maintained indefinitely, at least not under the same name’ (p. 210).
Our objective in this paper is to lay out an analytical scaffold that reduces ambiguity and provides a backbone for empirical analysis. The scaffold is grounded in realist perspectives on causality based on mechanism-based explanation as the basis for theorizing (Bhaskar, 1997; Bunge, 2006; Harré & Madden, 1975; Sayer, 2000). In the next section we reflect on our own journey of studying organizations that tackle social problems. For the organizations we studied, organizing system change cannot be decoupled from tackling social problems. Social problems are embedded in social systems that give rise to, shape and perpetuate problems. Studying this nexus of organizations, social problems and systems made us painfully aware of the limitations of the well-vetted theoretical and methodological toolkit on which we relied. This journey forced us to reflect on our roles, biases and effectiveness as researchers and instilled a healthy dose of respect for complex social realities vis-a-vis our comparably simplistic theories. In subsequent sections, we will reflect on contemporary research on organizations, social problems and systems. We then develop an analytical scaffold of systems that supports third-mandate research on organized system change and conclude with implications for productive theorizing.
Our Learning Journey
In the early 2000s, as faculty members in a strategy department at a European business school, we were fascinated by the relatively new phenomenon of social entrepreneurship (Mair & Martí, 2006; Seelos & Mair, 2005). We discovered several organizations that seemed atypical and extreme in how they addressed social problems in unusual contexts. In an early study, based on field work in Bangladesh, we wanted to learn more about a joint venture between the Grameen Bank (a nonprofit organization that provided micro loans to poor villagers in Bangladesh) and the IT company Telenor of Norway. The two organizations set out to accomplish a puzzling set of objectives: to build a telecommunications infrastructure that bridged the urban–rural divide, to put technology in the hands of poor rural women as an enabler of microbusinesses, to change the economic and social fabric in Bangladesh, to contribute to the development of Bangladesh as a country, and to build a profitable growth market for Telenor. In retrospect, this phenomenon represented many characteristics of organized system change and a great opportunity for third-mandate research. To circumvent the complexity of ambitions, structural, operational, financial and political arrangements involved in the phenomenon, we opted for relatively narrow and mainstream theoretical perspectives including the resource-based view and organizational collaboration. These perspectives provided focus by specifying analytically relevant factors, which in turn enabled us to model complex phenomena in the form of simple visual sketches (Seelos & Mair, 2007b). Looking at this organized system change through first- and second-mandate lenses ensured that our paper fit the criteria of editors and reviewers of a management journal. But this focus came at the cost of reducing rich phenomena to narrow perspectives. The paper clearly had impact: It remains highly cited and won two prestigious awards. But the more we engaged with Bangladesh and understood social realities in rural villages, the more we questioned whether our theoretical and methodological choices did justice to the phenomena.
Experimenting on the methodological front, we looked at and started to appreciate formal perspectives on the role and validity of models in science (Azevedo, 1997; McKelvey, 2002; Morgan & Morrison, 1999). We learned how to move from modelling that resulted in ‘naive sketches’ to modelling where we could be transparent about the empirical and theoretical choices involved in creating the model and about criteria for assessing the model’s validity. We relied on modelling to explain how Aravind, an eye hospital in India, organized its activities in a way that represented a fundamental change in India’s health system (Seelos, 2014). The transparent modelling around explicit validity criteria advanced our research. But the model specification created a template that did not allow us to integrate rich observations and insights related to the phenomenon collected in numerous field visits over several years. We had again taken parsimony too far. We understood that we needed to build an analytical scaffold for our research that would help us pay attention to important elements in complex social settings and to consistently enact epistemic principles, the ontological and epistemological positions from which we looked at the world and created knowledge. We adopted scientific realism, which Baum and Rowley (2002) referred to as ‘the most widely accepted epistemology among current philosophers . . . and the primary scientific logic-in-use in organization science’ (p. 21) and we started an inquiry on the implications of realism for our research. Tsoukas (2000) had argued earlier that realist ontological commitments and epistemological principles can constitute a metatheory for theorizing of managerial efforts in open systems. Our engagement with realist metatheory was informed and inspired by Harré’s causal realism (Harré & Madden, 1975), Bhaskar’s (1997) transcendental realism (sometimes referred to as critical realism), the pragmatic realism of Sayer (2000) and Bunge’s (2006) scientific realism and perspective on systemism. This elaboration of a realist metatheory was essential for developing an analytical scaffold in the form of a coding scheme that significantly increased the explanatory power of our research (Seelos & Mair, 2014). We applied this analytical scaffold and re-analysed data from the Aravind eye hospital to probe deeper into its