an examination into the causal logic of rent generation ... · strategic management journal strat....

Strategic Management JournalStrat. Mgmt. J., 22: 907–934 (2001)

DOI: 10.1002/smj.174

AN EXAMINATION INTO THE CAUSAL LOGIC OFRENT GENERATION: CONTRASTING PORTER’SCOMPETITIVE STRATEGY FRAMEWORK AND THERESOURCE-BASED PERSPECTIVE

YIANNIS E. SPANOS1* and SPYROS LIOUKAS2

1Athens University of Economics and Business, Athens, Greece2Athens University of Economics and Business, Athens, Greece and Organization forEconomic Cooperation and Development, Greek Delegation, Paris, France

In this study we revisit some fundamental questions that are increasingly at the heart of currentstrategic management discourse regarding the relative impact of industry and firm-specificfactors on sustainable competitive advantage. We explore this issue by referring to respectiveassertions of two major perspectives that dominate the literature over the last two decades: thePorter framework of competitive strategy and the more recent resource-based view of the firm.A composite model is proposed which elaborates upon both perspectives’ divergent causal logicwith respect to the conditions relevant for firm success.

Empirical findings suggest that industry and firm specific effects are both important butexplain different dimensions of performance. Where industry forces influence market performanceand profitability, firm assets act upon accomplishments in the market arena (i.e., marketperformance), and via the latter, to profitability. The paper concludes with directions for futureresearch that will seek to integrate both content and process aspects of firm behavior. Copyright 2001 John Wiley & Sons, Ltd.

INTRODUCTION

The field of strategic management has undergone,in the 90s, a major shift in focus regarding thesources of sustainable competitive advantage: fromindustry to firm specific effects. Williamson (1991)presents these two different perspectives undertwo general headings: strategizing and economiz-ing. The first underlines a market power impera-tive. The second is fundamentally concerned withefficiency.

Key words: competitive strategy; resource based view;industry forces; sustainable competitive advantage; pathanalysis*Correspondence to: Y. E. Spanos, Athens University of Eco-nomics and Business, 76 Patision Street, Athens 104 34, Greece.

Research drawing from traditional IndustrialOrganization and more specifically from Porter’s(1980, 1985, 1990, 1991) framework of com-petitive strategy adopts an “outside-in” perspec-tive regarding market structure and its effect onperformance. Within this framework the firm isviewed as a bundle of strategic activities aimingat adapting to industry environment by seeking anattractive position in the market arena. The sus-tainability of rents stemming from such a positionis critically dependent on the relative influence ofcompetitive forces encountered by the firm (McGa-han and Porter, 1997).

On the other hand, the more recent resource-based perspective (Barney, 1986a, b; 1991;Rumelt, 1991; Wernerfelt, 1984) redirects attention

Copyright 2001 John Wiley & Sons, Ltd. Received 16 November 1998Final revision received 15 December 2000

908 Y. E. Spanos and S. Lioukas

into idiosyncratic firm capital and postulates thatperformance is ultimately a return to unique assetsowned and controlled by the firm.

The question of the relative impact of indus-try vs. firm-specific effects on performance is stillopen, and has a theoretical as well as a man-agerial value of its own. In the former sense, itrelates to calls for complete dismissal of one or theother perspective, or instead of adopting a morebalanced stance between the two (Mahoney andPandian, 1992). In the latter sense, it relates toissues of managerial importance such as with thechoice between market maneuvering and capabili-ties building.

An important body of literature has sought tocompare and contrast divergent premises betweenthe RBV and other perspectives emphasizing mar-ket power types of rents, including Porter’s frame-work (see for example Teece, Pisano and Shuen,1997; Mahoney and Pandian, 1992; Conner, 1991;Peteraf, 1993). Another stream of research, directlyrelated to the present study, sought to empiricallydecompose performance variation amongst firmsand thus examine the relative impact of industryvs. firm’s resources and capabilities.

The present study attempts to extend the rele-vant empirical literature by proposing a compos-ite framework whereby both perspectives’ causallogic as to the mechanisms of rent generation isexplicitly modeled. In particular, three distinct butalso complementary “classes” of effects on per-formance are identified. These include (i) strategy,(ii) industry and (iii) firm-assets effects. Theframework is tested using empirical data. Resultsseem to support recent arguments in the literaturethat consider both industry and firm-level influ-ences as significant determinants of performance(Henderson and Mitchell, 1997). Furthermore, ourfindings seem to suggest that where industry forcesinfluence market performance and profitability,firm assets act upon accomplishments in the mar-ket arena (i.e., market performance) and via thelatter, to profitability.

The following section presents the theoreticalbackground of the two perspectives with respectto sustainable competitive advantage as well asthe rationale for the development of a compos-ite model. Subsequent sections present the modeldevelopment and hypotheses, and then the empiri-cal analysis and results. The paper concludes witha discussion of the findings and with directions forfuture research.

THEORETICAL BACKGROUND

The competitive strategy perspective

Within the classical industrial organization liter-ature scholars have typically assumed that firmmanagement can influence neither industry condi-tions nor its own performance. This view, reflectedby such works as Bain (1956) and Mason (1939),maintains that because firm conduct (i.e., strategy)is constrained by industry structural forces, it doesnot represent independent managerial action. Man-agement’s role can therefore be ignored.

This view was also supported by research inorganizational theory which emphasized the deter-ministic role of environment (population ecologyand natural selection, e.g., Hannan and Freeman,1977). Furthermore scholars within the IO tra-dition were primarily concerned with explainingand evaluating industry, as opposed to firm per-formance.

The modified framework advanced by Porter(1980; 1985; 1990; 1991) departs markedly fromtraditional IO theory in a number of importantways. First, Porter focuses on firm rather thanindustry performance, a characteristic of researchin the strategic management tradition. Second, forPorter industry structure is neither wholly exoge-nous nor stable, as commonly viewed in traditionalIO theory (Bain, 1968; Caves, 1972). Instead, in hismore recent writings, Porter (1991) views marketenvironment as partly exogenous and partly subjectto influences by firm actions, a notion similar to thatof “choice situation” advanced by Hrebriniak andJoyce (1985) or strategic choice (Child, 1972).

Finally, in Porter’s framework, the role of firm’sconduct in influencing performance, together withindustry structure, is explicitly recognized. Whileindustry structure still occupies a central role inexplaining firm performance, undoubtedly reflect-ing a heritage from traditional IO, Porter choosesto focus on the role of firm activities and posi-tioning as a fruitful venue for the development ofa dynamic theory of strategy (Porter, 1991). Thenfor Porter, holding industry structure constant, asuccessful firm is one with an attractive relativeposition. Either this position can arise from theselection of a cost base lower than the competitionor from the firm’s ability to differentiate its offer-ings and command a premium price that exceedsthe accumulation of the extra costs. Hence the twoprimary types of competitive advantage: differen-tiation or low cost.

Copyright 2001 John Wiley & Sons, Ltd. Strat. Mgmt. J., 22: 907–934 (2001)

Causal Logic of Rent Generation 909

Central to Porter’s view of strategy is the notionof activities. For Porter then, strategy is a con-sistent array or configuration of activities (Porter,1991: 102), aiming at creating a specific formof competitive advantage for which there existtwo fundamental types: differentiation or low cost.These in turn, together with the scope of operationsdefine the notion of generic strategies. Within thisframework, strategy choice is the product (andresponse to) of a sophisticated understanding ofindustry structure.

The resource-based view perspective

If for Porter’s competitive strategy framework,a firm is viewed as a bundle of activities, forthe resource-based scholars, firm is viewed as abundle of unique resources. As Barney (1991)notes, much of the empirical literature informedby Porter’s framework, chose to focus analysis onthe environment–performance relationship, plac-ing little emphasis on the impact of idiosyncraticfirm attributes on performance (Porter, 1990). Thiswas implicitly due to two main assumptions: Firstit was assumed that firms are identical in termsof strategically relevant resources and second, anyattempt to develop resource heterogeneity has nolong term viability due to the high mobility ofstrategic resources amongst firms.

In contrast, the Resource Based View of the firm(RBV) focusing on the relationships between firminternal characteristics and performance, advancestwo alternative assumptions: a) firms may be het-erogeneous in relation to the resources and capa-bilities on which they base their strategies, andb) these resources and capabilities may not beperfectly mobile across firms, resulting in hetero-geneity among industry participants.

Rooted in evolutionary economics and the workof Penrose (1959) the resource–based approachhas reestablished the importance of individual firm,as opposed to industry (or particular strategicgroups), as the critical unit of analysis.

Resources are defined as those tangible (or intan-gible) assets that are tied semi-permanently to thefirm (Maijoor and Witteloostuijn, 1996). Exam-ples of such resources are: brand names, in-houseknowledge of technology, skilled personnel, tradecontracts, efficient procedures, etc. (Wernerfelt,1984). In the early contributions, there was noexplicit distinction between resources and capabil-ities. According to Amit and Schoemaker (1993),

however, recourses are assets that either are ownedor controlled by a firm, whereas capabilities referto its ability to exploit and combine resources,through organizational routines in order to accom-plish its targets. In addition, Collis (1994) describedcapabilities as the socially complex procedures thatdetermine the efficiency with which organizationsare able to transform inputs into outputs.

More recently, Teece et al. (1997) offered a com-prehensive framework of dynamic capabilities thatreflect a firm’s ability to achieve new and innova-tive forms of competitive advantage. These encom-pass organizational and managerial processes (i.e.,coordination/integration, learning and reconfigura-tion), specific asset positions (i.e., technological,financial, reputational etc. assets) and path depen-dencies (i.e., the firm’s history).

The two perspectives’ causal logic regardingsustainable performance

While both perspectives have made significantand complementary contributions in the field ofstrategic management (Foss, 1996, 1997a; Amitand Schoemaker, 1993; Peteraf, 1993; Mahoneyand Pandian, 1992; Conner, 1991) they have beenat odds with each other regarding the origin ofsustainable competitive advantage.

Porter’s framework

In Porter’s framework firm performance is a func-tion of industry and firm effects (i.e., marketpositioning) (Grant, 1991; Porter, 1991). Becauseindustry structure is also, at least partly, suscep-tible to firm activities, these two determinantsof firm performance are ultimately interrelated.According to Porter, industry structure affects thesustainability of firm performance, whereas posi-tioning reflects the firm’s ability to establish com-petitive advantage over its rivals. Having gainedsuch an attractive position, a firm can exercisemarket power (Teece, 1984; Teece et al., 1997)and thus, gain “monopoly-type” rents. These rentsstem from the firm’s ability either to defend itselfagainst competitive forces (“defensive” effects), orto influence them in its favor (“offensive” effects)(Porter, 1980, 1985, 1991).

The difference between “defensive” and “offen-sive” (direct and indirect) industry effects is ofsome importance for the purpose of developinga composite model and deserves further atten-tion. The former type denotes a rather passive



stance against industry forces which in effect aretaken as given. In this case, strategy according toPorter (1980), can be viewed as creating defensesagainst the industry forces or as finding a pro-tected position. In contrast, the latter denote amore aggressive strategic posture where the firmseeks to alter the balance and underlying causesof industry forces. “Offensive” industry effectsthen, might more appropriately be called “puremonopoly–type” effects since they imply that thefirm is actively engaged in a conscious attempt toexercise market power.

Porter (1991) views resources occupying aninherently intermediate position in the chain ofcausality with respect to firm performance. Forhim, firm assets are built from either performingactivities (i.e., strategy) over time, or acquiringthem from environment, or both. In either case,the available stock of resources reflects prior man-agerial choices, the latter related to the choice ofstrategy. Thus, the argument goes, activities arelogically prior, since their successful implementa-tion requires different resources and skills, orga-nizational arrangements, control procedures andinventive systems (Porter, 1980).

In this vein resources are not valuable in andof themselves because they (and not vice versa)are attached to strategic activities. Maintainingor enhancing these assets demands reinvestmentthrough continuously performing these activities.Moreover, their significance critically depends onhow well they support the strategy pursued, andby extension how well they fit industry structure.

