voluntary turnover, social capital and organizational performance and shaw 2001... ·...

® Academy of Management Review2001, Vol. 26, No. 3, 446-456.

NOTE

VOLUNTARY TURNOVER, SOCIAL CAPITALAND ORGANIZATIONAL PERFORMANCE

GREGORY G. DESSJASON D. SHAW

University of Kentucky

We propose a supplemental perspective, based on organizational social capital, forexamining the voluntary turnover-organizational performance relationship. We viewexisting organizdtional-level theories as those focusing on cost or human capitalissues or, rarely, on a balance among these factors. But rapid changes in the natureof work, organizational structures, and interorganizational competitiveness increasethe importance of studying the role of social capital in the voluntary turnover-organ-izational performance relationship. We highlight areas of correspondence and diver-gence among the various perspectives, discuss implications for various performancemeasures, and outline several research directions.

In a recent Wall Street Journal article, Wysocki(2000) pinpoints a new type of Pied Piper effect—teams or networks of people leaving one com-pany for another—that is plaguing many organ-izations. The trend is to have employees at thecrux of social networks be recruited away fromorganizations, especially if they are seen by therecruiting organization as having the potentialto bring with them a raft of valuable col-leagues—a process referred to as "hiring viapersonal networks" (Wysocki, 2000: Al). Thereare several corporate victims (e.g., IBM, WaltDisney) of the departures en masse, while othercompanies (e.g.. Third Millennium, Healtheon/WebMD) have (at least temporarily) reaped thebenefits. This practice tops the list of unappre-ciated practices in the new economy. The trendmay be increasing, but Krackhardt and Porter(1986) found evidence of turnover clusters (i.e., asnowball effect) in a study of fast food restau-rants more than a decade ago. Interestingly, theseanecdotes and sparse empirical evidence reflect atrend in the organizational literature in which in-creasing importance is attached to social capitalor assets embedded in relationships that contrib-ute to the creation of valued outcomes (e.g.,Coleman, 1988; Leana & Van Buren, 1999). The loss

We thank Nina Gupta, Michelle Duffy, Scott Droege, DanBrass, Bruce Skaggs, and three anonymous reviewers forhelpful comments on earlier versions of this manuscript.

of key network members, especially when accom-panied by the loss of other key network members,can severely damage an organization's social fab-ric and perhaps eradicate its social capital alto-gether (Leana & Van Buren, 1999).

The consequences of these voluntary turnoverpatterns and the corresponding loss of socialcapital in terms of organizational performanceare unknown. Moreover, current organizational-level theory concerning voluntary turnover can-not adequately address these issues. In this pa-per we review existing theory and present asocial capital perspective designed to supple-ment current theory and advance the study ofvoluntary turnover and organizational perfor-mance. To accomplish this, we first briefly re-view current perspectives in which this issue isaddressed. Second, we discuss the implicationsof social capital losses in voluntary turnoversituations, highlighting areas of convergenceand divergence and suggesting critical outcomevariables. We then provide a departure point forfuture studies by promoting potentially fruitfulresearch avenues and addressing measurementissues.

Voluntary turnover—an employee's decisionto terminate the employment relationship—andinvoluntary turnover—an employer's decision toterminate the employment relationship—areboth fertile areas for research (e.g., Shaw, Del-ery, Jenkins, & Gupta, 1998). Here, we focus onvoluntary separations only, although we do not

446

2001 Dess and Shaw 447

discount the importance of involuntary turnoverand its implications for organizational perfor-mance. Our rationale is pragmatic and reflectedin the dynamics of today's knowledge economy.Over the next fifteen years, the demand for mid-career talent is projected to clearly outpacethe supply, creating a significant imbalance indemand versus supply. For example, the firmMcKinsey & Company concluded that "the mostimportant corporate resource over the next 20years will be talent: smart, sophisticated busi-ness people who are technologically literate,globally astute, and operationally agile" (quot-ed in Fishman, 1998: 104), and Capelli arguesthat "while the overall demand for labor willrise and fall, the war for talent will rage on"(2000: 105). Thus, on balance, the attraction andretention of talent in the knowledge economywill likely be more salient issues than downsiz-ing and restructuring. Clearly, many of ourideas also have important implications for in-voluntary turnover as the issue remains rele-vant in many volatile and competitive sectors.

CURRENT THEORETICAL POSITIONS

Whereas individual-level studies tend to fo-cus on the "problem" of voluntary turnover, witha focus on designing ways to prevent it (seeDalton, Krackhardt, & Porter, 1981, for an excep-tion), organizational-level theorists often recog-nize the potentially positive aspects of it. Thereare areas of overlap and distinction among thetwo key streams (cost and cost-benefit perspec-tives and human capital approaches). We viewthe key differences in terms of the conceptualpaths pursued in each perspective—that is, theydiffer in terms of why voluntary turnover is ex-pected to relate to organizational performanceand, in some cases, make differential predic-tions.

We label the first perspective the cost or cosf-benefit approach. For many years researchersand practitioners alike have attempted to iso-late the costs of voluntary turnover in terms ofseparation (e.g., exit interviews), replacement(e.g., advertising and selection costs; Darmon,1990), new-hire training (Smith & Watkins, 1978),and general administration (Dalton & Todor,1982) costs (see also Hom & Griffeth, 1995, for anextensive review). Widely varied ranges of thecosts associated with a single separation—andadditional speculation that there are hidden

costs not captured in these dimensions—haveresulted. The conceptual link to performance inthis approach is straightforward: higher inci-dents of turnover in organizations increase costsmonotonically, and financial performance isthereby lowered.