causal architecture (Seelos & Mair, 2014). What we carried over from this episode was a deep commitment to using analytical scaffolds that guided attention to an essential set of factors that was causally efficacious. And we learned to pay attention to exploring and analysing the unobservables in social systems. In the case of Aravind, this commitment was critical to uncovering behaviour or patterns unlikely to be revealed in interviews or by analysing survey data. For example, we discovered that the tradition of engineered marriages between members of the founding family of the Aravind hospital was a key factor in explaining unusual and extreme outcomes at the organizational level. We had missed this in an earlier paper (Seelos, 2014) where we applied narrow coding schemes and observational templates. We also started to engage counterfactual reasoning when analysing unobservables. Based on theories and experiences from similar situations, we constantly asked, ‘Why do we not see this?’ This attitude prompted a focused inquiry into unobservables illustrated by questions such as: ‘Why do Aravind’s eye doctors, who are among the best in the world, not leave for significantly higher paid jobs at other private hospitals?’ or: ‘Why are young women from rural India allowed to work at Aravind? Why are they not getting married instead?’ We termed these unobservables realist counterfactuals (Seelos & Mair, 2014). Realist counterfactuals are theoretically important as they allow researchers to go beyond scratching the surface of social systems and to explore less obvious aspects of a system’s workings. This style of inquiry necessitates a deep engagement with phenomena and sustained efforts to build trusted relationships with actors constitutive of phenomena. For example, it took ten years from our first field trip and multiple subsequent visits studying the organization Gram Vikas in Mohuda, a rural village in Odisha, India, to accumulate sufficient evidence to submit a manuscript on how Gram Vikas changed the causal architecture of villages as social systems to generate significantly lower levels of inequality and exclusion of lower-caste villagers (Mair, Wolf, & Seelos, 2016).
In their call to rehabilitate the third mandate Stern and Barley urge organizational scholars to pay more attention to social systems, but caution that ‘the study of social systems and social problems are different and should not be confused’ (Stern & Barley, 1996, p. 149). Instead of analytically separating social systems and problems, in this article we propose to conjointly explore the nexus of organizations, social problems and systems.
Literature on Organizations, Social Problems and Systems
Contemporary societies are made of organizations (Perrow, 1991). Organizations are involved in the creation of social problems and in the reproduction of social inequalities. Although this aspect of organizational life was largely sidelined in organizational theory by the 1990s and early 2000s (Hinings & Greenwood, 2002), organizational scholars have recently become more vocal and explicit about how organizations create new social problems and shape patterns of privilege and disadvantage in society. For example, Amis, Mair and Munir (2020) identified organizational myths such as meritocracy, efficiency and the positive effect of globalization as important but under-recognized factors in the reproduction of social inequities across gender, race and class. Castro, Phillips and Ansari (2020) revealed how corporations can become sites of corruption and propel systemic corruption. And Reinecke and Ansari (2016) problematized the role and responsibility of companies in a humanitarian crisis in a study on conflict minerals in the Democratic Republic of Congo.
However, organizations are also essential for societal progress (King, 2017).
It seems plausible that whatever progress our society has made over the last thirty years toward eliminating discrimination by race and gender can be traced primarily to the acts of organizations. Organizations, public and private, are largely responsible for widespread access to healthcare, although that access is currently under attack because of organizational interests. Organizations have been essential in raising the standard of living, if not the quality of life, in most Western nations. (Stern & Barley, 1996, p. 148)
Organizations that tackle and potentially solve social problems have become more central to management and organizational research. We briefly reflect on two popular research streams studying organizations and social problems to understand if and how they relate to organized system change.
Over the past decade, organizational theorists have effectively mobilized the label hybrid organizing to highlight internal challenges of organizations that address a social problem and simultaneously pursue commercial goals (Battilana, Besharov, & Mitzinneck, 2017). Hybrid organizing studies have covered a broad range of social problems, including the digital divide (Smith & Besharov, 2019), the integration of homeless people in the labour market (Tracey, Phillips, & Jarvis, 2011) and the exclusion of women from market-based activities (Venkataraman, Vermeulen, Raaijmakers, & Mair, 2016). However, these studies rarely include the social problems the organizations tackle as an object of analysis. Instead, the empirical setting (i.e. an organization tackling a social problem) is commonly used to legitimize and match the choice of theoretical perspective. Perspectives that foreground conflict and tensions inside organizations such as institutional logics, institutional work and paradox have become the preferred lens for studying these organizations, and scholars have largely applied methods and analytical strategies consistent with theses lenses (Smith & Besharov, 2019; Battilana, Sengul, Pache, & Model, 2015; Lawrence & Dover, 2015).