RBV perspective

In contrast, the resource-based perspective viewsthe issue of strategy–resources and the resources–performance relationships from exactly the oppo-site angle.

Within the traditional mainstream strategy re-search literature (see for example Andrews, 1971;Ansoff, 1965; Child, 1972), of which RBV incor-porates important concepts (Mahoney and Pandian,1992), strategy selection is based on careful eval-uation of available resources (strengths and weak-nesses). Over time, firms continue to follow strate-gies because of both the opportunities imposed bythe market environment and the constraints thatresult from their own accumulated asset base, orga-nizational structure, ownership and other firm spe-cific factors (Barney, 1991; McGee and Thomas,

1986). Current or future strategic decisions are con-strained by past resource deployments and result infurther reinforcement of strategic profile. This ofcourse should not be taken to imply a determinis-tic rigidity over firm’s strategic behavior. Becauseof constant environmental changes, managers dohave choices to make about strategic alternativesbut their options might be limited within the estab-lished framework of available resources.

Accordingly, then, and in sharp contrast to Por-ter’s contention, resources are valuable in and ofthemselves, driving the choice of strategy. WhereasPorter views strategy as being primarily industrydriven, the resource-based perspective posits thatthe essence of strategy is or should be defined by thefirm’s unique resources and capabilities (Rumelt,1984). Furthermore, the value creating potentialof strategy, that is the firm’s ability to establishand most importantly sustain a profitable marketposition critically depends on the rent generatingcapacity of its underlying resources (Conner, 1991).

In other words, this perspective’s contentionis that persistent differences in firm profitabilityrequire that either the firm’s product be distinc-tive (i.e., differentiated), or attain a low cost posi-tion relative to its rivals (Conner, 1991). This ofcourse is similar to Porter’s view. However forthe resource-based perspective, returns stemmingfrom such a position in the market place, result,unlike Porter’s and Bain-type IO, from acquiringand deploying valuable idiosyncratic assets ratherthan from industry structure. The underlying logicholds that the sustainability of effects of a com-petitive position rests primarily on the cost ofresources utilized for implementing the strategypursued. This cost can be analyzed with referenceto strategic factor markets (Barney, 1986a), that ismarkets where necessary resources are acquired. Itis argued that strategic factor markets are imper-fectly competitive, because of different expec-tations, information asymmetries and even luck,regarding the future value of a strategic resource.Should factor markets be perfectly competitive,then the cost of acquiring strategic resources wouldequal their going economic value in use for imple-menting this strategy, and hence no firm couldsustain its competitive advantage (Barney, 1986a).

Dierickx and Cool (1989) took the notion ofimperfectly competitive strategic factor marketsas their point of departure and further suggestedthat purchasable assets cannot constitute sourcesof sustainable rents, simply because they can be



traded in the market. Instead, critical resourcesare those that are built and accumulated withinfirm boundaries, their non-imitability and non-substitutability hinging on specific traits of theiraccumulation process.

Irrespective however of resources being acquiredor internally built, a fundamental premise of theresource-based perspective is the assumption ofsignificant and persistent firm heterogeneity interms of resource endowments. It is generallysuggested that this heterogeneity results becauseof barriers to imitation (Rumelt, 1991), and firms’inability to alter their accumulated stock of re-sources over time (Carrol, 1993). In this vein,unique assets are seen as exhibiting inherently dif-ferentiated levels of “efficiency”, in the sense thatthey are superior to others, hence producing Ricar-dian rents (Teece et al., 1997). Sustained profitsthen, are ultimately a return to the unique assetsowned and controlled by the firm.

The important point here is that a given strat-egy will generate sustainable performance differ-ential if and only if the resources used to conceiveand implement it are valuable, rare, non-imitableand non-substitutable (Barney, 1991). The implica-tion of this argument is that efficiency rents stem-ming from such assets could be categorized intotwo, interrelated dimensions: (a) rents stemmingdirectly from the efficient implementation of thegiven strategy currently pursued, and (b) indirectlyfrom enabling the firm to conceive and develop itsstrategy configuration.

The former can be denoted as a “pure” efficiencyeffect since it directly influences performance byvirtue of the efficiency (broadly construed, seeCollis, 1994) with which strategy is implemented.The latter, represents the ability (and the rentsindirectly stemming from this ability) to create astrategic position, and hence utility, as a resultof a strategy that is either entirely new relativeto rivals, or one that was not previously feasiblebecause of resource limitations. These latter effectsresult from firm assets that resemble Teece et al.’s(1997) notion of dynamic capabilities defined asthose that reflect the firm’s ability to achieve newand innovative forms of competitive advantage.

The complementarity between the twoperspectives

Besides the apparent conflicting views between thetwo perspectives outlined above, in reality both can

co-exist and shape actual firm behavior. It has beenrecently recognized that the “competitive strategy”and resource-base perspectives complement eachother in explaining a firm’s performance (Amit andSchoemaker, 1993; Peteraf, 1993; Mahoney andPandian, 1992; Conner, 1991). In fact, accordingto Wernerfelt (1984), Porter’s framework and theresource–based approach constitute the two sidesof the same coin. Intuitively, value creation stemsfrom the fit of internal capabilities to the strategypursued, and of strategy to competitive environ-ment (Barney and Griffin, 1992; Barney, 1992).As Barney and Zajac (1994) have argued, theexamination of strategy implementation skills (i.e.,resources and capabilities) cannot be understoodindependently of strategy content and the compet-itive environment within which the firm operates.

It could be argued that the resource-basedapproach, by emphasizing firm-specific effortsin developing and combining resources toachieve competitive advantage, provides the“Strength-Weaknesses” part of the overall SWOTframework, while industry analysis suppliesthe “Opportunities-Threats” part (Foss, 1996).In this respect then, the two approachesare complementary simply because they coverdifferent domains of application (Foss, 1997b;Barney, 1991) within the context of SWOTanalysis. While the resource-based approachemphasizes that focusing on firm effects isimportant in developing and combining resourcesto achieve competitive advantage, industry effectsare also critical. Environmental changes “maychange the significance of resources to the firm”(Penrose, 1959:79).

One important similarity between the RBV andPorter’s perspective is the shared view that per-sistent above-normal returns are possible, and thatto this end, an attractive strategic position is ofcrucial importance (Conner, 1991). However, asnoted above, related to this similarity a funda-mental difference arises, involving the nature ofrents a firm can achieve: monopoly-type of rentsfor the Bain-type IO (and Porter’s framework),and efficiency-type of rents for the resource-basedperspective. It is exactly this dual pattern of simi-larity—difference regarding firm performance, onwhich a composite model could be based. More-over, Porter’s perspective (especially as elabo-rated in his later work—Foss, 1996), despite beingclearly rooted in the tradition of the Bain-type IO,constitutes a definite attempt to reinstate the firm



as the critical unit of analysis. It follows then thatboth frameworks focus on the individual firm astheir subject matter, albeit their dissimilar viewson what is more or less important when examiningthe sources of competitive advantage.

All the above result in a fundamental compati-bility between these two modes of theorizing. Putdifferently, an attempt to compare and contrast thetwo perspectives’ causal logic within the context ofa composite framework is justified on the basis ofthree reasons: (a) the two perspectives are com-plementary as explicants of firms’ performance,in the sense that by drawing insights from both,one can gain a more balanced view on the sourcesof competitive advantage (“internal” and “exter-nal” determinants); (b) both perspectives seek toexplain the same phenomenon of interest (i.e., sus-tained competitive advantage), and (c) because theunit of analysis is identical in both cases (i.e., thefirm).

This however, should not be taken to denythat the Porter and the resource-based perspectivesdraw from two different, even antagonistic, theo-retical traditions. The principal aim of a compositemodel is not to attempt to resolve all the underly-ing theoretical tensions between the two perspec-tives. Instead, the principal aim is to identify therelative impact of industry vs. firm specific factorson firm performance, adding to the relevant empir-ical literature (see for example Schmalensee, 1985;McGahan and Porter, 1997; Hansen and Wer-nerfelt, 1989; Rumelt, 1991; Mauri and Michaels,1998). What distinguishes the present study is theattempt to explicitly model the dissimilar viewsas to the causal mechanisms deemed relevant forrent generation, whereas in the other studies thesemechanisms remain implicit.

MODEL DEVELOPMENT ANDHYPOTHESES

Effects and hypotheses

Following the arguments above, we can depictthe divergent causal logic of both perspectives ina composite model as in Figure 1. This modelincorporates the following effects: (i) strategy (or“utility” ’) effects that constitute the necessary con-dition for above average performance, (ii) indus-try and (iii) firm specific effects that providethe sufficient conditions for the sustainability ofperformance.

(i) Strategy or “utility” effects

Both the Porter and the resource-based perspec-tives acknowledge the importance of an attrac-tive strategic position (i.e., competitive advantage)viewed as an outcome of firm strategy activities.For both, the central issue is the creation of valuefor buyers, either in the form of differentiatedproduct, or one produced with lower costs. Irre-spective of the type of advantage however, thefundamental criterion for success is meeting mar-ket needs. Since offerings are not sold in and ofthemselves but for the utility they confer to theusers (Lancaster, 1966, 1971), we shall denote thisas “utility” type of effect, depicted as path ξ3 inFigure 1.

“Utility” effects then, result from the fit of theparticular offering to the particular needs of themarket segment addressed. Notwithstanding thefirm specific or industry effects (see below) afirm can enjoy, if there exists no sufficient marketdemand for its offerings, the firm cannot (arguably)achieve success. Then clearly, “utility” effects con-stitute a necessary condition for above–averageperformance, but not a sufficient one. Other typesof effects, acting independently or in combination,provide the sufficient conditions for the sustain-ability of rents. In this respect, “utility” effectsconstitute the base on which other types add onto condition the sustainability of performance.

(ii) Industry effects

Within Porter’s framework, industry occupies aninherently central role, either direct and/or indi-rect in determining the sustainability of strategicpositioning and hence of performance. In otherwords, industry forces affect the sustainabilityof above average performance against bargain-ing and against direct and indirect competition(Porter, 1991). Translating this form of theorizinginto causal modeling language, these effects aredepicted by paths ξ1 and ξ2. More specifically:

Direct industry effects on firm performance arethose that pertain to the firm’s given strategicposition in the market arena. These are repre-sented in Figure 1 as path ξ2 (industry forces →performance). Path ξ2 encapsulates the specificindustry effects to performance when a firm haschosen a defensive type of positioning.1 This

1 In principle path ξ2 would also represent overall indus-try attractiveness effects that pre-determine the inherent profit



Industry forces (barriers to entry, power oversuppliers, power over buyers, intensity of

competition, threat of substitutes)

ξ2: "

Dire

ct"

Indu

stry

Eff

ects

Market Performance

Profitability

Strategy

ξ1

ξ3: "Utility" Effects

Firm Assets

ξ4: Direct Efficiency Effects

ξ5

"Indirect" Efficiency Effects

"Indirect" Industry Effects

InnovativeDifferent.

MarketingDifferent.

Low Cost

Organizational

Marketing

Technical

Figure 1. Conceptual framework

position as noted earlier could result from defen-sive strategic actions that aim to a market posi-tioning that provides the best protection againstthe existing array of competitive forces.

Indirect industry effects on the other hand arethose that could result from “offensive” strategicmoves that aim at altering the balance of indus-try forces in the firm’s favor (Porter, 1980: 29).On this account, the product of paths ξ1∗ξ2 repre-sent “offensive” indirect industry effects. To putit in verbal terms, under the offensive type ofpositioning, strategy influences the relative bal-ance of the competitive forces the firm confronts(ξ1) and these forces in turn influence performance(ξ2).

potential of a given industry, applicable to all industry incum-bents (Collis and Montgomery, 1995). These effects are implic-itly considered here through ξ2.

