THE STRATEGIC CONTEXT OF EXTERNAL NETWORK TIES:...

® Academy of Management Journal2001. Vol. 4. No. 4, 639-660.

THE STRATEGIC CONTEXT OF EXTERNAL NETWORK TIES:EXAMINING THE IMPACT OF DIRECTOR APPOINTMENTS ON

BOARD INVOLVEMENT IN STRATEGIC DECISION MAKING

MASON A. CARPENTERUniversity of Wisconsin—Madison

JAMES D. WESTPHALUniversity of Texas at Austin

This study examines how external network ties determine a board's ability to contrib-ute to the strategic decision making process. Although the simple number of directorappointments to other boards does not affect board monitoring or advice on strategy,appointments that can provide directors with relevant strategic knowledge and per-spective do predict such involvement. In effect, the strategic context of social networkties, not simply the number of ties, is an important influence on corporate governance.

Increased research attention is being devoted tothe influence of corporate board members on im-portant organizational outcomes. According toagency theorists, effective boards independentlymonitor strategic challenges facing a firm and eval-uate management's performance in addressingthem (Beatty & Zajac, 1994; Fama & Jensen, 1983).Directors may overturn poor decisions or replace"underperforming" managers as a result of suchmonitoring (Brudney, 1982). Moreover, agency the-ory also suggests that directors are motivated toengage in competent monitoring activity by thereputational effects of poor decisions (Fama, 1980).From this perspective, poor management decisionswill be attributed, in part, to inadequate oversightby directors, which will ultimately hurt theirprospects iri the labor market for directors andmanagers.

In addition, although agency theorists emphasizethe board's role as an independent control mecha-nism, the broader literature on boards suggests asecond possible role for directors; specifically, insome cases they may provide ongoing advice to topmanagers on possible strategic changes or theimplementation of existing strategies (Demb &Neubauer, 1992; Lorsch & Maclver, 1989). In suchcases, boards serve as a strategic consultant to topmanagers, rather than (or in addition to) exercising

Both authors contributed equally and are listed inalphabetical order. We appreciate the helpful commentsand suggestions of Andrew Henderson and RakeshKhurana on an earlier version of this article. The articlehas also benefited from the contributions of Rita Kosnik,Gregory Northcraft, and three anonymous reviewers forthe Academy of Management Journal.

independent control. This view was advanced byPfeffer and Salancik (1978:170), who identified theprovision of advice and counsel and the exercise ofcontrol as two primary components of a board'sinternal administrative function (see also West-phal, 1999).

The governance literature also suggests that thereis considerable variance in the degree to whichdirectors make an actual impact on strategic deci-sion making, with some boards unable to monitoror advise management effectively (Mace, 1971;Wade, O'Reilly, & Chandratat, 1990). Prior empiri-cal studies have focused primarily on lack of boardpower as an explanation for limited board involve-ment. For example, it is often proposed that boardsare less likely to exert control over strategic deci-sion making on behalf of shareholders when theylack formal or social independence from manage-ment—as indicated by the percentage of outside ornonexecutive directors, or the prevalence of friend-ship ties or other social connections between man-agers and directors (Baliga, Moyer, & Rao, 1996;Boeker & Goodstein, 1993; Hill & Snell, 1988;Kesner, Victor, & Lamont, 1986; Mallette & Fowler,1992; Wade et al., 1990). However, as Westphal's(1998: 530) recent review of the literature suggests,there is little consistent evidence that board inde-pendence or "structural power" increases board in-volvement in strategic decision making.

Several authors have noted the persistent chal-lenges faced by directors in making meaningfulcontributions to corporate strategy, regardless oftheir power to do so (Demb & Neubauer, 1992;Lorsch & Maclver, 1989; Westphal, 1999; Westphal& Zajac, 1997). Such challenges have been held tostem in part from questions of whether many direc-

639

640 Academy of Management Journal August

tors have suitable knowledge or information to con-tribute meaningfully to strategy. For instance, it hasbeen repeatedly suggested that outside directors areoften inadequately prepared to participate in boarddiscussions because their time and attention aredivided and diluted by their other board appoint-ments; serving on boards at multiple companiesmakes it difficult for them to gain an adequateunderstanding of the issues facing any one firm.For these reasons, proponents of governance reformhave strongly advocated limits to the number ofboards upon which directors may sit [Business-Week, 1997).

A number of empirical studies have examinedhow board power and independence affect policyoutcomes, but very little research represents an at-tempt to identify factors that determine whetherboards have adequate knowledge and informationto make meaningful contributions to strategic deci-sion making. Moreover, research has not addressedthe specific question of how multiple board ap-pointments affect directors' ability to contribute tostrategy. To address these questions, we developeda sociocognitive perspective on how appointmentsto other boards affect the capability of the membersof a firm's board to monitor and advise its manage-ment in the strategic decision making process. Wetested our theoretical perspective with a uniquedata set that combines archival reports with pri-mary behavioral data obtained through surveys ofForbes 1000 outside directors and GEOs. The theo-retical perspective and empirical analysis pre-sented here may help answer recent calls for re-search that examines the "substantive context" ofboard appointments (Nohria, 1992: 14) and movesour understanding beyond the simple number ofsuch appointments or director independence aspredictors of board influence and decision making(Mizruchi, 1996; Pettigrew, 1992; Stinchcombe,1990).

THEORY AND HYPOTHESES

Director Appointments and Board Involvementin Strategic Decision Making

The sociocognitive perspective developed in thisstudy suggests the importance of directors' net-works of appointments to other boards in determin-ing whether they have the appropriate strategicknowledge and perspective to monitor and advisemanagement in the strategic decision making pro-cess. As noted above, critics of corporate gover-nance have typically argued that directors' appoint-ments to other boards reduce their ability tocontribute to decision making at a focal board.

Such an argument assumes that the knowledge andperspective gained on other boards are largely ir-relevant to decision making at the focal firm. Incontrast, our sociocognitive perspective indicateshow experience on other boards can enhance ordiminish directors' ability to contribute to strategy,by focusing their attention on relevant strategic is-

sues.The sociocognitive perspective on organizational

decision making suggests that individuals copewith complex decision making tasks by relyingupon the schemata or "knowledge structures" theyhave developed about their environment (Kiesler &Sproull, 1982; Walsh, 1995). In the absence of com-plete information, or given uncertainty regardingthe relevance of different pieces of information,individuals tend to follow a top-down or theory-driven approach to decision making, rather than abottom-up or data-driven approach based onpresent information (Abelson & Black, 1986; Nis-bett & Ross, 1980; Ocasio, 1997). Given the extremeinformation complexity facing directors in evaluat-ing strategic decisions (Lorsch & Maclver, 1989),they can be expected to rely heavily upon the im-plicit theories that they have developed regardingcorporate strategy and the competitive environ-ment. Moreover, from this perspective, the knowl-edge structures that individuals use to cope withinformation-processing demands are developedfrom experience in similar roles (Dearborn & Si-mon, 1958; Walsh, 1995).

In our framework, directors are likely to useknowledge structures developed from their experi-ence on other boards. The literature on interlockingdirectorates supports this view. This work demon-strates how the involvement of directors on otherboards provides an important source of informationabout business practices and policies (cf. Mizruchi,1996; Palmer, Jennings, & Zhou, 1993). For exam-ple, Useem (1982) observed that executives usetheir board appointments as a way to scan the en-vironment for timely and pertinent information. Hequoted several executives who suggest that boardappointments provide a vehicle for learning,making such statements as "Direct involvement inother companies' affairs replaces an awful lot ofreading . . . it's a hell of a tool for top managementeducation" (1982: 209-210). Similarly, directorscan learn about the efficacy of different practicesand how to implement them properly by observingthe consequences of management decisions (Haun-schild, 1993). Such learning is particularly vividbecause directors observe the decision-making pro-cess firsthand in their monitoring role, participateactively by giving advice to management, and thenwitness the consequences of those decisions.

2001 Carpenter and Westphal 641

Directors also learn about business practicesthrough their communication with other directorsin board and committee meetings. Information ac-quired from fellow directors may be particularlyinfluential because it often comes from a trustedsource (Davis, 1991; Useem, 1982; Weick, 1995).This information is typically more timely and up-to-date than that derived from secondary sources,and it may also be more salient because of its re-cency (Kahneman, Slovic, & Tversky, 1982). Thus,a sociocognitive perspective on board involvementemphasizes how directors' social structural con-text, including their ties to other boards, providesdirect strategic experience and indirect access tostrategic information through social contact withother directors. Such experience and informationcan, in turn, critically inform the knowledge struc-tures used to monitor decisions or give advice on afocal board. In the following section, we furtherdevelop our sociocognitive perspective to considervariation in the strategic context of director ties toother boards, in order to address whether and whenthose ties provide relevant strategic knowledge andperspective for monitoring and advising the man-agement of a focal firm.

Environmental Stability and DirectorInvolvement in Strategic Decision Making

Related board ties and involvement in stableenvironments. Environmental stability refers to theextent to which a firm's competitive environmentis complex, uncertain, and prone to strategicchange (Huber & McDaniel, 1986). It is emphasizedhere because such stability is a key determinant ofthe particular strategic issues facing a firm and itstop management (Duncan, 1972; Wiersema & Ban-tel, 1993). Accordingly, environmental stabilitymay be an important determinant of how directorscan contribute to strategic decision making. Theoryand research on environmental turbulence andstrategic decision making distinguish between twobasic strategic issues in the decision-making pro-cess—the development of new strategies and theimplementation of existing strategies—and suggestthat, in stable environments, the latter is more im-portant. Specifically, if there is less change in anenvironment, there is less need to regularly identifynew strategic alternatives in order to maintain fitwith the environment, so firms in stable environ-ments are more likely to compete primarily throughthe better implementation of existing strategies(Andrews, 1971; Fredrickson, 1984; Ginsberg,1990; Tushman & Romanelli, 1985). Moreover,Fama and Jensen (1983) explicitly recognizedboard monitoring of strategy implementation as an

important component of a board's obligation to pro-tect shareholders.

How then might directors' ties to other boardsenhance their ability to monitor and advise man-agement on the implementation of existing strate-gies in a stable environment? We suggest that di-rectors will be better able to contribute to strategyin such an environment when their other board tiesare strategically related to the focal firm. Strategi-cally related board ties refer to a director's appoint-ments to the boards of other companies that followsimilar corporate strategies and operate in similarproduct-market and international market contexts.A central tenet of the strategy literature is that theeffectiveness of strategy implementation is contin-gent on strategy content (Barney & Zajac, 1994;Galbraith & Kazanjian, 1986). For instance, re-search suggests that a strategy of diversificationrequires different kinds of corporate reward sys-tems and information systems than strategies thatfocus on a single market (Galbraith & Merrill, 1991;Gerhart & Milkovich, 1990; Gomez-Mejia, 1992).Thus, given that strategically related ties enabledirectors to observe firsthand the experiences ofother firms in implementing similar strategies, theyprovide directors with a more sophisticated under-

, standing of the combination of systems and struc-tures needed for successful implementation of thefirm's strategy. Moreover, strategically relatedbocud ties also help directors acquire relevantknowledge through social interaction with otherdirectors in board and committee meetings, asboard members evaluate management and raiseideas and suggestions for better strategy implemen-tation.