A second popular stream in research on organizations and social problems has explicitly highlighted the complex, uncertain and especially evaluative facets of social problems as posing difficulties for managing and organizing (Ferraro, Etzion, & Gehman, 2015; Levy, Reinecke, & Manning, 2016). Scholars and studies colonizing this stream (partly in cohabitation with hybrid organizing researchers) have used the label grand challenges, interpreted as ‘formulations of global problems that can be plausibly addressed through coordinated and collaborative efforts’ (George, Howard-Grenville, Joshi, & Tihanyi, 2016, p. 1880) to reclaim relevance for organizational research. (See Howard-Grenville, 2021 for a reflection.) However, scholars refer to grand challenges interchangeably as a theoretical concept (Ferraro et al., 2015), a specific type of managerial challenge (George et al., 2016), or a gap between theory and practice (Banks et al., 2016). Moreover, studies have used a wide range of empirical referents for what constitutes a grand challenge, ranging from Hilbert’s mathematical problems (George et al., 2016), to firm-related problems such as supply chains (Kim & Davis, 2016), to technological innovations (Grodal & O’Mahony, 2017). And finally, scholars have adopted the label grand challenges as shorthand for the complex social reality inherent in their empirical setting. Examples include the refugee crisis (Kornberger, Leixnering, Meyer, & Höllerer, 2018), natural disaster (Williams & Shepherd, 2016), global health (Vakili & McGahan, 2016), climate change (Wright & Nyberg, 2017) and inequality in rural India in our own study (Mair et al., 2016). Because of the multiple and often diverging uses of the label, organizational research on grand challenges might hamper the accumulation of knowledge (Munir, 2020).
Research on hybrid organizing and grand challenges has clearly helped to broaden the scope of organizational research and examine organizational practices that affect social life outside the boundaries of firms. So far, this research has not explicitly integrated social problems and social systems in the analytical and conceptual repertoire. Thus, we know relatively little about if and how social systems change because of organized efforts. Integrating social systems and knowledge on social problems is helpful to better understand (a) the objective conditions that give rise to the social problems, (b) how and by whom these conditions are problematized, (c) how these conditions affect organizational processes and outcomes, and (d) the variance in evaluations of the desirability of means and ends of organized efforts to change systems. Our objective in this paper is to build an analytical scaffold that helps to integrate social systems in the analytical strategy and theoretical ambition of this literature. To integrate existing perspectives more effectively, we provide a few insights into the vast and expansive literature on systems in adjacent disciplines.
System scholarship includes several disparate theoretical streams and traditions, and scholars have defined and used the term system inconsistently. Thus, the fragmented state of the systems literature might constitute an additional hurdle to theorizing and conducting empirical work on organized system change. The original promise of a grand theory of systems by pioneers such as Ludwig von Bertalanffy has produced little in terms of robust theories (Rousseau et al., 2016). Commentators have pointed to the invalid and often naive application of natural science metaphors such as equilibrium, entropy, homeostasis and organismic models, evolutionism, Darwinian functionalism, autopoiesis, or sociocybernetics to explain the lack of progress in theorizing social systems (Buckley, 1967; Castellani & Hafferty, 2009). Others have called out the assumption of holism as a major impediment, as it tends to diminish agency and the importance and place of individuals and their needs (Berman, 1996; Bunge, 2006).
Today, the conceptual and theoretical fragmentation imprinted in earlier literatures on systems prevails (Rousseau et al., 2016) with ‘dozens of small systems societies that speak to widely differing points of view as to what constitutes systems science for their members’ (Warfield, 2003, p. 508). In addition, methodological ambiguity stemming from too ‘many concepts, methodologies, methods and techniques . . . may well have inadvertently created a complex clutter of systems approaches’ (Reynolds & Holwell, 2010, p. 9). Thus, system perspectives do not provide us with a robust theoretical apparatus, nor do they specify clear empirical referents or criteria for evaluating truth claims.
In addition to epistemological challenges, we also lack a consistent ontological position on systems. Do systems exist in the real world, or are they merely social constructions? Plugging theoretical bits and pieces from disparate system literatures into existing organizational theories might create ontological and epistemological confusion (Lounsbury & Wang, 2020). Instead, we build a parsimonious analytical scaffold that facilitates integrating social systems into theorizing organized efforts to change systems.