(iii) Firm assets effects

Within the resource-based perspective, availableidiosyncratic assets resulting from strategic fac-tor market imperfections and/or internal processesconstrain strategy choice. Firm performance de-pends on strategic position (the type of “util-ity” effects discussed above) the sustainability ofwhich is defined by those unique resources andcapabilities. Efficiency rents refer to those thatstem from firm unique assets. Clearly then, andin contrast to industry driven influences, efficiencyeffects are obtained at the firm level.

Path ξ4 in Figure 1 represents the direct effi-ciency type effect. It denotes the direct influenceon performance resulting from the possession ofa superior stock of available resources, assumingthat the more the firm is endowed with resourcesrelative to competition, the higher will it be its



implementation efficiency and hence performance.Note that this path is independent of strategy, sincethe latter is taken as given.

On the other hand path ξ5 denotes the firm’sability to enhance and/or develop its activities inpursuit of a more complex and advantageous strat-egy configuration (Amit and Schoemaker, 1993).The more the firm is equipped with resources themore its ability to develop a utility creating strat-egy. As a result the product of effects involvingpaths ξ3 and ξ5, represents the combined influ-ence on performance resulting from the firm’s abil-ity to develop and/or modify its strategy posture,which ability is a consequence of available stockof resources.

Following the previous discussion two alterna-tive hypotheses could be advanced with respect tothe determinants of sustainable above average firmperformance:

Hypothesis 1: Firm performance depends oncompetitive advantage (as a necessary condi-tion) the sustainability of which depends ondirect and indirect industry effects.

This is tested against the alternative hypothesisadvanced by the resource-based perspective:

Hypothesis 2: Firm performance depends oncompetitive advantage (as a necessary condi-tion) the sustainability of which depends ondirect and indirect effects stemming from avail-able resources and capabilities.

Performance and controls

The present research relies on the use of perceivedmeasures to operationalize performance in termsof two dimensions, namely profitability and mar-ket performance (Venkatraman and Ramanujam,1986). Firm success is treated, therefore, as a two-dimensional phenomenon, where market perfor-mance reflects the external firm accomplishmentsin the market place, and profitability the inter-nal to the firm economic rents stemming fromits strategic activities. Moreover, because marketperformance has been shown to positively andoften significantly affect profitability in a num-ber of empirical studies2 we also hypothesize a

2 For example, the effect on profitability of market share, astandard indicant of market performance, has been widely

positive direct effect of market performance onprofitability.

We also include firm size, one of the mostfrequently studied contextual variables as a controlvariable in order to remove whatever effects thismay have on firm performance.

Finally as regards the model depicted in Fig-ure 1, it should be clear it is concerned with rela-tionships among composite constructs (i.e., strat-egy and firm assets) with respect to performance,rather than between individual generic strategies orspecific capabilities and performance. Strategy, inthis respect, is conceptualized as a “second order”construct that is being composed of all three “firstorder” strategic dimensions (i.e., marketing differ-entiation, innovative differentiation and low cost),on which a firm can score low or high. This treat-ment is in line with an established tendency in therelevant literature to consider cost and differenti-ation as compatible (see for example Miller andFriesen, 1986; Phillips et al., 1983). Moreover this“mixed” strategy is perhaps the most appropriatein an era of hyper-competition prevailing in mostindustries (see for example Miller, 1992).

In this same vein, the significance of firm capa-bilities (i.e., organizational, marketing and techni-cal) as sources of sustained competitive advantagedepends on their internal interconnectedness (Amitand Schoemaker, 1993; Dierickx and Cool, 1989),or put differently, on their mutual dependence(Teece, 1986). Under this view, the combinedvalue of the various firm resources and capabil-ities maybe higher than the cost of developing ordeploying each capability individually (Amit andSchoemaker, 1993).

METHODOLOGY

Sample and data collection

The exploratory nature of this research necessi-tated particular parameters for sample selection.First, the study focused on independent firms (orsingle business units) so that the effects of strat-egy and capabilities could be examined indepen-dent of the confounding effects of corporate levelconsiderations. Second, focusing on single firmsnecessitated an inter-industry sample to ensure

established by empirical research (see for example, Schoef-fler, Buzzell and Heany, 1974; Buzzell, Gale and Sultan, 1975;Rumelt and Wensley, 1981; Phillips, Chang and Buzzell, 1983;Prescott, Kohli and Venkatraman, 1986)



sufficient sample size and generalizability of theresults. Finally, only firms employing at leasttwenty employees were considered in the samplein order to ensure a minimum operating structureof each firm. Data were solicited from a popula-tion of 1090 Greek firms. These belong to variousmanufacturing industries such as food and bev-erages, wood and furniture products, chemicals,metal products, machinery, electric equipment andappliances. Data were collected through a struc-tured questionnaire dispatched to CEOs.

A number of approaches were used to ensureresponse quality and to enhance response rate.These collectively constitute a modified versionof Dillman’s (1978) “total design method”. Morespecifically, the process was organized as fol-lows: First, the research instrument was pretestedtwice. In its draft form, it underwent a pretestwith CEOs from three companies. A second pretestwas conducted after in depth discussions with aca-demics and questionnaire design experts. This sec-ond pretest was conducted in four firms, for twoof which in-depth case studies were developed.After some minor modifications the final question-naire was mailed to CEOs together with a letterexplaining the purpose of the study and assuringanonymity, together with a pre-stamped envelopeaddressed to one of the researchers. Four weeksafter the initial mailing we sent a follow-up mail-ing that included the same material as the first.

A total of 187 CEOs responded giving a res-ponse rate of 17%. Out of these, 147 questionnaireswere found usable. The average firm size is 160employees (median 67).

To test whether our respondents were differentfrom the non-respondents we examined if therewere any differences in the means of all vari-ables used in this study between early and laterespondents. The rationale behind such an analy-sis is that late respondents (i.e., sample firms in thesecond wave) are more similar to the general pop-ulation than the early respondents (Armstrong andOverton, 1977). The only statistically significantdifference found was for the “threat of substitutes”measure (F = 5.763, ρ < 0.05). Hence it appearsthat non-response bias is not a serious issue inthis study.

Measurement of constructs

As described above, a research instrument wasdeveloped to serve as the basis for collecting data

pertaining to industry structure, competitive strat-egy, resources and capabilities and performanceusing self-typing measures, a well accepted prac-tice in strategy research. Appendix 1 shows themeans, standard deviations and correlation matrixof the research variables. All constructs were mea-sured with multiple-item 5-point Likert scales (seeAppendix 2 for details), except for dimensions ofindustry structure.

Measures of Porter’s generic strategies werederived and adapted from Dess and Davis’ (1984)and Miller’s (1988) studies. The scale asks ques-tions regarding the extent of usage of specificcompetitive tactics relevant to marketing differen-tiation, innovative differentiation and low cost.

Empirical research on resources and capabil-ities has not yet reached maturity (Miller andShamsie, 1996), and thus original scales had tobe engineered based on theoretical contributionsfrom resource based scholars and extensive dis-cussions with academics and CEOs during thepre-testing phase of questionnaire development. Inthe present research, idiosyncratic firm assets weredetermined to include measures of organizational,marketing and technical capabilities. CEOs wereasked to indicate the extent to which these con-stitute particular strengths relative to competition.More specifically:

Organizational capabilities denote Teece et al.’s(1997) organizational and managerial processesencompassing managerial competencies, knowl-edge and skills of employees together with effi-cient organizational structure, organizational cul-ture, efficient coordinative mechanisms, strategicplanning procedures and ability to attract creativeemployees.

Marketing capabilities resemble Lado, Boydand Wright’s (1992) output-based competenciesand were measured with such items as build-ing of privileged relationships with customers andsuppliers, market knowledge, control over distri-bution channels, and strong “installed” customerbase.

Technical capabilities parallel Leonard-Barton’s(1995) technical systems, and Lado et al.’s trans-formation-based competencies, referring to thosecompetencies that are required to converting inputsinto outputs. These were measured with threeitems, that is, efficient production department,technological capabilities and infrastructure, andeconomies of scale and technical experience.



Following Confirmatory Factor Analysis (seeAppendix 2 for details) that corroborated the inter-correlation between the individual strategy as wellas between the firm assets dimensions, two com-posite measures of Strategy and Firm Assets weredeveloped by averaging the respective individualitems.

Measures of industry forces (i.e., barriers toentry, bargaining power of buyers, power of sup-pliers, and threat of substitutes) were representedby single item questions, except for compositecompetitive rivalry that was gauged by four itemsadapted from Achrol and Stern (1988). All thesemeasures were referring to the particular situa-tion confronted by each firm in its major marketserved.

As noted in the previous section, performancewas operationalised as a two-dimensional con-struct, including composite profitability, and com-posite market performance (Woo and Willard,1983; Venkatraman and Ramanujam, 1986). Theformer was gauged with three perceptual itemsreflecting return on equity, profit margin and netprofits relative to competition, whereas the laterwas measured with market share, absolute salesvolume, increase in market share and sales. Forall these items managers were asked to indicatetheir firm’s performance relative to competition,3

as is extensively common in similar studies. Fur-thermore, respondents were asked to indicate theirfirms’ relative performance over the last threeyears period in order to avoid bias from any tempo-ral fluctuations and also to proximate a notion ofsustainability of performance. To mitigate poten-tial autocorrelation effects, the items of subjectiveperformance were placed in a different part of the

3 As one of the anonymous reviewers noted, the relative com-parisons of CEOs’ responses are perhaps problematic due to thesubjectivity of perceptions. Compare this approach, however,with the alternative of collecting “objective” data (to the extentthat they are available) and treating them as belonging to a singlecoherent population. How can we compare, on the same variable,two firms operating into two distinct industries? This would cer-tainly necessitate some kind of normalization of the variable inquestion, in order to take into account the respective industry ref-erence point (usually the industry average) for this comparisonto be meaningful. But as many argue, industry is a rather vagueconcept, the boundaries of which are usually ill defined. Hencethe validity of such comparison may also be problematic. Evenin the case of a single industry study what one firm considers asits immediate domain of interaction(s) with its competitors doesnot necessarily coincide with that of the other. The “relativistic”comparability of our perceptual measures therefore may not beinferior to using “objective” data and “absolute” comparisons.

questionnaire relative to strategy and resources andcapabilities items.

Besides the fact that subjective performancemeasures have been widely used in strategy relatedresearch (see for example Dess and Robinson,1984; Robinson and Pearce, 1988; Venkatramanand Ramanujam, 1986, 1987), because the sampleincludes mostly SMEs it was anticipated that itwould be difficult to extract adequate and reliablefinancial information. Anonymity also precludedthe collection of such data from secondary sources.That was indeed the case since 80 out of theoriginal 187 sample firms did not provide com-pany name or balance sheet data. Besides, finan-cial data for SMEs are also criticized for beingunreliable and subject to varying accounting con-ventions or even to managerial manipulation fora variety of reasons (e.g., avoidance of corporateor personal taxes; see Dess and Robinson, 1984;Sapienza, Smith and Gannon, 1988; Powell andDent-Micallef, 1997).

Finally, organizational size was measured as thenatural logarithm of the number of employees.

The exclusive reliance of this study to subjec-tive responses was dictated by both practical andtheoretical considerations. From a practical viewpoint, the choice of perceptual data with respectto strategy and firm capabilities was necessitatedprimarily due to the non-availability of appropri-ate balance sheet data to capture such complexorganizational phenomena. For example, one com-monly used “objective” strategy indicator that iscentral to the measurement of innovative differ-entiation, namely product R&D/sales, could notbe computed since Greek firms are not obligedto report R&D expenses. The inadequacy of bal-ance sheet data is even more obvious in the caseof capabilities measures, since it appears impos-sible to capture the essence of valuable and hardto imitate idiosyncratic firm qualities from crudefinancial measures.

