Abstract
A commonly held view is that strategic organization theories progress as a result of a Popperian process of bold conjectures and systematic refutations. However, our field also witnesses vibrant debates or disputes about the specific assumptions that our theories rely on, and although these debates are often decoupled from the results of empirical testing, changes in assumptions seem closely intertwined with theoretical progress. Using the case of the resource-based view, we suggest that progressive change in theories of strategic organization may come about as a result of scholarly debate and dispute over what constitutes proper assumptions—even in the absence of corroborating or falsifying empirical evidence. We also discuss how changing assumptions may drive future progress in the resource-based view.
How does progressive change come about in theories of strategic organization? The perhaps dominant view in strategy, reflected in doctoral training, is the fundamentally Popperian (1959, 1963, 1994) notion that theories change as a result of empirical testing (see, for example, Bacharach, 1989; Camerer, 1985; Carlson and Hatfield, 2004; Miller and Tsang, 2010; Mizruchi and Fein, 1999; Montgomery et al., 1989; Newbert, 2007; Newbert et al., 2014; Powell, 2001). However, this view misses out on other possible sources of progressive change in theories of strategic organization. First, change may also be brought about by extra-scientific drivers, such as changes in the power or status of those who support (or oppose) specific theories (Ferraro et al., 2005). Second, strategic organization scholars may gravitate from one theory to another theory, modify an existing theory, or come up with a new theory because the new theory makes fewer and more elegant assumptions, aligns better with other theories, and can explain a larger set of phenomena. Thus, changing assumptions can drive progressive change. By “assumptions,” we mean what is in ordinary language covered by both “assumption” and “premise,” fundamental statements that may be both given (and typically seen as exempt from direct empirical testing; Lakatos, 1970) and non-given (i.e. that can be tested, easily changed, etc.). Assumptions are typically made about the fundamental units of analysis (e.g. individuals), how these units are related (interaction assumptions), and what are the consequences of such interaction (aggregation assumptions). Thus, assumptions are a key part of the skeleton of theorizing, and are fundamental to deriving explanations and predictions (see Mäki, 1994; Musgrave, 1981; Nagel, 1963: 211).
We develop and illustrate the point that progressive theoretical change may happen as a result of changes in assumptions, using examples drawn from the resource-based view (RBV) (Barney, 1991; Peteraf, 1993; Peteraf and Barney, 2003). 1 We suggest that progressive change in the RBV (i.e. expansion of the explanatory scope of a theory or improvement of its casual articulation) has largely come about by a process of refining assumptions, making the theory better capable of addressing hitherto imperfectly understood phenomenon. This has been accomplished by drawing on economics, the dominant source discipline of the RBV (Lockett and Thompson, 2001). For example, progress has been accomplished by refining the existing economics-based understanding of strategic factor markets by the use of information economics (e.g. Makadok and Barney, 2001). We refer to this as assumptional deepening. However, strategic organization scholars may also further theorize by broadening the assumptional base of theory, by drawing on other disciplines than the relevant source discipline. For example, psychologically grounded insights in managerial cognition may be brought to bear on the understanding of resource attributes and resource value (Danneels, 2011). Or, motivational psychology insights may be added to the understanding of the value-creating potential of human resources with a RBV framework (Gottschalg and Zollo, 2007; Lindenberg and Foss, 2011). We call this assumptional broadening, which we conjecture will be a major source of theoretical progress in the RBV.
Progressive change in theories of strategic organization
Progress
The production of novel conjectures that withstand rigorous testing has traditionally been seen as the main criterion for good strategic organization research (e.g. Bacharach, 1989; Camerer, 1985; Miller and Tsang, 2010). However, there are other criteria of good research. For example, Van de Ven (1989) argues that good theory is practical “because it advances knowledge in a scientific discipline, guides research towards crucial questions, and enlightens the profession of management” (p. 486). Weick (1995) points to the importance of generative capability of new theory by emphasizing the process of theorizing and creative activities such as “abstracting, generalizing, relating, selecting, explaining, synthesizing, and idealizing” (p. 389). In an attempt to more generally summarize accepted characteristics of good theory, Kuhn (1998: 103) states that good theory should possess predictive accuracy, be logically consistent in the sense of internal consistency as well as consistency with other currently accepted theories, be simple in order to bring order to complex and originally disjoint patterns, have a broad scope in terms of generating predictions that extend beyond individual observations, and possess the potential for generating new research findings in terms of predicting and explaining previously unobserved phenomena or relationships.
