firm-specific knowledge resources and competitive … · 2020. 1. 13. · 1266 h. c. wang, j. he,...

Strategic Management JournalStrat. Mgmt. J., 30: 1265–1285 (2009)

Published online EarlyView in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/smj.787

Received 22 February 2007; Final revision received 8 May 2009

FIRM-SPECIFIC KNOWLEDGE RESOURCES ANDCOMPETITIVE ADVANTAGE: THE ROLES OFECONOMIC- AND RELATIONSHIP-BASED EMPLOYEEGOVERNANCE MECHANISMS

HELI C. WANG,1* JINYU HE,1 and JOSEPH T. MAHONEY2

1 Department of Management of Organizations, School of Business and Management,Hong Kong University of Science and Technology, Kowloon, Hong Kong2 Department of Business Administration, College of Business, University of Illinoisat Urbana-Champaign, Champaign, Illinois, U.S.A.

The resource-based view of the firm emphasizes the role of firm-specific resources, especially firm-specific knowledge resources, in helping a firm to achieve sustainable competitive advantage.However, the deployment of firm-specific knowledge often requires key employees to makespecialized human capital investments that are not easily redeployable to other settings. Thus, inthe absence of effective safeguards and trust building devices, employees with foresight maybe reluctant to make such specialized investments. This study explores both economic- andrelationship-based governance mechanisms that might mitigate this underinvestment problem.Effective use of these governance mechanisms enables a firm to obtain greater performancefrom its efforts to deploy firm-specific knowledge resources. Empirical results further supportthese key arguments. Copyright 2009 John Wiley & Sons, Ltd.

INTRODUCTION

According to the resource-based view of the firm,a firm’s ability to achieve and sustain a competi-tive advantage is directly related to the strength of‘isolating mechanisms’ that protect the firm’s valu-able and rare resources from imitation by rivals(Mahoney and Pandian, 1992; Rumelt, 1984).An important isolating mechanism is the firm-

Keywords: Firm-specific knowledge resources; employeehuman capital; employee governance mechanisms;resource-based view of the firm*Correspondence to: Heli C. Wang, Department of Managementof Organizations, School of Business and Management, HongKong University of Science and Technology, Kowloon, HongKong. E-mail: [email protected]

specificity of resources. Resources with such afeature are not easily tradable or redeployable out-side the firm (Dierickx and Cool, 1989), makingit difficult for rivals to imitate them. Resource-and knowledge-based research generally maintainsthat among the types of firm-specific resourcesexamined, firm-specific knowledge has the great-est potential to serve as a source of sustainablecompetitive advantage (Coff, 1997; Grant, 1996;Kogut and Zander, 1992).

A firm’s knowledge base is the informationinputs, know-how, and capabilities that organi-zational members draw on when searching forinnovative solutions (Dosi, 1988). Firm-specificknowledge often results from a search for, andaccumulation of, new solutions that build upon

Copyright 2009 John Wiley & Sons, Ltd.

1266 H. C. Wang, J. He, and J. T. Mahoney

a firm’s established knowledge base (Cohen andLevinthal, 1989; Nelson and Winter, 1982; Teece,1986). Such firm-specific knowledge may includethe ability to operate and maintain customizedequipment and familiarity with specialized prac-tices for developing and manu-facturing uniqueproducts (Døving and Nordhaug, 2002).

However, rarely can a firm automatically achievesuperior economic performance from its firm-specific knowledge resources. Instead, a firm usu-ally requires its key employees to make com-plementary investments in human capital in theprocess of absorbing and deploying firm-specificknowledge. When a firm’s rent-generating knowl-edge resources are to a large extent firm-specific,the human capital investments required of itsemployees are also likely to increase in speci-ficity. Similar to other types of firm-specific invest-ments, the employees’ firm-specific human capi-tal is imperfectly redeployable, or is valued lessin the external labor market than within the firm(Williamson, 1985). As a result, in the absence ofcomplete contracting, the employees’ investmentsgive the firm opportunities to hold employees upex post (through, for example, lower compensationand extended work hours). Thus, key employeeswith foresight might be reluctant to make suchfirm-specific investments that would place themin a weak bargaining position. Without effectivesafeguards and/or substantial trust between thefirm and its key employees, a firm might not beable to realize the potential economic rents thatcan be generated from its firm-specific knowledgeresources (Cornell and Shapiro, 1987).

This scenario can be illustrated by the exam-ple of Intel employees. In 1970, Intel plannedto invest in developing the first semiconductorDRAM (dynamic random access memory), the 1kilobit ‘1103.’ Despite the economic attractivenessof the project, however, Intel’s engineers wereseriously concerned about the potential negativeconsequences of developing knowledge and skillsspecific to DRAM technology. According to Gor-don Moore, then CEO of Intel, ‘There was a lotof resistance to semiconductor technology on thepart of the core memory engineers. The engineersdidn’t embrace the 1103 until they realized thatit wouldn’t make their skills irrelevant’ (Coganand Burgelman, 1989: 2–3). The Intel exampleshows that while firm-specific knowledge can gen-erate competitive advantage for a firm, it is alsolikely to give rise to key employees’ reluctance to

invest in the necessary firm-specific human capitalbecause such investment can put them in a poten-tially vulnerable position. Thus, a firm needs toadopt effective employee governance mechanismsto reduce its key employees’ concerns and aligntheir goals with those of the firm.

This study considers two such governance mech-anisms: economic-based mechanisms such asemployee stock ownership, and relationship-basedmechanisms such as long-term firm-employee rela-tionships, and makes two interrelated arguments.First, a firm’s governance system is endogenous toits resource composition: the higher a firm’s levelof firm-specific knowledge resources, the greaterthe need to adopt economic- and relationship-basedemployee governance mechanisms. Second, effec-tive use of such governance mechanisms enables afirm to achieve better economic performance fromits firm-specific knowledge resources.

The existence of, and the solution to, employeegovernance problems are, thus, critically impor-tant for explaining the economic rents generatedby firm-specific knowledge resources. Accord-ingly, this study attempts to contribute to bothemployee governance and resource-based litera-tures. First, this study emphasizes a contingencyview of employee governance in which governancemechanisms are endogenous to the nature of firmresource composition, such as the heterogeneity ofits knowledge assets. Second, this study contributesto the resource-based theory by emphasizing thatfirm-specific knowledge can only influence poten-tial economic rents. The governance of employeeswho deploy such knowledge in the value creatingprocess moderates the de facto rents generated bythe firm. Indeed, the next generation of resource-based theory needs to place more emphasis onthe logic that a firm’s resource base and the effec-tiveness of its governance system jointly influenceits economic performance (Gottschalg and Zollo,2007; Kim and Mahoney, 2005; Makadok, 2003).

BACKGROUND AND HYPOTHESISDEVELOPMENT

Firm-specific resources and the potentialfor sustainable competitive advantage

The resource-based view of the firm considersfirms as bundles of heterogeneous resources thatinclude tangible and intangible assets, operationalprocesses, and products (Amit and Schoemaker,

Copyright 2009 John Wiley & Sons, Ltd. Strat. Mgmt. J., 30: 1265–1285 (2009)DOI: 10.1002/smj

Firm-specific Knowledge and Employee Governance 1267

1993). Among these, knowledge is often con-sidered a firm’s most important resource (Grant,1996; Kogut and Zander, 1992). Firm knowledgecan be broadly classified as either firm-specific orgeneral. If a firm pursues new knowledge closeto its existing knowledge base and with specificapplications to its own business setting, the firm islikely to develop firm-specific knowledge, whichis more useful to the firm and has less applica-bility across firm boundaries (Helfat, 1994; Pavitt,1991). Alternatively, some firms may emphasizethe development of general knowledge, which ismore often built upon the knowledge available inthe market and is less specialized to these firms’own settings. Different knowledge accumulationand development strategies result in key differ-ences among firms in terms of the degree of firm-specificity of their knowledge resources.

Although both firm-specific and general knowl-edge resources are critical for firm operations,accumulating firm-specific resources is generallyassociated with higher risk, since the economicvalue of such resources is ultimately influenced byexogenous factors such as market conditions (Bar-ney, 2001; Bowman and Ambrosini, 2000; Priemand Butler, 2001). Once firm-specific resourcesprove valuable in the market, however, the rar-ity and inimitability associated with firm-specificresources enable such resources to play a morestrategic role than general resources. Due to theidiosyncrasies in firm contexts, rival firms seek-ing to appropriate value through imitating anotherfirm’s firm-specific knowledge must gain accessnot only to the knowledge itself, but also tothe organizational routines and complementaryresources supporting its deployment. This charac-ter of firm-specific knowledge has been describedby Helfat (1994), who emphasizes that firm-specific R&D activities, as key inputs to firm-specific knowledge, are difficult for other firms toimitate:

The outcome of firm-specific R&D can provedifficult for other firms to imitate, if they donot have access to the assets to which theR&D was applied. Additionally, the firm-specific nature of the R&D process furtherimpedes imitation: when the R&D processhas an important tacit element, it is diffi-cult for others outside of the firm to repli-cate the process. If the firm has superior and

difficult to replicate dynamic routines to sup-port new process and product development,other firms may be left playing catch up, forexample, by trying to reverse engineer prod-ucts once another firm has introduced them.Essentially, the difficulty of imitating (orfinding effective substitutes for) other firms’R&D processes or outcomes retards the dif-fusion of technical knowledge across firms.This in turn enhances appropriability, sincedirect competition between firms in R&D isreduced (Helfat, 1994: 175).

In contrast, although general knowledgeresources have lower risk associated with theirvalue, they are less rare and more subject to readyimitation by other firms. This makes them anunlikely source of sustainable competitive advan-tage. So despite the risk associated with developingfirm-specific knowledge resources, firms seekingto gain and sustain competitive advantage oftenfind it crucial to do so through, for example, invest-ing in firm-specific R&D (Helfat, 1994).

