catching falling stars: a human resource...

CATCHING FALLING STARS: A HUMANRESOURCE RESPONSE TO SOCIAL CAPITAL’S

DETRIMENTAL EFFECT OF INFORMATIONOVERLOAD ON STAR EMPLOYEES

JAMES B. OLDROYDSHAD S. MORRIS

The Ohio State University

Because star employees are more visible and productive, they are likely to be soughtout by others and develop an information advantage through their abundant socialcapital. However, not all of the information effects of stardom are beneficial. Wetheorize that stars’ robust social capital may produce an unintended side effect ofinformation overload. We highlight the role of human resource management in min-imizing the effects of information overload for stars, and we discuss avenues for futureresearch.

Human resource (HR) scholars have arguedthat some employees are more valuable thanothers (Becker & Huselid, 2006; Hausknecht,Rodda, & Howard, 2009; Lepak & Snell, 1999).Consistent with a resource-based view of thefirm, the highest-performing employees createdisproportionate value, providing a rare but vi-tal opportunity for an organization to increaseits competitive advantage through human capi-tal (Barney, 1991; Barney & Wright, 1998; Lepak &Snell, 2002). Alternatives such as hiring agreater number of average performers or en-hancing nonhuman assets are not adequatesubstitutes for the value created by top perform-ers (Eccles & Crane, 1988; Kelley & Caplan, 1993;Lepak, Takeuchi, & Snell, 2003; Narin, 1993). Forexample, in professional service industries anorganization’s top performers both generate thebulk of that organization’s business and consti-tute its core knowledge assets (Eccles & Crane,1988). Studies of scientists and academic re-searchers have consistently found that employ-ees at the top of the performance distributionare many times more valuable than their lower-performing colleagues (e.g., Cole & Cole, 1973;

Ernst, Leptein, & Vitt, 2000; Narin & Breitz-man, 1995).

As a result of their uniquely valuable humancapital contributions, top performers are oftenthe most widely recognized employees in agiven organization (Trevor, Hausknecht, & How-ard, 2007; Trevor & Nyberg, 2008). Top performersare also more visible than their peers in internaland external labor markets (Groysberg, Lee, &Nanda, 2008), as evidenced by increased inter-est in hiring practices regarding the highest-performing employees (Gardner, 2005; Lazear,1986). Research on the professional service in-dustry has demonstrated that top performersreceive particular attention from competingorganizations, which tend to view their achieve-ments as valuable assets ripe for acquisition(Greenwood, Hinings, & Brown, 1990).

When top performers possess high internaland external visibility, they are considered to bestars. Following Groysberg and colleagues(2008), we define “stars” as employees who (1)demonstrate superior performance in relation toothers in their respective organizations and (2)are highly visible in the labor market. Theseunique characteristics often endow stars withwhat sociologists call a “cumulative advan-tage”—namely, their productive resources in-crease at a rate exponentially greater than theirless visible and less valuable peers (Cole &Cole, 1973; Zuckerman, 1977). For example, starsin science fields find it easier to acquire theresources necessary to facilitate research, such

We gratefully acknowledge the excellent support of DavidLepak throughout the revision process and thank our anon-ymous reviewers for their insightful comments and sugges-tions. James Oldroyd also thanks the generous support of aSungkyunkwan University summer research grant in sup-porting this project. Finally, we thank to Jay Anand, Tai-Young Kim, and Steffanie Wilk for their helpful comments onthis work.

� Academy of Management Review2012, Vol. 37, No. 3, 396–418.http://dx.doi.org/10.5465/amr.2010.0403

396Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyrightholder’s express written permission. Users may print, download, or email articles for individual use only.

as colleagues seeking collaboration, cadres ofhighly capable students, and access to data-bases (Zuckerman & Merton, 1972). Conse-quently, they are likely to find themselves em-bedded in a virtuous cycle: they meet withincreased access to information resources andopportunities to increase their productivity,which lead to increased visibility in the labormarket, which, in turn, results in even more re-sources and opportunities (Allison, Long, &Krauze, 1982).

From a relational perspective, one by-productof stardom is the abundance of social capital,defined here as the structure of relationship net-works and information available to an individ-ual (Bourdieu, 1986; Burt, 1992). In this regard, ahigher number of network connections or poten-tial sources of information increase an individ-ual’s social capital (Baker, 1990). Nahapiet andGhoshal state that “the central proposition ofsocial capital theory is that networks of relation-ships constitute a valuable resource” (1998: 243).In other words, the value of network ties, as wellas the social capital that comes with them, de-rives primarily from privileged access to infor-mation and opportunities (Burt, 1997). Becausestar employees are highly visible in the labormarket and others are likely to seek relation-ships with them, stars will likely develop expo-nentially high levels of social capital that, inturn, perpetuate and reinforce their positions(Burkhardt & Brass, 1990; Groysberg et al., 2008;Kang, Morris, & Snell, 2007). For instance,through a series of studies, Groysberg (2010) hasdemonstrated that one of the key factors in stars’success is not only their unique and uniquelyvaluable human capital but also their socialcapital—that is, the relationships they possesswith others in the organization.