Analytical Scaffold of Systems
Parsimonious scholarship on organized system change needs to specify the simplest set of system realms that has explanatory efficacy. We propose an analytical scaffold that specifies the realms of organized system change that, in turn, satisfy the condition of collective sufficiency. Because the realms are generic and enable different research efforts, the more stringent condition of our scaffold is the condition of collective necessity. Lack of attention to any element lowers explanatory efficacy. Our proposed analytical scaffold comprises three system realms. First is the socially constructed problem realm of systems. This realm is concerned with subjective processes of evaluating and problematizing situations. Second is the objectively constituted situational realm that attends to factual characteristics of situations and their dynamics of change. This realm is concerned with factual realities that serve as an objective arbiter of claims about social problems and constitute objective constraints to what organizations and people can and cannot do. And third is the realm of causality, understood as the causal mechanisms that generate the objective characteristics of situations and the criteria by which situations are subjectively evaluated as problems. We assume that organized system change intends to change this causal realm to create a different system state (situational realm) that is evaluated as less troubling and more desirable by relevant constituents (problem realm). When the Bangladesh-based NGO BRAC introduced antidiarrheal treatments based on a simple sugar and salt solution in rural villages, uptake of this new medicine was opposed by villagers who preferred medicines prescribed by representatives of the local health system. The effectiveness of the sugar-salt treatment (causal realm) created strong positive evidence (situational realm) that led people in thousands of Bangladeshi villages to change their minds and eventually adopt this treatment (problem realm). This intervention fundamentally changed the treatment of diarrhea in both local and public health systems, saving the lives of hundreds of thousands of children (Seelos & Mair, 2017). Alternatively, changes in the causal realm may trigger alternative problematizations. Changing norms and values (causal realm) have recently enticed political decision-makers to publicly problematize the unequal treatment of same-gender marriage. In some countries, this problematization enabled coordinated action to change important aspects of a system’s legal state (situational realm). Finally, understanding variance in problematizations among actors in a system or variance in observed situational characteristics may reveal important aspects of the less observable aspects of a system’s causal architecture. This insight may reveal opportunities and hurdles for effective interventions.
As an example of the application of this analytical scaffold to a socio-ecological system and an illustration of the interplay between the realms, consider an effort to organize system change around the topic of global warming. Because countries objectively vary in their liabilities and vulnerabilities toward the effects of global warming, a local system perspective is crucial even for such a global challenge. Reflecting on the situational realm involves data that indicate the extent, direction, volatility and pace of temperature changes. This focus grounds emotional and often biased disputes over the problématique of global warming (problem realm) in an objective set of facts. Shellenberger (2019) argued that failure to separate the objective facts of a situation from our subjective problematizations fuels misunderstandings and polarization, and stifles effective action. Devising effective change on global warming also needs to consider the causal realm. An objective account of observed temperature changes requires identification of the causal mechanisms that create these situational changes. An interpretation of whether these changes are troubling and ought to be dealt with may depend on one’s cognitive beliefs about whether the causal mechanisms are part of a natural fluctuation of the climate system that is of little concern or are human-made and require intervention (problem realm). This argument illustrates the interplay between assumptions in the causal and problem realms. Decisions about whether and how to organize change may also depend on economic aspects. People differ in their interpretation of whether the costs of global warming are justified by the economic benefits created by activities that generate warming as a side effect. Proposals for organized change also involve normative interpretations of their appropriateness. Slowing global warming can result from either changing behaviour that would otherwise contribute to global warming or from investing in technology that has a cooling effect. Although the results of either option may objectively be the same, different normative mechanisms generate tensions about the appropriateness of each option (Mathiesen, 2015). Finally, causal mechanisms related to power matter when they create imbalances of whose arguments get heard or create asymmetries in who wields control over critical resources that affect social problems. Recently, Ecuador launched a power gamble when it demanded that the global community pay the country for not extracting fossil fuel from a precious national park (Finer, Moncel, & Jenkins, 2010). Failure to pay would result in the destruction of a public good and the extraction of fossil fuels that would negatively impact global warming. This example illustrates how our analytical scaffold captures key elements of organized system change. We visualize the elements of our analytical scaffold in Figure 1.

Elements of the analytical scaffold for organized system change.
This scaffold is a versatile instrument because it seems to be largely issue- and scale-independent. It helps us to understand efforts at global and regional scales, such as in the case of global warming, but also guides research on system change at the scale of villages in rural India around human rights issues. The scaffold guides empirical work by specifying the central areas of attention to organized system change. In the following sections we will discuss in greater detail the three realms that compose our analytical scaffold.
Problem realm of systems
Research on social problems has a long tradition in sociology (Frank, 1925; Loseke, 2003; Merton & Nisbet, 1961; Spector & Kitsuse,1977), but this tradition has largely been sidelined by organizational theorists. We argue that engaging sociological literature on social problems provides a distinct and empirically useful perspective on organized system change. Social problems are defined as the products of social evaluative processes that determine the validity and the priority of social problems vis-a-vis competing problems under circumstances of limited attention capacity (Hilgartner & Bosk, 1988). Because people not only perceive different problems in the same situation but also perceive the same problems differently, claims about the legitimacy and priority of social problems are contested.
