‘OL Vols No Lae IS THE RESOURCE-BASED “VIEW” A USEFUL PERSPECTIVE FOR STRATEGIC MANAGEMENT RESEARCH? YES JAY B, BARNEY ‘The Ohio State University Hare examine each of the major issues relsed by Priem and Butler (thio iseue) cbout ‘my 1881 articie and eubsequent resource-besed research. While it turas out thet Priem cand Buller's direct erlticisme of the 1881 article are uniounded, they do remind resource-based researchers of some important requirements of this kind ol research. Taleo discuss jome important Sssues not raised by Priem and Butler—the resolutions fof whick will be necessary if « more complete resource-boved theory of strategic advantage is to be developed, Priem and Butler's (this issue) critique of my 1991 journal of Management article raises sev- eral important issues, about both the article and subsequent developmenis in the resource-based view (RBV) of the firm. While I disagree with most of these authors’ criticisms, they clearly provide a service by creating a forum within which the creation, development, and future of resource-based models of competition can be discussed and debated. Priem and Butler's criticisms fall into four broad catagories: (1) that the resource-based theory I develop in the 1981 paper is tautologi- cal, (2) that my argument fails to acknowledge that many different resource configurations could generate the same value for firms and, thus, would not be sources of competitive ad- vantage, (3) that the zole of product markets is underdeveloped in the argument, and (4) that the theory developed in the atticle has limited prescriptive implications. Idiscuss each of these criticisms in turn, At the end of this response, I also discuss several important issues in the Held of strategic management that are ad- Gressed neither in the 1981 paper nor in subse quent resource-based work. These issues, I think, constitute part of the research agenda that resource-based and other theorists must ad- Comments and suggestions from Asli Arikan, Valentine Della Corte, Konetamtina Kicusis, Michae! Leiblein, Doug ‘Miller, Mike Pong, Mauro Serelli, and Heli Wang have bbeon helpful in writing this article. 1 began writing this crticle whiie visiting the Markoting Deparment at Boconai University in Milon, ftaiy. Iam grateful for the space and Intellectuel elimave I was provided there. “ eee dress if the field of strategic management is te continue to progress. THE TAUTOLOGY CRITIQUE Priem and Butler's first and, in many ways, most important critique of the 1991 article is that the RBV presented ‘s tautological—that irs pri- mary assertions ate true by definition and, thus, not subject to empirical test (Williameon, 1989) Following Bacharach (1989), the authors attempt to demonstrate the tautological nature of the 1291 argument by substituting the definitions of value, rarity, and strategic advantage giver there into what they characterize as one of the cent:al empirica! assertions of the RBV: only valuable and rare resources can be sources of competitive advantage. The assertions thus de- rived are clearly tautological. However, the fact that Priem and Butler are able to restate parts of the 1991 argument in ways that make it tauto- logical is not the same thing as demonstrating that the argument is, ia fact, tautological. It is important to recognize that, at this defini- tional level, all strategic management theories are tautological ‘in the way Priem and Butler describe. For example, Porter's (1880) assertions bout the relationship between industry attrac- tiveness and firm performance can be reduced to tautology by observing that firms in attractive industries will outperform firms in unattractive industries and by defining industry attractive- ness in terms of the ability of firms to perform weil. Transaction cost economics also can be reduced to tautology: hierarchical forms of gov- emance will replace market forms of gover-2 Academy of Manogement Review nance when the costs of market governance are greater than the costs of hierarchical gover- nance. Indeed, this is known as the Coasian tautology. Thus, the ability to restate a theory in ways that make it tautological provides n0 in- sights about the empirical testability of the the- ory whatsoever. Of course, the critical issue is not whether c theory can be restated in such way as to make it toutological—since this con always be done— but whether ct least some of the elements of that theory have been parameterized in @ way that makes it possible to generate testable empirical assertions. For example, Porter's theory is clearly not tautological since he specifies the conditions that make an industry more or less attractive independent of the periormance of firms in that industry. Porter parameterizes in- justry attractiveness through the well-known, ve forces” framework, ct parameterization that enables Porter to mcke empirically testable as- sertions of the form, firms operating in industries characterized by high rivalry, high threat of sub- stitutes, high threat of entry, high buyer power, and high supplier power will pertorm at « lower level than firms operating in industries without these attributes. In a similar way, Williamson (1975) parame- terizes the attributes of transactions in ways that make it possible to specify conditions under which the costs of market governance will be greater than the costs of hierarchical gov nance. Willicmson has explored several ve: sions of this parameterization, but the most crit- ical transaction attribute he has identified seems to be transaction-specific investment. This porameterization enables Williamson to make empirically testable assertions of the form, transactions characterized by high trans- be less costly to "Moreover, because @ theory is tautological does ast ‘mean that it might not be insightful and even empizically {rojtful. For exarmple, cll game theoretic models are tauto- logical in the sense that the hypotheses they generate are completely determined by the assumptions adopted in the modele and the laws of mathematics applied to these 2i- sumptions. However, these tautological modals can some- ‘mes goncrate quite counterintuitive insights that cam, in principle, load to important empirical research. Again, the fscuo is not tautology, por se, but, rather, whether the prop- sitions derived from a tautology can be paremetrized in c way that makes empirical testing possible. Jemucry manage through hierarchical governance than through market governance. ‘Thus, the real theoretical chelienge presented by Priem and Butler is not "Can the RBV pre- sented in the 1691 paper be restated in a way that makes it :autological?” but is, rather, “Are some aspects of this resource-based theory pa- ramererized