Beyond the aforementioned practical considera-tions there exists also strong theoretical rationalesupporting the choice of subjective data. Lefeb-vre, Mason and Lefebvre (1997) refer to what theyterm the “influence prism” of CEOs’ perceptionsto note that CEOs’ diverging views of the envi-ronment may “override factual characteristics ofthe environment” (1997: 861). Within this line ofreasoning it is often argued that managers’ per-ceptions shape behavior and are more critical tostrategy making and firm performance than some



“mentally distant” objective indicators (Hambrickand Snow, 1977; Snow, 1976; Chattopadhyay etal., 1999). This premise is also supported by thesocial constructionist perspective that maintainsthat reality as such is socially constructed andhence, according to Weick (1979) there is no suchthing as an “objective” environment, but rather itis those parts of the information flows that thefirm “enacts” through attention and belief. Admit-tedly, other scholars object to this line of reason-ing (e.g., Aldrich, 1979; Dess and Beard, 1984;Keats and Hitt, 1988; Lawless and Finch, 1989),but along with Chattopadhyay et al. (1999) wecould argue that managerial perceptions shape to avery important extent the strategic behavior of thefirm. In this sense, the use of shelf-reported mea-sures might be justified, albeit not without potentialproblems.

Validation of proposed constructs

It is well known that survey research, if not prop-erly conducted, can provide misleading resultswith measurement errors representing one of themost significant sources of bias. While how-ever, measurement errors are almost inevitable, theextent to which these errors affect the findingsis a function of what particular efforts (a priori)and what checks (a posteriori) have been under-taken, in order to minimize and asses the potentialbias.

On this account construct validation is par-ticularly relevant. In effect it involves a mul-tifaceted process comprising three basic steps.The first, content validity, requires the identifi-cation of a group of measurement items whichare deemed to represent the construct of inter-est. The second step, construct validity, seeks toestablish the extent to which the empirical indi-cators actually measure the construct. The finalstep, nomological validity, involves the determi-nation of the degree to which a construct relatesto other constructs in a manner predicated by the-ory. These issues are dealt with in Appendix 2,with the exception of nomological validity whichis implicitly addressed in the context of the sub-stantive relations examined in this study. All anal-yses (see Appendix 2 for detailed description ofprocedures and results) provide reasonable con-fidence that the measures used are valid andreliable.

RESULTS

Model estimation and fit

The structural relations among the constructs inour conceptual model were examined with pathanalysis using the maximum likelihood estima-tion (MLE) procedure in statistical package EQS(Bentler and Wu, 1995). Results obtained fromfitting the model in Figure 1 are presented inTables 1, 2, and 3. Table 1 summarizes the resultsof direct effects (i.e., strategy or “utility”, industryand firm assets -paths ξ3, ξ2 and ξ4 respectively)as well as overall model fit statistics. Table 2presents the effects of strategy → industry (pathξ1), and of firm assets → strategy (path ξ5). Theseare the first components of the product terms thattogether with paths ξ2 and ξ3 of Table 1 ‘pro-duce’ the indirect effects on performance. Indi-rect effects are presented in Table 3 (paths ξ1∗ξ2and ξ5∗ξ3 for industry and firm assets effectsrespectively).

The overall model shows a chi-square valueof 19.257 (df = 12) having a p-value of 0.082,indicating an excellent fit to the data. However,considerable discussion has taken place in the

Table 1. Path analysis results: direct effects

Standardised estimate

Market Profit-Parameter (from → to) Performance ability

“Utility” Effects (ξ3 )Strategy → 0.406∗∗∗ 0.049

Direct Industry Effects (ξ2 )Threat of Substitutes → −0.037 0.017Barriers to entry → 0.107 0.035Power of Suppliers → −0.020 −0.318∗∗∗

Competitive Rivalry → −0.148∗ 0.074Power of Buyers → 0.033 −0.091

Direct Efficiency Effects (ξ4 )Firm Assets → 0.277∗∗∗ 0.083Market Performance → 0.306∗∗

ControlSize → 0.152∗∗ 0.095

% Explained Variance R2 0.521 0.368

Model statistics: X2(12) = 19.257 p = 0.082 CFI = 0.964,RMR = 0.069, RMSE = 0.084.∗∗ denotes p < 0.5; ∗∗∗ denotes p < 0.01



Table 2. Path analysis results: effects of strategy on industry and of firmassets on strategy

Standardised % Explainedestimate Variance (R2)

Strategy → Industry (ξ1 )Strategy → Threat of Substitutes −0.064 0.003Strategy → Barriers to entry −0.002 0Strategy → Power of Suppliers −0.372∗∗∗ 0.139Strategy → Competitive Rivalry −0.286∗∗∗ 0.082Strategy → Power of Buyers 0.036 0.001

Firm Assets → Strategy (ξ5 )Firm Assets → Strategy 0.704∗∗∗ 0.496

∗∗ denotes p < 0.5; ∗∗∗ denotes p < 0.01.

Table 3. Path analysis results: industry and firm—assets indirect effects toperformance

Standardised estimate1

Market ProfitabilityPerformance

Industry Indirect Effects2 (ξ1 ∗ξ2 )Threat of Substitutes 0.00237 −0.00115Barriers to entry −0.00021 −0.00007Power of suppliers 0.00744 0.118296∗∗

Competitive Rivalry 0.04233 −0.02059Power of Buyers 0.00119 −0.00331

Firm Assets Indirect Effects3 (ξ5 ∗ξ3 ) 0.285824∗∗∗ 0.032384

1 Standard Error of estimate of e.g., ξ1∗ξ2 is computed as ξ12sξ22 + ξ22sξ12−sξ22sξ12 wheresξ1 and sξ2 are standard errors of ξ1 and ξ2 respectively2 ξ1∗ξ2 from Tables 1 and 2 [e.g., (strategy → suppliers) −0.372∗−0.318 (suppliers →profitability)]3 ξ5∗ξ3 from Tables 1 and 2 [e.g., (firm assets → strategy) 0.704∗0.406 (strategy →performance)]∗∗ denotes p < 0.5; ∗∗∗ denotes p < 0.01

Structural Equation Modeling literature concerningthe validity of the X2 test as an index of modelfit, especially when the sample size is small (asis in our case). As a result a number of adjunctfit indexes have been proposed that reflect theimprovement in fit of a specified model, whichincludes fixed and free parameters, over the inde-pendence model, in which all parameters are fixedto zero. The usual cut-off point recommendedis 0.90. The most commonly used fit index isBentler’s (1990) Comparative Fit Index (CFI),which has a value of 0.964 in our case, withmost other fit indices exceeding 0.95. In addi-tion the examination of residuals via indices suchas Root Mean Square Residual (0.069) and Root

Mean Square Error of Approximation (RMSEA)with 0.084, also indicate acceptable model fit.

Strategy (or “utility”) effects

Firm strategy, consistent with hypotheses 1 and2 appears to influence positively and significantlyfirm success (see Table 1), but only with respectto market performance (0.406, p < 0.01) and notto profitability (0.049, ns). In fact, strategy’s effectto market performance appears to be the strongestrelative to all other hypothesized relationshipswhere the dependent variable is a performancedimension. This seems to support the notion thatstrategy (“utility”) effects constitute a prerequisitecondition for above normal firm performance.



Industry effects

In accordance with the Porter framework, Hypo-thesis 1 stated that firm success, except fromstrategic positioning (i.e., strategy driven com-petitive advantage), is also dependent on bothdirect and indirect industry effects. The signifi-cance of these effects is partially supported by ourfindings.

Concerning direct industry effects (see Table 1),only competitive rivalry and power of supplierswere found significant for market performance andprofitability respectively. More specifically, thedirect effects of competitive rivalry on market per-formance are negative and marginally significant(−0.148, p < 0.10). On the other hand, signifi-cant direct industry effects on profitability wereonly found for the bargaining power of suppliers(−0.318, p < 0.01).

Industry indirect effects to performance repre-sent the firm’s ability to influence structure in sucha way as to sustain its performance over rivals.As noted in the model development section theseeffects are operationalized as the product of twocausal relationships: strategy’s effect on industrystructure (ξ1), multiplied by the latter’s influenceon performance (ξ2) (i.e., ξ1∗ξ2).

From Table 3 it appears that the only significantindirect industry effect is that of power of supplierson profitability (0.12, p < 0.05).

With respect to the effects of strategy → indus-try structure (paths ξ1), our findings (see Table 2)suggest that the only significant influences invol-ved concern the decrease of competitive rivalry(−0.286, p < 0.01) and of bargaining power ofsuppliers (−0.372, p < 0.01). Interestingly, thesetwo aspects of industry structure are those that arealso found to directly influence the two dimen-sions of firm performance (i.e., market perfor-mance and profitability respectively). Furthermore,the decrease of power of suppliers as a conse-quence of “offensive” strategic activities, results inthe positive indirect effect (ξ1∗ξ2) to profitabilitypresented in Table 3.

At any rate, however, as shown in Table 2,results indicate that the extent of firms’ influenceon competitive forces is minimal, explaining only14% and 8% of variance of power of suppliersand rivalry respectively. Given these, it is notsurprising that the only significant indirect effectsfound concern the relationship between strategy →power of suppliers → profitability.

Firm assets effects

Hypothesis 2, in accordance with the resource-based perspective states that the sustainability ofrents arising from an attractive market positioncritically depends on firm unique assets. Thishypothesis is also partially supported since signif-icant positive direct and indirect effects are foundbut only with respect to market performance.

More specifically, Table 1 shows that firm assetsseem to directly influence market performance(0.28, p < 0.01). Indirect effects on the other handare operationalized as the product of resourcesinfluence on strategy and the latter’s influence onperformance (i.e., ξ5∗ξ3).

From Table 2 it can be seen that available firmassets significantly and positively determine theconfiguration of strategic activities (0.704, p <

0.01). As can be seen from Table 3, the combinedeffect of both influences, (i.e., firm assets’ indirectinfluence on performance) is also positive andsignificant. As noted above, firm assets exert adirect positive effect on market performance butnot on profitability.

In general our results show that profitability isonly affected from an element of industry struc-ture, both directly and indirectly, namely power ofsuppliers, and from market performance. This lat-ter is important since it seems to support the viewthat market performance is a key to profitability,at least for the firms in our sample.

The role of market performance

Within the traditional structure-conduct-perfor-mance paradigm, market performance representsalong with other industry characteristics (such asentry and mobility barriers, bargaining power etc.),an element of market structure that necessarilyresults to market power and hence to profitabil-ity (Imel and Helmberger, 1971; Shepherd, 1972;Gale, 1972). Porter instead argues that the effect ofmarket performance to profitability is not straight-forward, depending on how well a firm with highmarket performance is protected against competi-tive forces. If that holds, even low-share firms canachieve superior profitability (Porter, 1980).

Other researchers, especially those within theresource based perspective interpret the marketperformance-profitability association as causallyspurious. Put more simply they are both viewed asmanifestations of firm success jointly determined



by superior firm capabilities (Rumelt and Wensley,1981; Jacobson, 1990).

Our findings (see Table 1) however, seem toindicate that its association to profitability is notcausally spurious. Technically a variable’s X (i.e.,market performance) apparent influence on Y (i.e.,profitability) is spurious when it is being entirelyattributable to X’s relationship with another vari-able Z (i.e., firm assets) (Cohen and Cohen,1983). Our results instead show that although theeffects of Z (firm assets) → Y (profitability) andof Z → X (market performance) are controlled for,the effect of X → Y is still positive and signifi-cant. This seems to imply a different interpreta-tion for the Z → X → Y chain of causality. Infact X, that is market performance, appears tobe a significant intervening variable to the firm’sassets (Z) → profitability (Y) relationship. In otherwords, although firm assets’ direct influence onprofitability is weak and insignificant (0.083, p >

0.10), its effect via market share (i.e., assets →market performance → profitability, = 0.277∗

0.306) is significant (0.08, p = 0.059).4

Hence market performance seems to consti-tute a significant intervening variable in thefirm’s resources-profitability relationship. Thesame pattern holds5 for the strategy → marketperformance → profitability relationship (0.12,p = 0.035). In other words, while the direct effectof strategy on profitability is weak, its effect viamarket performance is strong and significant. Inview of the above then, it seems logical to arguethat market performance is an important predeces-sor to profitability, at least in the context of oursample.