While an abundance of social capital can pos-itively impact stars’ performance, not all of theeffects of abundant social capital are positive(Adler & Kwon, 2002). For instance, scholars arebeginning to explore some potential limitationsof social capital due to structural challenges(Arenas, Díaz-Guilera, & Guimera, 2001; Burt,1997; Dodds, Muhamad, & Watts, 2003; Guimera,Díaz-Guilera, Vega-Redondo, Cabrales, & Are-nas, 2002; Watts, 2004), differential effects of spe-cific network contexts (Cummings & Cross, 2003;Xiao & Tsui, 2007), links to negative groups(Lechner, Frankenberger, & Floyd, 2010), limitedcontrol (Buskens & van der Rijt, 2008; Ryall &

Sorenson, 2007), and overembeddedness (Gar-giulo & Benassi, 2000). All of these factors repre-sent some form of structural constraint in whichties are not as helpful as they might otherwisebe, either because they are redundant or be-cause they offer limited access to novel informa-tion. However, while such structural factors lim-iting the value of social capital are important toprovide a more robust picture of organizationaltrials and successes, they do not reflect a seri-ous challenge created by the abundant socialcapital amassed by star employees.

We uniquely highlight that because of theirhigh visibility and performance status, stars arelikely to build up an abundance of ties leadingto nonredundant information flows. But, by thevery nature of their unique positions, stars areless likely to face structural constraints and aremore likely to amass exponentially high levelsof social capital. As a result, they are much moresusceptible to another form of constraint thanare their average-performing peers: informationoverload. In a state of overload, cognitive limi-tations may constrain the value of a star’s socialcapital; if the information load goes unmanagedfor long periods of time, the star may stumbleand, ultimately, fall (Herbig & Kramer, 1994; VanGerven, Paas, Tuovinen, & Tabbers, 2003). Thus,it is important to generate a greater understand-ing of how and when social capital adverselyaffects the performance of both the star em-ployee and the organization in which he or sheis embedded. In so doing we posit that a star islikely to fall in the absence of specific individ-ual, organizational, and network-wide actionsthat help manage the continual increase in boththe information given to and the demands madeof the star. Using a process approach to theorydevelopment, we focus on this important butpreviously unexplored boundary condition ofthe value of social capital, one that applies prin-cipally to star employees.

The conditions under which some stars shineand others fall have long proven difficult to ex-plain, let alone produce (cf. Groysberg, 2010). Webegin by reviewing the link between social cap-ital and star employees, emphasizing that be-cause of the affiliatory nature of network forma-tion, stars are likely to have exponentiallyhigher levels of social capital than their lessvisible and less highly performing peers. Wethen discuss how this abundance of social cap-ital may unintentionally result in information

2012 397Oldroyd and Morris

overload for star employees. We go on to explorethe implications of a curvilinear theory of socialcapital on the performance of star employeesand their organizations, highlighting the bound-ary conditions of our theory and explicatingwhen information overload may or may not ad-versely affect star employees with abundant so-cial capital. We then highlight strategic HRmanagement responses, targeted at the individ-ual, organizational, and network level, thatmight help mitigate the potentially negativeside effects of social capital for stars. Finally,we conclude by outlining several ways in whichfuture research can link human resource man-agement, stars, social capital, and informationoverload.

STARS’ HUMAN AND SOCIAL CAPITAL

Stars are employees who consistently andsubstantially perform better than others in theirorganizations and are also highly visible intheir respective labor markets (Groysberg et al.,2008). Stars are common across industries andare frequent topics of discussion in knowledge-based industries where organizational value islargely tied to employees’ individual abilitiesand potential for coordination. The work differ-ential between stars and nonstars is vast. Forexample, as Groysberg notes:

The phenomenon of stardom— of performerswhose productivity massively outstrips that oftheir colleagues—is well documented. One studyfound that the top 1 percent of employees inhighly complex jobs outperform average perform-ers by 127 percent. Another reported an eight-to-one productivity difference between star com-puter programmers and average programmers.The top 1 percent of inventors was found to befive to ten times as productive as average inven-tors (2010: 616).

Zucker, Darby, and Armstrong (1998) found,similarly, that in the biotech industry stars rep-resent only three-quarters of 1 percent of scien-tists, but they account for 17.3 percent of pub-lished articles. Thus, star scientists publishalmost twenty-two times as many articles astheir average colleagues.