In the cost-benefit approach the costs of turn-over are consequential, but organizations canalso realize benefits from a certain level of vol-untary turnover. Payroll and related reductions(e.g., Dalton & Todor, 1982), voluntary separa-tions of poor performers (Dalton et al., 1981), im-provements in innovation, and reductions instagnation (e.g., Abelson & Baysinger, 1984;Schneider, 1987) are all noted as benefits of acertain level of turnover.

The cost-benefit approach differs from a strictcost-based formulation in that a curvilinear, asopposed to negative and monotonic, relation-ship between voluntary turnover and organiza-tional performance is predicted; also, it is pre-dicted that the relationship is positive in arestricted range between zero and an optimallevel, but negative thereafter. The conceptualfoundation for the curvilinear prediction differsdepending on the expected benefits of low lev-els of turnover. From a purely financial cost per-spective, low voluntary turnover levels are jus-tified, since the cost of retaining all employeesoutweighs the benefits of doing so, whereas in-novation and change arguments imply that verylow turnover creates stagnation and "trainedincapacity" (Dalton & Todor, 1979: 266).

The predominant theoretical approach to ex-amining organizational-level consequences ofvoluntary turnover is human capital theory,which suggests that organizational functioningis determined by the accumulation of firm-specific, valuable human capital (Strober, 1990).Human capital is seen as the primary determi-nant of productivity, and because voluntaryturnover diminishes human capital, productivityis weakened as turnover increases. Human cap-ital theorists make distinctions between firm-specific and more general skills and abilities(e.g., Becker, 1993; Pil & Leana, 2000). But from anorganization's perspective, both types of invest-ments are costly, and once an investment deci-sion (in specific or general skills) is made, theorganization has an incentive to continue theemployment relationship. That is, voluntaryturnover eliminates the organization's return oninvestment in the employee (i.e., a loss of a pro-

448 Academy of Management Review July

ductive worker), and productivity should de-crease in proportion to the lost skills. There issome evidence, however, that the loss of idio-syncratic, as opposed to general, human capitalmay be more damaging (e.g., Pennings, Lee, &van Witteloostuijn, 1998). The picture is morecomplicated, of course, when viewed from theindividual's perspective, since voluntary turn-over is considerably more costly when humancapital is solely firm specific (Becker, 1993).

In summary, a direct negative relationship be-tween voluntary turnover and productivity lev-els can be derived from human capital theory.The relationship between turnover and finan-cial performance is more complicated whenviewed through this lens. On the one hand, amediation model is suggested, where workforceproductivity mediates a negative relationshipbetween voluntary turnover and organizationalperformance. On the other hand, Osterman(1987) notes that firms may make human capitalcost tradeoffs, where some firms invest heavilyin human capital and expect high productivitywhile others pursue profitability through lowinvestments and low productivity.

In the cost and cost-benefit approaches, it ispredicted that voluntary turnover rates gener-ally erode financial performance by increasingcosts, although cost savings are sometimes ev-ident. In human capital theory, in contrast, twoalternatives are hinted at. Voluntary turnovermay weaken organizational performancethrough the mediating role of lower productivity,but firms may attempt to match the costs ofturnover with a corresponding level of invest-ment in human capital.

While several researchers have called formore theory testing in these areas, empiricalorganizational-level examinations of these per-spectives are exceedingly scarce. Regardingcost-based approaches, two empirical studiesprovide evidence that voluntary turnover in-creases costs and reduces efficiency in organi-zations. Kasarda (1973) found that turnover ratesamong public school teachers were associatedwith increases in the number of administratorsand overhead costs, and Alexander, Bloom, andNuchols (1994) found that nursing turnover di-minished hospital efficiency. Interestingly, al-most two decades after curvilinear cost-benefitpredictions began to appear (e.g., Abelson &Baysinger, 1984), no empirical support at the or-ganizational level has been found for this for-

mulation (see Shaw, Delery, & Gupta, 2000, for arare test), although Katz (1982) found a group-level curvilinear effect. With regard to humancapital theory, Osterman (1987) concluded that anegative relationship between organizational-level turnover and productivity was fairly wellestablished, although he relied mostly on indi-rect evidence (e.g., individual-level job mobilityresearch) to draw this conclusion.

In the next section we detail a supplementarysocial capital perspective on the relationshipbetween voluntary turnover and performance.

VOLUNTARY TURNOVER AND SOCIALCAPITAL CONSIDERATIONS

We follow Leana and Van Buren in examiningsocial capital at the organizational level—thatis, as "a resource reflecting the character of so-cial relations within the organization, realizedthrough members' levels of collective goal ori-entation and shared trust" (1999: 540). Social cap-ital is seen as a public good (organizationalresource), rather than a private good (individualresource). It is critical because knowledge-based resources, in contrast to property-basedresources (e.g., machinery, equipment, andland), are tacit in nature and cannot be pro-tected easily against loss or unauthorized trans-fer (Miller & Shamsie, 1996). Thus, combiningand leveraging knowledge-based resources andcreating firm-specific ties clearly add value(perhaps synergistically) to the organization.