An implication of the above broad views of the characteristics of “good theory” is that theories may change in a progressive way, not as a result of direct empirical tests of the theory, but because scholars modify an existing theory, or come up with a new theory that makes fewer and more elegant assumptions, aligns better with other theories, and can explain a larger set of phenomena. Making novel assumptions is associated with progressive change when the novel assumptions make the theory more general (it applies to a larger set of domains, that is, scope expansion) or increases the causal grain of the theory (it becomes more accurate in terms of explanation and prediction, that is, causal articulation), or both (see Kitcher, 1993; Mäki, 2004). Thus, explanatory progress may come in the form of “erotetic progress, that is, progress in explanatory questions” that are posed or enabled by the theory in question (Mäki, 2004: 320). Such progress may be of two kinds (Mäki, 2004). First, explanatory progress can occur as a result of scope expansion linked to increased unification, that is, expanding the number of phenomena that can be explained by the theory. Second, explanatory progress may also occur as a result of increased causal articulation, that is, increasing the causal “grain” of explanations. This form of progress primarily refers to a change in the theory’s set of explaining items that allows it to provide a better, or more detailed, causal account of a specific phenomenon.
Assumptions
Assumptions are of critical importance to theory since they ultimately determine a theory’s explanatory and predictive accuracy, logical consistency, explanatory scope, simplicity, and ability to generate new explanations and predictions. Theoretical discussion typically concerns the explanatory and predictive consequences of an assumption, and the question of against what explanatory aim the assumption should be evaluated. Assumptions come in many forms and can be dimensionalized in a number of ways. The overarching dimension is that of abstraction.
The role of simplifying theoretical assumptions is, of course, to make a complex reality cognitively, epistemically, and computationally manageable (Friedman, 1953; Krajewski, 1977; Mäki, 1994). Examples of unrealistic social science assumptions include (perfectly) rational economic man, symmetric information, and competitive markets. While management scholars may make less use of theories based on blatantly unrealistic assumptions than scholars in other disciplines (e.g. economics), such assumptions can certainly be found in our fields. A typical dividing line is between those who advocate realism and those who advocate simplicity and clear test implications (e.g. Friedman, 1953; Shugan, 2007). Rich discussions in strategic management (Foss and Hallberg, 2014; Lam, 2010; Tsang, 2006), economics (Foss and Foss, 2000; Friedman, 1953; Hallberg, 2015; Machlup, 1955; Mäki, 1992, 1994, 2000; Nagel, 1963; Samuelson, 1963), marketing (Shugan, 2007; Tsang, 2009), and sociology (Fararo, 1992; Lindenberg, 1992) reflect these different positions.
In general, assumptions involve representing real phenomena in “stylized” or “idealized” ways by means of suppressing some, potentially relevant, real phenomena, while highlighting others (Dubin, 1978; Krajewski, 1977; Poole and Van de Ven, 1989; Whetten, 1989). Hence, theorists abstract from reality and create theories by means of what they choose to include as explained and explaining factors. Thus, the precise form that theoretical abstraction takes is regulated by the assumptions made in theory building.
Three forms of assumptions
Assumptions can take three different forms (Musgrave, 1981). First, theorists use assumptions to abstract from specific factors that they think have a negligible effect on the explained phenomenon (negligibility assumptions). Second, assumptions may also be used to specify the domain of application of a theory, that is, its “boundary conditions” (domain assumption). Third, assumptions may be used to temporarily simplify theory building. For example, a theorist might temporarily retain an assumption that has been shown to be false in order to simplify the explanation and make the theory more tractable (heuristic assumption). Economists frequently use such heuristic assumptions. For example, few economists would claim that individuals are literally capable of forming “rational expectations” (i.e. forecasts based on the correct econometric specification of the economy) (Muth, 1961); yet, this assumption simplifies economic modeling greatly. These three forms of assumptions jointly determine the depth and breadth of theories, that is, the scope and causal articulation of theories.