Firm-specific knowledge resourcesand employee governance mechanisms

Firm-specific knowledge resources alone often donot automatically lead to superior economic per-formance. Instead, a firm often requires its keyemployees to make complementary human capi-tal investments crucial to absorbing and applyingfirm-specific knowledge. Thus, it is of strategicimportance for the firm to encourage key employ-ees to invest in such firm-specific human capital.

Becker’s (1964) human capital theory providesan early exposition of employee invest-mentsin specific human capital. Because firm-specifichuman capital has, by definition, limited economicvalue in alternative settings, employees whosehuman capital has a substantial firm-specificcomponent are constrained in their transactionswith the focal firm. Moreover, writing andenforcing contracts associated with firm-specificknowledge and human capital investments aregenerally difficult (Hart, 1995; Tirole, 1988;Williamson, 1975). As a result, employees’investment in firm-specific human capital likelyleaves them vulnerable to opportunistic behavioror the potential for holdup by the firm (Klein,Crawford, and Alchian, 1978; Schelling, 1960;



Williamson, 1985).1 For example, a firm ‘. . .maybe tempted to exaggerate financial difficulties inorder to justify paying lower wages to workers’(Milgrom and Roberts, 1992: 334).

If employees rationally expect that their firmmay hold them up in the future, this expectationwill have a direct negative effect on their currentincentives to make specific human capital invest-ments (Hart, 1995; Williamson, 1985). They willbe inclined to underinvest unless they are ensuredvia some effective forms that such opportunis-tic behavior by the firm will not ensue (Shleiferand Summers, 1988; Titman, 1984). Thus, firm-specificity in a firm’s knowledge resources isa double-edged sword. On the one hand, firm-specific knowledge is crucial for a firm’s gainingand sustaining competitive advantage; on the otherhand, such knowledge induces concerns among thefirm’s employees about economic appropriation.

It then follows that when it is critical for afirm to enlist its key employees to make firm-specific human capital investments, the firm mayneed to promise not to hold up the employeesex post. However, previous studies have shownthat such simple promises are generally unreli-able—after the employees have made their specificinvestments, the firm has an incentive to engagein economic holdup and break its promises (Kyd-land and Prescott, 1977; Laffont and Tirole, 1988).The employees may be aware of this possibilityand decline to make specific investments in thefirst place. Without credible economic safeguardsand/or sufficient trust, employees will be reluctantto commit to firm-specific investments in skills thathave low redeployability in other firms; instead,they will prefer to invest in more generic skillsthat can be applied elsewhere.

This study focuses on two general types ofemployee governance mechanisms that can helpfirms mitigate their employees’ reluctance to makefirm-specific human capital investments: aneconomic-based governance mechanism, and arelationship-based mechanism. The former mech-anism is intended to secure employees’ economicgains against ex post value appropriation; the latter

1 Although there are cases where employees may hold up thefirm by, for example, threatening to resign from some projectsor even leave the firm, generally speaking, the firm has morebargaining power in such a bilateral monopoly situation. So therisk of holdup often affects employees more severely than thefirm (Glick and Feuer, 1984; Rock and Wachter, 1999).

mechanism focuses on fostering trust between thefirm and its employees.

Economic-based governance mechanism:employee stock ownership

As an important human resource managementpractice, employee stock ownership has receivedconsiderable research attention (Blair, 1995; Kruseand Blasi, 1997). For example, a wide range of rea-sons have been proposed to explain firms’ adoptionof employee stock ownership, including provid-ing employee benefits, inducing employee effort,fulfilling management’s philosophical commitmentto shared ownership, gaining tax advantages, andfinancing capital acquisitions (Klein, 1987; Kruseand Blasi, 1997).

The current study builds on, but departs from,this research literature by considering a strate-gic situation in which stock ownership may beparticularly relevant to inducing employee effort:namely, when the firm relies on firm-specificknowledge resources to achieve sustainable com-petitive advantage and thus requires key employ-ees to make substantial firm-specific human capitalinvestments. In this setting, concerns about firm-level holdup and employee underinvestment issuesare especially prominent, and accordingly effectivegovernance solutions to employee underinvestmentare most needed.

The idea of using employee stock ownership asan economic-based governance solution toemployee underinvestment problems is built uponproperty rights theory (Barzel, 1989; Demsetz,1967; Libecap, 1989), which maintains that owner-ship rights are instrumental in allocating resourcestoward their highest economically valued uses,making these ownership rights crucial to achievingoperating efficiency. Based on this view, employ-ees will be more willing to invest in specializedhuman capital if they are able to effectively avoidholdup by the firm, and thus appropriate a fairshare of the rents generated from these invest-ments.

A particular feature of firm-specific knowledgeresources and employees’ accompanying humancapital investments is the often ambiguous natureof ownership (Coase, 1960; Milgrom and Roberts,1992). In general, employment contracts that gov-ern exchanges of employee services are costly towrite and enforce (Coase, 1937; Klein, 1980) andtherefore are typically incomplete (Hart, 1995).



The uncontracted dimensions of the exchange,which are in essence property rights or ownershiprights, are the ‘residual rights of control’ (Gross-man and Hart, 1986; Hart and Moore, 1990). Whenthese residual rights of control do not accrue to thetransacting party who invested in the resource, thisparty has less incentive to devote the full effortneeded to maximize the rent-generating potentialof the resource. Thus, the residual rights or owner-ship of resources involved in a transaction shouldgo to the party whose firm-specific investmentshave the most influence over the resources, butare the most difficult to contract over (Grossmanand Hart, 1986; Rajan and Zingales, 1998). Specif-ically, when the key resource involved is firm-specific knowledge, granting some equity own-ership to the key employees who must absorband deploy such knowledge becomes economicallydesirable. Such equity ownership provides keyemployees greater power to bargain ex post overthe economic rents generated from the deploymentof firm-specific knowledge assets.

In addition, firms can also use employee stockownership to serve as a control mechanism suit-able to resource compositions characterized byhigh levels of firm-specific knowledge resources,where the behavior of the key employees inabsorbing and deploying firm-specific knowledgeis often difficult to observe and measure. Theemployer’s insufficient knowledge of the transfor-mation process, or imperfect task programmabil-ity, may force reliance on outcome control—asopposed to behavior control—to induce coopera-tive effort among the firm’s key employees (Eisen-hardt, 1985; Ouchi, 1980). Linking the finan-cial reward of key employees to the firm’s eco-nomic performance through granting key employ-ees equity ownership is likely to best align thefirm’s interests with those of the key employees(Shleifer and Vishny, 1997). In sum, employeestock ownership provides both a measure of resid-ual control and a vehicle for profit-sharing toencourage productive effort.

Granting key employees such ownership rightsdirectly, however, is costly to the firm. In additionto the direct reduction of the share of the eco-nomic profit accrued to the firm, other sharehold-ers’ investment incentives are lessened becausethey will gain only a fraction of the marginal prod-uct for the full marginal cost of their investment(Roberts and Van den Steen, 2000). Meanwhile,employee stock ownership can potentially give

rise to dysfunctional behavioral responses from theemployees due to potential mismatches betweenemployees’ and employers’ time horizons and riskpreferences (Miller, 1992). For example, becauseemployees with stock ownership often hold anundiversified investment portfolio, their potentialaversion to the downside risks of financial diffi-culty or bankruptcy may reduce their incentivesto invest in firm-specific human capital. In addi-tion, employee equity ownership may not fullysafeguard against employees shirking if value dis-tribution is not fine-tuned to individual perfor-mance (Holmstrom, 1982). Moreover, while equityownership gives employees more opportunities toparticipate in the firm’s decision-making process,greater involvement of employees may also leadto administrative inefficiency (Blasi, Kruse, andBernstein, 2003). Therefore, as Kruse and Blasi(1997) emphasize, in order for employee own-ership to serve as an effective mechanism ofgovernance, firms need to give attention to ‘. . .the circumstances in which employee ownershipis implemented, the history of employee rela-tions in the company, and other company policiesthat may support or work against positive effectsof employee ownership’ (Kruse and Blasi, 1997:114).

Due to these costs and downside risks associ-ated with employee stock ownership, firms oftenfind it necessary to use extra caution when con-sidering adopting such policies, unless employeeconcern for firm opportunistic behavior is severeand the employee underinvestment problem iscritical. Thus, when a firm has a high level offirm-specific knowledge, where the concern foremployee underinvestment is likely to be the great-est, the firm is more likely to grant its key employ-ees ownership rights. Therefore,

Hypothesis 1: Everything else equal, a firm’slevel of firm-specific knowledge resources is pos-itively associated with its use of employee stockownership as a governance mechanism.

Relationship-based governance mechanism:firm-employee relationships

Building trusting relationships with key employ-ees may also serve as a governance mechanismthat a firm can adopt to encourage employees to



make specialized human capital investment accom-panying firm-specific knowledge resources. Sinceemployees’ concerns about holdup by the firm maybe based on perceptions that the firm is in a posi-tion to unfairly expropriate their investments infirm-specific human capital, the firm’s efforts tobuild trust may help reduce the threat of such per-ceptions. Although it may generally be difficult foremployees to trust a firm’s simple promise not toengage in unfair expropriation ex post, the employ-ees may be more willing to commit if the firmshows its trustworthiness through practices thatcan help build its reputation as a fair and caringemployer (Barney and Hansen, 1994; Rousseau,1995; Zaheer, McEvily, and Perrone, 1998).