From a sociocognitive perspective, these socialconnections and opportunities for vicarious learn-ing can lead to more highly developed knowledgestructures for implementing the focal firm's strat-egy; Research on sociocognition has shown thatindividuals who have experience concentrated in arelated domain, rather than dispersed across differ-ent contexts, have more highly developed knowl-edge structures for that domain, with fewer schemacategories and more information units per category(Day & Lord, 1992; Lurigio & GarroU, 1985; Sujan,Sujan, & Bettman, 1988). As a result, individualsengaging in problem solving for domains (such asstrategy) where they have concentrated experiencenot only have more information at their disposal,but also have more efficiently structured informa-tion, and this leads to faster and more accurateinformation processing (Day & Lord, 1992).

Thus, to the extent that firms in stable environ-ments compete primarily through better imple-mentation of existing strategies (rather than iden-


tification of new strategies to fit a changing envi-ronment), strategically related board ties shouldenhance directors' ability to contribute to strategicdecision making in stable environments throughmonitoring and advice interactions. This point sug-gests several related hypotheses. The first hypoth-esis addresses the effect of a director's appoint-ments to other boards on his or her perceivedability to contribute to a focal board. Involvementwith other firms following strategies similar tothose of a focal firm better equips a director tomonitor management decision making and providerelevant information and advice to managers on thefocal board. Thus, such an individual may perceivehim or herself to be better able to contribute tostrategic decision making. In effect, a director's per-ceived ability to contribute to decision making isexpected to mediate the effect of appointments toother boards on involvement in monitoring andadvice interactions. The first hypothesis, whichpredicts that strategically related board appoint-ments will increase a director's perceived ability tocontribute to decision making, concerns one por-tion of this mediated relationship.

Two additional hypotheses address the overall,board-level relationship between the portfolio of alldirectors' appointments and actual board behavior.As discussed above, boards may contribute to stra-tegic decision making by regularly monitoring thedecision-making process, as suggested by agencytheorists, or by providing advice to top managerson strategic issues. The theoretical perspective de-veloped here suggests that in stable environments,board appointments to strategically related firmsshould enhance the capacity of a focal firm's boardmembers to contribute to strategic decision makingthrough increased monitoring activity or the provi-sion of more advice to management on strategicissues. Thus,

Hypothesis la. In a stable environment, theappointment of a director to the boards ofother firms that are strategically related to thefocal firm will increase the director's perceivedability to contribute to board discussions ofstrategic issues.

Hypothesis lb. In a stable environment, theappointment of a director to the boards ofother firms that are strategically related to thefocal firm will increase the level of board mon-itoring of strategic decision making.

Hypothesis lc. In a stable environment, theappointment of a director to the boards ofother firms that are strategically related to the

focal firm will increase the level of board ad-vice interactions on strategic issues.

Heterogeneous board ties and involvementin unstable environments. As noted above, unsta-ble environments are characterized by a relativelyhigh level of unpredictable change or volatility(Aldrich, 1979; Duncan, 1972; Sharfman & Dean,1991), which places considerable information-processing demands on corporate leaders (Carpen-ter & Fredrickson, 2001; Fredrickson & Mitchell,1984; Priem, 1990; Wiersema & Bantel, 1993). Insuch environments, organizational success oftendepends on the ability of top managers to identifynew strategic alternatives that maintain an organi-zation's fit with its changing environment (Hale-blian & Finkelstein, 1993; Tushman & Anderson,1986). Thus, leaders must not only attend to thecurrent strategy, but must also recognize when andhow that strategy should be changed.

Our sociocognitive perspective suggests that inturbulent environments, directors' ability to con-tribute to strategy is enhanced by a combination ofstrategically related board ties and ties to firms thatfollow different strategies than does a focal firm. Asin stable environments, board involvement withsimilar companies provides information andknowledge that can enhance a director's ability tomonitor the firm's current strategy. Moreover,board ties to similar companies can also help adirector stay abreast of changes in the businessenvironment.. However, board involvement with companiesfollowing different strategies and operating in dif-ferent business environments typically provides di-rectors with greater knowledge and insight about abroad range of potential strategic alternatives. Asdiscussed above, our perspective suggests thatboard ties to firms with related strategies can pro-vide firsthand experience and indirect informationthrough social contact with other directors thatleads to more highly developed knowledge struc-tures related to implementing a firm's current strat-egy. Research has also shown that well-developedknowledge structures can hinder effective informa-tion processing outside an individual's area' of ex-pertise. For instance, individuals with more con-centrated exposure to a particular problem-solvingapproach or strategy are less likely to notice orconsider alternative approaches (Dutton & Duncan,1987; Ocasio, 1997). Thus, when directors' otherappointments are concentrated among similar com-panies, they are less likely to notice strategic alter-natives emerging in other environments. In effect,directors' schemata for major strategic alternativesare highly focused on a limited range of options.


Similarly, the social cohesion perspective on net-work ties suggests that if the appointments of thedirectors of a firm's board are highly concentratedamong firms with very similar strategies, these in-dividuals are more likely to become socialized intoaccepting the firm's current strategy (Burt, 1987;Palmer et al., 1993). Social network ties, includingboard interlock ties, channel social influence aswell as information (Burt, 1987; Davis,- 1991;Walker, 1985). Thus, through their participation inmonitoring implementation of similar strategies atother firms, directors with strategically relatedboard ties should tend to develop beliefs that jus-tify those strategies and accept them as appropriate.As a result, when external appointments are largelyconcentrated among firms with strategies similar tothe focal firm's, this similarity may reinforce exist-ing managerial commitment to the current strategyand lead directors to ignore environmental changesthat threaten its long-term viability (note that stra-tegically related ties may not increase commitmentto the current mode of strategy implementation,because such ties link the focal board with firmsthat have similar strategy content, but not necessar-ily similar implementation).

Thus, in turbulent environments, directors canbenefit from interlocks that expose them to possiblestrategic alternatives. Such ties effectively broadenthe schemata or knowledge structures that directorsuse in monitoring and advising management, sothat boards are not only more likely to identifyappropriate strategic alternatives to a current strat-egy, but are also more willing to change the currentstrategy in order to maintain the organization's fitwith its changing environment. In effect, board tiesto firms with different strategies can help preventor counteract excessive managerial commitment toa firm's current strategy (Hambrick, Geletkanycz, &Fredrickson, 1993; Weick, 1995). AUernatively,when director appointments are concentratedamong strategically dissimilar firms, with very fewboard ties to firms following similar strategies, aboard may lack sufficient expertise to monitor thefirm's current strategy or to assess the organiza-tional implications of abandoning the strategy. In-deed, a common dilemma facing firms is how todevelop new strategies while simultaneously im-plementing current strategies (Kazanjian & Drazin,1987). Therefore, in unstable environments, an op-timal portfolio of outsider board appointments mayinclude a heterogeneous mix of ties to strategicallysimilar and dissimilar firms (that is, both firmswith strategies similar to those of the focal firm andthose with dissimilar strategies). Such heterogene-ity is likely to wed the knowledge and expertiserequisite to monitor ongoing implementation of the

current strategy with information and advice aboutpossible strategic alternatives that would allow thefirm to maintain fit with its changing environment(Ginsberg, 1990).

This perspective is consistent with the view, pre-sented in the top management team literature, thatexposure to different beliefs about means-ends re-lationships (different beliefs about what strategieslead to high performance) through greater diversityof backgrounds and experience can facilitate adap-tation in turbulent environments by stimulating de-bate about the appropriateness of a current strategyand about the feasibility of strategic alternatives(Boeker, 1997; Bourgeois, 1980; Haleblian & Finkel-stein, 1993; Hambrick & Mason, 1984). Moreover,the larger literature on demographic diversity sug-gests that diversity in job-related experience (inthis case, experience in strategic decision makingon other boards) tends to have beneficial effects ongroup process and performance. In contrast, diver-sity on visible, or non-job-related, characteristics(such as age) is less likely to have such beneficialoutcomes (Carpenter & Fredrickson, 2001; Jehn,Chadwick, & Thatcher, 1997: 300; O'Reilly, Snyder,& Boothe, 1993; Pelled, 1996; Williams & O'Reilly,

• 1998). Research on small group decision makingsuggests that the positive effects of job-related di-versity on group performance are mediated by task-related conflict, and the more negative effects ofnon-work-related diversity are mediated by affec-tive or relationship conflict. In summarizing thisliterature, Jehn and colleagues stated that "visibleindividual demographic differences increase rela-tionship conflict, while informational demographicdifferences increase task-focused conflict [and]conflict researchers . . . have recently found thatwhile relationship conflicts based on personalityclashes and interpersonal dislike are detrimental togroup functioning, task conflicts based on disagree-ments regarding the specific task content are bene-ficial in many situations" (Jehn et al., 1997: 287-288). Similarly, Williams and O'Reilly stated that"In general, research shows that functional diver-sity [diversity on a job-related characteristic] haspositive effects on group performance" (1998: 100),and they also attributed this overall finding tohigher levels of task conflict in diverse groups.

Moreover, job-related diversity is particularlylikely to enhance group functioning when individ-uals know each other (Pelled, Eisenhardt, & Xin,1999; Williams & O'Reilly, 1998). According toWilliams and O'Reilly, "Ironically, laboratory stud-ies have shown that groups may be more able to useunique information when group members are fa-miliar with one another instead of being strangers"(1998: 99). Pelled and colleagues (1999) showed


that such familiarity reduces emotional conflict re-sulting from demographic diversity. For these rea-sons, heterogeneous experiences among boardmembers may not enhance affective conflict, givenabundant evidence that directors of Forbes 1000firms are not only familiar with one another, butoften form cohesive social bonds in board and com-mittee meetings and through social interaction out-side of formal meetings (Useem, 1982).

Thus, in our sociocognitive framework, diver-sity in the strategic experience of board membersthrough network ties to other firms should en-hance their ability to contribute in turbulent en-vironments by engendering debate or task-relatedconflict about a firm's current strategy. Such in-teraction should facilitate adaptation to environ-mental changes (Wiersema & Bantel, 1992). More-over, given the social cohesion characterizingmost boards, the benefits of debate and task-related conflict among directors should not beoverwhelmed by affective conflict. These bene-fits from diversity in strategy experience are lessrelevant in stable environments where there isless need to engage in strategic change to main-tain fit with the environment.

This argument leads to a second set of hypothe-ses. Hypothesis 2a, which parallels Hypothesis la,addresses how an individual director's board ap-pointments may enhance her or his perceived abil-ity to contribute to decision making, given thatsuch perceptions are expected to mediate the effectof appointments to other boards on involvement inmonitoring and advice interactions. Our theoreticalargument that heterogeneous board ties can en-hance involvement in turbulent environments im-plies that, to the extent that an individual director'sboard ties complement the ties of other directorswith respect to strategic relatedness, thus increas-ing board heterogeneity, the director should be bet-ter able to contribute to the board through monitor-ing and advice interactions. According to this logic,for instance, a director with many board appoint-ments to strategically dissimilar firms is particu-larly valuable to a focal board if the ties of otherdirectors are concentrated among strategically sim-ilar firms. Two additional hypotheses, 2b and 2c,which parallel Hypotheses lb and lc, address howboard-level ties may affect actual board behaviorsin unstable environments. The theoretical argu-ment developed above would suggest that in suchenvironments, boards with a heterogeneous mix oflinks to both strategically different and strategicallysimilar firms should be better able to contribute tostrategic decision making through either increasedmonitoring activity or more frequent advice tomanagement on strategic issues.