in ways that can generate testable hypotheses?” In the next sections I examine the extent to which each of the components of this resource-based theory are parameterized in ways that can generate testable propositions. Parameterizing Value Ctearly, of cil the theory elements in the 1981 article, the value variable is the least fully pa- rameterized. This is because, as Priem and But- ler correctly observe, the determinction of the value of a firm's resources is exogenous to the resource-based theory presented in the 198! az- ticle. In fact, the exogenous nature of value de- termination is stated in the 1981 article: ‘These environmenial models help isolate those firm attributes that exploit opportunities and/or noutralize threats, and thus specify which fi attributes can be considered as resources. The resource-based model then suggests what addi- tional charactoristics that these resources must possess if they aze to generate sustained corpet- tive advantage (Barney, 1991: 100; emphasis added). Since the determination of the value of a re- source is exogenous to the argument presented in the 1991 article, it is not surprising that the conditions under which resources will and will not be valucible are not fully specified there. ‘That said, it would be inappropriate to sug- gest that the 1991 aricle fails to give at least some guidance as to how the value of « resource can be determined. In particular, the article in- dicates that resource value must be determined by models of the competitive environment within which a firm competes. Indeed, since 1981, work has continued on using these kinds of models to estimate resource value. This work falls into wo large categories: (1) efforts to use structure-conduct-performance (S- C-P; Bain, 1956)-based theories to specify the conditions under which different firm resources will be valuable and (2) efforts to determine the value of firm resources that apply other theories Conwicht © 2001. All robe reserved: = cs200 Barney 4 derived from industrial organization models (UO) of perfect and imperfect competition (Con- ner, 1991). In my own work I acknowledge the insights that can be generated from applying the S-C-P framework to understanding the value of firm attributes (Barney, 1991: 100), but I have focused more on non-S-C-P-based theories of the value of firm attributes. Consider, for excmple, my 1997 discussion of the ability of cost leadership strategies to gen- erate sustained competitive advantages @ar- ney, 1997: Chapter 6). I begin this discussion by describing several firm attributes that may be associated with cost leadership (eg., volume- derived economies of scale, cumulative volume- derived learning curve economies, policy choices, and so forth) and then show how these attributes can generate economic value in at least some market settings. The logic I use to demonstrate the value of these attributes is mazket structure logic that is consistent with traditional microeconomics (see Figure 6.4 in Barney, 1997), Only after identifying the condi- tions under which cost leadership can generate ‘economic value do I tum the discussion to the conditions under which cost leadership can be « source of competitive advantage (i.e. rare) and sustained competitive advantage (ie., rare and costly to imitate). Nor am i the only researcher that has followed up on the suggestions in the 198} article for how to value firm resources. Theoretically, progress on this front can be found in Leonaré-Barton (2982), Barney ard Hansen (1994), McWilliams and Smart (1985), end Hunt (1997, 2000), among others. Empirically, two of the papers cited by Priem and Butler (ie., Brush & Artz, 1899, and Miller & Shamsie, 1996) are important precisely because they address the value of resources question. Additional empirical work has been done by Barnett, Greve, and Park (1894), Makadok (1998, 1999), Poppo and Zenger (1998), and mony others. in ail high-quality resource- based work, researchers must begir: by address- ing the value of resources with theoretical tools that specify the market conditions under which different resources wil! and will not be valuable. Although additional work is required, I believe ‘we are developing a more complete understand- ing of these conditions. ‘Thus, although the value variabie in Barney (1981) is not fully parameterized, in the article there is recognition of the importance of doing Copyright: ©2004; Albtights reserved, «=e -« this and even a suggestion of some ways might be done. While, strictly speaking, Priem cand Butler's critique does not directly apply to the 1991 argument, it does apply to resource- based theorists who have tried to examine the implications of resource-based logic without considering the market conditions under which @ firm's resources will and will not be valuable. Indeed, If I were to write the 1991 article today, ! would definitely enhance the discussion of value along the lines outlined here. The bri discussion of value in the 199i article could have indicated to some that determining the value of resources is less important than deter- mining the rarity and imitability of resources—« int of view with which I clearly disagree. Porameterising Rarity Prim and Butler also suggest that the term rare is not parcmeterized in the 1891 article and, thus, that ony assertions including “rare” must be tautclogical. I certainly agree that since the concept of rarity is act exogenous 1o the RBY developed in the 1991 paper, if rare was aot parameterized in that article, then aay assertions made with this term must remain tau- tological. However, in fact, rare is perameterized in the 1991 article. Although this parameteriza- tion is aot as complete as ! would like, it is nevertheless specific enough to generate om ically testable assertions. Toe parameterization of rare is discussed in the last paragraph of she section titled Rare Re- How rare a valuable firm resource must be in order to have the poteatial for generating « com- petitive advantage is a difficult question ...1n general, as long os the number of firms that pos- sess c particuicr valuable resource... is less than the number of firms needed to generate por fect competition dynamics in an industry ... thet resource has the poiential of generating a com- petitive advantage (Barney, 1981: 107)” Of course, a complete parameterization of rare would enable a researcher to specify the maxi- * As will bocome clear later, [with Chad not used the term industry in this parametrization of the concept of rarity Rather, I should have focused simply on the number of firms that must possess ¢ resource wn order to generate perfect competition dyacmics, Independant of whather those firms operated in c particular industry