The research literature has often considered thefirm-employee relationship as an exchange of trustand commitment (Blau, 1964; Etzioni, 1961; Mow-day, Porter, and Steers, 1982). When employeesconsistently receive favorable treatment from theirfirm, they are likely to develop a sense of trustthat the firm will not take advantage of theirvulnerability ex post and thus are more likelyto reciprocate the favorable treatment by engag-ing in firm-specific human capital investments(Gouldner, 1960). Further, employees form per-ceptions of the extent to which they are valuedand cared about; and they use such perceptions asa basis for evaluating the degree of commitmentthey will provide to the firm (Vandenberghe et al.,2007). Research has also generally shown thatrelational governance is associated with greateremployee cooperation and trust, which in turnreduces employee concerns about holdup by thefirm and increases employees’ acceptance of thefirm’s goals as close to their goals. For exam-ple, Ouchi (1980) maintains that in clan organi-zations: ‘a variety of social mechanisms reducesdifferences between individual and organizationalgoals and produces a strong sense of community’(Ouchi, 1980: 136). The shared values embed-ded in well-functioning firm-employee relation-ships often inspire key employees to self-enforcetheir psychological contracts with the firm and todevote sufficient effort to cooperation.

Nevertheless, the development and maintenanceof relational governance are costly because theyoften require substantial long-term investment bythe firm in various human resource practices, suchas recruitment benefits and training, team-basedperformance appraisal, and long-term employment

policies (Collins and Smith, 2006). This cost fac-tor implies that the use of relational governanceis more beneficial when substantial exchange haz-ards are present, as in the context of firms withhigh levels of firm-specific knowledge resources.Therefore,

Hypothesis 2: Everything else equal, a firm’slevel of firm-specific knowledge resources is pos-itively associated with the extent to which itplaces emphasis on establishing good relation-ships with its key employees.

Employee governance and performanceadvantage based on firm-specific knowledge

Although firm-specific knowledge resources are apotential source of a firm’s superior economic per-formance, if employees’ concerns for holdup bythe firm lead to a tendency for underinvestment infirm-specific human capital, the firm’s actual eco-nomic performance may diverge substantially fromits potential (Kim and Mahoney, 2005). Thus, bothexplicit and implicit governance mechanisms, suchas employee stock ownership and firm-employeerelationships, can be important in influencing theperformance impact of firm-specific knowledge(Gottschalg and Zollo, 2007; Wang and Barney,2006). Consistent with this view, some researchershave recognized organizational capabilities andmotivation as two key drivers for firm competitivebehavior, and consequently firm-level economicperformance (Chen, 1996; Gimeno, 1999). Thesetwo different but coexisting drivers jointly influ-ence the economic outcome of the firm.

Viewed in this way, the current resource-basedtheory of the firm focuses primarily on the poten-tial of a firm’s resources—especially firm-specificknowledge-based resources—for generating eco-nomic rents. What has been far less examined fromthis perspective is the willingness of a firm’s keyemployees to deploy these firm-specific resources.Indeed, the governance of key employees whoactually deploy and utilize a firm’s resourcestoward productive uses moderates the de factoeconomic rents generated by the firm. There-fore, to better explain and predict how superiorfirm-level resources bring sustainable competitiveadvantage, we need to simultaneously consider afirm’s resource base and the effectiveness of itsgovernance system (Makadok, 2003).



To reduce the gap between potential economicrents and de facto economic rents, a firm mustmitigate its employees’ tendency to underinvest infirm-specific human capital and align their eco-nomic interests with those of the firm. Other-wise, the capability for firm-specific knowledgeto deliver superior economic performance will besubstantially compromised. Put differently, a keybarrier to the realization of economic rent basedon firm-specific knowledge resources is employ-ees’ underinvestment in firm-specific human capi-tal; and employee governance mechanisms need toincorporate appropriate solutions to the employeeunderinvestment problem. Hypotheses 1 and 2 sug-gest that an effective governance system for afirm with a high level of firm-specific knowledgeresources will more likely grant key employeesequity ownership and/or build trusting relation-ships with them. This logic suggests that thereshould be an interaction effect between firm-specific knowledge resources and employee gov-ernance mechanisms on firm-level economic per-formance. A firm’s governance system might beconsidered analogous to a valve, the inappropri-ate functioning of which will hinder the flow offirm-specific knowledge resources toward full real-ization of economic rents. Therefore, the followingare hypothesized:

Hypothesis 3a: Everything else equal, the rela-tionship between the level of firm-specific knowl-edge resources and firm-level economic per-formance is positively moderated by employeestock ownership.Hypothesis 3b: Everything else equal, the rela-tionship between the level of firm-specific knowl-edge resources and firm-level economic perfor-mance is positively moderated by firm-employeerelationships.

It should be noted however that unlike firm-employee relationships, employee stock ownershipoften influences employee incentives in a morecomplex manner (Kruse and Blasi, 1997). Whileemployee equity ownership directly addressesemployees’ concerns about holdup by ensuringthem bargaining power concerning ex post rentappropriation, it may not fully safeguard againstemployees’ shirking behavior, which often occurswhen value distribution is not fine-tuned to indi-vidual performance. This shirking or free-rider

problem can be especially severe if key employ-ees need to work in a group. Such a problemis sometimes referred to as the ‘1/N problem.’With N employees in a company or a group, eachemployee will get an average of only 1/N of anyextra economic surplus generated by their bet-ter performance (Holmstrom, 1982). Furthermore,when a firm or a team is large, the contributionof a particular employee to team output often can-not be precisely measured and rewarded accord-ingly (Alchian and Demsetz, 1972). Therefore, insuch settings, purely economic incentives can giverise to dysfunctional behavioral responses (Miller,1992). So employee stock ownership may be lesseffective in resolving the underinvestment prob-lem in large firms. Such a limitation of employeeownership-based governance is captured in the fol-lowing hypothesis:

Hypothesis 4: The moderating role of employeestock ownership in the relationship between thelevel of firm-specific knowledge resources andfirm-level economic performance will be weakerin larger firms.

DATA AND METHODS

Data and sample

Our sample was compiled from several datasources: Standard & Poor’s (S&P) Compustatseries, United States patent data, the U.S. Securi-ties and Exchange Commission’s (SEC) EDGARdata, and data from KLD Research and Analyt-ics Inc. Since all of the datasets covered multipleyears, we were able to construct an unbalancedpanel dataset based on their overlapping periods.

We began our sample selection with the group offirms in manufacturing industries (four-digit Stan-dard Industrial Classification [SIC] codes 2000 to3999) that were also listed in the KLD databasebetween 1994 and 2002. The focus on manu-facturing firms enabled us to construct a sam-ple of firms that shared some common charac-teristics in terms of their innovation processes,but provided sufficient variation in terms of thelevel of knowledge specificity across firms. TheKLD data were used to construct the measurefor firm-employee relationships. KLD data havebeen widely used in business and society research,



and are considered to be the best data avail-able for a comprehensive measure of corporatesocial relationships and stakeholder management(Hillman and Keim, 2001; Sharfman, 1996; Wad-dock and Graves, 1997). The data have been usedto compile profiles and social ratings evaluatingeach company’s strengths and concerns in severaldimensions including community, diversity, firm-employee relations, environment, products, and soforth. We obtained our firm-employee relationshipsmeasure from the ‘employee relations’ dimensionof the KLD data (see Appendix).

Information on employee stock ownership wascollected from the SEC data. The SEC requiresevery registered firm to file a definitive proxy state-ment (DEF14-A) annually, which discloses thebeneficial ownership of common stock holdingsin excess of five percent. Note that for firms inwhich employees do not hold at least five percentcollectively, this information is not reported andwas thus coded as zero. Roughly 12 percent of theobservations in our sample had nonzero values.2

The combined firm-employee relationships andemployee stock ownership data were then mergedwith patent citation data and data from S&P’sCompustat series, where we obtained our measuresof firm-specific knowledge resources and firm-level economic performance respectively. Hall,Jaffe, and Trajtenberg (2001) have created a datafile that contains detailed information about almostthree million U.S. patents granted between 1963and 1999, and over 16 million citations made ofthese patents for those granted between 1975 and1999. Moreover, Professor Bronwyn Hall has com-piled updated information on patents and their cita-tions up to 2002.3 Because our unit of analysiswas the firm, we aggregated the patents and theircitation counts to the firm level (Rosenkopf andNerkar, 2001). After merging all the datasets anddeleting observations with missing values for keyvariables, the final panel data contained 211 firmsand 1,329 firm-year observations between 1994

2 According to the latest estimate of the National Center forEmployee Ownership (www.neco.org), about 15 percent to 20percent of all public companies today have adopted employeestock ownership plans. The somewhat lower percentage ofnonzero values in our sample may be because a firm had adoptedemployee stock ownership but the total percentage holding wasbelow five percent, or because the percentage of firms has beengrowing over the years.3 http://elsa.berkeley.edu/∼bhhall/patents.html (last accessed 20March 2008).

and 2002. The firms were distributed within 18two-digit and 65 three-digit SIC codes.

Measures

Firm economic performance (Tobin’s Q)

Tobin’s Q was employed to capture each firm’seconomic performance, as it reflects the market’sexpectations of the firm’s future growth and profitpotential (Lindenberg and Ross, 1981). Tobin’s Qwas approximated using the market-to-book ratio.The market value numerator was the year-endmarket value of the firm’s common stock, plusthe book value of its preferred stock and debt;the book value denominator was the year-end totalassets figure. In addition, to be consistent with ourstatistical model, which shall be discussed in thenext section, we conducted a log transformation ofour proxy of Tobin’s Q.

Firm-specificity of knowledge resources (FS)

Patent citations provide direct evidence of thepath of knowledge flow and knowledge spillovers,since each patent normally identifies several oth-ers constituting the state-of-the-art technology onwhich it builds. It is possible to tabulate the fre-quency with which patents cite previous patentsof the same firm and patents assigned to otherfirms. Previous research has used such tabulationsto explore questions involving spatial diffusion(Jaffe, Trajtenberg, and Henderson, 1993), inter-national knowledge flows (Jaffe and Trajtenberg,1999), and spillovers from public research (Jaffeand Lerner, 1999).