On what basis do organizations select social problems of concern when, as Loseke (2003) claimed, ‘there is no necessary relationships [emphasis in original] between any objective indicators (statistics, results of tests) of social problem conditions and what [Americans] worry about, what politicians focus on, or what television, newspapers, or magazines tell us about’ (p. 8)? This perspective focuses attention on claim-making processes: Who claims that a situation is troubling and warrants change, and on what grounds are such claims made?
One example from our own research illustrates this research dimension. Observing the work of Gram Vikas, an organization that constructs water infrastructure and toilets for households in poor villages in Odisha, India, may be framed as one of many programmes under President Modi’s ‘Clean India’ initiative. Under this initiative, millions of toilets were constructed in India to address the social problem of open defecation practices and their negative health and environmental impacts. Many donors that financially support Gram Vikas evaluate the organization’s performance in terms of number of toilets built or improvements in a community’s health state. From this perspective, the performance of Gram Vikas seemed less successful in scale compared to other initiatives that offered free toilets to villages. Closer engagement with the organization and the villages in which it operated, however, revealed other aspects of the system that are causally important but difficult to observe. We identified a concealed agenda by Gram Vikas, where building toilets was merely a distraction for addressing what the organization considered to be a troubling situation of inequality based on high levels of social exclusion of lower-caste people as well as women of all castes (Mair et al., 2016). Privileged and influential villagers did not perceive inequality as a problem and resisted any change that overtly addressed inequality. Even those villagers who were discriminated against accepted their fate as ‘given’ and found comfort in the belief that they would be rewarded for their suffering in another life. The result of change efforts would benefit some villagers but remove valued privileges from others, therefore the latter would not support any change efforts. The variance in how people perceive the identity and priority of problems and the legitimacy of having them addressed creates challenges for those who organize system change. Can they organize change that is technically and culturally feasible and does not trigger levels of resistance that stifle progress? Can they generate alternative situations that people with different roles, status and preferences could live with even if these situations were also not ideal? Gram Vikas learned that people were reluctant to ‘change their minds and hearts’ unless their fundamental material needs were satisfied first. Attending to local priorities around problem claims, the organization pragmatically decided to work on high-priority problems such as health. This decision built momentum for change and established trusted relationships with villagers: As we stuck through thick and thin with them, they began to confide in us. When we started the medical programme—that is the one which started their confidence in us. So slowly they gained faith in us. Then they started telling us about all these problems.
5
This anecdote illustrates an analytically interesting challenge inherent in organized system change: How do actors interpret problem realms to identify effective pathways for engaging with systems that do not close opportunities for change from the start due to conflict and lack of trust? Furthermore, many problems in social systems are difficult to observe. Changing systems around problems that organizations perceive and prioritize may not match local problem perspectives and perceived priorities. An example of this challenge is organizations in the philanthropic and development sectors that have long been criticized for an attitude of Western neocolonialism that defines and frames problems according to templates that are ideologically vetted (Easterly, 2006).
Attending to the problem realm of systems helps researchers trace how organized system change overtly and covertly enables progress, how these actions are legitimized, and how organizations maneuvre the minefields of conflict and resistance due to differences in problem evaluations and priorities. When we explored the question, ‘what legitimizes covertly addressing inequality in villages against their will?’ Gram Vikas shared their criteria for evaluating legitimacy and progress. The organization claimed that no village that they have worked with wanted to reverse the changes once they had experienced them, which usually takes about three years of an intense system change effort. Gram Vikas also observed that more and more women from villages did not allow their daughters to get married to men from villages that were not ‘changed’ by Gram Vikas.
Situational realm of systems
Decades ago, Robert Merton (1976) warned us against falling victim to a naïve subjectivism about social problems and to ignore objective constraints that ‘affect both the choices that people make and the personal and social consequences of these choices’ (p. 22). For example, high levels of illiteracy, pollution, discrimination, or a lack of infrastructure constitute objective realities (situations) that critically effect people’s choices, actions, and relations. Without this factual understanding, social problems may be dismissed as mere social constructions without any objective basis. When Gram Vikas collected health data from villages, it identified malaria as a key disruptor of community health. This opened an opportunity to create goodwill and trust by building a small medical team that offered malaria medication to villagers. Gram Vikas learned to use objective data and factual arguments to overcome the resistance of powerful villagers by ‘de-emotionalizing’ contested topics and framing them in terms of logic. For example, it decided to frame its intervention around inequality as a health intervention. This framing helped justify the inclusion of lower caste people, as well as women of all castes, in Gram Vikas’s programs of building water and sanitation infrastructure. The founder of Gram Vikas remembered: We use water and sanitation as an entry point program, and water and sanitation is maybe the only program where we are able to reach the whole community, the rich, the poor, the literate, the illiterate, the high caste, the lower caste. Because everybody is in need of water.