For most of the twentieth century, the key fo-cus of managerial effort was toward a more ef-ficient allocation of the factors of production:labor and capital. The salient resources of con-cern to managers were tangible resources suchas natural resources, land, and money, as wellas intangibles such as brands, reputation, andcustomer loyalty. Today, however, more than 50percent of the gross domestic product (GNP) indeveloped economies is knowledge based, in-cluding such notable industry sectors as com-puters, software, pharmaceuticals, education,and so on (Drucker, 1997; The Economist. 1996).Intellectual and information processes now cre-ate most added value for firms in large serviceindustries. According to the Bureau of Labor Sta-tistics, the service-producing sector continues tolead projected employment growth, and the tenleading industries—accounting for 60 percent ofprojected job growth between 1996 and 2006—


are all in service industries (Franklin, 1997, citedin Meyer & DeTore, 1999). Moreover, intellectualactivities, such as research and development,process design, and product design, generatethe bulk of added value in manufacturing indus-tries as well (Quinn, Anderson, & Finkelstein,1996).

Interestingly, the discussion of the role of so-cial capital in the turnover-performance rela-tionship is virtually nonexistent. When compar-ing the existing approaches to the voluntaryturnover-performance relationship with a socialcapital approach, we see clear areas of congru-ence and divergence. For example, although theprecise form of the relationship could be de-bated, in the cost-benefit, human capital, andsocial capital approaches, very high levels ofvoluntary turnover that would hinder the organ-ization's ability to function effectively would bepredicted. The perspectives also complementone another in that they focus on factors neces-sary for organizational functioning. That is, re-gardless of whether or not an organization isknowledge-based, controlling replacementcosts, retaining highly skilled employees, andpreserving social capital are important goals.But as the nature of work rapidly changes, therelative weights that these factors should begiven in assessing turnover's relationship toperformance are markedly different. Moreover,there are differences between human and socialcapital losses in terms of the degree of perfor-mance erosion (e.g., additive or exponential)likely to result from high turnover levels. Fi-nally, the type of intermediate performance oroutcome variables each is likely to affect is ex-pected to be different.

Weight of Human and Social CapitalConsiderations in the Turnover-PerformanceRelationship

In knowledge-based organizations, develop-ing and retaining human capital become lessimportant as the key players (talented profes-sionals in particular; see Pink, 1998) take the roleof "free agents"—that is, they bring with them tothe table the requisite skill levels in most cases.The development of social ties in organizationsmay help tie key knowledge workers (i.e., free-agent professionals) to the firm (Capelli, 2000).Increasingly, it is becoming widely acceptedthat knowledge workers exhibit greater loyalty

to their colleagues and their profession than totheir employing organization (Capelli, 2000). AsFeldman notes, workgroup loyalty is often amuch stronger force among professionals thancommitment to "an amorphous, distant, andsometimes threatening corporate entity" (2000:179). This type of loyalty is associated withgreater relational ties (in terms of number andstrength) and may serve as the backbone of ef-fective performance. The long-term performanceconsequences of voluntary turnover may be lessattributable to a skill (human capital) deficitthan to the accumulated social capital lostthrough voluntary turnover. Thus, addressingthese issues in the knowledge economy alsoentails the challenge of managing and retain-ing organizational social capital among em-ployees in general and professionals in partic-ular.

These arguments tie in quite well with the keytenets of the resource-based view of the firm—that is, advantages accrue from the creation ofunique bundles of resources that competitorsare unable to imitate (Barney, 1991; Wernerfelt,1984). Teece, Pisano, and Shuen (1997) distin-guish between "core competence" and "dynamiccapability" by emphasizing the managerial pro-cesses involved in combining resources for ad-vantage(s). Typically, such imitation is problem-atic because of the inherent specialization,scarcity, and tacit knowledge in human assets(Lippman & Rumelt, 1982). The role of an organi-zation's culture, processes, structures, and soforth is critical in combining and leveraging in-dividual talents and competencies. Prahaladand Hamel have also articulated the idea of"informal networks and practices" and "links."They posit that "the real sources of competitiveadvantage are to be found in management'sability to consolidate corporate-wide technologyand production skills into competencies thatempower individual businesses to adaptquickly to changing opportunities" (1990: 82).

Although constructed from the private-goodview of social capital, Burt's (1997) distinctionbetween human capital and social capital ishelpful in addressing resource combination andleveraging of knowledge resources. Burt arguesthat "while human capital refers to individualability, social capital refers to opportunity"(1997: 339). When viewed from the organization'sperspective, this view suggests that managersprimarily add value by coordinating people—by


brokering relationships between individuals inorder to get the right people together to developthe opportunities. Thus, the importance of organ-izational social capital in the functioning of or-ganizations increases in knowledge-based or-ganizations, where highly skilled free-agentprofessionals are available in the market andpositively synergistic relationship resourcestake on increasing importance. Moreover, socialcapital losses resulting from voluntary turnoverwill likely be more predictive of organizationalperformance than human capital losses in suchorganizations.

Additive and Exponential Performance Erosionfrom Voluntary Turnover

The second area of divergence between exist-ing approaches to the voluntary turnover-performance relationship concerns the magni-tude of the loss of human and social capitalresulting from voluntary turnover. Cost-basedand human capital approaches to explainingthe voluntary turnover-performance relation-ship are essentially aggregated individual-level perspectives—that is, one can isolate the"costs" of turnover to organizations by aggregat-ing replacement costs or the skills lost by themultiple voluntary separations. Therefore, theexpected diminishment in organizational per-formance can reasonably be expected to bemonotonic; indeed, in some of the rare empiricalresearch addressing these issues, just such arelationship is predicted (e.g., Alexander et al.,1994). Social capital, by way of contrast, is cre-ated through combining and leveraging re-sources. As such, it may yield exponential per-formance benefits for organizations, but it alsoincreases the potential downside risk exposureshould something go wrong (e.g., voluntary turn-over of key network members). Several authors(e.g., Gerhart, Trevor, & Graham, 1996) havenoted that the combining and leveraging of re-sources is an effective way to increase upsideperformance potential; it simultaneously ex-poses the organization to the potential for neg-ative synergy should the central figure and/orother key players in the network voluntary exitfor other opportunities.