Hypothesis 2a. In an unstable environment,the greater the extent to which a director'sboard appointments to other firms comple-ment the appointments of other directors intheir strategic relatedness to the focal firm, thegreater the director's perceived ability to con-tribute to board discussions of strategic issues.

Hypothesis 2b. In an unstable environment,the greater the extent to which directors' boardappointments are heterogeneous in their stra-tegic relatedness to the focal firm, the higherthe level of board monitoring of strategic deci-sion making.

Hypothesis 2c. In an unstable environment, thegreater the extent to which directors' boardappointments are heterogeneous in their stra-tegic relatedness to the focal firm, the higherthe level of board advice interactions on stra-tegic issues.

METHODS

Sample and Data Collection

The sample frame for this study consisted of 600large and medium-sized companies randomly se-lected from the Forbes 1000 index of U.S. industrialand service firms. To gauge the behavioral pro-cesses that characterize board involvement in stra-tegic decision making, we sent a questionnaire sur-vey to all 600 CEOs from these companies. Inaddition, to assess directors' perceptions abouttheir involvement, we sent a second survey to eachindividual serving as an outside director at a com-pany whose CEO had responded to the first survey(n = 1,312). All surveys were distributed in April1995.

Although surveys have been used frequently tomeasure behavioral processes at lower levels oforganizations, surveys of top managers have oftensuffered from low response rates (less than 25 per-cent). To ensure the highest possible response inthis case, we took the following steps (Fowler,1993; Groves, Cialdini, & Couper, 1992): (1) Anin-depth pretest was used to streamline the survey,making it easier and more appealing to complete(see further discussion below); (2) requests for par-ticipation linked the current study with an ongoingseries of surveys on top management issues con-ducted by a major business school (to which hun-dreds of these CEOs' peers had responded), empha-sized the need for research on CEO-board relations,and engaged respondents' natural interest in thetopic (see Groves et al., 1992); and (3) about 21 daysafter the initial mailing, nonrespondents were sent

2001 Carpenter and 'Westphal 645

a second letter with a new questionnaire. In total,263 GEOs and 564 outside directors responded,representing response rates of 44 percent and 43percent, respectively. These response rates are highin comparison to those of other top managementsurveys (cf. Pettigrew, 1992). Data on diversifica-tion, internationalization, or board interlocks wereunavailable for 35 of the responding companies;thus, survey data from 228 GEOs and 492 outsidedirectors are used in the analyses, numbers repre-senting 38 percent of all GEOs and directors in thesample frame (that is, on the average, the finalsample includes 2.2 directors per company in theGEO survey).

To check for nonresponse bias, we collected ar-chival data for companies in the larger sampleframe. For the 531 companies with complete data,we examined whether respondents and nonrespon-dents differed significantly on several differentvariables derived from archival sources using theKolmogorov-Smirnov two-sample test (Siegel &Gastellan, 1988). This test assesses whether signif-icant differences exist in the distribution of respon-dents and nonrespondents for a given variable, in-cluding differences in central tendency, dispersion,skewness, and so forth. The results of this testprovide consistent evidence across multiple vari-ables that respondents and nonrespondents comefrom the same population. Moreover, separate anal-yses also showed that directors in the survey sam-ple were representative of directors in the largersample frame (all outside directors at the 531 com-panies for which complete archival data were avail-able) with respect to (1) the independent variables(such as relatedness of individual directors' boardties), and (2) director characteristics included ascontrol variables in the study (such as managementexperience, education level, board tenure, andfunctional background). Thus, it appears that sam-ple selection bias was not present in the data.

Data on board interlocks, director characteristics,ownership, and board structure were obtained fromthe following sources: Standard &• Poor's Registerof Corporations, Directors, and Executives, the Dun&• Bradstreet Reference Book of Corporate Manage-ments, Who's Who in Finance and Industry, andcorporate proxy statements. Data on firm diversifi-cation and international operations were obtainedfrom the PG-GOMPUSTAT Business SegmentDatabase and from Gompact Disclosure. Size andperformance data were also obtained from PG-GOMPUSTAT. All independent and control vari-ables were gauged in the period t - 1, and depen-dent variables were gauged at time t.

Dependent Measures

Data for our three dependent variables, directors'perceived ability to contribute to board discussions,board monitoring, and board advice interactionswere gathered through surveys. To enhance theconstruct validity of the survey measures, we con-ducted a pretest involving in-depth pilot inter-views with 22 top managers and board members(cf. Fowler, 1993: 102). Following Judge andZeithaml (1992), in the questionnaire we definedstrategic issues as nonroutine, resource allocationdecisions that should affect the performance of anorganization. We further sought to develop surveyquestions that would assess a board's involvementin either the implementation or the considerationof new strategies. We used feedback from partici-pants in the pretest to ensure that the items wouldcapture both areas of strategic issue involvementand, in the survey, included additional questionsthat focus on each area of involvement. After com-pleting the pilot questionnaire, each individualwas asked to identify questions that were unclear,difficult to answer, or potentially subject to bias.These interviews were also used to ensure thatquestions were interpreted as expected, to identifyimprovements to the format of the survey, and tomodify its length. Multiple response formats wereused to reduce response bias, and items measuringeach construct were scattered throughout the sur-vey (DeVellis, 1991). Moreover, we carefullyworded questions to minimize the likelihood ofsocial desirability bias, using input from the pilotinterviews.

Director's perceived ability to contribute to boarddiscussions was measured with a multi-item scalein the director survey. Specific items in this scaleassess the degree to which directors perceive thatthey have sufficient knowledge on relevant strate-gic issues to contribute to board discussions andthe degree to which they feel adequately preparedto contribute. Gronbach's alpha for this scale was.88, suggesting acceptable interitem reliability(Nunnally, 1978). After a factor analysis was ap-plied to the survey items using the iterated princi-pal factors method, a scree test indicated one com-mon factor, and promax rotation verified that allitems loaded on the same factor as expected, withloadings for each item greater than .5. Thus, weestimated factor scores using the Bartlett method(Harman, 1976).

Board advice interactions and board monitoringwere also assessed with multi-item scales. Thewording of each question was developed fromavailable qualitative research (Alderfer, 1986;Demb & Neubauer, 1992; Lorsch & Maclver, 1989)


suggesting how top managers and directors de-scribe GEO-board interaction and the board's rolevis-a-vis management (for instance, as a soundingboard on strategic issues); in addition, we usedfeedback from the pilot interviews to further im-prove the clarity and face validity of each question.For instance, questions about advice interactionsasked about the extent to which the GEO solicitedboard input on corporate strategy and the fre-quency of advice and council discussions withboard members, and monitoring questions askedabout the extent to which the board monitoredstrategic decision making or evaluated GEO perfor-mance. The Appendix gives specific items. Thesemeasures assess monitoring and advice interac-tions at the board level. We also conducted separateanalyses of individual director involvement inmonitoring and advice interactions; the results ofthese analyses were consistent with those reporjtedbelow for boards.

The iterated principal factors method was thenagain applied to the survey items. A scree testshowed two common factors, and promax rotationindicated that the monitoring and advice itemsloaded on different factors as expected, with load-ings for each item greater than .5 on one factor andless than .2 on the other. Furthermore, Gronbach'salpha was .92 for the monitoring scale and .89 forthe advice interactions scale, again indicating ac-ceptable reliability (Nunnally, 1978). Accordingly,given encouraging evidence regarding interitem re-liability and discriminant validity (that is, the fac-tor loadings suggesting that GEOs discriminatedbetween the two constructs as expected), we alsoestimated the monitoring and advice factors usingthe Bartlett method (Harman, 1976).

Further analyses were conducted to assess theinterrater reliability of these measures. Specifi-cally, we compared GEO and outside director re-sponses by calculating kappa coefficients for themonitoring and advice items. Kappa is a correlationcoefficient that corrects for the expected level ofcorrelation between raters (chance correlation).Values exceeding .75 are typically thought to indi-cate excellent agreement beyond chance, and val-ues between .40 and .75 are considered indicativeof fair to good agreement beyond chance (Fleiss,1981; Landis & Koch, 1977). The sample for thisanalysis included companies with a respondingGEO and at least one responding director [n = 188).Kappa coefficients exceed .75 for all survey itemsbut one, which achieved .73, and the overall kappais .82. Given these high levels of interrater reliabil-ity, it is perhaps not surprising that the hypothe-sized effects presented below were substantivelyunchanged when monitoring and advice interac-

tions were measured with director responses ratherthan GEO responses, or vice-versa.

Independent Measures

Related board appointments in stable environ-ments. Hypotheses l a - l c predict the effects of in-dividual directors' board appointments to otherfirms that are strategically related to a focal firm. Inour measures of relatedness, inside and outsidedirectors were combined, as separate analyses re-vealed no significant differences in results whenwe distinguished between the two.

We measured relatedness across four differentstrategic dimensions that have been studied exten-sively in the strategy literature (e.g., Kim, 1989;Porter, 1986, 1998; Rumeh, 1974): First, board ap-pointments related by product market were mea-sured as the number of a director's appointments tothe boards of companies in a primary business sim-ilar (having the same four-digit Standard IndustrialGlassification [SIG] code) to that of a focal firm,divided by the director's total number of appoint-ments. Second, board appointments related by for-eign market were measured by counting appoint-ments to the boards of companies with the sameprimary foreign market as a focal firm, normalizedby the number of appointments. For example, twofirms that reported France as their primary foreignmarket would exhibit a high degree of foreign mar-ket relatedness. In separate analyses, we weightedappointments according to the difference betweena focal firm's presence iri each of its markets (as aportion of total sales) and the appointed-to firm'spresence. The results presented below were sub-stantively unchanged, demonstrating that theywere robust to different measures of foreign marketrelatedness.

For the third dimension of strategic relatedness,we assessed the diversification of a focal firm andof the other firms to which a director was con-nected using the entropy measure (Palepu, 1985).This measure takes into account the number ofsegments in which a firm operates and weightseach segment according to its contribution to totalsales. It is defined as follows: SP,- X ln(l/P,), whereP is the sales (dollar value) attributed to segment iand ln(l/P/) is the weight for each segment i, or thelogarithm of the inverse of its sales. We calculatedthe absolute difference between the diversificationof a focal firm and the diversification of each of theother firms to which a director was connected. Thedifference scores were then added and normalizedby the total number of appointments. This figurewas then subtracted from the highest value of di-versification dissimilarity in the sample, so that


higher values indicate greater relatedness. The rea-son for calculating relatedness as an average can beillustrated with the following example: say onefirm's board (A) has 18 ties with firms that arerelatively dissimilar to the focal firm (the related-ness is .1 for each tie), and another board (B) hasthree ties with firms that are very similar to thefocal firm (the relatedness is .6 for each tie). Failingto normalize these measures would give each boarda comparable relatedness score, when in fact B'sdirectors have more related experience.

Appointments related by degree of international-ization, our fourth dimension, were measured witha variation of the degree of internationalization[DOI] composite measure validated by Sullivan(1994). This measure gauges internationalization interms of three important and theoretically distinctcharacteristics. The first characteristic, foreignsales, is calculated as a ratio of foreign sales to totalsales and reflects a firm's dependence on sales toforeign markets. The second characteristic, foreignproduction, reflects a firm's reliance on owned for-eign asset stocks and is measured by foreign assetsas a percentage of total assets. As in research oninternational business, here the sales and assetcharacteristics address a firm's dependence on for-eign consumer markets and dependence on foreignproduction resources, respectively. The third char-acteristic, geographic dispersion, gauges a firm'snumber of country subsidiaries as a percentage ofthe highest number of country subsidiaries repre-sented in our sample. Sullivan (1994) found thatthis characteristic provided a rough indication ofthe cultural variety associated with the previoustwo dimensions (Johansen & Vahlne, 1977). Allthree DOI variables range theoretically from 0 to 1.