Firm-specific knowledge often results from firmssearching and accumulating new knowledge on thebasis of their established knowledge bases (Cohenand Levinthal, 1989; Teece, 1986). If patents repre-sent knowledge creation, and patent citations rep-resent knowledge flows, the frequency with whicha firm cites its own previous patents indicates thedegree to which the firm is building upon its ownknowledge base. The higher the percentage of suchinternal accumulation, the more likely the firm’sinnovative knowledge is firm-specific. This logicsupports constructing a measure of firm-specificknowledge resources using patent citations. Thepatent citation data for 1975 to 2002 were ana-lyzed to generate two proxies for the level of firm-specific knowledge resources. The first was the



share of self-citations made, calculated by count-ing all citations made in a firm’s new patents in acertain year that cited the firm’s previous patents,then dividing this by the total number of citationsmade in all of the firm’s new patents in that year.The second proxy was a weighted count measurecalculated as follows:

Firm-specif ic

knowledge= Number of prior self -

citations made (adjusted by

f irm size)∗ extent to which

prior self -cited patents are

subsequently cited by

the f ocal f irm

The weight was added to take into account thefirm-specificity of the firm’s prior patented knowl-edge: even though a firm cites its own previouspatents, if these previous patents are also widelycited by other firms (which makes the weight rathersmall), the degree of firm-specificity in knowledgeas measured by the count of self-citations madeshould be discounted.

Firm-employee relationships

Our measure of firm-employee relationships isbased on the ‘employee relations’ dimension of theKLD data. The employee relations dimension hasa total of seven ‘strengths’ (strong union relations;no layoff policy; cash profit-sharing; employeeinvolvement; strong retirement benefits; health andsafety strength; and other strengths) and three ‘con-cerns’ (poor union relations; health and safetyconcern; and workforce reduction). A firm’s totalnumber of ‘concerns’ was subtracted from its num-ber of ‘strengths’ to arrive at a net score for firm-employee relationships. Note that we excluded the‘cash profit-sharing’ dimension from the list, as itis a financial-based governance mechanism. Theseremaining items take into consideration long-termemployee well-being, and therefore may serve as areasonable proxy for firm-employee relationships.4

4 ‘Employee involvement’ is not a pure relationship-based item,since it includes granting employees shares and options. More-over, the ‘union relations’ items are geared toward lower-levelemployees, whereas our theoretical focus is on key knowledge

Employee stock ownership

Employee stock ownership was measured in termsof the percentage of beneficial ownership of thefirm’s common stock held by employees as a col-lective.5 The data were extracted from the EDGARdatabase.

Control variables

Previous empirical studies (Hall, 2000; Hall, Jaffe,and Trajtenberg, 2005) have considered R&Dexpenditure as an innovation input and an impor-tant determinant of the intangible component ofmarket value. To be consistent with our statisticalmodel, we scaled each firm’s yearly R&D expen-diture by the firm’s total assets (RD/A). Patentingintensity represents another important aspect of theknowledge resources that contribute to a firm’smarket value, since a firm’s patents were taken asthe output of the firm’s investment in knowledgecreation (Griliches, 1981; Hall, 2000; Hall et al.,2005). Patenting intensity was calculated by divid-ing the aggregate number of a firm’s patents by itstotal assets (PAT/A).

The R&D expenditure and patenting intensityindicators were included in the equations withemployee stock ownership and firm-employee rela-tionships as dependent variables. It is reasonable toexpect that firms with greater knowledge resourcesin general (as indicated by R&D expenditure andpatenting activity) have greater difficulty measur-ing employee effort, and are therefore more likelyto adopt the governance mechanisms under con-sideration.

In addition, we controlled for firm size, firm age,and financial slack in these equations. Previous

workers. Thus, as a robustness test, we constructed a more con-servative measure of firm-employee relationship by removingthese two items. Our key empirical results remained consis-tent. In addition, we conducted robustness tests using a dummymeasure of firm-employee relationships based on the ‘100 BestPlaces to Work For’ list (http://www.greatplacetowork.com/best/list-bestusa.htm). Consistent results were again generated usingthis alternative measure. Details of these robustness tests areavailable from the authors upon request.5 Because of the relatively large number of zeros in this vari-able, we conducted a supplementary analysis coding employeestock ownership as a dummy variable (those having collectiveemployee ownership of five percent or above were coded as one,otherwise, zero). In addition, we constructed a third measure ofemployee stock ownership using the ‘cash profit-sharing’ dimen-sion from the KLD data. The main results with these alternativeapproaches were largely consistent with those of the main anal-ysis. Details of these alternatives are available from the authorsupon request.



Equation 3

Firm-employeerelations

Exchangehazards

Firm-specificknowledgeresources

Firm economicperformance(Tobin’s Q)

Superiorresourceposition

Employee stockownership

Equation 1

Equation 2

Figure 1. The conceptual model

empirical studies (Gregg and Machin, 1988; Kruse,1996) have found that larger companies have agreater level of employee stock ownership. Firmsize may also affect firm-employee relationshipspositively, due to the presence of economies ofscale in relationship-building activities. Firm agehas previously been shown to be related to theadoption of certain employee governance mecha-nisms (Gomez-Mejia, Larraza-Kintana, and Makri,2003; Schulze, Lubatkin, and Dino, 2003). Finan-cial slack may affect the extent to which a firmis willing to adopt employee stock ownershipplans (Jones and Kato, 1993; Poole and Jenkins,1990) and engage in relationship-building with itsstakeholders, including employees (Waddock andGraves, 1997).

Firm size was proxied by the natural logarithmof the total number of employees in a firm. Fol-lowing the methods of Bourgeois (1981) and Singh(1986), we use each firm’s current ratio (currentassets divided by current liabilities) to representits financial slack. In addition, year dummies andindustry dummies6 at a three-digit SIC level wereincluded in all of the equations to control for timeand industry effects.

Estimation method

Overall model

Our hypotheses require that we empirically test asystem of equations simultaneously (see Figure 1).

6 We conducted additional tests by running the regressions basedon individual industries. The results were largely consistent withthose of the analyses with industry controls. Details are availablefrom the authors upon request.

The first and second equations were formulatedto test whether the level of firm-specific knowl-edge resources was associated with the degree ofemployee stock ownership and with the cultiva-tion of good firm-employee relationships. The thirdequation examined whether or not the level of firm-specific knowledge resources affects firm perfor-mance and whether the relationship between firm-specific knowledge resources and performance ismoderated by employee stock ownership and firm-employee relationships.

First-stage equations:

Employee Ownership = α′0

+ β ′1Lagged Employee Ownership

+ β ′2FS + β ′

3Employee Relations

+ β ′4RD/A + β ′

5PAT/A + β ′6Size + β ′

7Age

+ β ′8Slack + β ′

9Other Controls + ε′

Employee Relations = α′′0

+ β ′′1 LaggedEmployee Relations

+ β ′′2 FS + β ′′

3 Employee Ownership

+ β ′′4RD/A + β ′′

5PAT/A + β ′′6Size

+ β ′′7Age + β ′′

8Slack

+ β ′′9Other Controls + ε′′

Second-stage equation7:

log Q = α0 + β1RD/A + β2PAT/A + β3FS

7 Please see the subsection below for the detailed derivation ofthe second-stage equation.



+ β4Employee Ownership

+ β5Employee Relations

+ β6Size + γ7FS∗ Employee Ownership

+ γ8FS∗ Employee Relations

+ δ9FS∗ Employee Ownership∗ Size + ε

Specifying and testing the third equation inde-pendently would introduce significantly biasedestimates due to the endogeneity of the employeestock ownership and firm-employee relationshipvariables (Hamilton and Nickerson, 2003). Thus,the equations were tested using a series of two-stage models that considered the endogeneity ofgovernance mechanisms. Two-stage least squares(2SLS) analysis is a popular approach for estimat-ing simultaneous equations (Greene, 2002;Kennedy, 1998). A 2SLS estimation combinesall exogenous variables in the first-stage equation(but not in the second-stage equation) to createa combined variable to act as a ‘best’ instru-mental variable. In this analysis, the exogenousvariables were the lagged employee governancevariables, firm age, and financial slack. The first-stage tests had employee stock ownership and firm-employee relationships as the dependent variables.New variables were obtained by regressing thefirst-stage models. The second-stage equation hasfirm-level economic performance (Tobin’s Q) asthe dependent variable. We then used as regressorsthe newly estimated employee stock ownership,firm-employee relationships, their interactions withfirm-specific knowledge resources, and other fac-tors that were thought to affect Tobin’s Q.

The second-stage performance equation

Following the pioneering work of Waugh (1928)and Griliches (1961), and subsequent studies(Griliches, 1981; Hall, 2000; Hall et al., 2005), wedetermined the economic returns on firm-specificknowledge resources using the hedonic regressionmethod, which regresses a firm’s market valueagainst bundles of resources that compose the firm,including measures relating to intangible resourcessuch as knowledge and innovations. This approachis often used when no direct economic valuation ofa resource is available and the resource can onlybe valued when bundled with other resources ofthe firm.