Assumptions and progressive change in the RBV
Assumptions are inherently contestable. Scholarly debates are often the result of different conceptions of the location of two boundary lines drawn up by the particular assumptions that are made in a theory: (a) between a theory’s current explaining items and all other potentially explaining items and (b) between a theory’s current explained items and all other potentially explained items (Mäki, 2004). In turn, the location of these boundary lines is determined by the negligibility, domain, and heuristic assumptions that scholars make (Musgrave, 1981). To make these distinctions concrete, consider Table 1, which maps key assumptions of the core RBV model and how these have changed in later papers.
Changing assumptions in the RBV.
RBV: resource-based view.
The RBV makes negligibility, domain, and heuristic assumptions on a number of different domains to which the theory applies, namely, the three domains of factor markets (A), internal organization (B), and product markets (C). Table 1 lists key assumptions made in the core RBV contributions, highlighting the assumptions that are necessary for the existence of sustained competitive advantage, and shows how these assumptions have been changed or modified in later RBV papers.
The RBV is an interesting case in the context of the assumption-centered framework laid out in the previous sections. First, it is often taken to be strategy’s dominant approach (Acedo et al., 2006; Newbert, 2007; Peteraf et al., 2013; Powell, 2001; Priem and Butler, 2001). Second, although it is characterized by some heterogeneity and has somewhat fuzzy boundaries to other views and perspectives (e.g. it is overlapping with the knowledge-based capabilities, core competence, and dynamic capabilities views), it is possible to identify and characterize an analytical core in the RBV (i.e. Barney, 1986, 1991; Lippman and Rumelt, 1982; Peteraf, 1993). Third, the RBV has drawn much nourishment from the soil of economics, notably price theory (e.g. Alchian and Allen, 1969), the theory of the firm and contract theory (e.g. Alchian and Demsetz, 1972; Hart, 1995; Williamson, 1996), information economics (e.g. Akerlof, 1970; Hayek, 1945), cooperative game theory (e.g. Neumann and Morgernstern, 1944), and financial economics (e.g. Jensen and Meckling, 1976). As such, it illustrates how drawing assumptions from a key source discipline may be both enabling and constraining. Fourth, the RBV also provides an effective illustration of the role of assumptions in the process of theoretical progress. Thus, perhaps because of its somewhat heterogeneous character and the different kinds of economics it has drawn from (see Acedo et al., 2006; Newbert, 2007; Peteraf et al., 2013), the RBV has to a large extent progressed because scholars have continually revisited and modified the basic assumptions of the theory.
In the following, we dig deeper into the assumptional changes that have been proposed in later RBV papers. Thus, we discuss specific patterns of how assumptions in the core RBV papers have changed in later papers, such as assumptional deepening and assumptional broadening, and how these patterns may lead to progressive change in terms of expanding a theory’s scope and better articulating the causal mechanisms proposed by the theory; see Table 2.
Examples of changing assumptions and progressive change in the RBV.
RBV: resource-based view.
Assumptional deepening
So far, most changes in the assumptional base of the RBV have been in the form of assumptional deepening through refining existing assumptions (e.g. being more precise about asymmetric information conditions) and by adding new ones, usually grounded in economics.
Scope expansion: product markets
The perhaps best example of how assumptional deepening has led to progressive change in the RBV through scope expansion is how original idealizing assumptions about competitive product markets in the core RBV model have been relaxed in later papers building on cooperative game theory. According to Peteraf and Barney (2003), the assumption of frictionless competition in product markets (i.e. domain C in Table 1) in early RBV papers (e.g. Lippman and Rumelt, 1982; Peteraf, 1993) may be seen as a kind of ceteris paribus assumption aimed at holding other (product market frictions) factors constant. This ensures that RBV predictions can be examined in a clean manner and uniquely linked to factor market imperfections, the latter essentially being nullified. However, while such assumptions may be justified for reasons of tractability and simplicity, they may also pose threats to the robustness of the theory if the effects of the idealized or omitted variables are, in fact, not negligible. The product market side of the RBV has been addressed in a string of cooperative game theory papers (e.g. Brandenburger and Stuart, 1996, 2007; Chatain, 2014; Chatain and Zemsky, 2011; MacDonald and Ryall, 2004; etc.). In the pioneering contribution to this stream, Brandenburger and Stuart (1996) deliver new insight into how heterogeneity at different stages of the value chain affects firms’ chance of capturing value (and how the prospect for capturing value in turn affects the possibility for value creation). MacDonald and Ryall (2004) analyze how the structure of product markets affects a firm’s chance to appropriate value from, for example, a resource-based advantage.