Foreign sales, foreign production, and geo-graphic dispersion are summed to form our com-posite measure of DOI, which therefore has a theo-retical range of 0 to 3. Like Sullivan (1994), wefound that these variables demonstrated high inter-item reliability [a = .86) and loaded on one factorwith a high eigenvalue and high explained varianceand that the composite measure was normally dis-tributed. There are several single-indicator mea-sures of DOI, including number of foreign subsid-iaries and ratios such as foreign sales to total sales,foreign assets to total assets, and number of foreignemployees to total employees, but Sullivan showedthat these other measures are highly correlated. Togauge relatedness on this dimension of strategy, wecalculated the absolute difference between the DOIof a focal firm and the DOI of each of the other firmsto which a director was connected. Again, the dif-ference scores were added and normalized by thenumber of appointments. We then subtracted this

figure from the highest value of DOI dissimilarity inthe sample to create an index of relatedness.

Hypotheses lb and lc predict the effects ofboard-level appointments to other firms that arestrategically related to a focal firm on board behav-ior. To test these hypotheses, we developed a set ofaggregate variables for the relatedness of individualdirector appointments across all directors on aboard. For instance, we calculated product marketsimilarity by counting the number of companies towhich the focal firm was connected by a board tiein a primary business (as reflected in a four-digitSIG code) similar to that of the focal firm, dividedby the total number of appointments. For analysesconducted at the level of the individual director,predicting the director's perceived ability to con-tribute (Hypotheses la and 2a), the variables werecalculated for the particular director.

Heterogeneous board appointments in unstableenvironments. To test Hypotheses 2a through 2c,we developed separate measures of the heterogene-ity of board appointments. First, we assessed theextent to which an individual director's board ap-pointments to other firms differed from the ap-pointments of other directors in their strategic re-latedness to a focal firm by calculating the absolutevalue of the difference between the director's relat-edness score on a given strategic dimension and theaverage relatedness scores of other directors on theboard (the measure of relatedness is defined above).High values indicated that a director's appoint-ments differed from the appointments of otherboard members on a given dimension of related-ness. For example, if other directors had many ap-pointments to firms with very similar internationalpresences, a director with more appointments tofirms with international strategy that differed fromthe focal firm's would have a relatively high heter-ogeneity score on that dimension.

Finally, we developed a set of variables measur-ing strategic heterogeneity at the board level toassess the effect of board appointments on involve-ment in turbulent environments. We measured theextent to which a board's portfolio of director ap-pointments is heterogeneous in its relatedness on agiven dimension as (Slr̂ - r\)/n, where r, is therelatedness score for director i, f is the averagerelatedness for the board, and n is the number ofboard members (note that this board-level measureexcludes duplicate ties to the same firm). Highervalues indicated greater heterogeneity in strategicrelatedness; conversely, smaller values indicatedthat a firm's appointments were concentratedamong firms that were either relatively similar orrelatively dissimilar to the focal firm on a givenstrategic dimension.


Other Measures

Environmental instability, the change rate of en-vironmental factors relevant to strategic decisionmaking (Duncan, 1972), is often a function of anindustry's competitive dynamics (Ghen, 1996;Wiersema & Bantel, 1993). Following Wiersemaand Bantel (1993), we measured environmental in-stability as changes in the industry concentrationratio. This ratio is calculated as the percentage of anindustry's sales, at the four-digit SIG level, ac-counted for by the four largest firms. Large absolutechanges in the concentration ratio indicate highenvironmental instability. Such change reflects"shifts in market share due to new entrants, exits,consolidations, and erosion of market share, thusthey capture the dynamic nature of a firm's indus-trial environment" (Wiersema & Bantel, 1993: 493).We observed change over the three-year periodprior to the survey date (1992-94). Moreover, weused (monthly) stock price volatility as an alterna-tive measure of instability (calculated as the aver-age annual price coefficient of variation across allfirms in the industry in a particular year [Hauns-child & Miner, 1997]). The results presented belowwere substantively unchanged, demonstrating thatthey were robust to alternate measures of environ-mental instability.

Several control variables were also included inthe models. Following Wiersema and Bantel's(1992) classification scheme for level of education,we controlled for directors' education level usingnumber of years of schooling (cf. Kosnik, 1987). Adirector's years of higher-level education may indi-cate knowledge pertinent to strategic decision mak-ing, which may affect ability to contribute in po-tentially competing ways to appointment to otherboards. Similarly, more extensive general manage-ment experience could also be associated with di-rectors' ability to monitor and advise management.Therefore, we also controlled for management ex-perience, measured as the number of years duringwhich a director had previously worked on a topmanagement team. In addition, we also controlledfor the number of functional areas in which direc-tors had prior experience, because their ability tocontribute may be enhanced if they have a broaderbase of prior experience (see Finkelstein, 1992).Directors may acquire firm-specific expertise overtime; therefore, their tenure may influence the abil-ity of directors to monitor and advise management.Thus, we controlled for the number of years direc-tors had served on a board. Each of these fourvariables was averaged across directors in theboard-level models.

We also controlled for the prior level of diversi-

fication and internationalization, since the com-plexities involved in managing highly diversifiedand/or international firms may tend to requiregreater monitoring and/or advice capabilities(Sanders & Garpenter, 1998; Zajac & Westphal,1994). For similar reasons, we controlled for firmsize, measured as the natural logarithm of firmsales. In addition, we controlled for prior firm per-formance in models of board monitoring and ad-vice interactions. An adaptation perspective wouldsuggest that relatively poor firm performance couldprompt GEOs to appoint directors with relevantappointments to other boards to enhance theboard's capability to monitor and/or provide advice(Gyert & March, 1963). Two recent studies haveprovided evidence for a negative relationship be-tween firm performance and board involvement instrategic decision making (Johnson, Hoskisson, &Hitt, 1993; Judge & Zeithaml, 1992). Accordingly,we included two measures of firm performance:prior return on equity, an accounting-based mea-sure, and prior market-to-book value, a market-based measure.

Several studies contain the argument that GEOsco-opt boards, rendering them passive by appoint-ing their friends as directors (e.g., Finkelstein &Hambrick, 1988). Some researchers have also sug-gested that a GEO secures the passivity of outsidedirectors simply by appointing them to the board(Alderfer, 1986; Boeker, 1992; Wade et al., 1990).Moreover, recent research has shown that top man-agers' openness to board advice is largely deter-mined by the portion of a board composed of aGEO's personal friends and the portion of the boardcomposed of directors appointed after the GEO wasappointed (Westphal, 1999). Thus, we controlledfor these variables (CEO-board friendship ties andappointments after the CEO) in the analyses, mea-suring friendship ties with questions in the CEOsurvey. We also included CEO ownership as a con-trol variable in the monitoring and advice models.Agency theorists have argued that incentives canprovide an alternative or substitute for board mon-itoring, and incentives can also motivate top man-agers to seek advice from a board (Beatty & Zajac,1994; Rediker & Seth, 1995). Moreover, incentivescould motivate other directors to monitor strategicdecision making more actively and to offer theirexpertise on strategic issues (Bergh, 1995). Thus,we controlled for director ownership in analyses ofboard monitoring and advice interactions. In addi-tion, we controlled for board size in these models,given that large boards may be less cohesive andthus less able to monitor decision making effec-tively (Johnson et al., 1993). We also controlled forboard size in models of directors' perceived ability


to contribute to decision making because each di-rector might see himself or herself as having lessopportunity to contribute on a large board.

Prior studies have also examined whether a board'sleadership structure is related to indicators of theboard's ability to control management (e.g., Baliga etal., 1996; Finkelstein, 1992; Mallette & Fowler, 1992).Thus, we controlled for leadership structure in mod-els of board monitoring, using a dichotomous variablecoded 1 if the CEO and board chair positions wereseparate and 0 otherwise. Although we did not expectboard centrality to independently affect board in-volvement, in separate analyses we included differ-ent measures of centrality, such as in-degree central-ity and the Bonacich measure (Wasserman & Faust,1994); these measures yielded consistently nonsignif-icant results, and the hypothesized effects on boardmonitoring and advice interactions were unchanged.We also did not expect the hypothesized effects onadvice interactions to be affected by the level of mon-itoring, or vice-versa, because the dependent vari-ables were lagged and negatively correlated; separateanalyses confirmed that the results were unchangedwhen the monitoring was included in the advicemodels and advice was included in the monitoringmodels.

RESULTS

Table 1 provides the means, standard deviations,and bivariate correlations for all data used to

analyze predictions of directors' perceived abilityto contribute to board discussions. Table 2 providesthese descriptive statistics for all data used to ana-lyze predictions of board monitoring and adviceinteractions. Given that all our dependent variableswere continuous and that the independent vari-ables were continuous or categorical, multiple or-dinary least squares (OLS) regression analysis wasthe primary statistical technique employed. Theentire sample was dichotomized at the median ofthe environmental stability measure, with thosefirms falling below the median accordingly classi-fied as in relatively stable industries and those fall-ing above it classified as in relatively unstable in-dustries. Thus, we ran separate sets of regressionsfor each subsample (stable and unstable) and an-other set using the product-term approach to testinteraction effects. Analyses of directors' ability tocontribute to board discussions were conducted forthe sample of responding directors, and analyses ofboard monitoring and advice interactions wereconducted for all companies in the survey sample.Tables 3 and 4 present regression results.

Results of the multiple regression analysis testingHypothesis la support the prediction that relatedappointments, will be positively associated withdirectors' perceptions of their ability to contributeto board discussions in stable environments (Table3). For example, the coefficients for product-mar-ket, diversification, and internationalization relat-edness were all positive and significant. And,

TABLE 1Descriptive Statistics and Pearson Correlation Coefficients for Analyses of Directors' Perceived Ability to

Contribute to Board Discussions'*

Variable

Board ties weighted by:la. ftoduct-market relatednesslb. Foreign market relatednesslc. Diversification relatednessId. Internationalization relatedness

Mean

0.390.510.520.46

Board ties weighted by complementarity in:2a. Product-market relatedness2b. Foreign market relatedness2c. Diversification relatedness2d. Internationalization relatedness3. Education level4. Tenure on the board5. Functional areas6. Management experience7. Focal firm diversification8. Focal firm internationalization9. Logarithm of sales

10. Board size11. Perceived ability to contribute

0.320.300.280.35

15.969.681.29

21.320.760.687.62

12.410.00

s.d.