We thus followed the basic approach employedin the market valuation literature that relates themarket value of a company to the economic valueof its tangible assets and various measures ofits intangible assets (Griliches, 1981; Hall, 2000).Specifically, MVi was defined as the market valueof companyi. The assets of the firm were regardedas made up of two parts: physical or tangibleassets that are recorded and are measured as partof the total assets of the firm, and knowledgeor intangible assets that largely go unrecordedand do not appear in the accounts as part of thetotal assets. The recorded economic value of thetangible assets was represented by Ai , and thecorresponding measures of the economic value ofthe intangible assets were represented by Ki . Themodel was

MVi = q(Ai + γKi)σ (1)

Taking the logarithms of both sides,

log MVi = log q + σ log Ai + σ

log(1 + γKi/Ai) (2)

where log(1 + γKi/Ai) is generally approximatedby γKi/Ai in the literature (Hall et al., 2005). γ

measures the shadow value of knowledge assetsrelative to the tangible assets of the firm. σ is unityif returns to scale are constant, which should be thecase in a cross-sectional setting. log Ai , the log ofbook assets, can then be moved to the left side ofthe equation:

log(MVi/Ai) = log q + γKi/Ai. (3)

Since Tobin’s Q is defined as a firm’s marketvalue relative to the replacement value of its assets,and the replacement value is approximately thevalue of the tangible assets of the firm, MVi/Ai ,the ratio of a firm’s market value to the value ofits tangible assets is in fact Tobin’s Q. Therefore,the above equation can be estimated with thelogarithm of the conventional Tobin’s Q (Qi) asthe dependent variable, as we did here. Therefore,the estimation model becomes:

log Qi = log q + γKi/Ai + εi (4)

where εi is a normally distributed error term.



On the basis of the model shown above, wecame up with its expanded version:

log Q = α0 + β1RD/A + β2PAT/A + β3FS

+ γ4FS∗ Employee Ownership

+ γ5FS∗ Employee Relations + δ6FS∗

Employee Ownership∗ Size + ε (5)

Equation (5) is the second-stage equation, whichwas estimated in this study. The β coefficientsrefer to intangible assets including both generaland firm-specific knowledge assets: the ratio ofR&D to total assets (β1), the ratio of patent countsto total assets (β2), and the level of firm-specificknowledge assets (β3). The γ coefficients refer tothe variables hypothesized to affect market valuethrough motivating employees to cooperate withthe firm, including interactions of the level of firm-specific knowledge assets with employee stockownership (γ4) and with firm-employee relation-ships (γ5). The δ parameter incorporates any three-way interaction among firm size, employee stockownership, and firm-specific knowledge, takinginto consideration the extent to which the mod-erating effect of employee stock ownership alsodepends on firm size (as specified in Hypothe-sis 4).

To address concerns about autocorrelation andthe unobserved heterogeneity that are likely tobe present when estimating a panel dataset, weapplied the standard techniques of panel dataanalysis. Hausman’s (1978) test was conducted

to determine the appropriate estimation method.The results demonstrated significant correlationsbetween the error terms and the regressors, sug-gesting that firm fixed-effects models should bepreferred over random-effects models in this sta-tistical analysis. We thus applied 2SLS with fixedeffects to test the models. In particular, this wasimplemented by reshaping the second-stage vari-ables into difference form, and then using ProcSyslin within SAS to do 2SLS on the differencedvariables.

RESULTS

Table 1 shows the descriptive statistics and cor-relation matrix for the main variables. The twomeasures of the level of firm-specific knowledgeresources (share of self-citations made and theweighted number of self-citations made) werehighly correlated. In addition, both measures werepositively correlated with the performance mea-sure, logged Tobin’s Q. Consistent with our expec-tations, both measures of firm-specific knowledgeshowed significant, positive correlations with themeasures of the two employee governance mech-anisms: ‘employee stock ownership’ and ‘firm-employee relationships’ (except that the correla-tion between weighted number of self-cites andemployee stock ownership were not significant).R&D spending and patenting intensity were alsopositively correlated with logged Tobin’s Q, aswould be expected.

Table 1. Descriptive statistics and correlation matrix

Variables Mean s. d. 1 2 3 4 5 6 7 8 9

1. Log (Tobin’s Q) 0.77 0.56

Firm-specific knowledgeresources

2. Share of self-cites 0.10 0.14 0.15∗

3. Weighted number ofself-cites (×10−2)

0.22 0.46 0.08∗ 0.33∗

4. Employee stock ownership 1.32 4.08 0.02 0.07∗ 0.015. Firm-employee relationships 0.19 1.12 0.21∗ 0.19∗ 0.20∗ −0.056. R&D intensity 0.05 0.04 0.39∗ 0.10∗ 0.16∗ −0.04 0.20∗

7. Patenting intensity 0.06 0.15 0.13∗ 0.21∗ 0.20∗ −0.03 0.15∗ 0.28∗

7. Firm size 8.45 1.38 −0.04 0.04 0.22∗ 0.18∗ 0.08∗ −0.09 0.29∗

9. Firm age 23.36 13.73 −0.00 0.03 −0.02 0.03 0.07∗ 0.01 −0.03 0.39∗

10. Financial slack 0.03 0.09 0.39∗ 0.01 0.03 −0.03 0.15∗ 0.07∗ −0.06∗ −0.09∗ −0.19∗

N = 1329.∗ p < 0.05.



Table 2. The determinants of employee stock ownership and firm-employee relationships: results from first-stagemodels with firm fixed effect

Labels DV: Employee stock ownership DV: Firm-employee relationships

Model 1 Model 2 Model 3 Model 4Lagged DV 0.74∗∗∗ 0.74∗∗∗ 0.81∗∗∗ 0.81∗∗∗

(0.02) (0.02) (0.02) (0.02)

Firm-specific knowledge resources

Share of self-cites 3.41∗∗ 2.55∗∗∗

(1.25) (0.35)Weighted number of self-cites (×10−2) 0.04+ 0.09∗∗∗

(0.02) (0.02)Employee stock ownership −0.03∗∗ −0.03∗∗

(0.01) (0.01)Firm-employee relationships −0.32∗∗ −0.27∗

(0.11) (0.11)R&D intensity 1.77 1.59 3.15∗∗∗ 3.07∗∗

(3.98) (4.16) (1.01) (1.12)Patenting intensity 1.98 1,23 2.35 1.99

(8.23) (7.95) (3.55) (2.78)Firm size 0.44∗∗∗ 0.42∗∗∗ 0.10∗ 0.09∗

(0.11) (0.11) (0.04) (0.04)Firm age 0.00 0.00 0.01∗ 0.01∗

(0.00) (0.00) (0.00) (0.00)Financial slack 0.49 0.68 1.77∗∗∗ 1.80∗∗∗

(1.57) (1.61) (0.49) (0.49)R2 0.68 0.67 0.79 0.79

N = 1329.Standard errors are shown in parentheses.Industry and year dummies are included but not reported.Significant at the + p < 0.10; ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001 level.

Table 2 shows the empirical results of the first-stage models, which included regressions ofemployee stock ownership and firm-employee rela-tionships against factors thought to affect theextent to which a firm adopts the employee gover-nance mechanisms.

The empirical results from the first-stage anal-ysis are largely consistent with the prediction ofHypotheses 1 and 2, which indicate that firms withhigher levels of firm-specific knowledge resourceswere more likely to adopt appropriate gover-nance mechanisms to align key employees’ effortswith the interests of the firm. Employee stockownership was found to be positively associatedwith both measures of firm-specific knowledgeresources (Models 1 and 2), although the rela-tionship was only marginally significant when theweighted number of self-cites was used as the firm-specific knowledge measure (Model 2). Consistentpatterns are also found in the equations with firm-employee relationships as the dependent variable.The coefficients on both measures of firm-specific

knowledge resources were positive and statisti-cally significant (Models 3 and 4). Thus, we foundsupport for both Hypothesis 1 and Hypothesis 2,providing evidence for our argument regarding theendogeneity of employee governance mechanisms.

Table 2 also shows that employee stock owner-ship and firm-employee relationships were nega-tively associated with each other, indicating thatthese two governance mechanisms may be substi-tutive.8 Further, large firms were found to be morelikely to use both forms of employee governancemechanisms, while older firms were more likelyto rely on good relationships with their employ-ees. Financial slack was positively associated withfirm-employee relationships, consistent with the

8 To further explore the interrelationship between the two gov-ernance mechanisms in influencing the performance effectsof firm-specific knowledge, we conducted additional analysisby including a three-way interaction term among firm-specificknowledge, employee stock ownership, and firm-employee rela-tions. However, no statistically significant effect was found forthe interaction term. Details of these empirical results are avail-able from the authors upon request.



idea that cash-rich firms are more likely to spendon activities that enhance their relationships withemployees (Waddock and Graves, 1997). But thesame effect was not found when employee stockownership was used as the dependent variable.Similarly, consistent with our prediction, R&Dintensity was positively related to firm-employeerelationships; but the effect is not statistically sig-nificant for employee stock ownership.

Second-stage financial performance estimates

Table 3 presents the empirical results from thesecond-stage models. For the models applying2SLS, we used the estimated employee stock own-ership and firm-employee relationships variablesgenerated in the first stage, their interactions withfirm-specific knowledge, and other factors thatwere thought to affect Tobin’s Q as regressors.

The effects of the level of firm-specific knowl-edge, other intangible resources, including R&Dand patenting intensities, and firm size, are shownin Models 1 and 6. Both measures of firm-specificknowledge showed positive and statistically signif-icant effects on firm performance, indicating thatfirms with higher levels of firm-specific knowledgeassets are more likely to achieve better economicperformance. Also, as expected, both R&D spend-ing and patenting intensity were positively andsignificantly related to performance.

In Models 2 and 7, employee stock ownershipand its interactions with measures of the level offirm-specific knowledge assets were added. Theinteraction term was positively related to firmperformance for both measures of firm-specificknowledge, although the coefficients on the inter-action terms were only marginally significant.These results provide some support for Hypothesis3a, which proposes that the effect of firm-specificknowledge assets on firm performance should bepositively moderated by employee stock owner-ship.