The value that stars create for knowledge-based organizations determines both their hu-man and social capital (Groysberg et al., 2008).The rare, intangible resource of a star’s humancapital is a key source of competitive advantage

for his or her organization (Barney & Wright,1998; Huselid, 1995; Lepak & Snell, 1999). A star’sfirm-specific human capital may include knowl-edge about how to accomplish complex tasks ina particular organization, how to develop trustamong a team of employees, and how to createa sense of commitment to a firm’s success, alongwith other vital, valuable information (Barney &Hansen, 1994; Conner & Prahalad, 1996).

Strategic HR management researchers recog-nize the importance of stars’ social capital in thecreation of organizational values (e.g., Dess &Shaw, 2001; Wright, Dunford, & Snell, 2001). Be-cause stars have greater access to information,as well as access to avenues for sharing it, theytend to create more value for their organizations(Dess & Shaw, 2001). For instance, in a study ofR&D project managers, Allen and Katz (1985)found that star employees are a key source oftechnical knowledge. Because of this knowl-edge, organizational colleagues consulted mostfrequently with star employees, and stars spentsignificantly more time than their colleaguesconferring with those within and outside of theirown technical specialties. Similarly, Kang et al.(2007) have argued that top employees createvalue for organizations via their social relation-ships with colleagues across departments andfunctions. These relationships provide access tonew information and opportunities that the starsthen use to create greater organizational value.

THE UPWARD SPIRAL OF STARDOM ANDINFORMATION OVERLOAD

The Affiliatory Nature of Social Capital

Recent research in physics (Barabasi & Cran-dall, 2003), information technology (Ebel,Mielsch, & Bornholdt, 2002), and biology (Jeong,Tombor, Albert, Oltvai, & Barabasi, 2000) indi-cates that network formation often follows anaffiliatory pattern. Rather than forming ran-domly between actors, associations form bychoice, based on the actors’ preferences andwhat they are searching for or hoping to gain(Newman, 2002). Numerous studies show that af-filiatory networks represent trends in social andwork life, in which people tend to gravitate to-ward a few individuals who are key to achiev-ing their objectives (e.g., Newman & Park, 2003).The people who are the “recipients”—that is, thekey individuals to whom others gravitate—are

398 JulyAcademy of Management Review

often high-performing and visible actors inthese networks.

Because of affiliation, organizational net-works are not composed of randomly generatedconnections, Rather, once stars emerge in anorganization, others will tend to gravitate tothem, and these new associations will result ina virtuous circle that further increases stars’ im-portance and visibility (Newman, 2002). As a re-sult of their high visibility and strong perfor-mance, star employees are likely to befrequently asked for advice and to have influ-ence over and association with others(Burkhardt & Brass, 1990). Because this processis an upward spiral, other employees in the or-ganization are likely to seek out stars once theyhave achieved visibility, heightening the up-ward spiral effect. In other words, when employ-ees seek to build new relationships, they aremost likely to build those relationships withhigh-performing, highly visible stars.

Such affiliatory patterns create cumulativeadvantages for stars. The stars’ prior successincreases their future endowments of social cap-ital. This is a mechanism in which social capitaland human capital are recursive, with each re-inforcing and increasing the other. As other em-ployees form ties with star employees, they, too,try to access their unique human capital or gainaccess to their robust relationship; thus, thestars’ organizational power and influence in-crease concurrently (Emerson, 1962).

Using a social network structure approach,Tichy and Tushman (1979) showed that starsplay a “linking pin” role since they occupy thecore of the organization’s network structure. Ad-ditional research has shown that employeeswho have the highest organizational recogni-tion have greater access to resources, includingsocial prestige and knowledge, from which theymight attract others seeking status, information,or money (Bacharach & Lawler, 1980; Foa & Foa,1974). For instance, as Allen and Katz (1985)noted, stars have robust connections and so cankeep up with new developments in their fields,further increasing their human capital. As a re-sult, stars are frequently embedded in a virtuouscycle of social capital development, in whichhigher social capital (i.e., more associations) in-creases their appeal, causing more and morepeople to gravitate toward them.

Because of the nature of affiliatory tie forma-tion, stars are likely to be connected to many

more individuals than average employees; thisconnection pattern follows a power-law (Pareto)distribution of ties. Thus, rather than simplyhaving a few more ties than average employees,stars are likely to have exponentially more as-sociations and the concomitant social capitalthan average employees. As a consequence,stardom does not provide a marginal increase insocial capital over that of the average employeebut, rather, an exponential increase.