0.290.280.250.31

0.280.240.210.261.307.720.91

12.990.520.481.503.540.89

l a

.16

.19

.23

-.09.11.03.06.07.13.03.01.18.12.04.06.16

lb

.10

.16

.05

.22

.01-.12

.05

.07

.16

.16

.12

.15-.03

.09

.09

lc

.09

.04

.02-.10

.06

.05

.07

.12

.09

.20

.15-.15

.02

.19

Id

-.10.07.04.03.12.05.03.05.08.19

-.08.03.16

2a 2b

.16

.16 .14

.21 .16

.04 .04

.07 .11

.05 .08

.01 .04

.01 .08

.09 .07-.01 .03-.05 .06

.15 .10

2c

.07-.07

.03-.04

.16

.20

.02

.01-.04

.17

2d

-.03-.04

.02

.06

.07

.18

.02

.08

.19

3

.08

.07

.02-.01

.17

.01-.01

.07

4

-.02.15.07.04.09.04.26

5

.19

.09

.05-.01

.03

.16

6

.07

.06

.08-.02

.04

7

.27

.25

.13-.16

8 9 10 11

.16

.10 .11-.12 .05 -.13

' 71 = 492 directors; for correlations greater than .13, p < .05.

u

•a<

o

-.10

CD txCM O

CDCM

.04

,

CDO

.16

m

.03

,

CMO

.10

.08

.02

,

o

.02

.13

-.10

1

CO

o

-.02

-,

1

o

CMCO

CM

q

-.04

-.22

-

1

CO

-.10

-.

1

CM

-.05

.01

--.

06-.1

7 -

-.07

-.

1

orH

C O r t t S . C O C M T I O ^O r H O O r H r H O C M r H

o

cd

o

envCA

^ ^

Ia;

o

CM

ISCM

o

"^ in

o o1 1tx C3i

o o

CM

o1CMO

CMO

o

CM

COO

ino1

o

CT)rH

in

o

CMO

1rHO

COo

CDO

O

rHO

ino1

o

CO

o

CO

o

rH

CM

CMO

o

CMo

rH

'

o1rHO1

o

o1

rH

CDO

O

(NT-*

CO

o

COo

oo

o1CO

o

o

CO

o

COrH

1

cno

o

CO

o

o1rHrH

COO

ino

o1

o

CMo1CM

o

o

c->rH

O

cno

CO

o

rH

o1

o

inCM

rH

CO

o

o

ino

o

inrH

rH

COO

ino1

• o

o>o

o1orH

O

CJ!Oi

o

o

CMo1

o1CO

o

in

o

CMO1mO

o

lOCM

'

o

o

o1CM

o

oCM

CDCM1

cn

1CO

o

in

rH

CM

COCM

CMCM

CO

smrH

mrH

nCM

CM

CO

oC O C O C M O a 5 C M ( N C M i n C D C O C OO O r H O O r H r H O O O O C M

CM O5 ID tx CMO O O rH rH

^ C O •!)< rHO O O O

rH O C7>O rH O

I Itx CD CO CO

o o o ocj) ^ CO m CO COo o o o o o ^ in CD o

O O O CM

^ IN IX CO

o o o o O O O O O O O O O O C MrH ^ rH CO (35rH O O CM CM

CO COO C\I

rH OO <H

CO O) tv rH CO CM CMO O rH O O O rH

^ (N CM CM CD CMO O rH rH rH rH

CM CO CO CO inO O O rH rH

Vn.co"

o o o o

o o o o

op g

H-a ^

inCM

d rH CO O O OC O C O r H O i n O J r H C O C O C O C M I vO 5 C O C O t N q c O C O q r H C D I X C D

riiiiddbdd

O COCO CO

O OO O

.•S o o

aXIa

: rel

ai

o

mar

reig

n

(UaXI

rela

t

ao

cati

vers

ifi

s

atio

n

N

iona

tern

ati cn

m

ated

n(

£CO ^ (J XJ i- cd ^ u T S

O CN] CM CM C^CQ


TABLE 3Results of Multiple Regression Analysis for Stable Environments"

Independent Variable

1. Board ties weighted by:Product-market relatednessForeign market relatednessDiversification relatednessInternationalization

relatedness

2. Board ties weighted bycomplementarity/heterogeneity in:

Product-market relatednessForeign market relatednessDiversification relatednessInternationalization

relatedness

3. Education level4. Tenure on the board5. Functional areas6. Management experience7. Focal firm diversification8. Focal firm international-

ization9. Logarithm of sales

10. Board size11. CEO ownership12. CEO-board friendship ties13. Appointments after the CEO14. Director ownership15. Prior return on equity16. Prior market-to-book value17. Board leadership structure

Constant

F

Hypothesis la:Director Contribution

Model 1

0.07 (0.031*0.02 (O.Oll*0.05 (0.04)0.00 (0.00)

-0.11 (0.07)-0.13 (0.08)

-0.03 (0.02)-0.02 (0.01)

1.33 (0.53)*

5.17***.16

Model 2

0.44 (0.17)**0.07 (0.18)0.66 (0.21)**0.33 (0.15)*

0.18 (0.14)0.26 (0.17)0.12 (0.36)0.13 (0.15)

0.07 (0.03)*0.02 (0.05)*0.05 (0.05)0.00 (0.00)

-0.13 (0.07)-0.12 (0.08)

-0.03 (0.03)-0.02 (0.01)

1.38 (0.53)**

10.70***.47.31***

Hypothesis lb:Board Monitoring

Model 1

0.10 (0.09)0.02 (0.01)*0.20 (0.09)*0.00 (0.01)

-0.31 (0.13)*-0.37 (0.14)**

-0.01 (0.05)-0.03 (0.03)-1.33 (0.77)-0.09 (0.17)-0.31 (0.27)

2.19 (0.85)*-1.03 (0.60)

0.07 (0.07)0.46 (0.16)**

-1.26(1.44)

4.59***.21

Model 2

0.57 (0.28)*0.07 (0.33)1.98 (0.59)***0.78(0.25)***

0.36 (0.28)0.49 (0.32)0.15 (0.32)

-0.00 (0.27)

0.10 (0.09)0.03 (0.01)*0.20 (0.09)*0.00 (0.01)

-0.31 (0.13)*-0.35 (0.13)**

-0.01 (0.05)-0.02 (0.03)-1.32 (0.77)-0.09 (0.17)-0.29 (0.27)

2.13 (0.84)*-1.03 (0.60)-0.07 (0.07)

0.45(0.16)**

-1.56(1.51)

6.22***.39.18***

Hypothesis lc:Board Advice Interactions

Model 1

0.12 (0.09)0.03 (0.01)**0.11 (0.09)0.00 (0.01)

-0.28 (0.13)-0.27 (0.14)

-0.01 (0.05)-0.04 (0.02)

0.84 (0.76)0.38 (0.16)0.73 (0.25)**0.81 (0.88)

-1.04 (0.58)-0.01 (0.06)

0.07 (1.46)

5.34***.22

Model 2

0.71 (0.29)**0.44 (0.34)2.27(0.61)***0.87(0.25)***

-0.04 (0.30)0.07 (0.34)0.55 (0.34)

-0.30 (0.28)

0.11 (0.09)0.03 (0.01)*0.11 (0.10)0.00 (0.01)

-0.28 (0.13)*-0.27 (0.14)

-0.02 (0.05)-0.05 (0.03)

0.86 (0.77)0.38 (0.16)*0.73 (0.25)**0.77 (0.89)

-1.04 (0.58)-0.01 (0.06)

0.25 (1.59)

7.36***.42.20***

" Unstandardized coefficients are reported with standard errors in parentheses; n = 250 for directors' perceived ability to contribute, andn = 114 for board monitoring and advice interactions.

* p < .05** p < .01

*** p < .001r-tests were one-tailed for hypothesized effects, two-tailed for control variables.

although the coefficient for foreign market related-ness was not significant, it was positive, as hypoth-esized. Conversely, the results indicate that havingmore board appointments to firms with differentstrategies tends to reduce directors' perceived abil-ity to contribute to board discussions. These resultsheld after we controlled for complementarity instrategic relatedness, which is consistently unre-lated to directors' perceived ability to contribute instable environments. Several control variables sig-

nificantly predicted the ability to contribute, in-cluding a director's education level and tenure onthe board and the number of functional areas inwhich the director had prior experience; board sizewas negatively related, as expected.

A similar pattern of results provided robust sup-port for Hypotheses lb and lc (Table 3). Hypothesislb predicts that, in stable environments, board ap-pointments to other strategically related firms will bepositively associated with the level of board monitor-


TABLE 4Results of Multiple Regression Analysis for Unstable Environments"

Independent Variable

1. Board ties weighted by:Product-market relatednessForeign market relatednessDiversification relatednessInternationalization

relatedness

2. Board ties weighted bycomplementarity/heterogeneity in:

Product-market relatednessForeign market relatednessDiversification relatednessInternationalization

relatedness

3. Education level4. Tenure on the board5. Functional areas6. Management experience7. Focal firm diversification8. Focal firm international-

ization9. Logarithm of sales

10. Board size11. CEO ownership12. CEO-board friendship ties13. Appointments after the CEO14. Director ownership15. Prior return on equity17. Prior market-to-book value18. Board leadership structure

Constant

FR^Afl̂

Hypothesis la: DirectorContribution

Model 1

0.01 (0.03)0.02 (0.01)**0.06 (0.04)0.00 (0.00)

-0.06 (0.08)-0.14 (0.08)

-0 .01 (0.03)-0.02 (0.01)

-1.20 (0.57)*

5.08***.15

Model 2

0.14 (0.18)-0.10 (0.21)

0.25 (0.22)0.23 (0.15)

0.39(0.15)**0.44 (0.18)**1.30(0.38)***0.59(0.16)***

0.01 (0.03)0.02 (0.01)**0.06 (0.04)0.00 (0.00)

-0.16 (0.08)*-0.15 (0.08)

-0 .01 (0.03)-0.02 (0.01)

-1.05 (0.56)

13.36***.5944***

Hypothesis lb:Board

Model 1

0.09 (0.08)0.02 (0.01)0.20 (0.09)*0.01 (0.01)

-0.92 (0.41)*-0.29 (0.14)*

-0.02 (0.04)-0 .01 (0.03)-1.37(0.73)-0.05 (0.19)-0.46 (0.30)

1.89 (0.83)*-1.14 (0.61)-0 .07 (0.07)

0.45 (0.18)*

-1.59(1.44)

4.05***.19

1 Monitoring

Model 2

0.11 (0.27)-0 .08 (0.32)

1.38 (0.58)*0.13 (0.24)

0.96 (0.31)**0.64 (0.33)*1.11 (0.32)***0.59 (0.26)*

0.11 (0.09)0.02 (0.01)0.20 (0.09)*0.01 (0.01)

-0 .91 (0.41)*-0 .28 (0.14)*

-0.02 (0.04)-0 .01 (0.03)-1.36(0.73)-0 .05 (0.19)-0.46 (0.30)

1.88 (0.82)*-1.13(0.61)-0.07 (0.07)

0.45 (0.18)*

-2.75 (1.47)

5.59***.36.17***

Hypothesis lc: BoardAdvict

Model 1

0.04 (0.09)0.02 (0.01)0.17 (0.09)0.01 (0.01)

-0 .73 (0.39)-0.27(0.13)*

-0 .01 (0.04)-0.04 (0.02)

1.21 (0.70)0.41 (0.19)*0.74 (0.30)*0.83 (0.80)

-1 .07 (0.60)-0.02 (0.07)

0.13 (1.36)

4.61***.21

! Interactions

Model 2

0.33 (0.26)0.34 (0.31)0.33 (0.56)0.29 (0.23)

0.67 (0.30)*0.66 (0.32)*1.05 (0.32)***0.94 (0.26)***

0.04 (0.09)0.03 (0.01)*0.17 (0.09)0.01 (0.01)

-0 .71 (0.40)-0 .27 (0.13)*

-0 .01 (0.04)-0.04 (0.03)

1.18 (0.71)0.41 (0.19)*0.73 (0.30)*0.81 (0.80)

-1.06 (0.60)-0.02 (0.07)

0.13 (1.47)

6.49***.41.20***

° Unstandardized coefficients are reported with standard errors in parentheses; n = 242 for directors' perceived ability to contribute, andn = 114 for board monitoring and advice interactions.