DISCUSSION OF RESULTS

Overall our results above seem to support theneed for a composite framework that will seek tosynthesize premises from both perspectives. Themodel depicted in Figure 1, is not only based onthe notion that the two perspectives are supple-mentary in explaining firm performance, but also

4 Standard error computed as in Table 3, note 1.5 Note that because the effects of elements of industry structureto market performance are generally weak (see Table 1), theintervening status of market performance with respect to industryforces regarding profitability is not supported by our data.

extends this mode of theorizing by explicitly treat-ing the mechanisms through which industry andfirm assets influence performance.

More specifically the results imply the co-existence of three distinct but also complemen-tary classes of effects, which supplement oneanother in determining performance. These are(i) “utility” type of effects (i.e., strategy configu-ration), (ii) industry (direct–“defensive” and indi-rect–“offensive”) and (iii) firm specific (direct andindirect) effects on performance. They are dis-tinct first because, they represent different con-ditions for achieving above–normal performance;second, because industry and firm effects oper-ate in different domains: through structural marketforces the former and through idiosyncratic stockof resources the latter. Concurrently however, theyare complementary as both “internal” and “exter-nal” conditions represent the two sides of the samecoin (Wernerfelt, 1984).

All types of effects are, to a greater or lesserextent, subject to firms’ strategic choices. Thecase of internal capabilities notwithstanding, evenindustry structure characteristics are at least par-tially endogenously determined by the firms’actions, especially in the case of “offensive” strate-gic maneuvering (Porter, 1991).

Returning to the issue of the relative impact ofclasses of effects on performance, our results showthat strategy activities are significant direct deter-minants of market performance, and indirectly (viathe latter) of profitability. This seems to confirmthe first part of Hypotheses 1 and 2 which positthat strategy driven competitive advantage definesthe necessary condition for above average perfor-mance, albeit not a sufficient one, as suggested bythe significance of both industry and firm specificeffects.

Regarding industry effects, it was found thatcompetitive rivalry directly influences market per-formance, while power of suppliers affects directlyand indirectly profitability. This is in line with thePorter framework and Hypothesis 1 that states thatthe sustainability of rents is dependent on indus-try effects directly (i.e., defense against direct andindirect competition and against bargaining, Porter,1991) and indirectly, through firm’s actions alter-ing the balance of the same industry forces in itsfavor.

More specifically, competitive rivalry capturesthe perceived threats to the very survival of thefirm that originate from harsh competition, thus



constituting a key characteristic of the externalenvironment (Lefebvre et al., 1997). In this veinit generally represents a fearsome characteristicof the external environment, constituting a seri-ous threat to the survival of firms. According toKhandwalla (1977: 335) hostile environments are“risky, stressful and dominating “. Our results arein line with the traditional industrial organisationliterature that suggests that in the absence of fiercecompetition, firms can exert a greater degree ofcontrol in managing price and, therefore, profits(Bain, 1951). Furthermore, a number of studieshave shown a positive relationship between indus-try concentration, which is directly related to lowenvironmental hostility and lack of competition,and firm performance (Hill and Hansen, 1991).For example Capon, Farley and Hoenig (1990),in a meta-analysis of empirical studies investi-gating strategy–performance relationships, foundthat industry concentration was positively relatedto firm-level returns.

It is interesting to note however that our resultsdo not seem to indicate that competitive rivalryhad any significant (direct or indirect) negativeeffects to profitability. Rather its significant influ-ence relates only to market performance. As Porter(1980: 17) notes, fierce competition and hence thepattern of competitive moves and counter–movesmay or may not adversely affect the firm or theindustry as a whole. For example price compe-tition, unless price elasticity of demand is highenough, will normally result to lower revenues andprofitability. On the other hand, advertising bat-tles may well expand demand for the benefit of allfirms.

It should be noted that price competition did notsurvive the preliminary scale purification proce-dure and therefore was not included as a mea-surement component of the competitive rivalryconstruct in this study. Hence, apart from theintuitively justified negative influence to marketperformance it seems that non-price competitiondoes not affect the profitability of the firms in oursample.

The direct negative effect found of power of sup-pliers on profitability seems to confirm Porter’s(1980) argument that suppliers can reduce prof-itability if they can exert bargaining power byeither raising prices or reducing the quality ofpurchased goods they supply. The decrease ofpower of suppliers as a result of “offensive” strat-egy positioning (indirect effect), in line with this

argument, will have positive effects on profitabil-ity, as indeed was the case with our findings.

With regards to firm assets effects, as arguedin the model development and hypotheses sectionabove, the rents stemming from such assets couldbe categorized into direct (“pure”) and indirectefficiency effects.

The significance of the former effects to marketperformance seems to indicate that implementationefficiency, stemming from a superior asset baserelative to rivals, is critically important in enablingthe firm to sustain a high share position in the mar-ket. Indirect efficiency effects on the other handdenote the firm’s ability to succeed not as a resultof implementation efficiencies but rather becauseof developing and strengthening its strategic pos-ture, which in turn is a consequence of its availablestock of resources. As before, these indirect effectswere found significant with respect to market per-formance, but not to profitability.

On this account two related issues are worth not-ing. First, the significant effects of firm’s assets onstrategy, is in line with the contention of resource-based scholars that a firm should develop andnurture its strategy profile building upon its avail-able stock of resources (e.g., Rumelt, 1991). AsGrant (1991) notes, the resources and capabilitiesof a firm are the central consideration in formu-lating its strategy. Moreover, they represent theprimary constraints upon which a firm can estab-lish its identity as well as the primary sources of itsprofitability. This latter contention however is notsupported by our data. Secondly, although assetsdo not affect profitability either directly or indi-rectly (i.e., via strategy), they do influence prof-itability via market performance. In other words,our results seem to suggest that the firm’s availablestock of resources is critical for (a) developing itsstrategy configuration, and (b) for achieving highmarket performance and through the latter, highprofitability.

Hence, our results are in line with Hypothesis 2and the resource-based perspective that views thesustainability of firm performance stemming fromspecific assets (either externally acquired in imper-fect markets or internally built), that are imper-fectly mobile, inimitable and non-substitutable(Peteraf, 1993).

To summarize then, our findings seem to indi-cate that together with strategic activities bothindustry and firm asset effects significantly con-tribute to firm success. This is in accordance with



Hansen and Wernerfelt’s (1989) remark that firmsthat can demonstrate excellence in both arenaswill do significantly better than those that strivefor more uni-dimensional concepts of excellence.Interestingly these effects concern different dimen-sions of performance. Where industry forces influ-ence directly and indirectly market performanceand profitability, firm assets act upon accomplish-ments in the market arena (i.e., market perfor-mance), and only via the latter, to profitability.

This result is perhaps most critical, since, ifconfirmed by future studies, it opens a researchquestion that needs to be explicitly addressed theo-retically: If firm performance is a multidimensionalphenomenon, then why are the different dimen-sions of performance affected differentially by itsvarious hypothesized determinants? Our resultsseem to suggest that industry and firm effects arenot only both potentially significant, but instead,they need to complement each other given thatthey affect distinct but strongly linked dimensionsof performance.

IMPLICATIONS AND FUTURERESEARCH

The findings of the present study should be con-sidered under the prism of the recent debate con-cerning the relative importance of industry vs. firmspecific effects on performance. As argued in theintroduction, the question between industry andfirm specific effects has a value of its own, bothin theoretical and practical terms.

In the theoretical front, relevant empirical find-ings in the quest to identify the “ultimate” sourcesof sustainable above average firm performanceare at best equivocal since no conclusive resultsseem to have emerged. Schmalensee (1985) andMcGahan and Porter (1997) conclude that indus-try effects explain an important portion of profitsvariability, whereas Hansen and Wernerfelt (1989),Rumelt (1991) and Mauri and Michaels (1998)report that firm effects are more important thanindustry effects on firm performance. Our ownresults point to the significance of both classes ofeffects that in addition appear to influence differentdimensions of firm performance.

Admittedly, these findings are not directly com-parable because of critical differences in researchdesign and statistical models employed. However,it could be argued, that there exists another, and

perhaps more important reason underlying theseinconsistencies in research findings: sources of per-sistent firm success might prove fundamentallycontext specific (Collis, 1994).

It should also be noted however that irrespec-tive of differences in relative magnitude estimatesattributed to different samples, operationalisationof measures and econometric specification emplo-yed (McGahan and Porter, 1997), these as well asour results seem to point to a simple fact; botheffects seem to be real and to coexist (Mauri andMichaels, 1998). Perhaps more important, strate-gizing and economizing are not mutually exclusive(Teece et al., 1997; Williamson, 1991). Sustainablecompetitive advantage may result form both indus-try and firm specific effects which in fact may alsobe complementary (Mauri and Michaels, 1998).

If then, industry and firm effects are real, coex-ist and are not incompatible, exclusive reliance ononly one of these might be argued to result in fail-ure. Interestingly, proponents of each of the per-spectives attack the other on this same ground. Forexample Porter (1991) argues that the competitivevalue of resources can be enhanced or eliminatedby changes in technology, rivals’ behavior or buy-ers’ needs, which, an exclusive focus on resourcesmight overlook. On the other hand, proponentsof the resource-based perspective point out thatPorter’s framework misdirects managers to focuson industry level characteristics, encouraging themto expend assets on influencing structure, eventhough their firm cannot uniquely benefit from thechanges (potentially) incurred (McWilliams andSmart, 1993).

A critical and in our opinion valid argument herestates that strategizing, that is, employing actionsthat aim at impeding competition, is only rele-vant to firms that already possess market power,which are only a small fraction of the total popu-lation (Williamson, 1991). However, our findingsseem to indicate that even in the case of smallermarkets consisting mainly of SMEs, as is ourcase, at least some degree of exercise of poweris possible (Chamberlin, 1962) as manifested withthe significant indirect effect involving power ofsuppliers.

Managerial implications follow directly from theforegoing discussion regarding the coexistence andcomplementarity of “utility”, industry and firm-specific effects on sustainable competitive advan-tage. Given this study’s finding that different, but



strongly linked, performance dimensions are dif-ferentially affected by these classes of effects, thenmanagement’s strategic choices need to carefullybalance strategy, industry and firm-specific factors.Exclusive reliance on only one or two of thesemight prove problematic.

The results show that strategy configuration (i.e.,“utility” effects) is a direct determinant of mar-ket performance and, indirectly, of profitability. Itrepresents, however, a necessary but not sufficientcondition for above average performance. The sus-tainability of firm’s success is dependent uponindustry and firm-specific effects. With respectto the former, the findings show that defensivemaneuvering (i.e., protection against competitiverivalry) is critical for securing market performance,whereas offensive maneuvering (i.e., decrease ofpower of suppliers) can significantly affect prof-itability. The firm’s available stock of resourcesand capabilities, on the other hand, is critical fordeveloping its strategy configuration as well as forachieving market performance, and via the latter,profitability.

Moreover, the role of market performance onprofitability adds another dimension on the afore-mentioned managerial implications. If, as our re-sults appear to suggest, market performance is acritical intervening factor in the firm assets →profitability and the strategy → profitability rela-tionships, then management needs to carefully con-sider the question of how to build and sustainhigh market share. Again, both industry and firm-specific factors appear relevant.

Taken overall, these findings seem to imply thatinstead of treating market performance and prof-itability on the one hand, and maneuvering anddeveloping firm-specific assets on the other, as sep-arate, even antagonistic choices, it would perhapsbe better if they were considered as componentsof a holistic framework. Managers should adopta more integrative strategic posture by viewingthem as inextricably linked, forming the funda-mental pillars on which sustainable success couldbe established.