Underscoring the difference between the mar-ginal increase and the exponential increase insocial capital is the difference between randomand affiliatory networks. This is a vital pointwhen investigating how social capital affectsinformation flow to star employees. Figure 1 il-lustrates this point. In Figure 1 we model fourdifferent networks of 100 people each. Two ofthese networks have an average of five ties peractor, and two have an average of ten ties peractor. We compare the difference in both typesof network structures for networks formed in arandom versus affiliatory manner. Both the five-tie and ten-tie networks consist of 100 people,and each of the networks has an equal averagenumber of ties. However, the distribution of tiessignificantly differs between the networks. Onthe one hand, in the first network (which has anaverage of five ties per actor) and third network(which has an average of ten ties per actor), thedistributions follow a random pattern of rela-tionship formation. On the other hand, in thesecond network (which has five ties) and fourthnetwork (which has ten ties), the distributionsfollow an affiliatory, or preferential, model.1 Inthe affiliatory networks the majority of the rela-tionships are linked to a small number of stars,concentrated in the center of the network. In therandom networks the ties are more evenly dis-persed across the network.

Figure 2 quantifies this difference betweenthe random and affiliatory networks. In this fig-ure we construct a measure of “in-degree cen-trality” for each actor. In-degree centrality is ameasure of the incoming ties directed toward anactor. Our results show that in affiliatory net-works the most central star employees havemore than two-and-a-half times as many asso-

1 Network theorists also refer to affiliatory networks asscale-free networks (e.g., Barabasi & Crandall, 2003).


FIGURE 1Random versus Affiliatory Network Graphs

Random 100-node network with average in-degree centrality of 5

Random 100-node network with average in-degree centrality of 10

Affiliatory 100-node network with average in-degree centrality of 5

Affiliatory 100-node network with average in-degree centrality of 10

FIGURE 2Centrality in Random and Affiliatory Networks

Comparison of the in-degree centralitya for each node in the

random and affiliatory network of 100 actors with an average in-

degree centrality of 5 ties. The highest in-degree centrality for

actors in the random network is 10, while the highest in-degree

centrality for actors in the affiliatory network is 26.

Comparison of the in-degree centralitya for each node in the

random and affiliatory network of 100 actors with an average in-

degree centrality of 10 ties. The highest in-degree centrality for

actors in the random network is 27, while the highest in-degree

centrality for actors in the affiliatory network is 46.

30

25

20

15

10

5

0

Random

Affiliatory

50

Random

Affiliatory

454035302520151050

1 9 17 25 33 41 49 57 65 73 81 89 97

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99

a In-degree centrality is a sum of the ties that are directed toward an actor.


ciations as those most central in randomnetworks.

Figure 3 takes this analysis even further, dem-onstrating that the stars’ network ties in affilia-tory networks are often exponentially higherthan others’ ties in the same network. For boththe random and affiliatory examples, we havecalculated the percentage of all ties connectedto each actor in the network. In our comparison,for instance, we found that the top 10 percent ofemployees’ in-degree centrality in the randomnetwork accounts for 18 percent of the network’sin-degree ties. The top 10 percent of actors in theaffiliatory network, however, account for over40 percent of ties. We see that the magnitude ofthis effect intensifies as we narrow our defini-tion of stars to fewer and fewer actors. For in-stance, the most central actor in a random net-work receives just 2 percent of the in-degree ties,whereas in an affiliatory network the individualoccupying that position receives 6 percent ofties, or three times the number of ties. Becauseorganizational networks follow an affiliatorypattern and stars are likely to be key players inthese networks, we posit the following.

Proposition 1: Because of their highperformance and high visibility, staremployees are likely to have exponen-tially more social capital than aver-age employees.

Stars and Information Flow

Recent research in the information scienceshas repeatedly demonstrated that communica-tion networks follow a power-law distribution—namely, key actors in a given network are likelyto both receive and send more information thannonkey actors.2 Since stars are likely to have anabundance of social capital, they have access tonumerous contacts, increasing the averageamount of information they receive (Groysberg& Lee, 2008; Lechner et al., 2010). As a result,stars are not only more likely to have more as-sociations than average employees but aremore likely to actively communicate using theseties. This may be due to their colleagues’ need toaccess their human capital, an effect demon-strated through empirical studies of actors withhigh human capital. For instance, Burkhardtand Brass (1990) found that experts are morelikely to be sought out in organizations thannonexperts.

Furthermore, once a star has obtained infor-mation, abundant social capital may enable

2 Baeza-Yates, Boldi, and Castillo (2006) have demon-strated the affiliatory nature of web page links, showing thatregardless of the function used to identify page rank, thedistribution is likely to follow a power-law function indicat-ing that a few key nodes (or sources of information) areexponentially more likely to be active.