To illustrate, as an individual shares knowl-edge with others, those others can reap the ben-efits of the information (leading to lineargrowth). Furthermore, when those people then

go on to share the knowledge with others andfeed back questions, amplifications, and modi-fications that add further value for the originalsender, such accumulation of knowledge cre-ates exponential total growth (Quinn et al.,1996). Not too surprisingly, "star" performers atBell Labs were identified as those scientists whoplaced a high priority on and were successful innetworking activities (Kelley & Caplan, 1993).They understood that networking was a bartersystem and, after accumulating credits, had lit-tle problem in receiving help when they askedothers for advice. Clearly, such individualshave, among other things, access to a broaderrange of informational sources and tend to cre-ate more value for their organizations. Voluntaryturnover among such individuals will create arather significant gap in the knowledge-sharingand knowledge-generating activities within theorganization. These problems are magnified bythe potential for simultaneous departures ofother network members.

The preceding discussion highlights criticalissues in organizations today that are not com-pletely captured theoretically by existing per-spectives. In existing approaches the effect ofthe departure of a long-tenured individual onorganizational performance would be estimated(1) by comparing the human capital accumula-tions of the departing employee with the re-placement and estimating the correspondingshort-term productivity loss and/or (2) by esti-mating the short-term savings (e.g., lower pay)of the new employee versus replacement costs(e.g., recruiting and selecting a replacement).Although valuable—and certainly these havean impact and should be addressed—these ap-proaches neglect to consider the value of thedeparting individual's social capital, his or herplacement in the key social networks, and thecorresponding and possibly long-term disrup-tions in these systems.

Research demonstrates that individuals inlong-standing groups develop transactive mem-ory—a shared memory phenomenon for encod-ing and storing information in the social system(Wegner, 1987). Individuals cultivate externalmemory and information aids through othersand, in doing so, become part of the larger sys-tem (Wegner, Erber, & Raymond, 1991). The pro-cess reduces the cognitive load of any one indi-vidual and gives the group a large pool ofinformation resources across domains. Volun-


tary turnover dramatically damages transactivememory, since transactive memory is based onthe relative knowledge of others.

Moreover, trust is a key facet of social capitalaccumulation (Leana 8f Van Buren, 1999) and canbe considered an alternate control or gover-nance mechanism (Floyd & Wooldridge, 2000).Trust lowers transaction costs (Nahapiet &Ghoshal, 1998), since norms of reciprocity arestrengthened and the uncertainty associatedwith potentially opportunistic behavior is miti-gated (Provan, 1993). Such trust in exchangerelationships is vital, given the salient role ofresource combinations in knowledge organiza-tions. Therefore, organizations suffer dispropor-tionate losses when individuals who are suc-cessful in creating social capital via themaintenance and augmentation of network re-lationships throughout the organization leavethe organization. As Argote (1999) notes, themore an organization's performance depends onthe knowledge residing in employees, thehigher the rate of depreciation because of vol-untary turnover. To summarize, the predictedeffects of voluntary turnover on performancefrom existing theoretical perspectives are gen-erally negative but monotonic. In contrast, so-cial capital losses from voluntary turnover mayhave exponential negative effects on organiza-tional performance.

In this paper we discuss how voluntary turn-over may erode an organization's social capital,as well as the consequent performance implica-tions. We recognize, however, that the directionof this relationship may be logically reversed—the extent of an organization's social capitalmay strongly affect voluntary turnover levels(Brass, 1995). As noted earlier, professionals tendto be more loyal to their immediate workgroupthan to their employing organization (e.g.,Capelli, 2000; Feldman, 2000). Thus, althoughtheir loyalty to the organization may be low,people who have strong interpersonal relation-ships with peers are less likely to terminate theemployment relationship. Peteraf (1993) pro-vides the example of a Nobel Prize-winning sci-entist—a resource approaching perfect mobility.If this individual developed social capitalthrough firm-specific ties and synergistic rela-tionships with talented managers, or identifiedclosely with colleagues, the likelihood of his orher maintaining the employment relationshipwould be greater.

Moreover, trust, a key facet of social capital(Tsai & Ghoshal, 1998), can be an important fac-tor in determining whether people will voluntar-ily leave the organization. Leana and Van Buren(1999) make a distinction between fragile andresilient trust. Whereas the former is based on astrategy of reciprocity and the immediate likeli-hood of rewards, the latter refers to deeper, on-going reciprocal norms and, in effect, creates"expectations that bind" people to organizations(Kramer & Goldman, 1995, cited in Nahapiet &Ghoshal, 1998: 255).

We recognize that the causal direction be-tween voluntary turnover and social capital isdifficult to isolate and that the relationship isreciprocal, or one of repeated causal sequences.

The Voluntary Turnover-OrganizationalPerformance Relationship: Reconsidering theDependent Variable

Drawing on Burt's (1997) earlier distinction be-tween human capital and social capital, we sug-gest that the performance outcomes of voluntaryturnover will be quite different depending uponwhether one takes a human capital or socialcapital perspective. As noted by Becker and Ger-hart (1996), human capital, viewed either as la-bor or as a business function, typically is re-garded as a cost to be minimized or as apotential for increases in efficiency. As noted,existing perspectives, generally speaking, donot provide organizational-level predictionsthat vary from an aggregation of individual-level predictions. For the most part, resourcesare considered discrete, and the summation ofthe corresponding benefits and costs associatedwith such resources at lower levels in the organ-ization yields an outcome measure for the entireorganization. In essence, there is implicitly nochance for aggregation errors since interdepen-dencies among resources are not considered. InThompson's (1967) terminology, pooled interde-pendence is assumed, wherein each resourcerenders a discrete contribution to the whole.Therefore, outcome measures such as efficiencyand productivity—at the subunit or organization-al level—would be most appropriate and thegeneral means to this end would be to optimallyadjust incoming factors of production.