In light of the above arguments, it might beconcluded that “intellectual isolating mechanisms”(Mahoney and Pandian, 1992: 374) between thetwo perspectives might prove unprofitable forstrategy related research. Instead, a conversationbetween the two might provide insights that willfurther our understanding of the sources of sustain-able competitive advantage.

The findings and implications of the researchshould also be considered in light of its limita-tions. As noted in the methodology section, self-reported data were used to test the model. Despiteconsiderable efforts devoted both at the data andconstruct validation phases to ensure data qual-ity with encouraging results, the potential of sur-vey biases cannot be excluded. Admittedly, therespondents’ perceptions regarding the issues cen-tral to this study might not necessarily coincideexactly with objective reality. This could resultin potential biases. First, subjective biases maystem from an implicit tendency of respondents torationalize their firms’ competitive behavior basedon received wisdom about what constitutes effec-tive management praxis. Second, respondents’ sub-jective definition of industry boundaries couldaffect the accuracy of comparisons with rivals asregards their firms’ strategies, unique assets andperformance. This may be particularly problem-atic with respect to international rivals as it isnot entirely clear whether respondents were ableto accurately compare their firms’ situation withthat of international competitors represented (asopposed to those actively present) in the localmarket.

Apart from the disadvantages related with per-ceptual measures, the time period used for assess-ing the sustainability of performance is admittedlyshort (i.e., previous three years) to account forany business cycle effects or transient problems.It is important to note, however, that a longertime-frame (e.g., five instead of three years) couldendanger the reliability of responses. Another lim-itation of the research is related to the relativelysmall sample size which may have led to non-response bias, despite evidence pointing to thecontrary. In addition, our sampling frame excludedthose international competitors that are not oper-ating locally. This might have introduced somedegree of sample selection bias. It should be noted,however, that including these firms in the sam-ple through their local representatives, would haveintroduced problems with respect to data qualityand comparability. Finally, the research design wascross-sectional, not longitudinal. As such, cause-effect relationships may not be definitively inferredfrom the results.

At the theoretical level, it should be emphasizedthat while the study seeks to identify the relativeimpact of industry vs. firm-specific factors onperformance, further theoretical work is needed



to extend the inherently static character of thisresearch program, by bringing together an explicitexamination of the dynamics of the processes bywhich firms build their competencies and engagein strategic maneuvering within a given industry.To be sure, this mode of theorizing pertains to aconceptually rich, fine grained theory of the firmthat will seek to integrate both “content” (i.e.,conditions for achieving sustainable competitiveadvantage) and “process” (i.e., an account of thedynamics of firm growth and change) aspects ofstrategy research.

Such an attempt to develop a model that tran-scends the theoretical underpinnings of both per-spectives in a dynamic sense, is beyond the scopeof this study. Instead, our main emphasis here wasto statistically examine effects (industry and firm-specific) within a static equilibrium context, notthe dynamics of the processes by which they havebeen evolved, or better, by which they continu-ously develop.

On this account, it could be argued that thePorter, and to a lesser extent, the RBV perspec-tives lack a clearly delineated theory of the firmwithin their conceptual apparatus. For one, Porter’sframework has relatively little to say about pro-cessual and behavioral issues (Foss, 1996) that bynecessity should be accounted for in any attemptto understand firms as dynamic and ever chang-ing entities. This is because Porter, despite thefact that his later work connects in importantways to RBV, argues that firm performance canbe ultimately traced back to successive managerialchoices and external (to the firm) initial conditions(Porter, 1991), saying nothing about the internalorganizational processes that produced such cre-ative choices.

On the other hand, RBV is admittedly bettersuited to deal with organizational and behavioralprocesses in comparison to the IO inspired Porterframework, since it views resources as encompass-ing a broad range of organizational, social and indi-vidual phenomena which have traditionally beenthe subject of organization theory and organiza-tional behavior (Barney, 1991). But resource-basedscholars have only recently began to systemati-cally explore these dynamic issues. The “processbranch”, focusing on the process side (Mahoney,1995), represents the dynamic capabilities turn inthe RBV school of thought.

Within this latter line of research there appearsa gradual progression in focus from an account of

which resources (and why) may be valuable, tohow these may be created (Galunic and Rodan,1998). In this respect, a number of resource basedscholars have recently begun to explore the gen-eral processes by which organizational capabilitiesare developed (see for example Dierickx and Cool,1989; Henderson, 1994; Leonard-Barton, 1995;Pisano, 1994; Teece et al., 1997) to form higherorder complex collectives (i.e., core competen-cies). It could be argued however that althoughpromising, the “process branch” of the resource-based theorizing is still underdeveloped. It is per-haps this underdevelopment that underlies whatFoss (1997b) notes as a lack of integration betweenan analysis of firm growth and change that wouldprovide an account of the sources of sustainedcompetitive advantage.

Within this discourse, one perhaps most promis-ing line of research is the emerging Knowledge-based theory of the firm which posits that thefundamental input and primary source of value inbuilding organizational capabilities is knowledge(Grant, 1996a). As Kogut and Zander (1992) putit, the theoretical challenge is to understand theknowledge base of a firm as resulting to a set ofcapabilities that constitute its sources of compet-itive advantage. Interestingly enough, this appre-ciation of knowledge as a fundamental factor, isalready anticipated (as with a large part of currentresource-based thinking) in the work of Penrose.In particular, Penrose (1959) argues that the poten-tial uses of services rendered by firm resourcesvary along with changes in firm knowledge. Fur-thermore, Penrose, by introducing the distinctionbetween “objective” knowledge and experience,and by emphasizing the critical role of the latterin firm growth, preempts the now common dis-tinction between explicit and tacit knowledge (seefor example Grant, 1996a,b; Nonaka, 1994) intro-duced by Polanyi (1966).

But why organizational knowledge? Becauseknowledge may be argued to constitute the keyconceptual lens for understanding the firm as adynamic and ever evolving social micro-system.For it is through shared and socially embeddedknowledge that organizational members interpretenvironmental stimuli (e.g., competitors’ moves,customers’ changing needs, emerging technologiesetc). These interpretations may in turn igniteinternal adaptive responses to build appropriateskills and to synthesize these into core compe-tencies, that are themselves (these responses) the



product of knowledge exploitation and integra-tion. Within this line of reasoning then, it isthe quality of knowledge created (Nonaka, 1994)and of integration mechanisms (Grant, 1996a,b)that constitute the ultimate sources of sustainedcompetitive advantage. While the links betweenknowledge and firm capabilities on the one hand,and between knowledge and competitive advan-tage on the other, have been further elaboratedin the work of Fiol (1991), Hamel (1991), Grant(1996a,b), Nonaka (1994) and Reed and deFillipi(1990) among others, no detailed explanationsare offered as to the origin of organizationalcapabilities (Verona, 1999). Further theoreticalwork is needed to address the systemic subtletiesinvolved in organizational knowledge creation andexploitation.

It also seems that this theoretical perspectiveis not yet mature enough to allow for empiricaltesting, at least not in the manner of large scalesurveys that are common in strategy research.Instead, more qualitative (e.g., ethnographic)methodological approaches are warranted here,given the inherently abstract nature of conceptssuch as knowledge creation and exploitationthat would necessarily be involved in such aresearch program. How, for example, managerialexperience and the related “team capital” (Penrose,1959), or idiosyncratic firm history (Teece et al.,1997) can be evaluated and “measured” againstcomplex processes of transformational change(e.g., repositioning)?

ACKNOWLEDGEMENTS

The authors wish to acknowledge the contribu-tion of Associate Editor Karel Cool and reviewerswhose critical comments have led them to substan-tially improve the paper. We also would like tothank D. Bourantas, N. Lambroukos, A. Athanassopoulos, N. Vonortas and Y. Caloghirou for theirassistance and comments in various stages of theresearch.

REFERENCES

Achrol R, Stern L. 1988. Environmental determinants ofdecision making uncertainty in marketing channels.Journal of Marketing Research 25: 36–50.

Aldrich HE. 1979. Organizations and Environments .Prentice-Hall: Englewood Cliffs, NJ.

Amit R, Schoemaker P. 1993. Strategic assets and orga-nizational rent. Strategic Management Journal 14(1):33–46.

Andrews KR. 1971. The Concept of Corporate Strategy .Dow Jones-Irwin: New York.

Ansoff I. 1965. Corporate Strategy . McGraw Hill: NewYork.

Armstrong JS, Overton T. 1977. Estimating nonresponsebias in mail surveys. Journal of Marketing Research14: 396–402.

Bain J. 1951. Relation of profit rate to industry concentra-tion: American manufacturing, 1936–1940. QuarterlyJournal of Economics 65: 293–324.

Bain J. 1956. Barriers to New Competition . Harvard Uni-versity Press: Cambridge, MA.

Bain J. 1968. Industrial Organization . Wiley: New York.Barney J. 1986a. Strategic factor markets: expectations,

luck and business strategy. Management Science 32:1231–1241.

Barney J. 1986b. Organizational culture: can it be asource of competitive advantage? Academy of Man-agement Review 11(3): 656–665.

Barney J. 1991. Firm resources and sustained competitiveadvantage. Journal of Management 17: 99–120.

Barney J. 1992. Integrating organizational behavior andstrategy formulation research: a resource based anal-ysis. In Advances in Strategic Management , Vol. 8,Shristava P, Huff A, Dutton J (eds). JAI Press: Green-wich, CT; 39–62.

Barney J, Griffin R. 1992. The Management of Organi-zations: Strategy, Structure, and Behavior . HoughtonMifflin: Boston, MA.

Barney J, Zajac E. 1994. Competitive organizational be-havior: toward an organizationally-based theory ofcompetitive advantage. Strategic Management Jour-nal , Winter Special Issue 15: 5–9.

Bentler P. 1990. Comparative fit indexes in structuralmodels. Psychological Bulletin 197: 238–246.

Bentler P, Wu E. 1995. EQS for Windows User’s Guide.Multivariate Software: Encino, CA.

Buzzel R, Gale B, Sultan R. 1975. Market share: a key toprofitability. Harvard Business Review 53(1): 97–106.

Capon N, Farley J, Hoenig S. 1990. Determinants offinancial performance: meta-analysis. ManagementScience 36: 97–107.

Carrol GR. 1993. A sociological view on why firms dif-fer. Strategic Management Journal 14(4): 237–249.

Caves EE. 1972. American Industry: Structure, Conduct,and Performance (3rd edn). Prentice-Hall: EnglewoodCliffs, NJ.

Chamberlin E. 1962. The Theory of Monopolistic Compe-tition: A Re-examination of the Theory of Value, (8thedn). Harvard University Press: Cambridge, MA.

Chattopadhyay P, Glick W, Miller CC, Huber G. 1999.Determinants of executive beliefs: comparing func-tional conditioning and social influence. StrategicManagement Journal 20(8): 763–789.

Child J. 1972. Organizational structure, environment andperformance: the role of strategic choice. Sociology6(1): 1–22.



Cohen J, Cohen P. 1983. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences . Law-rence Erlbaum: Hillsdale, NJ.

Collis D. 1994. How valuable are organizational capabil-ities? Strategic Management Journal , Winter SpecialIssue 15: 143–152.

Collis DJ, Montgomery CA. 1995. Competing on resour-ces: Strategy in the 1990s. Harvard Business Review73(4): 118–128.

Conner K. 1991. An historical comparison of resource-based theory and five school of thought within indus-trial organization economics: do we have a new theoryof the firm? Journal of Management 17: 121–154.

Dess G, Beard D. 1984. Dimensions of organizationaltask environments. Administrative Science Quarterly29: 52–73.

Dess G, Davis P. 1984. Porter’s (1980) generic strategiesas determinants of strategic group membership andorganizational performance. Academy of ManagementJournal 27: 467–488.

Dess G, Robinson R. 1984. Measuring organizationalperformance in the absence of objective measures:the case of the privately-held firm and conglomer-ate business unit. Strategic Management Journal 5(3):265–273.

Dierickx I, Cool K. 1989. Asset stock accumulation andsustainability of competitive advantage. ManagementScience 35: 1504–1511.