FIGURE 3Percent of Actors with Ties to Star Employees in Random and Affiliatory Network Conditions

Percent of ties in random and affiliatory networks linked to star

employees in 5 average ties network

Percent of ties in random and affiliatory networks linked to star

employees in 10 average ties network

Random network Random networkAffiliatory network Affiliatory network

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0Top 1% Top 5% Top 10% Top 25% Top 1% Top 5% Top 10% Top 25%


him or her to leverage his or her structural po-sition, facilitating the flow of new and valuableinformation across structural boundaries orgaps in the network space. Burt notes that thosewith abundant social capital often have a “sayin whose interests are served,” and such an in-dividual will act as an “entrepreneur in the lit-eral sense of the word—a person who addsvalue by brokering the connection between oth-ers” (1997: 342).3

Here we compare the volume of informationstars are likely to receive with the volume ofinformation average employees in the affilia-tory networks are likely to receive. In this casewe hold flow constant per actor (a conservativemeasure), finding that stars (who constitute thetop 1 percent of actors) have nearly nine timesas many ties as average employees. Assumingthat there is an average of five ties present inthe network, that each tie generates the sameinformation load, and that both incoming andoutgoing information flows emerge from eachtie, we project that stars will receive eighteentimes as much information as average employ-ees. As the number of ties increases, this loadincreases in a power-law manner. In addition, itis likely that attention paid to a star is evengreater, because stars are not only more likely tohave more connections but are also more likelyto be involved in active ties (in the form of re-quests, questions, etc.). This interactive effect ofcombining the number and volume of flowclearly indicates that stars are likely to receiveand send an abundance of information. Hence,we posit the following.

Proposition 2: Because of their abun-dant social capital, star employeesare likely to send and receive expo-nentially more information than aver-age employees.

Stars and Information Overload

Individuals process information, according toSternberg (1977: 317), by turning information intointelligence through a process of “components.”Component processing includes activities suchas coding, inferring, mapping, applying, and re-sponding to information. While individuals varyin their ability to process information owing totheir cognitive abilities (Ackerman, 1986; Kanfer& Ackerman, 1989; Locke, 1965), instructional de-sign scholars note that all individuals rely onshort-term or working memory to process infor-mation on a day-to-day basis (Baddeley, 1986;Miller, 1956). Individuals are conscious of andcan monitor only the content of their workingmemories. Because information processing re-quires working memory, information loads thatexceed its capacity may overwhelm an individ-ual’s information processing activities. Theselimitations of working memory are widelyknown and accepted (Sweller, van Merrienboer,& Paas, 1998). As stated by Sweller et al.:

Because working memory is most commonly usedto process information in the sense of organizing,contrasting, comparing, or working on that infor-mation in some manner, humans are probablyonly able to deal with two or three items of infor-mation simultaneously when required to processrather than merely hold information. . . . it is thisfactor that provides a central claim of cognitiveload theory (1998: 252–253).

Much of the research on social capital as-sumes a linear relationship between informa-tion flow and stardom or highly central posi-tions such that the more information anemployee receives, the better his or her perfor-mance will be (e.g., Burt, 1992, 1997). Informationprocessing theory further clarifies our under-standing of the potential risks of informationoverload on those employees with high levels ofsocial capital. This theory asserts that individu-als benefit from the receipt of information, butonly until they reach a point at which they areunable to process additional incoming informa-tion (O’Reilly, 1980; Tushman & Nadler, 1978).Beyond this point, additional information be-comes a liability (Eppler & Mengis, 2004). In astate of overload, an individual’s ability to per-form rapidly declines (Chewning & Harrell,1990). In any given context, then, if the amount ofinformation an individual receives exceeds hisor her information processing ability, the extrainformation may harm performance (Boone, van

3 Several studies demonstrate a strong link between per-formance and actors who facilitate the connection of otheractors in the network. Some of these advantages includeprestige (Allen & Cohen, 1969), early job promotion for them-selves and their subordinates (Burt, 1997; Katz & Tushman,1983; Podolny & Baron, 1997), higher salaries (Seibert,Kraimer, & Liden, 2001), and better job performance. Each ofthese advantages is distinct to star employees (Brass,1984, 1985).


Olffen, & van Witteloostuijn, 2005; Carpenter &Fredrickson, 2001; Wadhwa & Kotha, 2006).