When the effect of voluntary turnover isviewed from a social capital perspective, therelevant performance outcomes are quite differ-


ent. Rather than an emphasis on pooled interde-pendence, as the human capital perspectivesuggests, sequential and reciprocal interdepen-dence become salient (Thompson, 1967). Here,the focus is on relationships among individualsand how they are able to combine resources inunique ways that are difficult for competitors toimitate, thus enhancing the potential for sus-tainable competitive advantages (Barney, 1991;Wernerfelt, 1984). Nonaka and Takeuchi makethe important point that "knowledge is createdand expanded through social interaction be-tween tacit and explicit knowledge" (1995: 61).Also, unlike other forms of capital, the returnsfrom knowledge—facilitated by social capital—are subject to increasing, as opposed to decreas-ing, returns (Evans & Wurster, 1997). That is, asan employee shares a skill or competence withothers, the potential returns to the organizationexpand, not contract.

Given the potential for knowledge creationand innovation via the relationships inherent inthe social capital approach, a much broaderconceptualization of performance, as opposed tothe human capital perspective with its narrowemphasis on efficiency and productivity, isneeded. To this end, Kaplan and Norton's (1996)"balanced scorecard" provides multiple per-spectives on performance, enabling a richer as-sessment of a firm's performance, as well as itsstrengths and weaknesses. The four perspec-tives are financial, customer value, operations,and organizational. For example, in addition toproviding superior financial returns, organiza-tional social capital could serve to augment afirm's leadership and ability to engage in organ-izational learning (organizational); core busi-ness processes, such as product developmentand order fulfillment (operations); and innova-tion and quick response (customer value). Simi-larly, based on Wooldridge and Floyd's (1990)work on middle-level involvement in strategyformulation, outcomes of social capital could beboth instrumental (e.g., improved decision qual-ity) and affective (e.g., commitment).

FUTURE RESEARCH DIRECTIONSAND CONCLUSIONS

Drawing on Burt's (1997) earlier distinction be-tween human capital and social capital, we sug-gest that research on the organizational perfor-mance consequences of voluntary turnover

include a focus on the retention of social capital(in addition to one emphasizing the retention ofhuman capital). As noted in the opening vi-gnette and illustrated throughout, individualswith a strong network of relationships are valu-able in terms of having access to both informa-tion and resources for their employing firm, at-tracting other high-performing individuals intotheir organization, and maintaining strong net-work ties to external stakeholders such as cus-tomers, suppliers, and alliance partners. Thus,they are in a position to create "learning plat-forms" (Grenadier & Weiss, 1997) from whichnew resource combinations may emerge. Clearly,their value goes far beyond their base of individ-ual knowledge and skills.

Based on our analysis, several potentially in-teresting areas of research are evident. In futureresearch scholars should explore two contradic-tory perspectives on how networks create socialcapital and the performance implications forvoluntary turnover. Coleman (1988) has empha-sized the role of cohesive ties in facilitating co-operation. Burt (1992), however, contends thatsuch ties lead to inflexibility and inhibit com-ple>9-task coordination. He draws upon struc-tural hole theory to argue for the benefitsachieved via brokerage opportunities formed byweak ties, which result in less network closureand greater access to a wider variety of infor-mational sources. Therefore, researchers couldexplore contingencies associated with the rela-tive benefits of each potential outcome—that is,cooperation versus timely information regard-ing new opportunities. One could argue, for ex-ample, that in complex and dynamic environ-ments (Child, 1972), an organization wouldbenefit more from the bridging of structuralholes than from network closure because of theneed for a variety of timely sources of informa-tion. In stable and simple environments, how-ever, cohesive ties would help to ensure moretimely implementation of strategies.

Similarly, the strategy that a firm or businessunit follows could also be a contingency to con-sider. For example, managers with social net-works characterized by closure would be hy-pothesized to be rnore advantageous in a firmfollowing cost leadership strategies (Porter, 1980)because of the need for control and coordinationto implement rather constrained strategies. Alter-natively, the ambiguities and uncertainties gener-ally associated with differentiation strategies


(Porter, 1980) would require a broader range ofinformation sources and inputs, enabling manag-ers with social networks characterized by weakties a greater opportunity to add value. Thus, inthe context of our paper and from a normativeperspective, firms should develop retention strat-egies for talent based partly on the type of socialnetworks managers or professionals have devel-oped.

Another intriguing extension would be the ex-ploration of the possibility of cross-over ef-fects—that is, the impact of voluntary turnoverin one organization on the performance of otherclosely linked organizations. Interorganization-al linkages are becoming more intense and or-ganizational boundaries less distinguishable(e.g., through strategic alliances, outsourcing,sole supplier relationships, increased customerinvolvement in product design, and so forth).The result is that voluntary turnover among in-dividuals who have created valuable socialcapital across organizational boundaries mayerode the performance of alliance partners aswell as the focal organization. Such researchquestions have not been addressed to ourknowledge, but a social capital approach couldbe used as a platform for pursuing these ideas.