* p < .05** p < .01

*** p < .001T-tests were one-tailed for hypothesized effects, two-tailed for control variables.

ing of strategic decision making. Hypothesis lc pre-dicts that such appointments will be positively asso-ciated with the level of advice interactions onstrategic issues. Three different kinds of strategic re-latedness (product-market, diversification, and inter-nationalization relatedness) were positively associ-ated with both monitoring and advice interactions.Conversely, having more board appointments tofirms with unrelated strategies was negatively associ-ated with both kinds of involvement. A separate anal-ysis provided evidence that directors' perceived abil-

ity to contribute effectively mediated theserelationships: when ability to contribute is added tomodels of monitoring and advice interactions, theeffects of the relatedness variables become nonsignif-icant, and the coefficient for ability to contribute isstrongly and positively significant in both models(Baron & Kermy, 1986). Again, as shown in the table,we found that the hypothesized efifects were signifi-cant even after controlling for heterogeneity in strate-gic relatedness, which is consistently imrelated toeither form of board involvement.


Hypothesis 2a is the first test of our theoreticalframework in unstable competitive environments.Specifically, in such environments we predictedthat the more a director's board appointments com-plemented the appointments of other directors intheir strategic relatedness to the focal firm, thegreater would be the director's perceived ability tocontribute to board discussions. Table 4 shows theresults of regression analyses predicting directors'perceived ability to contribute in unstable environ-ments. Complementarity in strategic relatedness ispositively related to directors' perceived ability tocontribute for all four dimensions of corporatestrategy, supporting Hypothesis 2a. At the sametime, the results show that simple relatedness doesnot predict the ability to contribute.

The results for tests of Hypotheses 2b and 2c arealso presented in Table 4. Consistent with our predic-tions for unstable environments, the greater the het-erogeneity of directors' board appointments in termsof product-market, diversification, and international-ization relatedness, the higher the level of board mon-itoring and advice interactions. Heterogeneity inforeign market relatedness was also marginally asso-ciated with advice interactions in such environments.These results held after the simple relatedness ofboard ties was controlled for; simple relatedness wasgenerally unrelated to either kind of involvement.Moreover, separate analyses again suggested that di-rectors' perceived ability to contribute to board dis-cussions mediated the effects of complementary ap-pointments on director monitoring and adviceinteractions. Thus, the results consistently show thatthe strategic relatedness of board ties increases direc-tor involvement in stable environments but does notdo so in imstable environments. The opposite patternemerges in unstable environments. We also con-ducted separate analyses using the product term ap-proach to test interactions between environmentalstability and the independent variables, and the in-teractions were significant, consistent with the split-sample findings.

To supplement the primary analyses of overallboard involvement presented above, we conductedfurther analyses that focused on board involvementin either strategy implementation or the consider-ation of new strategies using additional survey items(such as "To what extent does the board monitor theimplementation of strategic decisions?"). In general,the relatedness of board ties was positively associatedwith reported involvement in implementation, and itwas negatively associated with advice on new strate-gic alternatives in turbulent environments (this effectwas not significant in stable environments). Theseresults appear to further support our theoretical argu-ments, as we would expect directors who have expe-

rience with related strategies to be more capable ofcontributing insight on implementation of a firm'scurrent strategy. In contrast, directors who have ex-perience with different strategies should be more ca-pable of contributing insight about alternatives. Also,the nonsignificant effect of relatedness on adviceabout new strategic alternatives in stable environ-ments is consistent with the view that strategicchange is relatively less important in stable environ-ments than in turbulent environments. The implica-tions of our combined results are discussed below.

DISCUSSION

Overview of Findings

Overall, the findings provided strong support forour theoretical framework. Empirical analysesyielded a consistent pattern of results suggestingthat the monitoring and advising behavior of direc-tors depends on the strategic perspective and baseof expertise provided by their appointments toother boards. For example, strategically relatedboard ties were found to enhance board involve-ment in firms facing relatively stable environments,and strategically heterogeneous board ties werefound to enhance involvement in firms facing rel-atively unstable environments. Therefore, it wouldappear that the strategic context of director ap-pointments, not simply the presence or number ofsuch appointments, is an important influence oncorporate governance.

A growing body of research on boards recognizescorporate director appointments as an indicator ofaccess to information network fiows (Mizruchi, 1996;Palmer et al., 1993), but much of it has emphasized afirm's number of interlocks (centrality) as indicativeof the extent to which it is integrated into a commu-nity of information. Critics of interlock research haveemphasized the need to examine the "substantivecontext" of interlock appointments if their conse-quences are to be understood (Nohria, 1992: 14; Pet-tigrew, 1992; Stinchcombe, 1990). Our findings ad-dress this issue by showing how board interlock tiesaffect a firm's corporate governance to the degree thatsuch ties are aligned with the strategic needs of thefirm. Thus, although the literature on interlockingdirectorates has tended to emphasize a firm's socialcontext as an important determinant of organizationadbehavior, the findings of this study suggest that theeffects of a firm's social structural context are moder-ated by its strategic context.

Accordingly, this study extends network theoriesof corporate governance. The findings are consis-tent with a sociocognitive perspective on boardinvolvement in which director ties to other boards


provide direct strategic experience and indirect ac-cess to strategic information through social contactwith other directors, which in turn can criticallyinform the knowledge structures used to monitordecisions or give advice on a focal board. Thoseknowledge structures can enhance board capabilityto contribute to strategic decision making if theyaddress the primary strategic issues facing firms,such as implementation of existing strategies instable environments versus the development ofnew strategies and implementation in turbulentenvironments. Conversely, knowledge structuresconditioned by board ties to other firms can reducea board's capability to contribute to strategy if thesestructures do not match the strategic contingenciesfacing the firm. This view is consistent withCranovetter's (1992) and Weick's (1995) social net-work embeddedness perspectives on organiza-tional behavior, which suggest that individual ac-tion is conditioned by information spread throughsocial structural relations, rather than strictly by anindividual's personal decision-making capabilitiesor by widespread social norms. In extending thisperspective to explain board involvement in strate-gic decision making, our study pinpoints the spe-cific kinds of network ties that can influencebehavior.

Moreover, research on board power and controlhas typically focused on structural board indepen-dence as a critical determinant of a board's abilityto protect shareholder interests. Such research hasnot addressed whether and when directors havethe appropriate strategic perspective aud knowl-edge to exercise control effectively. In the behav-ioral literature on boards, low levels of directorinvolvement have typically been attributed to thesocial and political influence of top managers overoutside directors (cf. Finkelstein & Hambrick,1996); the findings of this study suggest that direc-tors may also abstain from monitoring and advisingmanagement to the extent that they lack relevantstrategic information and knowledge needed tocontribute to the decision-making process. As moreboards acquire the structural power needed to in-fluence organizations (Useem, 1993; Westphal &Zajac, 1997), it becomes increasingly important todevelop models of board effectiveness that predictwhether boards also have the ability to exercisetheir influence.

Although agency-based views on boards of direc-tors attribute lower levels of director involvementto the significant costs associated with monitoringactivities, such as the time required to prepare formeetings (Beatty & Zajac, 1994), the perspectivedeveloped in this study suggests how and when theinformation and expertise directors gain from serv-

ing on other boards can permit them to economizeon governance costs, by raising the "returns" frominvesting a given amount of time and attention inthe monitoring and advising of management at aparticular firm. Our findings are consistent withthe view that if directors' appointments to otherboards provide them with relevant strategic infor-mation and expertise and focus their attention onrelevant strategic issues (such as strategic changeversus implementation), then directors are likely torealize positive sociocognitive "externalities" fromtheir appointments to other boards. That is, invest-ments in monitoring and advising at one firm pro-vide directors with knowledge and perspective thatraise the quality of their contributions at otherfirms with similar strategic priorities. Such exter-nalities permit economies of monitoring and advis-ing across firms. Conversely, directors' appoint-ments may provide them with inappropriatestrategic information and focus their attentionaway from critical strategic issues facing a focalfirm. Under those conditions, directors may insteadexperience negative sociocognitive externalitiesfrom their other board memberships, thus leadingto diseconomies of monitoring and advising.

Finally, some prior research on the diffusion ofinnovations has distinguished between a director'sexperience as an inside board member—at her orhis home company—and experience as an outsidedirector, at another company; these statuses aresometimes referred to as "directional" and "nondi-rectional," respectively (Palmer et al., 1993). Forinstance. Palmer and his colleagues suggested thatinside director ties may have stronger effects on thediffusion of innovations than outside director ties,because insiders are likely to devote more timethan outsiders to implementing innovations, lead-ing to greater psychological commitment to them(Palmer, Barber, Zhou, & Soysal, 1995; Palmer etal., 1993). Similarly, directors' beliefs about corpo-rate strategy might be influenced more by theirexperience as top managers at their home compa-nies than by their experience as outside directors atother firms. In order to test this possibility empiri-cally, we conducted separate analyses in which therelatedness of manager-director ties was measuredseparately from the relatedness of outside directorties. The hypothesized effects were supported forboth sources of experience, and the effects of man-ager-director ties were not consistently strongerthan the effects of outside director ties.

Limitations, Implications, and Conclusion

Limitations. Although the findings of this studyare consistent with a sociocognitive perspective on


board involvement in which board ties to otherfirms help determine whether directors have theappropriate knowledge and perspective to contrib-ute to strategic decision making, our empirical ap-proach does not permit a direct examination of thecognitions that mediate these relationships. Thislimitation is shared by most empirical research inthe top management team literature, which is alsolargely rooted in a sociocognitive perspective(Finkelstein & Hambrick, 1996; Hambrick & Mason,1984). In addition, although we did measure thedegree to which directors advised and monitoredmanagement, we were unable to assess the qualityof the advice given or the extent to which it ulti-mately improved firm performance. Finally, thefindings of this study may be less applicable torelatively small firms, whose directors tend to havefewer board appointments at other companies.

Future researcb directions. These limitationssuggest several directions for future research. Per-haps most importantly, research is still needed thatgoes beyond our relatively direct measures of boardbehavior to gauge the strategic expertise and per-spectives of corporate board members, assessinghow cognitions mediate the effects of board ties onstrategic decision making. For instance, researcherscould examine whether directors who have strate-gically related board ties demonstrate more nu-anced understanding of the requirements for im-plementing that strategy. Content analysis mightbe used to measure the complexity of director rec-ommendations regarding information systems, re-ward systems, and other systems and structuresneeded to implement strategies effectively. Thisapproach has been used in research on group deci-sion making to measure the complexity of groupmember cognitions (e.g., Tetlock, Peterson,McGuire, Chang, & Feld, 1992).

In addition, to the extent that the strategic relat-edness or heterogeneity of board ties predicts di-rector behaviors, such behaviors may have partic-ular implications for firm performance. Forinstance, our theoretical argument would suggestthat heterogeneous board ties could enhance per-formance for firms operating in turbulent environ-ments. In partial support of this view, Uzzi (1996)found that firms in the highly turbulent women'sfashion industry were most likely to survive whenthey possessed a combination of strong and diverseties to other organizations. Thus, future researchshould study whether or not the congruence ofenvironmental stability and director appointmentssimilarly affects performance differentials amongfirms.