This framing allowed the organization to extract an agreement for 100% inclusion before Gram Vikas would start working with a village: ‘We say that unless [lower caste households] also have a toilet you will be ultimately eating the shit of those people because the fly which sits on the shit of those people won’t distinguish between you and them.’ Framing the intervention around objective data and logical arguments also enabled the establishment of objective facts. Once a village agreed to 100% inclusion, Gram Vikas then also changed the situation by demanding an upfront payment from every household (where richer households were required to cover for poor ones) or there would be no water infrastructure. This new situation created a sunk-cost constraint that prevented villagers from withdrawing from the programme. This new situation also helped keep the villagers’ attention focused on building a water and sanitation infrastructure. This focus provided objective indicators for progress and distracted villagers from changes with respect to norms, conventions and power relations that the organization orchestrated in the background: There are remarks one still hears from men in villages where work has just begun: ‘Gram Vikas says we must involve women to make decisions, but this is just for RHEP [the name of Gram Vikas’ change program], not for other activities.’ Little do they realize that once the dent is made, the ‘cracks’ open wider and wider. (Mair et al., 2016)
Studying social systems requires that we integrate subjective and objective dimensions of reality and that we explore the relations between the realms. Coupled with this concomitant attention to both realms are evaluations of implicit and explicit system boundary criteria in use by organizations. These evaluations constitute important analytical questions, such as: Where do organizations draw boundaries around situations and problems, and how do they draw boundaries around their responsibilities as agents of change? Which problems and which constituencies are within or outside the scope of an organized system change effort?
Causal realm of systems
The third realm of our analytical scaffold—the causal realm—focuses our attention on the causal mechanisms that generate troubling situations and that therefore are also the targets of organized system change. Differences in causal assumptions generate different propositions about the effectiveness and appropriateness of organized system change. Our earlier example about global warming illustrates this evaluative process. As realist scholars, we are committed to an objective reality and this position necessitates that only some assumptions about the causal architecture of systems are objectively true and can be an effective basis for intended system change. Sayer (2000) justifies this position: I would argue that it is the evident fallibility of our knowledge—the experience of getting things wrong, of having our expectations confounded, and of crashing into things—that justifies us in believing that the world exists regardless of what we happen to think about it. (p. 2)
From this position, realists validate knowledge in terms of its practical adequacy, ‘that is in terms of the extent to which it generates expectations about the world and about results of our actions which are realized’ (Sayer, 2000, p. 43). This realist principle of practical adequacy couples our analytical scaffold with an ambition for third-mandate research to generate knowledge that is relevant for practice.
Establishing criteria for exploring the causal black box of systems is an important avenue for scholarly work on the third mandate. Davis and Marquis (2005) advocated for mechanism-based theorizing to explain organizational phenomena. But the authors also lamented the inconsistencies of defining the concept of mechanism that range from ‘sometimes true theories’ to ‘universal social laws’ (Davis & Marquis, 2005). For this causal realm, we are engaging recent philosophical and methodological advances in mechanism-based theorizing grounded in a realist philosophy of science (Bunge, 2006; Mahoney, 2001; Seelos & Mair, 2014). We side with Bunge (2006) who stated that proper explanation requires identifying the mechanism(s) that make a system of concern tick. Mechanisms are events or processes that change the state of a system, or that sustain the state of a system that otherwise would change. For example, system change in villages organized by Gram Vikas is a complex causal process that unfolds over several years, and this process can be further analysed into its constitutive mechanisms that collectively generate change in villages (Mair et al., 2016). Gram Vikas leverages the power of village elders to appoint women in decision-making bodies in a village. This power mechanism changes the state of decision-making in this village system in a fundamental manner. The requirement of prepayments by all villagers before work is initiated is a mechanism that sustains the state of active participation in building water infrastructure. From experience with earlier prototype villages, Gram Vikas learned that the absence of this mechanism would create too many opportunities for ceasing to collaborate. This analytical work of explicating the causal mechanisms that make systems tick distinguishes description of organized system change from causal explanation.
We agree with organizational scholars that realism provides a productive metatheory for theoretical and empirical work (Baum & Rowley, 2002; Tsoukas, 2000). Realist inquiry attends to the objective and subjective realms of systems as the context in which causal mechanisms are enabled and triggered, and whether and how mechanisms realize their effects in specific situations (Seelos & Mair, 2014). In practice, this analytical work requires a deep engagement with phenomena and efforts to identify the relevant observable and unobservable factors that make a system tick. Explicitly focusing on the problem realm of the villages where Gram Vikas operated enabled us as researchers to probe more deeply into this system’s causal apparatus, whose effects included sustained high levels of inequality, visible in entrenched patterns of how villagers interact with each other (Mair et al., 2016). This deeper causal understanding enabled an exploration of the causal mechanisms underpinning Gram Vikas’s organized system change, which resulted in lower levels of inequality, manifested in altered patterns of social interaction, and persisted well beyond the intervention.