While all organizational actors face some riskof information overload, a star’s robust and con-stantly increasing social capital places him orher in a unique position that is likely to lead toinformation overload if not carefully managed.Moreover, a star’s robust social capital not onlylikely burdens him or her with extreme levels ofinformation flow but also places the star in aposition where other employees are likely toseek advice and expertise and, in so doing,cause frequent interruptions that compromisethe star’s ability to complete tasks (Rudolph &Repenning, 2002). Grove (1983: 67), for example,described the constant request for informationand advice received by managers as “theplague of managerial work.” Similarly, Perlow(1999) showed that the frequent coworker inter-ruptions experienced by highly visible softwareengineers ultimately led to “a time famine,”wherein engineers had too many informationrequests and could not properly performtheir jobs.

Such frequent information requests requireextra information processing activity and oftennecessitate immediate attention. These requestscan also interrupt information processing fo-cused on the task at hand (Cellier & Eyrolle,1992; Kirmeyer, 1988). In an organizational set-ting, additional attention and visibility can in-crease the amount of information requests, de-

creasing an employee’s attention and ability toconcentrate on the specific requests themselves(Oldham, Kulik, & Stepina, 1991; Perlow, 1999).Moreover, scholars such as Jett and George(2003) have shown that information technologyhas increased the number of interruptions frominformation requests, with email and otherforms of electronic communication heighteningthe frequency with which people can interactand interrupt one another at work (e.g., Cutrell,Czerwinski, & Horvitz, 2001; Speier, Valacich, &Vessey, 1999).

In line with other studies of network structureand load (e.g., Watts, Dodds, & Newman, 2002),4

we suggest that even when star employees haverelatively low cognitive costs associated withprocessing information for each message, theyare frequently overloaded by the cumulativeburden created by their exponential amounts ofsocial capital. In Figure 4 we show star employ-ees’ cumulative information burden for an infor-mation flow of five and ten messages per net-work contact. Because of the nature of affiliatorynetworks, and assuming only five incoming and

4 Network scholars have demonstrated that affiliatory net-works are robust against random breakdowns. However,they are easy to disrupt with focused attacks (of overload) onthe most highly connected actors. In other words, when starsare overloaded, their failure to process information can eas-ily disrupt the network (Cohen, Havlin, & ben-Avra-ham, 2002).

FIGURE 4Information Load in an Affiliatory Network with Average Centrality of Five Actors at Five and Ten

Messages per Actor

1000

900

800

700

600

500

400

300

200

100

0

Information load 5messages per tie

Information load 10messages per tie

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99


outgoing communications per tie, we find thatstars carry an information burden of over 450messages, as compared with 25 messages forthe average employee.

Because exponentially higher levels of socialcapital are likely to burden star employees withinformation, we posit the following.

Proposition 3: Because of their ex-treme level of information flow, starsare likely to experience informationoverload.

The Individual Performance Effects of Stars’Information Overload

A star’s performance is likely to be hinderedby this deluge of information. Information pro-cessing studies have clearly demonstrated alink between effective information processingand employee performance (Eppler & Mengis,2004). Scholars exploring information overloadhave emphasized that the consequences of in-formation overload not only act as a limit to theemployees’ performance but actually may de-crease the overall performance of the individualexperiencing overload (e.g., Jacoby, 1977; Meier,1963). For instance, Malhotra noted that

although consumers develop mechanisms forlimiting their intake of information, their limitedprocessing capacity can become cognitivelyoverloaded if they attempt to process “too much”information in a limited time, and this can resultin confusion, cognitive strain, and other dysfunc-tional consequences (1984: 437).

Similarly, in his study of library workers,Meier (1963) found that overwhelmed individu-als had to completely stop the information flowthey received until they could catch up on theirprocessing tasks. Oskamp (1965) also found thatinformation improves decision-making abilityup to a certain point, but when the flow of infor-mation exceeds that point, additional informa-tion diminishes the person’s decision outcomes.Likewise, Connolly (1977) found that excessiveinformation leads to a decreased accuracy indecision making. Schick, Gorden, and Haka(1990) noted that the burden of information over-load leads to confusion, an inability to set pri-orities, and a deficit in information recall. Over-load has also been shown to reduce decisionmakers’ ability to identify relevant information(Hodge & Reid, 1971; Streufert, 1973). Hence, weposit the following.

Proposition 4: As star employees expe-rience information overload, their per-formance is likely to decrease.

The Organizational Performance Effects ofStars’ Information Overload

Because stars are required to share theirknowledge with others, they are likely to receivemany requests for advice and information. Pro-fessional service organizations often identifystars as “thought leaders” or “knowledge ex-perts”—people others can turn to for help. Starsare not only identified but often actively put incontact with others, across business and geo-graphic lines, to ensure visibility and accessi-bility by peers (Lorsch & Tierney, 2002). They arelikely to be singled out for formal and informalmentoring responsibilities (Noe, 1988). In fact, asPhillips-Jones (1983) pointed out, most mentoringrelationships are informal, incited by admira-tion for the star or by job demands that require astar’s expertise. Thus, the very mentoring oppor-tunities meant to energize employees can feellike a punishment for success if the programsare not designed to consider the potential forinformation overload in the case of stars.