In today's environment firms must also lookbeyond their organizational boundaries to cre-ate value for the firm and strengthen competi-tive advantage(s). Porter (1985) and others sug-gest that managers view their firm as part of asystem of value chains and analyze where theycan add value for the firm, as well as for suppli-ers, customers, and alliance partners. Dyer andSingh (1998) argue that given efficient marketsfor factor inputs, they are either readily avail-able to all competing firms, or acquiring them isapproximately equal to the economic value thatthey create (Barney, 1986). Recent studies haveindicated that productivity gains in the valuechain are possible when trading partners arewilling to make relation-specific investmentsand combine resources in unique ways (e.g..Dyer, 1996).

The above contrasting perspectives of cohe-sive ties (Coleman, 1988) versus the brokerageopportunities provided by weak ties (Burt, 1992)raise an important methodological consider-ation. We are aware of a level-of-analysis issuewith Burt's (1992) work on structural holes andthe concept of organizational social capital, butthe former may have implications for the public-

good approach. As an example. Brass andBurkhardt (1993), in their review of the socialnetwork literature, propose three alternate cen-trality measures that are relevant to the presentdiscussion. The first is "in-degree" centrality,which can provide an indicator of a person'sprestige, since it is an indicator of how often anindividual is chosen. The second measure,"closeness," indicates both the direct and indi-rect links a person has to others. Therefore, itcan measure the extent of an actor's indepen-dent access to others. "Betweenness," the thirdmeasure of centrality, refers to the extent towhich the focal person falls between pairs ofother persons on the shortest path connectingthe pairs (Brass & Burkhardt, 1993). Thus, be-tweenness provides an indicator of the extent towhich others are dependent on an actor.Whereas closeness is similar to Coleman's(1998) cohesive ties, betweenness parallelsBurt's concept of bridging structural holes. Aninteresting research question becomes this:When individuals voluntarily exit an organiza-tion, which indicators of centrality have thegreatest potential to explain and predict dys-functional organizational outcomes?

We applaud attempts by researchers to de-velop specific measures applicable to the re-search context (e.g., Pennings et al., 1998; Tsai &Ghoshal, 1998), since they may allow more spe-cific and powerful testing of social capital hy-potheses. Such measures, however, also maysacrifice generalizability, since it is not possible"to be simultaneously general, accurate, andsimple" (Weick, 1979: 35). We encourage re-searchers to consider the possible tradeoffs ofaccuracy and generalizability when evaluatingalternative measures of social capital. Further-more, an important methodological consider-ation may be, in part, illustrated via Rosen-berg's (1968) distinction between two types ofgeneralization: descriptive and theoretical.Whereas the former involves generalizing "afinding based on a smaller number of cases to abroader population" (1968: 222), the latter gener-alizations occur when "variables are seen asindicators or indices of broader concepts" (1968:223). Thus, social capital research may providemisleading results and interpretations if re-searchers do not take care when conceptualiz-ing and measuring the construct. In this case ofcentrality, an example would be when a broadvariable (e.g., centrality) is operationalized as a


single dimension (e.g., closeness) but generali-zations are made to the broader construct.

We have focused on the benefits of social cap-ital to organizations in the context of voluntaryturnover, but we should also briefly address thepotential costs. In an ironic twist, the downsideof social capital dovetails with some ideas pre-sented in the cost-benefit approaches we out-lined earlier. In the extreme, organizational so-cial capital can be potentially costly, as well asserve to restrict the infusion of new ideas andinnovation (Nahapiet & Ghoshal, 1998). Leanaand Van Buren argue that individuals are "so-cialized in the norms, values, and ways of work-ing inherent to the workgroup and the organiza-tion" (1999: 550). Such socialization processescan be expensive in terms of both financial re-sources and managerial commitment and mayrepresent a significant opportunity cost thatshould be evaluated in terms of the intendedbenefits. If such expenses become excessive,profitability may be eroded.

Furthermore, if accepted behaviors and beliefsystems become institutionalized, innovationwill become stifled, because tacit social pres-sures may prevent individuals from deviatingfrom established procedures and practices(DiMaggio & Powell, 1983). This could result inlower revenue growth, for example, because afirm would be unable either to recognize andcommercially exploit new technologies or suc-cessfully enter new product markets. Similarly,such failures could reduce profitability becausemanagers might not have the foresight to envi-sion new ways to effectively combine and lever-age resources. Moreover, deeply rooted mind-sets, which serve to inhibit a firm's ability toproperly respond to new environmental threatsand opportunities, may lead to the hiring, re-warding, and promotion of like-minded peoplewho tend to further intensify organizational in-ertia and erode innovative processes. As notedby Nahapiet and Ghoshal, "Organizations highin social capital may become ossified throughtheir relatively restricted access to diversesources of ideas and information" (1998: 260).Such homogeneity of perspectives tends toerode the effectiveness of decision-making pro-cesses.

Finally, the study of "best practices" by organ-izations, which serve to promote social capital,should be explored as a means of further devel-oping both normative and descriptive theory.

For example, at Novell, reward structures gobeyond the assessment of human capital (i.e.,individual skills and competencies) and serve toreinforce the value of collaborative and coUe-gial behavior on the part of knowledge workers.In order to attain the rank of "distinguished en-gineer," for instance, a professional must beelected by his peers. As noted by Eric Schmidt,CEO: "It's a standard that encourages tech peo-ple to be good members of the tech community.It acts to reinforce good behavior on everyone'spart" (quoted in Mitchell, 1999: 176).