Advancing Research on Organized System Change
In the previous section we argued that attention to both objective conditions and subjective interpretations of social problems and a commitment to specify the (constellation of) mechanisms that constitute the causal apparatus of social systems are critical for advancing knowledge on organized system change. For researchers, parsimony is an important principle of effective empirical and theoretical work on social systems when it provides focus without underspecifying phenomena.
We are not the first organizational scholars to link the study of organizations, social problems and systems. Building our analytical scaffold provides a platform for conversations and debate among scholars who over the years have conducted research within the different system realms and have recently started to integrate realms (e.g. Ansari, Gray, & Wijen, 2011; Bansal, Kim, & Wood, 2018; Dorado & Ventresca, 2013; Etzion, Gehman, Ferraro, & Avidan, 2017; Waddock, Meszoely, Waddell, & Dentoni, 2015). The scaffold also enables conversations about transformative mechanisms of interest to scholars who study how organizations alter and shape institutional arrangements in fields (Hoffman, 1999; Mair & Hehenberger, 2014; Scott, Ruef, Mendel, & Caronna, 2000). Finally, a third-mandate perspective with an explicit focus on organized system change complements contemporary research on organizations that address social problems in two ways. First, a perspective on organized system change shifts the theoretical and empirical attention toward how transformation processes are initiated and orchestrated by organizations and on the intended and unintended consequences of such efforts. And second, this perspective turns societal challenges into objects of analysis rather than treating them as operational context or an unusual or extreme setting for the study of organizations. We believe that rejuvenating the third mandate, infusing the mandate with analytical ambition and clarity, and integrating the study of social systems and social problems will offer a constructive and productive path toward developing a body of scientific knowledge that values truth, proves useful and strives for societal progress (Davis, 2015). Invigorating the third mandate might also help regain the voice and relevance that organizational scholarship seems to have lost over the past several decades (King, 2017). Our analytical scaffold could advance evidenced-based policy making and complement analytical approaches such as randomized controlled trials. Such trials are prevalent in economics to test effects of organizational interventions but can only speculate about transformative mechanisms and processes. As a community, organizational scholars could develop a repertoire of transformative mechanisms that elucidate the causal apparatus of organized system change. In our own work, for example, we have characterized the transformative mechanism of scaffolding that allowed an organization to change a social system from state A (higher levels of social inequalities) to state B (lower levels of social inequalities) (Mair et al., 2016).
Theorizing Opportunities on and Around Organized System Change
Theories are simplifications and capture assumptions about complex realities. Current theorizing in organizational research often relies on ‘uncritical, sanitized and dangerously misleading simplification of messy, complex social phenomena’ (Munir, 2020, p. 1). The research we advocate in this article exposes rather than tames social complexity. Enacting the third mandate along the ontological guideposts we introduce might imply that we put aside the noble aspiration of developing grand theory and more willingly endorse idiographic explanations. Idiographic studies and explanations grounded in concrete situations and particular evaluations of such situations located in time and space allow us to understand what is possible and to uncover the causal tendencies of objects and phenomena in particular contexts (Tsoukas, 1989). When theorizing centres on possibilities rather than universal laws, and on projective tendencies rather than predictable outcomes, we might be able to capture reality more accurately and to generalize knowledge claims more carefully to well-specified contexts (Ruzzene, 2011).
We hope our analytical scaffold encourages a stronger and more productive focus on mechanism-based explanations in organizational research (Davis & Marquis, 2005) and a curiosity about mechanisms associated with change in social systems discussed in adjacent disciplines (Stephan, Patterson, Kelly, & Mair, 2016). Opportunities for theoretical work on organized system change are ample. Here we briefly refer to the issues of boundaries and power as particularly promising foci for future theorizing.
Social systems as complex social realities do not have obvious boundaries. Yet, organized system change requires continuous attention and decisions to define and refine system boundaries to provide scope, scale and focus. These decisions include temporal and spatial aspects and vary greatly among phenomena. For example, Gram Vikas expects to intervene in a village for a period of three years (Mair et al., 2016), whereas the organization Sekem operates from a 200-year plan for enacting a holistic transformation of Egypt in its economic, cultural and social spheres (Seelos & Mair, 2007a). Gram Vikas targets small-scale societies in the form of rural villages in India; Sekem targets the entire nation of Egypt. Understanding how organizations reflect, decide and act on boundary criteria is an important opportunity for future theorizing. Knowing how boundaries get defined, and by whom, helps translate an unbounded social system into a concrete phenomenon for investigation. Over the past decades organization theorists have made important advances in understanding boundary work and the symbolic aspects of this work. (See Langley, Lindberg, Mørk, Nicolini, Raviola, & Walter, 2019, and Lamont & Molnár, 2002, for reviews.) In parallel, and based on a tradition of reflective professional practice, system theorists have developed categories of boundary judgements that include (a) boundary decisions about purpose and legitimacy, (b) boundary decisions about governance and control, and (c) boundary decisions about knowledge and expertise (Ulrich, 2014). Integrating boundary judgements—how an organization carves out its own system of concern or those aspects of the real world it considers relevant—is highly relevant for studying organized system change. For example, boundary decisions about purpose and legitimacy involve decisions about the intended beneficiaries. How do organizations justify and legitimize benefiting some and not others? Addressing these questions will also be helpful for examining contestation over means and ends in organized system change projects.