A second line of research explicitly breaks with the assumption that product market influences can meaningfully be nullified. This stream addresses product market assumptions in the RBV related to the potential endogeneity and heterogeneity of demand (Adner and Zemsky, 2006; Priem, 2007; Priem and Butler, 2001). The origin is Priem and Butler’s (2001) point that nullifying product market influences is problematic because the value creation potential of resources is ultimately determined by demand-side characteristics, so that assuming competitive product markets implies the risk of inaccurate estimates of resource value.
Causal articulation: strategic factor markets
Important progress has also been made in the analysis of strategic factor markets (i.e. domain A) with respect to relaxing and clarifying assumptions. For example, Denrell et al. (2003) argue that idiosyncratic information as implied by firm-specific capabilities combined with incomplete markets creates new insight into strategic factor market opportunities (compared to Barney, 1986). The added causal articulation stems from assuming that firm-specific capabilities are a source of superior insight in the value of complementary resource bundles. Adegbesan (2009) develops a coalitional game theory model that shows that firms can profit from engaging in strategic factor markets even in the absence of superior expectations about resource value. This can be the case when acquiring firms exhibit superior complementarity to target resources in strategic factor markets. Additional explanatory power stems from articulating in more detail assumptions about complementarity between resources.
Assumptional broadening
There is, however, a different form of change in the assumptions of a theory that may also be an important driving force behind progress. Specifically, increased explanatory scope and causal articulation may also come about by drawing on assumptions and explanatory mechanisms grounded in other than the traditional source disciplines of a specific field, a pattern of theory development that we refer to as assumptional broadening. In the case of the RBV, the traditional source discipline is economics, while potentially under-exploited disciplines may be psychology or sociology.
Scope expansion: the psychology and sociology of organizing resources
Lippman and Rumelt (2003) model firms as bundles of complementary, co-specialized resources. They argue that the “heart of business management and strategy concerns the creation, evaluation, manipulation, administration, and deployment of unpriced specialized resource combinations” (Lippman and Rumelt, 2003: 1083). While identifying new resources and discovering novel ways of combining resources are important avenues to superior performance, the motivation and cognition of the human resources in the firm bundle may also significantly affect competitive advantage. For example, Baron and Kreps (1999) point to the importance of achieving internal fit between human resource practices so that these are perceived as mutually reinforcing and consistent. The importance of consistency is grounded in the insight from psychology that “messages are more salient and recalled better when the multiple stimuli being transmitted are simple and support the same theme” (Baron and Kreps, 1999: 33).
Huy (2012) also points to realistic assumptions about human cognition and emotions as an important topic for future research in strategy. According to Huy (2012), insight from cognitive and social psychology may be crucial for a better understanding of how emotional responses develop and influence performance in small and large groups. He even suggests that “emotional capability” can be a source of competitive advantage and that strategy scholars should develop more rigorous theory around this theme. Gottschalg and Zollo (2007) pursue a similar agenda by linking a realistic and psychologically grounded theory of human motivation to an economic understanding of organizational performance. Specifically, they suggest that strategic management and theories of competitive advantage should be complemented with the concept of “interest alignment rents” which they define as “the improvement in organizational performance due to changes in employee behavior obtained through increased interest alignment that accrues to the organization, net the costs required to establish interest alignment” (Gottschalg and Zollo, 2007: 419). As another example, Lindenberg and Foss (2011) contribute to the microfoundations of organizational performance by drawing on goal-framing theory to better explain how motivation for joint production can be explained in terms of cognitive/symbolic management and organizational design. Finally, Bridoux et al. (2011) draw on social psychology and behavioral economics to analyze employee incentives to engage in cooperative behaviors and explore how this impacts value creation based on human resources.