Although such initiatives may serve to en-hance social capital and enable firms to reapthe potential benefits that we have addressed,exploratory research could also address theother side of the coin, so to speak. On the onehand, do such endeavors yield a favorable cost-benefit tradeoff and lead to decreased voluntaryturnover, the enhancement of employees' firm-specific ties, the lowering of transaction costs,and so on? Or, on the other hand, do such initi-atives lead to dysfunctional outcomes, such asgamesmanship, and an erosion in innovativeand novel decision making resulting from suchfactors as a lack of diversity in perspectives,backgrounds, and decision-making styles?

REFERENCES

Abelson, M. A., & Baysinger, B. D. 1984. Optimal and dysfunc-tional turnover: Toward an organizational level model.Academy of Management Review, 9: 331-341.

Alexander, J. A., Bloom, J. R., & Nuchols, B. A. 1994. Nursingturnover and hospital efficiency: An organizational-level analysis. Industrial Relations, 33: 505-520.

Argote, L. 1999. Organizational learning; Creating, retaining,and transferring knowledge. Boston: Kluwer.

Barney, J. B. 1986. Organization culture: Can it be a source ofsustained competitive advantage? Academy of Manage-ment Review. 11: 856-865.

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

Becker, B., & Gerhart, B. 1996. The impact of human resourcemanagement on organizational performance: Progressand prospects. Academy of Management Journal. 39:779-801.

Becker, G. S. 1993. Human capita?; A fheoreticai and empir-ical analysis with special reference to education (3rded.). Chicago: University of Chicago Press.

Brass, D. J. 1995. A social network perspective on humanresources management. Research in Personnel andHuman Resources Management. 13: 39-79.

Brass, D. J., & Burkhardt, M. E. 1993. Potential power and


power use: An investigation oi structure and behavior.Academy of Management Journal, 36: 441-472.

Burt, R. S. 1992. Stmctural holes: The social structure of com-petition. Cambridge, MA: Harvard University Press.

Burt, R. S. 1997. The contingent value of social capital.Administrative Science Quarterly, 42: 339-365.

Capelli, P. 2000. A market-driven approach to retaining tal-ent. Harvard Business Review, 78(1): 103-113.

Child, J. 1972. Organization, structure, environment, and per-formance: The role of strategic choice. Sociology, 6: 1-22.

Coleman, J. S. 1988. Social capital in the creation of humancapital. American Journal of Sociology, 94(Special Is-sue): 95-120.

Coleman. J. S. 1990. Foundations of social theory. Cambridge.MA: Belknap Press of Harvard University Press.

Dalton, D. R., Krackhardt, D. M., & Porter, L. W. 1981. Func-tional turnover: An empirical assessment. Journal of Ap-plied Psychology. 66: 716-721.

Dalton, D. R., & Todor, W. D. 1979. Turnover turned over: Anexpanded and positive perspective. Academy of Man-agement Review. 4: 225-235.

Dalton, D. R.. & Todor, W. D. 1982. Turnover: A lucrative harddollar phenomenon. Academy of Management Review,7: 212-218.

Darmon. R. Y. 1990. Identifying sources of turnover costs: Asegmental approach. Journal of Marketing, 54: 46-56.

DiMaggio, P. ]., & Powell, W. W. 1983. The iron cage revisited:Institutional isomorphism and collective rationality inorganizational fields. American Sociological Aeview. 48:147-160.

Drucker, P. F. 1997. The future that has already happened.Harvard Business Review, 75(5): 20-24.

Dyer. J. H. 1996. Specialized supplier networks as a source ofcompetitive advantage: Evidence from the auto indus-try. Strategic Management Journal. 17: 271-292.

Dyer, J. H., & Singh, H. 1998. The relational view: Cooperativestrategy and sources of interorganizational competitiveadvantage. Academy of Management Review, 23: 66-79.

The Economist. 1996. An acknowledged trend: The worldeconomy survey. September 28: 25-29.

Evans, P. B., & Wurster. T. S. 1997. Strategy and the neweconomics of information. Harvard Business Review,75(5): 71-82.

Feldman, D. C. 2000. From the me decade to the flee decade.In C. R. Leana & D. M. Rousseau (Eds.). fleJafionaJwealth: The advantages of stability in a changing econ-omy; 169-182. Oxford: Oxford University Press.

Fishman. C. 1998. The war for talent. Fast Company, 16(Au-gust): 104-108.

Floyd, S. W., & Wooldridge, B. 2000. Buiiding strategy fromthe middle: Reconceptualizing strategy process. Thou-sand Oaks. CA: Sage.

Franklin, J. 1997. Industry output and employment projec-tions to 2006. JVfonJhly tabor Review, 120(11): 39-57.

Gerhart, B., Trevor, C. O., & Graham, M. E. 1996. New direc-

tions in compensation research: Synergies, risk, andsurvival. Research in Personnel and Human ResourcesManagement, 14: 143-203.

Grenadier, S. R., & Weiss, A. M. 1997. Investment in techno-logical innovations: An option pricing approach. Journalof Financial Economics, 44: 397-416.

Hom, P. W., & Griffeth, R. W. 1995. Employee turnover. Cin-cinnati, OH: South-Western Publishing.

Kaplan, R. S., & Norton, D. P. 1996. The balanced scorecard.Boston: Harvard Business School Press.

Kasarda, J. D. 1973. Effects of personnel turnover, employeequalifications, and professional staff ratios on adminis-trative intensity and overhead. Sociological Quarterly,14: 359-358.

Katz, R. 1982. Effects of group longevity on project communi-cation and performance. Administrative Science Quar-terly, 27: 81-104.

Kelley, R., & Caplan, J. 1993. How Bell Labs creates starperformers. Harvard Business Review, 71(4): 128-139.