Models 3 and 8 added a three-way interac-tion term among employee stock ownership, firm-specific knowledge, and firm size. This was todirectly test Hypothesis 4, which states that themoderating role of employee stock ownership inthe relationship between the level of firm-specificknowledge resources and firm-level economic per-formance should be weaker for larger firms. Con-sistent with our prediction, the coefficients on the

three-way interaction terms were significant andnegative in both models.9

Models 4 and 9 in Table 3 provide the resultsafter adding the firm-employee relationships mea-sure and its interaction with measures of firm-specific knowledge assets to Models 1 and 4. Forboth firm-specific knowledge measures, the coef-ficients on the interaction terms were positive andstatistically significant. These results suggest thatthe relationship between firm-specific knowledgeassets and performance is positively moderated byfirm-employee relationships, supporting Hypothe-sis 3b.

Models 5 and 10 are the full models with the twogovernance mechanisms and their associated inter-action terms included in one model. The results arelargely consistent with those of the other modelspresented.

DISCUSSION AND CONCLUSIONS

The resource-based view of the firm emphasizesthe role of firm-specific resources, especially firm-specific knowledge assets, in achieving superioreconomic performance (Barney, 1991; Kogut andZander, 1992). However, relatively little researchhas explored employee governance mechanismsand how they may influence the actual economicbenefits that can be obtained from firm-specificresources (Gottschalg and Zollo, 2007; Makadok,2003). This study was based on the idea thatthe features of firm-specific knowledge resourcesthat constitute potential performance advantagessimultaneously give rise justifiably to employeeconcerns about holdup by the firm ex post, and thusa reluctance to invest in firm-specific human cap-ital. Moreover, adopting governance mechanismsthat help resolve the problem of employee underin-vestment in firm-specific human capital will enablethe firm to achieve a greater level of economic per-formance from its efforts to deploy firm-specificknowledge assets by reducing the gap betweenpotential and realized economic rents.

9 As an alternative test, in a supplementary analysis we splitthe sample into two subsamples according to firm size (withthe mean size as the cutoff point) and ran the regressions againusing employee stock ownership as a moderator. The coefficientson the interaction terms were positive and statistically strongerfor small firms than for large firms, a result consistent with thatusing the three-way interactions. These results are available uponrequest.



Tabl

e3.

The

dete

rmin

ants

offir

mec

onom

icpe

rfor

man

ce(D

V:

log

(Tob

in’s

Q))

:re

sults

from

seco

nd-s

tage

mod

els

with

firm

fixed

effe

ct

Lab

els

Mod

el1

Mod

el2

Mod

el3

Mod

el4

Mod

el5

Mod

el6

Mod

el7

Mod

el8

Mod

el9

Mod

el10

Bas

e2S

LS

2SL

S2S

LS

2SL

SB

ase

2SL

S2S

LS

2SL

S2S

LS

R&

Din

tens

ity4.

04∗∗

∗4.

13∗∗

∗3.

97∗∗

∗4.

28∗∗

∗4.

29∗∗

∗4.

22∗∗

∗4.

51∗∗

∗4.

17∗∗

∗5.

15∗∗

∗5.

11∗∗

∗

(0.4

3)(0

.46)

(0.4

5)(0

.43)

(0.4

2)(0

.44)

(0.4

7)(0

.45)

(0.4

3)(0

.44)

Pate

ntin

gin

tens

ity3.

12∗∗

∗3.

30∗∗

∗3.

25∗∗

∗2.

91∗∗

∗2.

83∗∗

∗3.

33∗∗

∗3.

27∗∗

∗3.

09∗∗

∗3.

23∗∗

∗3.

17∗∗

∗

(0.7

2)(0

.73)

(0.7

1)(0

.70)

(0.6

9)(0

.71)

(0.6

9)(0

.72)

(0.7

1)(0

.70)

Firm

size

−0.0

3∗−0

.02∗

−0.0

2∗−0

.02∗

−0.0

2∗−0

.04∗∗

−0.0

4∗∗−0

.03∗∗

−0.0

4∗∗−0

.02∗

(0.0

1)(0

.01)

(0.0

1)(0

.01)

(0.0

1)(0

.01)

(0.0

1)(0

.01)

(0.0

1)(0

.01)

Fir

m-s

peci

fickn

owle

dge

reso

urce

s

Shar

eof

self

-cite

s0.

57∗∗

∗0.

55∗∗

∗0.

62∗∗

0.58

∗∗∗

0.46

∗∗

(0.1

7)(0

.16)

(0.1

6)(0

.16)

(0.1

7)W

eigh

ted

num

ber

ofse

lf-c

ites

(×10

−2)

0.06

∗∗0.

04∗

0.05

∗∗0.

11∗∗

0.13

∗∗

(0.0

3)(0

.02)

(0.0

2)(0

.04)

(0.0

5)E

mpl

oyee

stoc

kow

ners

hip

(ESO

)0.

010.

010.

000.

010.

010.

00(0

.00)

(0.0

0)(0

.00)

(0.0

1)(0

.01)

(0.0

0)E

SO∗

Firm

-spe

cific

know

ledg

e0.

12+

0.13

∗0.

08∗

0.04

+0.

03+

0.03

+

(0.0

7)(0

.06)

(0.0

4)(0

.02)

(0.0

2)(0

.02)

ESO

∗Fi

rm-s

peci

fickn

owle

dge

∗Fi

rmsi

ze−0

.10∗

−0.0

9+−0

.07∗

−0.0

6∗

(0.0

5)(0

.05)

(0.0

3)(0

.03)

Firm

-em

ploy

eere

latio

nshi

ps(E

R)

0.06

∗∗∗

0.07

∗∗∗

0.05

∗∗∗

0.06

∗∗

(0.0

1)(0

.02)

(0.0

1)(0

.02)

ER

∗Fi

rm-s

peci

fickn

owle

dge

0.29

∗∗∗

0.32

∗∗0.

08∗∗

0.06

∗

(0.1

0)(0

.12)

(0.0

3)(0

.03)

R2

0.71

0.72

0.74

0.74

0.75

0.70

0.71

0.73

0.74

0.76

�R

2(v

s.M

odel

1or

5)0.

010.

030.

030.

040.

010.

020.

030.

05

N=

1,32

9.St

anda

rder

rors

are

show

nin

pare

nthe

ses.

Indu

stry

and

year

dum

mie

sar

ein

clud

edbu

tno

tre

port

ed.

Sign

ifica

ntat

the

+p

<0.

10;

∗p

<0.

05;

∗∗p

<0.

01;

∗∗∗p

<0.

001

leve

l.



The study’s key results broadly support thisorganizational economics logic. The empiricalfindings show that firms with greater firm-specificknowledge resources are more likely to adopt gov-ernance mechanisms appropriate for reducing keyemployees’ concerns about holdup by the firm. Thespecific governance mechanisms discussed in thisstudy include both an economic-based governancemechanism of granting employee stock ownership,and a relationship-based governance mechanismof building firm-employee relationships. Further,the increased use of these governance mechanismsstrengthens the relationship between the level offirm-specific knowledge and a firm’s economicperformance.

This paper thus contributes to the existingresearch literature in three key aspects. First, whileprevious research has investigated employee gov-ernance primarily in general terms, this study con-sidered interfirm differences in knowledge assetsexplicitly and emphasized a contingency view ofemployee governance systems. We maintain thatemployee governance mechanisms are endogenousto the nature of firm knowledge resources. Sec-ond, this paper contributes to resource-based the-ory by emphasizing that firm-specific knowledgeonly defines a firm’s potential for generating eco-nomic rents, and that the rents actually generatedwill also be influenced by the effectiveness of thefirm’s governance mechanisms. These efforts thusextend the resource-based view by emphasizingthat a firm’s resource base and the effectivenessof its governance mechanisms jointly influence itsprofitability. Third, the current paper provides thefirst systematic empirical tests of these arguments.

Despite these contributions, this paper has somelimitations that require future research to advanceits key arguments. First, this study followed anequilibrium approach by considering the firm as abundle of idiosyncratic knowledge resources. Bydesign, it was not able to address how differencesin firm-specific knowledge across firms arise inthe first place. Future research can fill this gapby taking a dynamic, process-focused approach inexploring firm-specific knowledge creation and therelated employee governance issues.

Second, although employee stock ownership andfirm-employee relationship were proposed in thisstudy as the most important employee governancemechanisms, there are many other alternativesthat may also be applied to motivate a firm’semployees to make firm-specific human capital

investments. For instance, employees may be morewilling to invest in firm-specific human capital ifthey are promised future promotion opportunities(Carmichael, 1983) or board membership (Robertsand Van den Steen, 2000). Due to data limitations,we are not able to directly incorporate these alter-native mechanisms in this study. Future researchcan take into consideration a broader range ofgovernance or motivating mechanisms. Moreover,this study has implicitly treated the two employeegovernance mechanisms as independent, alterna-tive channels that help a firm achieve performanceadvantages based on its firm-specific knowledgeresources. However, the two governance mecha-nisms may be interrelated in that they can be eithercomplementary or substitutive. Although we con-ducted some preliminary analysis exploring suchrelationships and found some mixed evidence,future research might consider examining the inter-relationship of the governance mechanisms moredeliberately.

Third, our study has focused mainly on nonexec-utive employees, but it could be argued that execu-tives’ motivation and governance may also play acritical role in exploiting firm-specific resources.Executives, after all, make the major decisionsabout resource accumulation and allocation. Thegovernance of executives in larger firms operat-ing in complex environments may be particularlyimportant. Therefore, although it was beyond thescope of this study, it would be interesting forfuture research to study the conditions under whichexecutive-level ownership and nonexecutive-levelemployee ownership may have similar or differentmotivating roles.