Organized system change is intractably related to power. Adding a system lens can advance existing research on organizations and power in important ways. For example, integrating insights from soft- or critical-system perspectives (see Jackson, 2003 for an overview) helps to shift the theoretical attention on power to pursue personal or organizational goals (Pfeffer, 1981) toward power as an attribute of social systems. Critical and emancipatory system perspectives (Jackson, 2003; Ulrich, 2014) connect power in systems with how boundary decisions are made by organizations, as well as the nature of social boundaries that include and exclude people in social systems. This lens complements and informs efforts to unpack the causal apparatus underpinning systemic forms of power that naturalize domination and that are largely unseen and often unobservable (Hehenberger, Mair, & Metz, 2019). As Munir (2020) pointed out, when studying grand challenges institutional theorists are ‘inclined to overlook larger structures of domination in favor of focusing on smaller, more manageable issues’ (p. 1). Applying our analytical scaffold and integrating insights from the literature onto soft systems might help to untie the knot that has limited progress on systemic forms of power in organizational and management research (Fleming & Spicer, 2014).
Implications for researchers and for practice
How do we define the boundaries of our scholarship? Boundary decisions are implied in selecting the phenomena that we study and how we go about studying them. Boundary criteria include time, space and ambition. For example, we may decide to study organized system change in contexts that are familiar and well understood or in settings that require us to develop new templates to guide observations and interpretations. These decisions reflect implicit boundary criteria such as how frequently we need to gather data and over which timescale, whether we can gather data remotely or are required to spend time closely engaging with an unfamiliar social system, or whether we aspire our research to illustrate and contribute to existing theories or ambitiously penetrate the causal black box of complex change efforts to develop new theory. These boundary decisions define our roles as scholars, the nature of contributions we can possibly make and the evolution of third-mandate research for organizational science.
We deliberately centred third-mandate research on organized efforts to change social systems because we perceive this focus as a timely opportunity for organizational scholarship. System change has become a popular buzzword in the world of philanthropists, impact investors, policy makers and development organizations. Current efforts such as those led by Co-Impact reflect funders’ ambitions and confidence that the change they desire can be organized and will lead to progress. However, organized system change often lacks a deep engagement with the three realms of systems: problem, situation and causal (Seelos, 2020; Seelos & Mair, 2018). Simplifying assumptions about social reality increases the likelihood that interventions will be prematurely stopped when the expected outcomes are not realized within anticipated time frames. Decision makers may not be conscious of the tendency of failed and prematurely aborted interventions to deeply affect and change social systems. Therefore, even well-intended interventions may have severe unintended and undesired consequences for the people inhabiting the social systems they target. We can therefore not separate third-mandate research from an explicit inquiry into the values and responsibilities of those who are involved in organized system change and those who examine their efforts. The analytical scaffold we present does not offer prescriptions. Yet, it is intended to help decision makers explicitly reflect on the choices they make and most of all to provide a useful scaffold for their own learning journey. For example, reflections on our framework might encourage funders to invest in a deeper understanding of the causal architecture of the system they intend to change instead of designing and implementing solutions based on experience from unrelated contexts. To intervening in a system it is necessary to learn about it and to understand how the system operates. However, intervening without an explicit effort of learning is irresponsible and not conducive to progress (Seelos & Mair, 2017). Doing research on organizations that intervene in social systems requires that we learn from and with these organizations, and that we do so without drinking their Kool-Aid or falling in love with their promotional stories. Paying attention to the three system realms helps to corroborate our own interpretations and to trace the trajectory of how objective conditions and subjective problematizations change over time. This attention also helps to unpack the mechanisms that underpin transformative processes over time.
To conclude, the potential for third-mandate research, and especially for research that focuses on organized system change, might be unnecessarily constrained if we limit ourselves to narrow theoretical perspectives. Equally, there exist no formulas or templates for writing theory on organized system change. All we have are scholarly ambition and curiosity and a commitment to truthful explanation.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