Dillman D. 1978. Mail and Telephone Surveys: The TotalDesign Method . Wiley: New York.

Fiol C. 1991. Managing culture as a competitive resource:an identity-based view of sustainable competitiveadvantage. Journal of Management 17: 191–211.

Fornell C, Larcker DF. 1981. Evaluating structural equa-tion models with unobservable variables and measure-ment error. Journal of Marketing Research 18: 39–50.

Foss N. 1996. Research in strategy, economics, andMichael Porter. Journal of Management Studies 33:1–24.

Foss N. 1997a. Resources and strategy: a brief overviewof themes and contributions. In Resources, Firms andStrategies , Foss N (ed.). Oxford University Press:New York; 3–18.

Foss N. 1997b. Resources and strategy: problems, openissues, and ways ahead. In Resources, Firms andStrategies , Foss N (ed.). Oxford University Press:New York; 345–365.

Gale BT. 1972. Market share and rate of return. Reviewof Economics and Statistics 54(4): 412–423.

Galunic DC, Rodan S. 1998. Resource recombinationsin the firm: knowledge structures and the potentialfor Schumpetarian innovation. Strategic ManagementJournal 19(12): 1193–1201.

Gerbing DW, Anderson JC. 1988. An updated paradigmfor scale development incorporating unidimensionalityand its assessment. Journal of Marketing Research 25:186–192.

Grant RM. 1991. Contemporary Strategy Analysis . Black-well: Oxford.

Grant RM. 1996a. Toward a knowledge-based theoryof the firm. Strategic Management Journal , WinterSpecial Issue 17: 109–122.

Grant RM. 1996b. Prospering in dynamically-competitiveenvironments: organizational capability as knowledgeintegration. Organization Science 7: 375–387.

Hambrick D, Snow C. 1977. A contextual model ofstrategic decision making in organizations. Academyof Management Proceedings : 108–112.

Hamel G. 1991. Competition for competence and inter-partner learning within international strategic alliances.Strategic Management Journal , Summer Special Issue12: 83–104.

Hannan MT, Freeman J. 1977. The population ecology oforganizations. American Journal of Sociology March:929–964.

Hansen G, Wernerfelt B. 1989. Determinants of firm per-formance: the relative importance of economic andorganizational factors. Strategic Management Journal10(5): 399–411.

Hart S, Banbury C. 1994. How strategy-making processcan make a difference. Strategic Management Journal15(4): 251–270.

Henderson R. 1994. The evolution of integrative capa-bility: innovation in cardiovascular drug discovery.Industrial and Corporate Change 3: 607–630.

Henderson R, Mitchell W. 1997. The interactions oforganizational and competitive influences on strat-egy and performance. Strategic Management Journal ,Summer Special Issue 18: 5–14.

Hill C, Hansen G. 1991. A longitudinal study of the causeand consequences of changes in diversification in theU.S. pharmaceutical industry, 1977–1986. StrategicManagement Journal 12(3): 187–199.

Hrebriniak L, Joyce W. 1985. Organizational adaptation:strategic choice and environmental determinism.Administrative Science Quarterly 35: 336–349.

Imel B, Helmberger P. 1971. Estimation of structure—profit relationships with application to the food pro-cessing sector. American Economic Review 62:614–627.

Jacobson R. 1990. What really determines business per-formance? Unobservable effects— the key to prof-itability. Management Science 9: 74–85.

Keats B, Hitt M. 1988. A causal model of linkages amongenvironmental dimensions. Macro organizational char-acteristics and performance. Academy of ManagementJournal 31: 570–598.

Khandwalla P. 1977. The Design of Organizations . Har-court, Brace, Jovanovitch: New York.

Kogut B, Zander U. 1992. Knowledge of the firm, com-binative capabilities and the replication of technology.Organization Science 3: 383–397.

Lado A, Boyd N, Wright P. 1992. A competency basedmodel of sustainable competitive advantage: towarda conceptual integration. Journal of Management 18:77–91.

Lancaster KJ. 1966. A new approach to consumer theory.Journal of Political Economy 74(2): 132–157.

Lancaster KJ. 1971. Consumer Demand: A New Appro-ach . Columbia University Press: New York.

Lawless M, Finch L. 1989. Choice and determinism: atest of Hrebriniak and Joyce’s framework on strategy-environment fit. Strategic Management Journal 10(4):351–365.



Lefebvre L, Mason R, Lefebvre E. 1997. The influenceprism in SMEs: the power of CEOs’ perceptionson technology policy and its organizational impacts.Management Science 43: 856–878.

Leonard-Barton D. 1995. Wellsprings of Knowledge:Building and Sustaining the Sources of Innovation .Harvard Business School Press: Boston, MA.

Mahoney JT. 1995. The management of resources and theresource of management. Journal of Business Research33: 91–101.

Mahoney J, Pandian J. 1992. The resource-based viewwithin the conversation of strategic management.Strategic Management Journal 13(5): 363–380.

Maijoor S, Van Witteloostuijn A. 1996. An empirical testof the resource-based theory: strategic regulation in theDutch audit industry. Strategic Management Journal17(7): 549–569.

Mason E. 1939. Price and production policies of large-scale enterprises. American Economic Review 29:61–74.

Mauri AJ, Michaels MP. 1998. Firm and industry effectswithin strategic management: an empirical examina-tion. Strategic Management Journal 19(3): 211–219.

McGahan A, Porter M. 1997. How much does industrymatter, really? Strategic Management Journal , Sum-mer Special Issue 18: 15–30.

McGee J, Thomas H. 1986. Strategic groups: theory,research and taxonomy. Strategic Management Jour-nal 7(2): 141–160.

McWilliams A, Smart DL. 1993. Efficiency v. struc-ture conduct performance: implications for strategyresearch and practice. Journal of Management 19:63–79.

Miller D. 1988. Relating Porter’s business strategies toenvironment and structure: analysis and performanceimplications. Academy of Management Journal 31:280–308.

Miller D. 1992. The generic strategy trap. The Journal ofBusiness Strategy Jan–Feb, 37–41.

Miller D, Friesen P. 1986b. Porter’s (1980) generic strate-gies and performance: an empirical examination withAmerican data (Part II). Organization Studies 7:255–261.

Miller D, Shamsie J. 1996. The resource-based view ofthe firm in two environments: the Hollywood filmstudios from 1936 to 1965. Academy of ManagementJournal 39: 519–543.

Naman JL, Slevin DP. 1993. Entrepreneurship and theconcept of fit: a model and empirical tests. StrategicManagement Journal 14(2): 137–154.

Nonaka I. 1994. A dynamic theory of organizationalknowledge creation. Organization Science 5: 14–36.

Nunnally JC. 1978. Psychometric Theory , 2nd edn.McGraw-Hill: New York.

Penrose ET. 1959. The Theory of the Growth of the Firm .Oxford University Press: Oxford.

Peteraf M. 1993. The cornerstones of competitive advan-tage: a resource-based view. Strategic ManagementJournal 14(3): 179–191.

Phillips L, Chang D, Buzzell R. 1983. Product quality,cost position and business performance: a test of somekey hypotheses. Journal of Marketing 47: 26–43.

Pisano G. 1994. Knowledge, integration, and the locusof learning: an empirical analysis of process develop-ment. Strategic Management Journal , Winter SpecialIssue 15: 85–100.

Polanyi M. 1966. The Tacit Dimension . Routledge: Lon-don.

Porter M. 1980. Competitive Strategy: Techniques forAnalyzing Industries and Competitors . Free Press:New York.

Porter M. 1985. Competitive Advantage: Creating andSustaining Superior Performance. Free Press: NewYork.

Porter M. 1990. The Competitive Advantage of Nations .Macmillan: New York.

Porter M. 1991. Towards a dynamic theory of strategy.Strategic Management Journal , Winter Special Issue12: 95–117.

Powell TC, Dent-Micallef A. 1997. Information technol-ogy as competitive advantage: the role of human, busi-ness, and technology resources. Strategic ManagementJournal 18(5): 375–405.

Prescott JE, Kohli AK, Venkatraman N. 1986. The mar-ket share—profitability relationship: an empirical ass-essment of major assertions and contradictions. Strate-gic Management Journal 7(4): 377–394.

Reed R, DeFillippi R. 1990. Causal ambiguity, barriersto imitation, and sustainable competitive advantage.Academy of Management Review 15: 88–102.

Robinson R, Pearce J. 1988. Planned patterns of strategicbehavior and their relationship to business unit perfor-mance. Strategic Management Journal 9(1): 43–60.

Rumelt R. 1991. How much does industry matter? Strate-gic Management Journal 12(3): 167–185.

Rumelt R. 1984. Toward a strategic theory of the firm.In Competitive Strategic Management , Lamb R (ed.).Prentice-Hall: Englewood Cliffs, NJ; 556–570.

Rumelt R, Wensley R. 1981. Market share and the rateof return: testing the stochastic hypothesis. Workingpaper, University of California, Los Angeles.

Sapienza HJ, Smith KG, Gannon MJ. 1988. Using sub-jective evaluations of organizational performance insmall business research. American Journal of SmallBusiness , Winter: 45–53.

Schmalensee R. 1985. Do markets differ much? AmericanEconomic Review 75(3): 341–351.

Schoeffler S, Buzzell R, Heany D. 1974. Impact of strate-gic planning on profit performance. Harvard BusinessReview 52(3): 137–145.

Shepherd WG. 1972.The Economics of Industrial Orga-nization . Prentice-Hall: Englewood Cliffs, NJ.

Snow C. 1976. The role of managerial perceptions inorganizational adaptation: an exploratory study. Aca-demy of Management Proceedings: 249–255.

Teece D. 1984. Economic analysis and strategic mana-gement. California Management Review 26(3):87–110.

Teece D. 1986. Profiting from technological innovation.Research Policy 15: 285–305.

Teece D, Pisano G, Shuen A. 1997. Dynamic capabili-ties and strategic management. Strategic ManagementJournal 18(7): 509–533.



Venkatraman N. 1989. Strategic orientation of businessenterprises: the construct, dimensionality, and mea-surement. Management Science 35(8): 942–962.

Venkatraman N, Ramanujam V. 1986. Measurement ofbusiness performance in strategy research: a compar-ison of approaches. Academy of Management Review11: 801–814.

Venkatraman N, Ramanujam V. 1987. Measurement ofbusiness economic performance: an examination ofmethod convergence. Journal of Management 13:109–122.

Verona G. 1999. A resource-based view of product devel-opment. Academy of Management Review 24:132–142.

Weick K. 1979. The Social Psychology of Organizing .Addison-Westley: Reading, MA.

Wernerfelt B. 1984. A resource-based view of the firm.Strategic Management Journal 5(2): 171–180.

Widaman KF. 1985. Hierarchicaly nested covariancestructure models for multitrait-multimethod data. Ap-plied Psychological Measurement 9(1): 1–26.

Williamson OE. 1991. Strategizing, economizing, andeconomic organization. Strategic Management Jour-nal , Winter Special Issue 12: 75–94.

Woo C, Willard G. 1983. Performance representation inbusiness policy research: discussion and recommen-dation. Paper presented at the 23rd Annual NationalMeetings of the Academy of Management, Dallas, TX.