Krackhardt, D., & Porter, L. W. 1986. The snowball effect:Turnover in embedded communication networks. Jour-nal of Applied Psychology. 71: 50-55.

Krame, R. M., & Goldman, L. 1995. Helping the group orhelping yourself? Social motives and group identity inresource dilemmas. In D. A. Schroeder (Ed.), Socialdilemmas: 49-68. New York: Praeger.

Leana, C. R., & Van Buren, H. J., III. 1999. Organizationalsocial capital and employment practices. Academy ofManagement Review, 24: 538-555.

Lippman, S. A., & Rumelt, R. P. 1982. Uncertain imitability: Ananalysis of interfirm differences in efficiency undercompetition. Bell Journal of Economics, 13: 418-438.

Meyer, M. H., & DeTore, A. 1999. Project development forservices. Academy of Management Executive, 13(3):64-76.

Miller, D., & Shamsie. J. 1996. The resource-based view of thefirm in two environments: The Hollywood film studiosfrom 1936-1965. Academy of Management Journal, 39:519-536.

Mitchell, R. 1999. How to manage geeks. Fast Company.25(Iune): 174-180.

Nahapiet, J., & Ghoshal, S. 1998. Social capital, intellectualcapital, and the organizational advantage. Academy ofManagement Review, 23: 242-266.

Nonaka, L, & Takeuchi, H. 1995. The inowledge-creatingcompany. New York: Oxford University Press.

Osterman, P. 1987. Turnover, employment security, and theperformance of the firm. In M. Kleiner (Ed.), Humanresources and the performance of the firm: 275-317. Mad-ison, WI: Industrial Relations Research Association.

Pennings, I. M., Lee, K., & van Witteloostuijn, A. 1998. Humancapital, social capital, and firm dissolution. Academy ofManagement Journal, 41: 425-440.

Peteraf, M. A. 1993. The cornerstones of competitive advan-tage. Strategic Management/ournal. 14: 179-191.

Pil, F. K., & Leana, C. R. 2000. Free-agency versus high-


involvement approaches to skill development: Enhanc-ing relational wealth. In C. R. Leana & D. M. Rousseau(Eds.), Relational wealth: The advantages of stability ina changing economy: 116-129. Oxford: Oxford UniversityPress.

Pink, D. H. 1998. Free agent nation. Fast Company, 12(Janu-ary): 131-163.

Porter, M. E. 1980. Competitive strategy. New York: FreePress.

Porter, M. E. 1985. Competitive advantage. New York: FreePress.

Prahalad, C. K., & Hamel, G. 1990. The core competence of thecorporation. Harvard Business Review, 68(3): 79-91.

Provan, K. G. 1993. Embeddedness, interdependence, andopportunism in organization supplier-buyer relation-ships. Journal of Management, 19: 841-857.

Quinn, J. B., Anderson, P., & Finkelstein, S. 1996. Leveragingintellect. Academy of Management Executive, 10(3): 7-27.

Rosenberg, M. 1968. The logic of survey analysis. New York:Basic Books.

Schneider, B. 1987. The people make the place. PersonneJPsychology, 40: 437-453.

Shaw, J. D., Delery, J. E., Jenkins, G. D., Jr., & Gupta, N. 1998.An organizational-level analysis of voluntary and invol-untary turnover. Academy of Management Journal, 41:511-525.

Shaw, J. D., Delery, J. E., & Gupta, N. 2000. Voluntary turnover,workforce productivity, and organizational performance.Working paper. University of Kentucky, Lexington.

Smith, H. L., & Watkins, W. E. 1978. Managing manpowerturnover costs. Personnel Administrator, 23(4): 46-50.

Strober, M. H. 1990. Human capital theory: Implications forHR managers. Industrial Relations.. 29: 214-239.

Teece, D., Pisano, G., & Shuen, A. 1997. Dynamic capabilitiesand strategic management. Strategic Management Jour-nal, 18: 509-534.

Thompson, J. D. 1967. Organizations in action. New York:McGraw-Hill.

Tsai, W., & Ghoshal, S. 1998. Social capital and value cre-ation: The role of intrafirm networks. Academy of Man-agement Journal, 41: 464-476.

Wegner, D. M. 1987. Transactive memory: A contemporaryanalysis of the group mind. In B. Mullen & G. R. Goeth-als (Eds.), Theories of group behavior: 185-208. New York:Springer-Verlag.

Wegner, D. M., Erber, R., & Raymond, P. 1991. Transactivememory in close relationships. Journal of Personalityand Social Psychology, 61: 923-929.

Weick, K. E. 1979. The social psychology of organizing. Read-ing, MA: Addison-Wesley.

Wernerfelt, B. 1984. A resource-based view of the firm. Stra-tegic Management Journal, 5: 171-180.

Wooldridge, B., & Floyd, S. W. 1990. Strategic process effectson consensus. Strategic Management Journal, 10: 295-302.

Wysocki, B., Jr. 2000. Yet another hazard of the new economy:The Pied Piper effect. Wail Street Journal, March 30: Al.

Gregory G. Dess is a professor of strategic management and holds the Carol MartinGatton Endowed Chair in Leadership and Strategic Management in the Gatton Col-lege of Business and Economics at the University of Kentucky. He received his Ph.D.from the University of Washington. His current research interests are in the areas ofentrepreneurship, knowledge management, and social capital.

Jason D. Shaw is an assistant professor of management in the Gatton College ofBusiness and Economics at the University of Kentucky. His current research interestsinclude workforce stability and organizational performance, compensation system fit,and personality-environment congruence issues.

voluntary turnover, social capital and organizational performance and shaw 2001... ·...

Documents