Fourth, this study also suffered from certain datalimitations. For example, the study was concernedwith key employees who are directly involved inthe deployment of firm-specific knowledge andthus need to invest in firm-specific human cap-ital. But the items compiled in the ‘employeerelations’ dimension of the KLD data relate toa much broader range of employees, includinglower-level employees. Although we conductedrobustness tests by removing some items espe-cially related to lower-level employees and appliedan alternative measure of firm-employee relation-ships, this does not resolve all of the concerns.Similarly, employee stock ownership may be tar-geted to various levels of employees, includinglower-level employees. Although some previous



empirical studies have shown that key employ-ees are likely to have greater stock ownership(Brickley and Hevert, 1991) and that knowledge-intensive firms have greater employee ownership(Chen and Huang, 2006; Ryan and Wiggins, 2002),data on the specific shareholding of key knowl-edge workers would enhance the validity of theempirical findings.

Further, the effectiveness of using stock own-ership as an employee governance mechanismshould be affected by the specific nature of theownership. For example, many employee stockownership programs involve a vesting process. Asvesting schedules vary enormously, the effective-ness of employee stock ownership as a governancemechanism is likely to be influenced by the vestingperiod. This study could not directly incorporatethis information because of the unavailability ofsystematic archival data. In addition, some previ-ous research studies have maintained that it is notemployee ownership per se, but employees’ votingrights and participation that motivate them to makefirm-specific human capital investments (Caramelliand Briole, 2007; Klein, 1987; Kruse and Blasi,1997). Again, we did not have detailed data onemployees’ voting rights and bargaining channels.Our percentage ownership was only a proxy foremployees’ ex post bargaining power and incen-tives for participation. Future research may be ableto collect additional data to consider these dif-ferent features and dimensions of employee stockownership and examine the effects of employeeownership arrangements at a finer-grained level.

The patent data utilized in this study providedrich information about technological knowledgestocks and flows, but they also had inherent lim-itations when applied to measuring firm-specificknowledge assets. Future research might use sur-vey data to explore other types of unpatentedknowledge. The exploration of the role of firm-specific knowledge assets can be extended to firm-specificity in heterogeneous knowledge creationprocesses. For example, superior economic perfor-mance may come not only from the firm specificityof a firm’s innovative output, but also from thefirm’s unique configurational capability for con-tinuously generating new innovations ahead of itscompetitors.

In summary, the resource-based view placeshuman resources among the most importantresources available to a firm, and central to the

debate about how firms achieve sustainable com-petitive advantage (Coff, 1999; Mahoney, 2005).Thus, the willingness of a firm’s key employees toinvest in essential firm-specific knowledge shouldbe an important area of study in the evolving sci-ence of organization. This study was conceived asanother step toward a better understanding at thestrategic level of the effect of employee motivationand governance on firm behavior and performance.

ACKNOWLEDGEMENTS

We thank Amy Hillman, Evan Offstein, and sem-inar participants at Hong Kong University of Sci-ence and Technology for their helpful commentsand suggestions, and Associate Editor ConstanceHelfat for her editorial leadership. We also thankKozhikode Rajiv Krishnan for excellent data assis-tance. This research was supported by the HongKong Research Grants Council through RGC grantHKUST6251/03H.

REFERENCES

Alchian AA, Demsetz H. 1972. Production, informationcosts, and economic organization. American EconomicReview 62: 777–795.

Amit R, Schoemaker PJH. 1993. Strategic assets andorganizational rent. Strategic Management Journal14(1): 33–46.

Barney JB. 1991. Firm resources and sustained com-petitive advantage. Journal of Management 17(1):99–120.

Barney JB. 2001. Is the resource-based ‘view’ a usefulperspective for strategic management research? Yes.Academy of Management Review 26: 41–57.

Barney JB, Hansen MH. 1994. Trustworthiness as asource of competitive advantage. Strategic Manage-ment Journal , Winter Special Section 15: 175–190.

Barzel Y. 1989. An Economic Analysis of Property Rights.Cambridge University Press: Cambridge, UK.

Becker GS. 1964. Investment in Human Capital: ATheoretical and Empirical Analysis with SpecialReference to Education . National Bureau of EconomicResearch: New York.

Blair MM. 1995. Ownership and Control: RethinkingCorporate Governance of the Twenty-First Century.Brookings Institution: Washington, DC.

Blasi J, Kruse D, Bernsterin A. 2003. In the Companyof Owners: The Truth about Stock Options (and WhyEvery Employee Should Have Them). Basic Books:New York.

Blau P. 1964. Exchange and Power in Social Life. Wiley:New York.



Bourgeois LJ. 1981. On the measurement of organiza-tional slack. Academy of Management Review 6(1):29–39.

Bowman C, Ambrosini V. 2000. Value creation versusvalue capture: towards a coherent definition of value instrategy. British Journal of Management 11(1): 1–15.

Brickley JA, Hevert KT. 1991. Direct employee stockownership: an empirical investigation. FinancialManagement 20(2): 70–84.

Caramelli M, Briole A. 2007. Employee share ownershipand job attitudes: does culture matter? HumanResources Management Review 17(3): 290–304.

Carmichael L. 1983. Firm-specific human capital andpromotion ladders. Bell Journal of Economics 14(1):251–258.

Chen HL, Huang YS. 2006. Employee stock ownershipand corporate R&D expenditures: evidence fromTaiwan’s information technology industry. AsiaPacific Journal of Management 23(3): 369–384.

Chen MJ. 1996. Competitor analysis and interfirmrivalry: toward a theoretical integration. Academy ofManagement Review 21(1): 100–134.

Coase RH. 1937. The nature of the firm. Economica 4:386–405.

Coase RH 1960. The problem of social cost. Journal ofLaw and Economics 3: 1–44.

Coff RW. 1997. Human assets and managementdilemmas: coping with hazards on the road toresource-based theory. Academy of ManagementReview 22(2): 374–402.

Coff RW. 1999. When competitive advantage doesnot lead to performance: resource-based theory andstakeholder bargaining power. Organization Science10: 119–133.

Cogan GV, Burgelman RA. 1989. Intel Corporation (A):The DRAM decision. Stanford Business School, Case9-BP2-56A, Stanford, CA.

Cohen WM, Levinthal DA. 1989. Innovation and learn-ing: the two faces of R&D. Economic Journal 99:569–596.

Collins CJ, Smith KG. 2006. Knowledge exchange andcombination: the role of human resource practices inthe performance of high-technology firms. Academy ofManagement Journal 49: 544–560.

Cornell B, Shapiro AC. 1987. Corporate stakeholdersand corporate finance. Financial Management 16(1):5–14.

Demsetz H. 1967. Toward a theory of property rights.American Economic Review 57: 347–359.

Dierickx I, Cool K. 1989. Asset stock accumulation andsustainability of competitive advantage. ManagementScience 35(12): 1504–1511.

Dosi G. 1988. Sources, procedures, and microeconomiceffects of innovation. Journal of Economic Literature26: 1120–1171.

Døving E, Nordhaug O. 2002. Learning firm specificknowledge and skills: Conceptual issues and empiricalresults. Proceedings of the 3rd European conference onOrganizational Knowledge, Learning and Capabilities.Athens, Greece.

Eisenhardt KM. 1985. Control: organizational andeconomic approaches. Management Science 31(2):134–149.

Etzioni A. 1961. A Comparative Analysis of ComplexOrganizations . Free Press: New York.

Gimeno J. 1999. Reciprocal threats in multimarketrivalry: staking out ‘spheres of influence’ in the U.S.airline industry. Strategic Management Journal 20(2):101–128.

Glick HA, Feuer MJ. 1984. Employer-sponsored trainingand the governance of specific human capitalinvestments. Quarterly Review of Economics andBusiness 24(2): 91–104.

Gomez-Mejia LR, Larraza-Kintana M, Makri M. 2003.The determinants of executive compensation infamily-controlled public corporations. Academy ofManagement Journal 46(2): 226–237.

Gottschalg O, Zollo M. 2007. Interest alignment andcompetitive advantage. Academy of ManagementReview 32(2): 418–437.

Gouldner A. 1960. The norm of reciprocity. AmericanSociological Review 25(2): 161–178.

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

Greene WH. 2002. Econometric Analysis . Prentice-Hall:Upper Saddle River, NJ.

Gregg PA, Machin SI. 1988. Unions and the incidenceof performance-based pay in Britain. InternationalJournal of Industrial Organization 6: 91–107.

Griliches Z. 1961. Hedonic prices for automobiles: aneconometric analysis of quality change. Reprinted inPrice Indexes and Quality Change: Studies in NewMethods of Measurement, Griliches Z (ed). HarvardUniversity Press: Cambridge, MA; 55–87.

Griliches Z. 1981. Market value, R&D and patents.Economic Letters 7: 183–187.

Grossman SJ, Hart OD. 1986. The costs and benefits ofownership: a theory of vertical integration. Journal ofPolitical Economy 94: 109–122.

Hall BH. 2000. Innovation and market value. InProductivity, Innovation and Economic Performance,Barrell R, Mason G, O’Mahony M (eds). CambridgeUniversity Press: Cambridge, UK; 177–198.

Hall BH, Jaffe AB, Trajtenberg M. 2001. The NBERpatent citation data file: lessons, insights andmethodological tools. NBER Working paper No. 8498,National Bureau of Economic Research: Cambridge,MA.

Hall BH, Jaffe AB, Trajtenberg M. 2005. Market valueand patent citations. Rand Journal of Economics 36:16–38.

Hamilton B, Nickerson J. 2003. Correcting for endogene-ity in strategic management research. Strategic Orga-nization 1: 53–80.

Hart O. 1995. Firms, Contracts, and Financial Structure.Clarendon Press: Oxford, UK.