Causal

Logic

ofR

entG

eneration929

APPENDIX 1: DESCRIPTIVE STATISTICS AND PEARSON CORRELATIONS AMONG VARIABLES

Variable

Mean SD 1 2 3 4 5 6 7 8 9 10

(1) Number of employees (log) 4.36 1.22 1(2) Threat of substitutes 2.89 1.58 0.03 1(3) Entry barriers 3.63 1.33 −0.01 0.1 1(4) Power of suppliers 3.59 1.00 −0.22∗∗ 0.17 −0.37∗∗∗ 1(5) Competitive rivalry 3.02 1.08 0.12 0.27∗∗∗ 0.03 0.16 1(6) Power of buyers (log) 2.88 1.07 0.00 −0.07 0.11 −0.03 0.09 1.00(7) Strategy 3.43 0.74 0.28∗∗∗ −0.05 0.01 −0.40∗∗∗ −0.24∗∗ 0.04 1.00(8) Firm Assets 3.50 0.66 0.21∗∗ −0.01 0.09 −0.32∗∗∗ −0.34∗∗∗ −0.04 0.70∗∗∗ 1.00(9) Profitability 2.90 1.02 0.29∗∗∗ −0.04 0.19∗ −0.49∗∗∗ −0.11 −0.06 0.44∗∗∗ 0.41∗∗∗ 1.00(10) Market Performance 3.49 0.74 0.30∗∗∗ −0.09 0.14 −0.36∗∗∗ −0.32∗∗∗ 0.04 0.67∗∗∗ 0.64∗∗∗ 0.51∗∗∗ 1.00

∗p ≤ 0.10; ∗∗ p ≤ 0.05; ∗ ∗ ∗p ≤ 0.01.

Copyright

2001John

Wiley

&Sons,

Ltd.

Strat.M

gmt.

J.,22:

907–

934(2001)


APPENDIX 2: MEASURES ANDCONSTRUCT VALIDATION RESULTS

Content validity

Most of the scales employed have been adoptedfrom existing and validated scales used in theextant literature. With respect to capabilities mea-sures, for which no established relevant scalesexist, specific items were developed based on theo-retical contributions from resource-based scholarsas well as from in-depth discussions with aca-demics and CEOs during the pre-testing phase ofquestionnaire development. Moreover, as describedin the sample and data collection section, consider-able efforts were made during the field-based vali-dation of the research instrument to ensure contentvalidity via establishing relevance with practiceand the elimination of wording problems (such asbiased, ambiguous, inappropriate or double mean-ing items).

Construct validity

To establish construct validity, a series of empiricaltests were used to examine the measurement prop-erties of the indicators, namely unidimensionality,reliability and validity. After an initial examinationprocedure that sought to identify items exhibitinglow item-to-construct correlation or items loadingsignificantly to more than one construct dimension,we tested the construct validity of our measuresby employing confirmatory factor analysis (CFA)using EQS (Bentler and Wu, 1995). Unlike the tra-ditional and more commonly used exploratory fac-tor analysis (EFA), CFA contains inferential statis-tics that allow for hypothesis testing regarding theconstruct validity of a set of measures, leading to astricter and more objective interpretation of valid-ity than does EFA (Gerbing and Anderson, 1988).

Unidimensionality

Unidimensionality in our case means that, forexample, the set of indicators gauging innovativedifferentiation strategy, relate exclusively to thisconstruct and not to another, say, marketing differ-entiation. Two sets of statistics were used for theverification of the unidimensionality hypothesis:(a) the significance of the factor loadings, that isthe estimated correlation between a particular itemand the latent construct it represents, and (b) the

overall acceptability of the measurement model interms of the model’s fit to the data, using a X2 testand adjunct fit indexes (see Table I). In our casethe two first-order measurement models (i.e., Per-formance and Competitive Rivalry) exhibit accept-able model fit, and all item-to-construct loadingsare statistically significant, thus demonstrating theunidimensionality of the scales used.

The same holds for the two second-order mea-surement models of Strategy and Firm Assets. Asecond-order model of say, competitive strategy,is based on a hierarchical structure in which Strat-egy is assumed to affect more specific strategydimensions (i.e., innovative differentiation, mar-keting differentiation and low cost) which in turnare measured by the specific items. In this concep-tual view, Strategy is a higher (second) order, moreabstract construct that is not directly measured.In contrast, more specific strategy dimensions areviewed as lower (first) order factors that are pre-sumed to be caused by Strategy. The dimensional-ity of this hypothesized structure, of both Strategyand Firm Assets, was supported by our data asmanifested by the overall acceptability of the tworespective measurement models, in terms of theCFI (and Robust CFI) fit index that exceeds thecut-off point of 0.90, and also by the significanceof the first and second order factor loadings.

Reliability

With respect to reliability, we computed the com-posite reliability estimates (Fornell and Larker,1981) which are directly analogous to the com-monly used coefficient alpha statistics. As shownin Table II all are quite satisfactory (well abovethe customary cut-off level of 0.70) thus provid-ing confidence that the individual items used areall consistent in their measurements and reliable.In addition all coefficients exceeded Nunnally’s(1978) recommended 0.70 level of internal con-sistency.

Convergent validity

Convergent validity was examined by comput-ing the indexes of average variance extracted,that is the amount of construct variance rel-ative to measurement error. An average vari-ance extracted of at least 0.50 (i.e., 50 percent)provides support for convergent validity (Gerb-ing and Anderson, 1988; Fornell and Larcker,



Table I. Measures and tests of unidimensionality

STRATEGY

Please indicate the extent to which you use each of the following competitive methods (1: much less thancompetitors . . .5: much more than competitors)

First order Second orderMeasures loadings loadings

Innovative 0.991Differentiation R&D expenditures for product development 0.740a

R&D expenditures for process innovations 0.802Emphasis on being ahead of competition 0.671Rate of product innovations 0.818

Marketing 0.811Differentiation Innovations in marketing techniques 0.895a

Emphasis on marketing department organization 0.893Advertising expenditures 0.698Emphasis on strong sales force 0.716

Low Cost 0.846Modernization and automation of production processes 0.800a

Efforts to achieve economies of scale 0.717Capacity utilisation 0.647

Model Summary Statistics: X2(44) = 99.526; p < 0.001; CFI = 0.935; Robust CFI = 0.941All first and second order loadings significant at p < 0.01aLoading fixed to 1 for identification purposes

FIRM ASSETS

Please indicate for each of the following, you firm’s strength relative to competition (1: much weaker thancompetitors. . .5: much stronger than competitors)

First order Second orderMeasures loadings loadings

Organizational/ 0.685Managerial Managerial competencies 0.765a

Knowledge and skills of employees 0.763Firm climate 0.661Efficient organisational structure 0.773Coordination 0.608Strategic planning 0.646Ability to attract creative employees 0.617

Marketing 0.764Market knowledge 0.699a

Control and access to distribution channels 0.636Advantageous relationships with customers 0.789Customers “installed base” 0.748

Technical 0.893Efficient and effective production department 0.814a

Economies of scales and technical experience 0.759Technological capabilities and equipment 0.706

Model Summary Statistics: X2(75) = 141.138; p < 0.001; CFI = 0.922; Robust CFI = 0.920All first and second order loadings significant at p < 0.01a Loading fixed to 1 for identification purposes



Table I. (Continued )

PERFORMANCE

Please indicate for each of the following, your firm’s performance relative to competition for the last three years(1: much below the average. . .5: much above the average)

First orderMeasures loadings

Market Position Sales volume 0.803Growth in sales volume 0.755Market share 0.749Growth in market share 0.789

Profitability Profit margin 0.840Return on own capital 0.810Net profits 0.839

Model Summary Statistics: X2(13) = 64.897; p < 0.001; CFI = 0.924; Robust CFI = 0.930All first order loadings significant at p < 0.01

INDUSTRY FORCES

How would you evaluate the intensity of competition your firm is facing with respect to: (1: very weakcompetition. . .5: very fierce competition)

First orderMeasures loadings

Competitive Product characteristics 0.612Rivalry Promotional strategies among rivals 0.774

Access to distribution channels 0.707Service strategies to customers 0.679

Model Summary Statistics: X2(2) = 7.278; p = 0.026; CFI = 0.968; Robust CFI = 0.995All first order loadings significant at p < 0.01

INDUSTRY FORCES: Single item measures

Barriers to Entry (1: very easy to enter. . .5: very difficult)Threat of Substitutes (1: not at all. . .5: extreme)Bargaining Power over Suppliers (1: very weak. . .5: very strong) (reverse scored to indicate power of suppliers)Bargaining Power of Buyers (% of sales to three biggest buyers)

1981). In our case (see Table II) all constructsexceeded the cut-off point with the exception oforganizational/managerial assets and competitiverivalry constructs, which are nevertheless veryclose to being acceptable (0.48 in both cases).

As a further test of convergent validity of theperformance measures, the accounting measures ofthe 107 firms (out of the original 187) for whichfinancial data could be obtained were correlatedwith their subjective responses. In this subsample,

return on sales, return on own capital and net profitsaveraged over the three year period covered by thesurvey correlated significantly with the respectivesubjective measure (0.48, 0.44 and 0.53, ρ < 0.01respectively). These correlations between objec-tive and subjective performance measures grew inmagnitude when industry effects were controlledfor, by examining more homogenous subsets offirms. For example, when performance measuresconvergence was analyzed for firms in the food



Table II. Reliability and convergent validity tests

Construct VarianceReliabilitya Cronbach’sa Extractedb

Conceptual Domain: StrategyInnovative Differentiation 0.84 0.82 0.58Marketing Differentiation 0.88 0.86 0.65Low Cost 0.77 0.73 0.52

Firm AssetsOrganizational/Managerial 0.87 0.88 0.48Marketing 0.81 0.77 0.52Technical 0.80 0.80 0.58

PerformanceMarket position 0.86 0.85 0.60Profitability 0.87 0.87 0.60

Industry ForcesCompetitive Rivalry 0.79 0.83 0.48

a Construct reliability (see Fornell and Larcker, 1981 for computation formula)b Average variance extracted (i.e., the proportion of variance in the construct that is not dueto measurement error) (see Fornell and Larcker, 1981 for computation formula)

industry, the correlations increased to 0.71, 0.60and 0.65 respectively. These results are consistentwith recent research which indicates that subjec-tive assessments of business performance obtainedby senior managers correlate strongly, albeit notperfectly with objective measures (see for exam-ple, Dess and Robinson, 1984; Hart and Banbury,1994; Naman and Slevin, 1993; Venkatraman andRamanujam, 1986, 1987).

Moreover, in order to examine for the potentialof cyclical effects going on in the industries stud-ied, we collected data on indexes of productionover a six years period (i.e., for the 1991-93

period on which subjective performance evalua-tions were referring, and for the consecutive threeyears period 1994-96). The inspection of variabil-ity of industrial production showed that only for 3out of the 13 industries studied, cyclical effects ofrelatively medium magnitude were present. There-fore it appears that cyclical effects were not aninhibiting factor for the validity of results obtained.

Discriminant validity

Discriminant validity is established when it can beshown that two or more constructs pertaining to

Table III. Discriminant validity tests

Pair of Constructs (� = 1)

Conceptual Domain: Strategy X2(d.f. = 45)Innovative Differentiation vs. Marketing Differentiation 136.938 (p1 < 0.01)Innovative Differentiation vs. Low Cost 106.430 (p < 0.01)Marketing Differentiation vs. Low Cost 156.204 (p < 0.01)Base Model (unconstrained) X2 = 99.526 (d.f. 44)

Firm Assets X2(d.f. = 76)Organizational/Managerial vs. Marketing 224.850 (p < 0.01)Organizational/Managerial vs. Technical 188.389 (p < 0.01)Marketing vs. Technical 181.082 (p < 0.01)Base Model (unconstrained) X2 = 141.138 (d.f. 75)

Performance X2(d.f. = 20)Market position vs. Profitability 177.194 (p < 0.01)Base Model (unconstrained) X2 = 64.897 (d.f. 19)

1 Denotes the significance of X2 difference between the constrained and unconstrained model



the same conceptual domain (e.g., the three genericstrategies) are not perfectly correlated (Widaman,1985). A commonly used approach for testing thediscriminant hypothesis compares two CFA mod-els: one in which the correlation of a pair of latentvariables is constrained to equal 1.0, and another

in which the correlation is free to vary (Venkatra-man, 1989). A significantly lower X2 value for theunconstrained model, provides support for discrim-inant validity. As shown in Table III, all modelcomparisons support the discriminant validity ofour measures.


an examination into the causal logic of rent generation ... · strategic management journal strat....

Documents