Scholars have also argued that organizationsoften spend the majority of their efforts provid-ing stretch assignments, “special” projects, anddevelopment programs solely for star players(Huselid, Beatty, & Becker, 2005). When organi-zations target these employees, the employeesmay become overburdened with responsibili-ties, and this may cause a decrease in theirability to share information and mentor others.

When stars experience information overload,they are likely to become bottlenecks in the or-ganization (Cross & Parker, 2004). In other words,stars who receive too much information will notbe able to share their expertise with others inthe organization. In this way, information over-load may hinder an organization’s ability to le-verage a star’s human capital. Consequently,information overload may not only affect boththe star’s individual performance but also theorganization’s performance on the whole.

Studies of scale-free networks demonstratethe effect of overload for key network nodes.These studies note that an “attack [overwhelm-ing the actor] on the most highly connectednodes of the network” can result in a seriousdisruption in network flow (Cohen et al., 2002:


14). The effects of overload for stars ripplethrough the system, and a small disruption for astar “suffices to disrupt the net for all” (Cohen etal., 2002: 14).

While information flowing to stars may not bedirectly proportional to the information theyshare with others, it is likely to be correlated. Asa result, the volume of information sent out bystars can come back to them, adding exponen-tially to their information load. Figure 5 shows acomparison of the difference in in-flowing, out-flowing, and total flow for actors in random ver-sus affiliatory networks. This highlights that inaffiliatory networks stars are likely to shoulderthe information burden resulting from both in-coming and outgoing ties; overloaded stars willbe unable to process and share information.

Limiting stars’ ability to share informationalso likely limits their ability to provide adviceand mentoring to others, both of which are nec-essary elements in fostering human capitalwithin an organization. For instance, DeLong,Gabarro, and Lees (2007) found that top perform-ers in professional service firms tend to focustoo much on satisfying clients, to the point ofneglecting their own and others’ personal skillbuilding. Therefore, we propose the following.

Proposition 5: As stars experience in-formation overload, they are more

likely to decrease the amount of valu-able information they share withtheir peers, stifling organizationalperformance.

Next, we draw further distinctions betweenstar and average employees. Research hasshown that, unlike average workers, star em-ployees have high external visibility, whichmeans they are likely to be able to leave anorganization if they feel burdened. Trevor (2001)emphasized this point, noting that star employ-ees’ high performance and visibility lead to highportability. Spence (1973) argued that when em-ployees make investments in their organiza-tions, they increase productivity and are visibleto others; thus, competitors note “signals” thatthese employees possess skills generally appli-cable across organizations, and such signals in-vite competitors to attempt to hire these employ-ees (Lazear, 1986), increasing the employees’likelihood of leaving the organization they ini-tially worked for (Schwab, 1991). In fact, researchsuggests that star employees’ turnover rateis not significantly affected by nationwide un-employment rates, when firms typically engagein only limited hiring (Lee, Gerhart, Weller, &Trevor, 2008; Trevor & Nyberg, 2008)—their mo-bility is robust even during economic uncer-tainty. The “war for talent” literature shows that

FIGURE 5In-Degree, Out-Degree, and Total Degree Centrality for Random and Affiliatory Networks

60

50

40

30

20

10

0

Cen

tra

lity

9992857871645750433629221581

In-degreerandom

Out-degreerandom

Total degreerandom

In-degreefree-scale

Out-degreefree-scale

Total degreefree-scale


professional industries typically hire freely fromone another and that highly pursued employeesoften have little loyalty to their employers, con-ceptualizing loyalty instead as a duty to thelarger profession (Gardner, 2005; Greenwood,Oliver, Sahlin, & Suddaby, 2008). Much literaturefocuses on the mobility of these individuals (e.g.,Groysberg et al., 2008; Marx, Strumsky, & Flem-ing, 2009).

We argue here that when stars are in a stateof chronic information overload, they may noticethat their performance is suffering and mayseek to remedy the situation. With the difficultyof single-handedly managing information over-load and the relative ease of moving to anotherorganization, stars are likely to become frus-trated and exercise their options to work forother organizations. Moving can provide themwith the opportunity to access more resourcesand face fewer demands on their time and en-ergy.5 Stars experiencing overload and then“shooting” to competitors is a well-understoodphenomenon. For example, Tom Rath, head ofworkplace consulting for Gallup, noted thatcompanies that require stars to share informa-tion “can be perceived as piling on. And that’sthe quickest way to push that person out thedoor” (quoted in McGregor, 2010).