Research might also examine whether firms withappropriate board appointments rely less on alter-

native control mechanisms. For example, Johnsonand colleagues (1993) observed that boards in-fluence firms either through strategic controls(such as board monitoring) or through financialmechanisms (such as managerial incentives).Therefore, in future studies, researchers might ex-amine whether the addition of board appointmentsthat provide relevant strategic expertise lead firmsto reduce their reliance on financial controls, thusminimizing the costs associated with such controls(that is, the costs resulting from imposition of firm-specific risk on decision makers [Jensen & Meek-ling, 1976]). Conversely, firms may rely moreheavily on such financial controls where directors'appointments to other boards do not provide themwith the relevant strategic perspective and knowl-edge base.

Finally, it should be noted that reciprocal boardappointments are common in many countries(Demb & Neubauer, 1992). However, because re-search in non-U.S. settings has emphasized the so-cial rather than the strategic nature of reciprocalboard appointments (e.g., Stokman, Zeigler, &Scott, 1985), there is an opportunity to extend thesocial psychological perspective developed in thisstudy to research on corporate governance outsideof the United States. For instance, given that thereare markedly fewer restrictions on the types ofboard connections and the contents of communica-tion flows between managers of different firms inEuropean countries, the strategic context of boardappointments may be even more influential there.

Managerial implications. In addition to theirtheoretical importance, the results of this research,have significant implications for both public policyand business practice. For example, there is grow-ing external stakeholder pressure for the U.S. gov-ernment to legislate constraints on board memberappointments that go beyond the existing ClaytonAct (Section 8) prohibition of certain kinds of in-terlocking directorates among direct competitors(BusinessWeek, 1997). Indeed, the results pre-sented here confirm that boards are less likely to beeffective advisors and monitors when their mem-bers are appointed to the boards of other firms thatare strategically irrelevant to the needs of a focalfirm. Therefore, to the extent that firms are pre-vented from realizing benefits from the relatednessor the heterogeneity of appointments discussedin this research, additional legislated board con-straints may have unintended negative conse-quences for the corporate governance of public U.S.firms.

Similarly, there is evidence that corporate lead-ers give considerable weight to the number of otherboard memberships held by director candidates


when selecting new outside directors for theirboards (Davis, 1993). However, the results of thisstudy suggest the need to consider whether a direc-tor will connect a firm with other organizations thatcan furnish relevant strategic knowledge and per-spective, rather than simply focus on the number ofboard seats the candidate already holds. For exam-ple, if a firm in a turbulent industry has a boardcomposed primarily of directors who sit on theboards of other firms with similar strategies, such aboard's ability to monitor and advise may be con-siderably enhanced by appointing directors who siton the boards of other firms that follow differentdiversification and internationalization strategiesin different product and geographic markets.

In conclusion, the findings of this study suggestthe value of developing sociocognitive theories thataddress whether and when corporate leaders havethe appropriate strategic knowledge and perspec-tive to contribute to corporate governance. More-over, although little rigorous empirical research hasdirectly examined behavioral processes in CEO-board relationships, our findings also show the po-tential power of models that link the broader socio-structural context in which boards are embedded,as well as the environmental conditions that sur-round them, with the microbehavioral processesthat occur inside the "black box" of corporateboards.

REFERENCES

Abelson, R. P., & Black, J. B. 1986. Introduction. In R. P.Abelson & J. B. Black (Eds.), Knowledge structures:1-18. Hillsdale, NJ: Erlbaum.

Alderfer, C. P. 1986. The invisible director on corporateboards. Harvard Business Review, 64(6): 38-52.

Aldrich, H. 1979. Organizations and environments.Englewood Cliffs, NJ: Prentice-Hall.

Andrews, K. 1971. The concept of corporate strategy.Illinois: Dow-Jones-Irwin.

Baliga, B. R., Moyer, C, & Rao, R. S. 1996. CEO dualityand firm performance: What's the fuss? StrategicManagement Journal, 17: 41-53.

Barney, J., & Zajac, E. 1994. Competitive organizationalbehavior: Toward an organizationally-based theoryof competitive advantage. Strategic ManagementJournal, 15(wint8r special issue): 5-9.

Baron, R., & Kenny, D. 1986. The moderator-mediatorvariable distinction in social psychological research:Conceptual, strategic, and statistical considerations.Journal of Personality and Social Psychology, 51:1173-1182.

Beatty, R., & Zajac, E. 1994. Managerial incentives, mon-itoring, and risk-hearing: A study of executive com-

pensation, ownership, and board structure in initialpublic offerings. Administrative Science Quarterly,39: 313-335.

Bergh, D. D. 1995. Size and relatedness of units sold: Anagency theory and resource-based perspective. Stra-tegic Management Journal, 16: 221-239.

Boeker, W. 1992. Power and managerial dismissal:Scapegoating at the top. Administrative ScienceQuarterly, 17: 538-547.

Boeker, W. 1997. Strategic change: The influence of man-agerial characteristics and organizational growth.Academy of Management Journal, 40: 152-170.

Boeker, W., & Coodstein, J. 1993. Performance and suc-cessor choice: The moderating effects of governanceand ownership. Academy of Management Journal,36: 172-186.

Bourgeois, L. 1980. Performance and consensus. Strate-gic Management Journal, 1: 227-248.

Brudney, V. 1982. The independent director—Heavenlycity or potemkin village? Harvard Law Review, 95:597-659.

Burt, R. S. 1987. Social contagion and innovation: Cohe-sion versus structural equivalence. American Jour-nal of Sociology, 92: 1287-1335.

BusinessWeek. 1997. The best and worst boards. Decem-ber 8: 90-98.

Carpenter, M. A., & Fredrickson, J. W. 2001. Top man-agement teams, global strategic posture, and themoderating role of uncertainty. Academy of Man-agement Journal, 44: 533-545.

Chen, M. 1996. Competitor analysis and interfirm rivalry:Toward a theoretical integration. Academy of Man-agement Review, 21: 100-135.

Cyert, R. M., & March, J. C. 1963. A behavioral theory ofthe firm. Englewood Cliffs, NJ: Prentice-Hall.

Davis, C. E. 1991. Agents without principles? The spreadof the poison pill through the intercorporate net-work. Administrative Science Quarterly, 36: 583-613.

Davis, C. E. 1996. Who gets ahead in the market forcorporate directors: The political economy of mul-tiple board memberships. Paper presented at theannual meeting of the Academy of Management,Cincinnati.

Day, D., & Lord, R. 1992. Expertise and problem catego-rization: The role of expert processing in organiza-tional sensemaking. Journal of Management Stud-ies, 29: 35-47.

Dearborn, D., & Simon, H. 1958. Selective perception: Anote on the departmental identifications of execu-tives. Sociometry, 21: 140-144.

Demb, A., & Neubauer, E. 1992. The corporate board.New York: Oxford University Press.


DeVellis, R. E. 1991. Scale development: Theory andapplications. Newbury Park, CA: Sage.

Duncan, R. 1972. Characteristics of organizational en-vironments and perceived environmental uncer-tainty. Administrative Science Quarterly, 17:313-327.

Dutton, J., & Duncan, R. 1987. The creation of momentumfor change through the process of strategic issuediagnosis. Strategic Management Journal, 8: 279-296.

Eama, E. 1980. Agency problems and the theory of thefirm. Journal of Political Economy, 68: 288-307.

Eama, E., & Jensen, M. 1983. Sepeiration of ownershipand control. Journal of Law and Economics, 26:301-325.

Einkelstein, S. 1992. Power in top management teams:Dimensions, measurement, and validation. Acad-emy of Management Journal, 35: 505-538.

Einkelstein, S., & Hambrick, D. 1988. Chief executivecompensation: A synthesis and reconciliation. Stra-tegic Management Journal, 10: 543-558.

Einkelstein, S., & Hambrick, D. 1996. Strategic leader-ship: Top executives and their effects on organiza-tions. St. Paul: West.

Eleiss, J. L. 1981. Statistical methods for rates and pro-portions. New York: Wiley.

Eowler, E. J. 1993. Survey research methods. NewhuryPark, CA: Sage.

Eredrickson, J. W. 1984. The comprehensiveness of stra-tegic decision processes: Extension, observation, fu-ture direction. Academy of Management Journal,27: 399-424.

Eredrickson, J. W., & Mitchell, T. R. 1984. Strategicdecision processes: Comprehensiveness and per-formance in an industry with an unstable environ-ment. Academy of Management Journal, 27:399-423.

Calbraith, C, & Merrill, G. 1991. The effect of compen-sation program and structure on SBU competitivestrategy: A study of technology-intensive firms.Strategic Management Journal, 12: 353-370.

Calbraith, J., & Kanzanjian, R. 1986. Strategy implemen-tation: Structure, systems, and processes. St. Paul:West.

Cerhart, B., & Milkovich, C. 1990. Organizational differ-ences in managerial compensation and financial per-formance. Academy of Management Journal, 33:663-691.

Cinsberg, A. 1990. Connecting diversification to perfor-mance: A sociocognitive approach. Academy ofManagement Review, 15: 514-535.

Comez-Mejia, L. 1992. Structure and the process of di-versification. Strategic Management Journal, 13:381-397.

Cranovetter, M. 1992. Problems of explanation in eco-nomic sociology. In N. Nohria & R. Eccles (Eds.),Networks and organizations: Structure, form andaction: 25-56. Boston: Harvard Business SchoolPress.

Croves, R. M., Cialdini, R. B., & Couper, M. P. 1992.Understanding the decision to participate in a sur-vey. Public Opinion Quarterly, 56: 475-495.

Halehlian, J., & Einkelstein, S. 1993. Top-managementteam size, CEO dominance, and firm performance:The moderating roles of environmental turbulenceand discretion. Academy of Management Journal,36: 844-863.

Hambrick, D., & Mason, P. 1984. Upper echelons: Theorganization as a function of its top managers. Acad-emy of Management Journal, 15: 514-535.

Hambrick, D., Celetkanycz, M., & Eredrickson, J. 1993.Top executive commitment to the status quo: Sometests of its determinants. Strategic ManagementJournal, 14: 401-418.

Harman, H. H. 1976. Modern factor analysis (3rd ed.).Chicago: University of Chicago Press.

Haunschild, P. 1993. Interorganizational imitation: Theimpact of interlocks on corporate acquisition activ-ity. Administrative Science Quarterly, 38: 564-592.

Haunschild, P., & Miner, A. 1997. Modes of interorgani-zational imitation: The effects of outcome salienceand uncertainty. Administrative Science Quarterly,42: 472-500.

Hill, C, & Snell, S. 1988. External control, corporatestrategy, and firm performance in research-intensiveindustries. Strategic Management Journal, 9: 577-590.

Huher, C, & McDaniel, R. 1986. The decision makingparadigm of organizational design. ManagementScience, 32: 572-589.

Jehn, K. A., Chadwick, C, & Thatcher, S. M. B. 1997.To agree or not to agree: The effects of value con-gruence, individual demographic dissimilarity,and conflict on workgroup outcomes. Interna-tional Journal of Conflict Management, 8: 287—305.

Jensen, M. C, & Meckling, W. H. 1976. Theory of thefirm: Managerial behavior, agency costs, and owner-ship structure. Journal of Financial Economics, 3:305-350.