The scope of the RBV can also be extended by highlighting, not so much the unique position brought about by unique resources, but rather how firms can gain advantages by virtue of their network positions (Burt, 1992). Thus, Zaheer and Bell (2005) linked insights from network theory (Burt, 1992) and the RBV. Specifically, they posited a complementarity between network structure and internal capabilities, such that “firms with superior network structures may be better able to exploit their internal capabilities and thus enhance their performance” (Zaheer and Bell, 2005: 809). Based on an empirical setting of Canadian mutual fund companies, they discerned the hypothesized complementarity in the data, and also found that innovative firms that also bridge structural holes get an additional performance boost.
Causal articulation: the psychology resource access
The basic assumption in the factor markets literature is that asymmetric information drives the differential expectations of actors (Barney, 1986; Denrell et al., 2003; Makadok and Barney, 2001). While actors on these markets are claimed to hold rational expectations concerning future resource value (Amit and Schoemaker, 1993) (i.e. they hold the same supply and demand model of the relevant factor market), they differ with respect to the information regarding resource value that they hold (Muth, 1961). When such information differs, there may be an opportunity for acquiring resources that are priced below their value to the acquiring firm.
However, cognitive and social psychology strongly suggests that valuations may differ even though individuals hold exactly the same information (Kahneman, 2003; Smith, 2008). In particular, psychology insight in confidence and optimism are potentially relevant to understanding factor market behaviors (e.g. Camerer and Lovallo, 1999; Capen et al., 1971; Dyer and Kagel, 1996; Thaler, 1988, etc.). As already suggested in the original statement of strategic factor market reasoning (Barney, 1986), many factor markets take the form of auctions and are therefore subject to the Winner’s Curse (i.e. overbidding caused by excessive confidence and optimism).
Amit and Schoemaker (1993) suggest that firms acquiring resources on strategic factor markets partly do this based on differential scenarios of possible futures. However, the literature suggests that scenarios are subject to a number of cognitive biases (Kahneman, 2003). In practice, this suggests that factor market expectations are “contaminated” by a host of non-economic considerations that are best understood through a psychology lens. Scholars address these questions in the emerging behavioral strategy field (see Powell et al., 2011). A key question here is how biases influence decision-makers’ efforts to gather information about the firm’s current resources. For example, Garbuio et al. (2011) describe this in terms of a resource structuring process subject to heuristics and biases such as the endowment effect, the familiarity effect, extremeness aversion, and endowed anchoring. Such resource analysis is of course also directly related to the specific abilities of each firm to accurately comprehend its competitive landscape and rivals’ perceived position within this landscape (Tsai et al., 2011).
Concluding discussion
An influential view of progress in strategic organization research is the combination of the hypothetico-deductive model and Popperian falsificationism. We argue that this view of progress should be complemented by the notion that theoretical progress may also result from changing a theory’s assumptions as a result of scholarly debate, quite independently of any empirical testing. In other words, while assumptions may be revised based on the results from empirical testing, assumptions may also change because scholars find the existing assumptional base of a theory wanting in various respects (e.g. the criteria for good theory suggested by Kuhn, 1998: 103). Thus, we point to a neglected mechanism that may drive progressive change even in the absence of empirical testing. Scientists may thus gravitate from one theory to another theory, modify an existing theory, or come up with a new theory because the new theory makes fewer and more elegant assumptions, aligns better with other theories, and can explain a larger set of phenomena.
An implication of our argument is a call to shift focus somewhat in doctoral training and journal publication policies from an arguably overly strong focus on empirical method and data to an emphasis on assumptions, disciplinary origins, and the potential benefits of combining insights from neighboring disciplines in a focused, rigorous manner. However, there are also dangers and pitfalls associated with more extreme forms of assumptional deepening and broadening. Thus, pushing a research agenda based on assumptional broadening too far might carry risks such as a loss of much of the simplicity, consistency, and elegance that a theory may obtain by virtue of relying mainly on one discipline. There are also corresponding pitfalls associated with assumptional deepening, which the reliance on economics in much strategy research may illustrate. Economics has in many respects completely different explanatory aims than management (e.g. market equilibrium/resource allocation vs firm heterogeneity/competitive advantage), so what is a fruitful assumption in economics is not necessarily a fruitful assumption in management (and vice versa).
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.