Hart O, Moore J. 1990. Property rights and the natureof the firm. Journal of Political Economy 98:1119–1158.

Hausman JA. 1978. Specification tests in econometrics.Econometrica 46: 1251–1272.



Helfat CE. 1994. Firm specificity in corporate appliedR&D. Organization Science 5(2): 173–184.

Hillman AJ, Keim GD. 2001. Shareholder value: stake-holder management, and social issues: what’s thebottom line? Strategic Management Journal 22(2):125–139.

Holmstrom B. 1982. Moral hazard in teams. Bell Journalof Economics 13(2): 324–340.

Jaffe AB, Lerner J. 1999. Privatizing R&D: patent policyand the commercialization of national laboratorytechnologies. NBER Working paper No. 7064.National Bureau of Economic Research: Cambridge,MA.

Jaffe AB, Trajtenberg M. 1999. International knowledgeflows: evidence from patent citations. Economics ofInnovation and New Technology 8: 105–136.

Jaffe AB, Trajtenberg M, Henderson R. 1993. Geo-graphic localization of knowledge spillovers as evi-denced by patent citations. Quarterly Journal of Eco-nomics 108(3): 577–598.

Jones DC, Kato T. 1993. The scope, nature, and effects ofemployee stock ownership plans in Japan. Industrialand Labor Relations Review 46: 352–367.

Kennedy P. 1998. A Guide to Econometrics. MIT Press:Cambridge, MA.

Kim J, Mahoney JT. 2005. Property rights theory,transaction costs theory, and agency theory: anorganizational economics approach to strategicmanagement. Managerial and Decision Economics26(4): 223–242.

Klein B. 1980. Transaction cost determinants of ‘unfair’contractual arrangements. American Economic Review70: 356–362.

Klein B, Crawford RA, Alchian AA. 1978. Verticalintegration, appropriable rents and the competitivecontracting process. Journal of Law and Economic 21:297–326.

Klein KJ. 1987. Employee stock ownership and employeeattitudes: a test of three models. Journal of AppliedPsychology 72: 319–332.

Kogut B, Zander I. 1992. Knowledge of the firm,combinative capabilities and the replication oftechnology. Organization Science 3(3): 383–397.

Kruse D. 1996. Why do firms adopt profit-sharingand employee ownership plans? British Journal ofIndustrial Relations 34(4): 515–538.

Kruse D, Blasi J. 1997. Employee ownership, employeeattitudes, and firm performance: a review of theevidence. In The Human Resources ManagementHandbook , Lewin D, Mitchell DJB, Zaidi MA (eds).JAI Press: Greenwich, CT; 113–151.

Kydland F, Prescott EC. 1977. Rules rather thandiscretion: the inconsistency of optimal plans. Journalof Political Economy 85: 473–491.

Laffont JJ, Tirole J. 1988. The dynamics of incentivecontracts. Econometrica 56(5): 1153–1175.

Libecap GD. 1989. Contracting for Property Rights.Cambridge University Press: Cambridge, UK.

Lindenberg EB, Ross SA. 1981. Tobin’s q ratio andindustrial organization. Journal of Business 54: 1–32.

Mahoney JT. 2005. Economic Foundations of Strategy.Sage: Thousand Oaks, CA.

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

Makadok R. 2003. Doing the right thing and knowingthe right thing to do: why the whole is greater thanthe sum of the parts. Strategic Management Journal ,October Special Issue 24: 1043–1055.

Milgrom P, Roberts J. 1992. Economics, Organizationand Management. Prentice-Hall: Englewood Cliffs,NJ.

Miller GJ. 1992. Managerial Dilemmas: The PoliticalEconomy of Hierarchy. Cambridge University Press:Cambridge, MA.

Mowday RT, Porter LW, Steers RM. 1982. Employee-organizational linkages: the psychology of commit-ment, absenteeism, and turnover. In Organizationaland Occupational Psychology , Warr P (ed). AcademicPress: New York; 219–229.

Nelson R, Winter SG. 1982. An Evolutionary Theoryof Economic Change. Harvard University Press:Cambridge, MA.

Ouchi WG. 1980. Markets, bureaucracies, and clans.Administrative Science Quarterly 25: 129–141.

Pavitt K. 1991. What makes basic research economicallyuseful? Research Policy 20(2): 109–119.

Poole M, Jenkins G. 1990. The Impact of EconomicDemocracy. Routledge: London, UK.

Priem RL, Butler JE. 2001. Is the resource-based theory auseful perspective for strategic management research?Academy of Management Review 26(1): 22–40.

Rajan RG, Zingales L. 1998. Power in a theory of thefirm. Quarterly Journal of Economics 113: 387–432.

Roberts J, Van den Steen E. 2000. Stakeholder interests,human capital investment and corporate governance.Stanford Graduate School of Business Working paperNo. 1631, Stanford, CA.

Rock EB, Wachter ML. 1999. Tailored claims and gov-ernance: the fit between employees and shareholders.In Employee and Corporate Governance, Blair MM,Roe MJ (eds). Brookings Institution Press: Washing-ton, DC; 121–159.

Rosenkopf L, Nerkar A. 2001. Beyond local search:boundary-spanning, exploration, and impact in theoptical disk industry. Strategic Management Journal22(4): 287–306.

Rousseau DM. 1995. Psychological Contracts in Organi-zations: Understanding Written and Unwritten Agree-ments. Sage: Thousand Oaks, CA.

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

Ryan HE, Wiggins RA. 2002. The interactions betweenR&D investment decisions and compensation policy.Financial Management 31: 5–29.

Schelling TC. 1960. The Strategy of Conflict. HarvardUniversity Press: Cambridge, MA.

Schulze WS, Lubatkin MH, Dino RN. 2003. Exploringthe agency consequences of ownership dispersionamong the directors of private family firms. Academyof Management Journal 46(2): 179–194.



Sharfman MP. 1996. The construct validity of the Kinder,Lydenberg, and Domini social performance ratingsdata. Journal of Business Ethics 15(3): 287–296.

Shleifer A, Summers LH. 1988. Breach of trust inhostile takeovers. In Corporate Takeovers: Causesand Consequences , Auerbach AJ (ed). University ofChicago Press: Chicago, IL; 33–56.

Shleifer A, Vishny R. 1997. A survey of corporategovernance. Journal of Finance 52: 737–783.

Singh J. 1986. Performance slack and risk takingin organizational decision making. Academy ofManagement Journal 29: 562–585.

Teece DJ. 1986. Profiting from technological innovation.Research Policy 15: 285–306.

Tirole J. 1988. The Theory of Industrial Organization.MIT Press: Cambridge, MA.

Titman S. 1984. The effect of capital structure on a firm’sliquidation decision. Journal of Financial Economics13(1): 137–151.

Vandenberghe C, Bentein K, Michon R, Chebat J-C,Tremblay M, Fils J-F. 2007. An examination of the

role of perceived support and employee commitmentin employee–customer encounters. Journal of AppliedPsychology 92(4): 1177–1187.

Waddock SA, Graves SB. 1997. The corporate socialperformance-financial performance link. StrategicManagement Journal 18(4): 303–319.

Wang H, Barney JB. 2006. Employee incentives to makefirm specific investment: implications for resource-based theories of corporate diversification. Academyof Management Review 31(2): 466–476.

Waugh FV. 1928. Quality factors influencing vegetableprices. Journal of Farm Economics 10: 185–196.

Williamson OE. 1975. Markets and Hierarchies: Analysisand Antitrust Implications. Free Press: New York.

Williamson OE. 1985. The Economic Institutions of Cap-italism: Firms, Markets and Relational Contracting.Free Press: New York.

Zaheer A, McEvily B, Perrone V. 1998. Does trustmatter? Exploring the effects of interorganizationaland interpersonal trust on performance. OrganizationScience 9: 141–160.



APPENDIX: THE CATEGORIES IN THE‘EMPLOYEE RELATIONS’ DIMENSIONOF THE KLD DATASET

STRENGTHS

Strong Union Relations (EMP-str-A): The Com-pany has a history of notably strong union rela-tions.

No-Layoff Policy (EMP-str-B): The Company hasmaintained a consistent no-layoff policy. KLD hasnot assigned strengths for this issue since 1994.

Cash Profit-Sharing (EMP-str-C): The Companyhas a cash profit-sharing program through which ithas recently made distributions to a majority of itsworkforce.

Employee Involvement (EMP-str-D): The Com-pany strongly encourages worker involvement and/or ownership through stock options available to amajority of its employees through gain sharing,stock ownership, sharing of financial information,or participation in management decision making.

Strong Retirement Benefits (EMP-str-F): The Com-pany has a notably strong retirement benefits pro-gram.

Health and Safety Strength (EMP-str-G): TheCompany is noted by the U.S. Occupational Health

and Safety Administration for its safety programs.KLD began assigning strengths for this issue in2003.

Other Strength (EMP-str-X): The Company has agood employee safety record or demonstrates othernoteworthy commitments to its employees’ well-being.

CONCERNS

Poor Union Relations (EMP-con-A): The Com-pany has a history of notably poor union relations.

Health and Safety Concern (EMP-con-B): TheCompany recently has either paid substantial finesor civil penalties for willful violations of employeehealth and safety standards, or has been otherwiseinvolved in major health and safety controversies.KLD changed the name of this rating from SafetyControversies in 2003.

Workforce Reductions (EMP-con-C): The Com-pany has reduced its workforce by 15 percentin the most recent year or by 25 percent dur-ing the past two years, or it has announcedplans for such reductions. Before 1994, the con-cern is only assigned to companies that have laidoff 15 percent of workers in the most recentyear.


firm-specific knowledge resources and competitive … · 2020. 1. 13. · 1266 h. c. wang, j. he,...

Documents