Johansen, J., & Vahlne, J. 1977. The internationaliza-tion process of the firm—A model of knowledgedevelopment and increasing foreign market com-mitments. Journal of International BusinessStudies, 8: 23-32.

Johnson, R., Hoskisson, R., & Hitt, M. 1993. Board ofdirector involvement in restructuring: The effects ofboard versus managerial controls and characteris-tics. Strategic Management Journal, 14(summerspecial issue): 14: 33-50.


Judge, W., & Zeithaml, C. 1992. Institutional and strategicchoice perspectives on board involvement in thestrategic decision process. Academy of Manage-ment Journal, 35: 766-794.

Kahneman, D., Slovic, P., & Tversky, A. 1982. Judgmentunder uncertainty: Heuristics and biases. Cam-bridge, England: Cambridge University Press.

Kazanjian, R., & Drazin, R. 1987. Implementing internaldiversification: Contingency factors for organizationdesign choices. Academy of Management Review,12: 342-354.

Kesner, I. E., Victor, B., & Lamont, B. T. 1986. Boardcomposition and the commission of illegal acts: Aninvestigation of Fortune 500 companies. Academyof Management Journal, 29: 789-799.

Kiesler, S., & Sproull, L. 1982. Managerial response tochanging environments: Perspectives on problemsensing from social cognition. Administrative Sci-ence Quarterly, 27: 548-570.

Kim, W. 1989. Developing a global diversification mea-sure. Management Science, 25: 376-383.

Kosnik, R. 1987. Creenmail: A study of board perfor-mance in corporate governance. Administrative Sci-ence Quarterly, 32: 163-185.

Landis, J. R., & Koch, C. C. 1977. The measurement ofobserver agreement for categorical data. Biometrics,33: 671-679.

Lorsch, J., & Maclver, E. 1989. Pawns and potentates:The reality of America's corporate boards. Boston:Harvard Business School Press.

Lurigio, A. J., & Carroll, J. S. 1985. Probation officers'schemata of offenders: Content, development, andimpact on treatment decisions. Journal of Personal-ity and Social Psychology, 48: 1112-1126.

Mace, M. 1971. Directors: Myth and reality. Boston:Harvard Business School Press.

Mallette, P., & Eowler, K. 1992. Effects of board compo-sition and stock ownership on the adoption of poi-son pills. Academy of Management Journal, 35:1010-1035.

Mizruchi, M. 1996. What do interlocks do? An analysis,critique, and assessment of research on interlockingdirectorates. In J. Hagan & K. Cook (Eds.), Annualreview of sociology, vol. 22: 271-298. Palo Alto, CA:Annual Reviews.

Nisbett, R., & Ross, L. 1980. Human inference: Strate-gies and shortcomings of social judgment. Engle-wood Cliffs, NJ: Prentice-Hall.

Nohria, N. 1992. Introduction: Is a network perspective auseful way of studying organizations? In N. Nohria &R. Eccles (Eds.), Networks and organizations:Structure, form, and action: 1-22. Boston: HarvardBusiness School Press.

Nunnally, J. 1978. Psychometric theory (2nd ed.). NewYork: McCraw-Hill.

Ocasio, W. 1997. Toward an attention-based view of thefirm. Strategic Management Journal, 18(summerspecial issue): 187-206.

O'Reilly, C, R. Snyder, & Boothe, J. 1993. Effects of execu-tive team demography on organizational change. In G.Huber & W. Glick (Eds.), Organizational change andredesign: 147-175. New York: Oxford UniversityPress.

Palepu, K. 1985. Diversification strategy, profit perfor-mance, and the entropy measure. Strategic Manage-ment Journal, 6: 239-255.

Palmer, D., Barber, B. M., Zhou, X., & Soysal, Y. 1995.The friendly and predatory acquisition of large U.S.corporations in the 1960s: The other contested ter-rain. American Sociological Review, 60: 469-499.

Palmer, D., Jennings, P., & Zhou, X. 1993. Late adoptionof the multidivisional form by large U.S. corpora-tions: Institutional, political, and economic ac-counts. Administrative Science Quarterly, 38:100-131.

Pelled, L. H. 1996. Demographic diversity, conflict, andwork group outcomes: An intervening process the-ory. Organization Science, 7: 615-631.

Pelled, L. H., Eisenhardt, K. M., & Xin, K. R. 1999. Ex-ploring the black box: An analysis of work groupdiversity, conflict, and performance. AdministrativeScience Quarterly, 44: 1-28.

Pettigrew, A. 1992. On studying managerial elites. Stra-tegic Management Journal, 13: 163-182.

Pfeffer, J., & Salancik, G. 1978. The external control oforganizations: A resource dependence perspec-tive. New York: Harper & Row.

Porter, M. E. 1986. Competition in global industries.Boston: Harvard Business School Press.

Porter, M. E. 1998. Competitive strategy (2nd ed.). NewYork: Eree Press.

Priem, R. 1990. Top management team factors, consen-sus, and firm performance. Strategic ManagementJournal, 11: 469-478.

Rediker, K., & Seth, A. 1995. Boards of directors andsubstitution effects of alternative governance mech-anisms. Strategic Management Journal, 16:85-100.

Rumelt, R. 1974. Strategy, structure, and economic per-formance. Cambridge, MA: Harvard UniversityPress.

Sanders, W. C, & Carpenter, M. A. 1998. International-ization and firm governance: The roles of CEOcompensation, top team composition, and boardstructure. Academy of Management Journal, 41:158-178.

Sharfman, M., & Dean, J. 1991. Conceptualizing and mea-suring the organizational environment: A multidi-mensional approach. Journal of Management, 17:681-700.


Siegel, S., & Castellan, N. J. 1988. Nonparametric statis-tics for the behavioral sciences (2nd ed.). NewYork: McGraw-Hill.

Stinchcomhe, A. 1990. Weak structural data. Contempo-rary Sociology, 19: 380-382.

Stokman, E., Ziegler, R., & Scott, J. (Eds.). 1985. Net-works of corporate power. New York: Polity Pressand Blackwell.

Sujan, H., Sujan, M., & Bettman, J. 1988. Knowledgestructure differences between more effective and lesseffective salespeople. Journal of Marketing Re-search, 25: 81-86.

Sullivan, D. 1994. Measuring the degree of internation-alization of a firm. Journal of International Busi-ness Studies, 25: 325-342.

Tetlock, P. E., Peterson, R. S., McGuire, C, Chang, S., &Eeld, P. 1992. Assessing political group dynamics: Atest of the groupthink model. Journal of Personalityand Social Psychology, 63: 403-425.

Tushman, M., & Anderson, P. 1986. Technological dis-continuities and organizational environments. Ad-ministrative Science Quarterly, 31: 439-465.

Tushman, M., & Romanelli, E. 1985. Organizational evo-lution: A metamorphosis model of convergence andreorientation. L. L. Cummings & B. M. Staw (Eds.),Research in organizational behavior, vol. 7: 171-222. Greenwich, GT: JAI Press.

Useem, M. 1982. Glasswide rationality in the politics ofmanagers and directors of large corporations in theUnited States and Great Britain. Administrative Sci-ence Quarterly, 27: 199-226.

Useem, M. 1993. Executive defense: Shareholderpower and corporate reorganization. Cambridge,MA: Harvard University Press.

Uzzi, B. 1996. Emheddedness and economic perfor-mance: The network effect. American SociologicalReview, 61: 4: 674-698.

Wade, J., O'Reilly, C. A., & Chandratat, I. 1990. Goldenparachutes: GEOs and the exercise of social influence.Administrative Science Quarterly, 35: 587-603.

Walker, G. 1985. Network position and cognition in acomputer software firm. Administrative ScienceQuarterly, 30: 103-130.

Walsh, J. 1995. Managerial and organizational cognition:Notes from a trip down memory lane. OrganizationScience, 6: 280-321.

Wasserman, S., & Eaust, K. 1994. Social network anal-ysis: Methods and applications. Cambridge, En-gland: Cambridge University Press.

Weick, K. 1995. Sensemaking in organizations. Thou-sand Oaks, CA: Sage.

Westphal, J. D. 1998. Board games: How CEOs adapt toincreases in structural board independence frommanagement. Administrative Science Quarterly,43: 511-537.

Westphal, J. D. 1999. Collaboration in the boardroom:The consequences of social ties in the CEO/boardrelationship. Academy of Management Journal, 42:7-24.

Westphal, J. D., & Zajac, E. 1997. Defections from the innercircle: Social exchange, reciprocity, and the diffusionof board independence in U.S. corporations. Adminis-trative Science Quarterly, 42: 161-183.

Wiersema, M., & Bantel, K. 1992. Top management teamdemography and corporate strategic change. Acad-emy of Management Journal, 35: 91-121.

Wiersema, M., & Bantel, K. 1993. Top management teamturnover as an adaptation mechanism: The role ofthe environment. Strategic Management Journal,14: 485-504.

Williams, K. Y., & O'Reilly, C. A. 1998. Demography anddiversity in organizations: A review of 40 years ofresearch. In L. L. Cummings & B. M. Staw (Eds.),Research in organizational behavior, vol. 20: 77-140. Greenwich, CT: JAI Press.

Zajac, E., & Westphal, J. 1994. The costs and benefits ofmanagerial incentives and monitoring in large U.S.corporations: When is more not better? StrategicManagement Journal, 15: 121-142.

APPENDIX

Survey Items and Factor Loadings

Perceived Ability to Contribute1. To what extent are you capable of .83

contributing to board discussions on strategicissues? ("not at all," 1; "moderately," 3; "verymuch so," 5)

2. To what extent do you have sufficient .78knowledge on relevant strategic issues tocontribute to hoard discussions? ("not at all,"1; "to some extent," 3; "very much so," 5)

3. To what extent are you able to add valuable .87insight to the hoard on strategic issues? ("notat all," 1; "to some extent," 3; "very muchso," 5)

4. I am capable of making important .76contributions to the strategic decision makingprocess, ("strongly disagree," 1; "neither agreenor disagree," 3; "strongly agree," 5)

Eor all the monitoring and advice items except forthe last one, the response scale was "minimally," 1;"moderately," 3; "very much so," 5.

Board Monitoring

1. To what extent does the board monitor top .81management strategic decision making?

2. To what extent does the board formally .80evaluate [the CEO's] performance?


APPENDIX(continued)

3. To what extent does the board defer to [the .85CEO's] judgment on final strategic decisions?

Board Advice Interactions

1. To what extent does [the CEO] solicit board .84assistance in the formulation of corporatestrategy?

2. To what extent are outside directors a .87"sounding board" on strategic issues?

3. [In the past twelve months:] How often have .81you provided advice and counsel to the CEOon strategic issues? ( times)

Mason A. Carpenter is an assistant professor in theSchool of Business and an associate of the WeinertCenter for Entrepreneurship at the University of Wis-consin—Madison. He received his Ph.D. from theUniversity of Texas at Austin. His research concernscorporate governance, top management teams, thestrategic management of global firms, and global newventures.

James D. Westphal is a professor of management in theMcCombs School of Business at the University of Texasat Austin. He received his Ph.D. from Northwestern Uni-versity. His current research interests include boards ofdirectors and executive compensation, institutional pro-cesses, and interorganizational networks.

THE STRATEGIC CONTEXT OF EXTERNAL NETWORK TIES:...

Documents

Transcript of THE STRATEGIC CONTEXT OF EXTERNAL NETWORK TIES:...