When a star employee decides to leave, theorganization suffers from the loss of valuablehuman capital, as well as the loss of socialcapital. Shaw, Duffy, Johnson, and Lockhart(2005) found that the loss of both human andsocial capital resulting from turnover is partic-ularly detrimental to an organization’s perfor-mance; even when a small number of star em-ployees leave an organization, such a loss has asharp negative effect on organizational perfor-mance. Hence, the extent to which a star’s infor-mation load can be effectively managed re-mains a key strategic concern. As a result, weposit the following.

Proposition 6: When stars experienceinformation overload, they are morelikely to leave a given organization,stifling organizational performance.

MANAGING STARDOM ANDINFORMATION LOAD

Prior studies have examined how to preserveinformation flow in scale-free networks; thesestudies show the importance of focusing on keyactors in the networks. For instance, simulationsof immunization strategies demonstrate thatwhen a system of random inoculation is utilized,a network remains contagious “even after im-munization of most of its nodes” (Cohen et al.,2002: 23). However, when key nodes are immu-nized, even though only a small fraction of ac-tors receive immunization, this strategic inocu-lation is sufficient to dramatically halt thespread of infection. Similarly, when HR manag-ers focus on increasing a star’s efficiency andeffectiveness, their efforts are likely to have aprofound effect in managing the side effects ofinformation overload for the organization as awhole. We suggest that because networks fol-low a nonrandom pattern, HR strategies shouldfollow suit and take a nonrandom approach tocurbing information overload. These strategiesshould focus on the key individuals in the orga-nization: the stars. We further suggest that whensuch strategies are absent, stars in these orga-nizations are likely to fall.

How can HR managers reduce the informationoverload side effects of amassing social capi-tal? To reduce the liabilities of a star’s abundantsocial capital, HR managers can use tactics thatincrease stars’ individual processing capacity,concentrate on the organization’s characteris-tics with respect to information flow, and bolsterthe structural foundation of the star’s network(Eppler & Mengis, 2004). To demonstrate the ef-fect of these practices, we turn to theories ofcognition, HR management, and social networksto examine (1) individual, (2) organizational, and(3) structural conditions that may influence thedegree to which stars experience informationoverload. We highlight that without active man-agement of the information load, star employeeswill tend to become overloaded with informa-tion, which will likely lead to either their fall ortheir decision to shoot off to other organizations.

Individual Conditions

While information overload is a likely out-come for employees with exponentially high so-cial capital, it is hardly a certainty. Stars have

5 We thank an anonymous reviewer for pointing us torelevant research demonstrating that star employees have ahigher likelihood of leaving their organization when theyexperience information overload.


the power to increase their information process-ing capacity.6 Instructional design scholarshave demonstrated that individuals’ ability toprocess information is restricted by limitationsin short-term memory (Baddeley, 1986), whichallows individuals to process only a few ele-ments before “overloading their capacity anddecreasing the effectiveness of processing” (Ka-lyuga, Ayres, Chandler, & Sweller, 2003: 23). Daftand Huber (1987), Sproull and Kiesler (1991),Whittaker, Swanson, Kucan, and Sidner (1997),Hansen and Haas (2001), and Kostova and Roth(2003) have argued that knowledge sharing islimited by an actor’s inability to act on sharedinformation and to distinguish between reus-able and nonusable information.

This limitation with short-term memory can befurther mitigated by developing long-term mem-ory, which increases information processing byapplying domain-specific knowledge that orga-nizes and categorizes information to aid in de-cision making. Decision criteria, also known asschemas, increase an individual’s ability to cat-egorize and prioritize information. Informationthat does not fit into existing schemas (Rumel-hart, 1975), scripts (Schank & Abelson, 1977),frames (Minsky, 1975), or categories (Lakoff, 1987)requires additional effort to process, and mayeven require the adaptation of existing linguis-tic frameworks or the creation of new ones. Onceschemas are created and understood, the infor-mation load vastly decreases. Sweller et al.highlight that

schemas both bring together multiple elementsthat can be treated as a single element and allowus to ignore myriads of irrelevant elements.Working memory capacity is freed, allowing pro-cesses to occur that otherwise would overburdenworking memory. Automated schemas both allowfluid performance on familiar aspects of tasksand— by freeing working memory capacity—permit levels of performance on unfamiliar as-

pects that otherwise might be quite impossible(1998: 258).

For example, Morris et al. (2009) found thatwhen people lack a shared vision or a sharedframework of what is important within the orga-nization, much of the information possessed byemployees is neither transferred nor processed.To increase stars’ information processing capa-bilities, then, HR managers should provide themwith the opportunity to have diverse workplaceexperiences, which will enable them to quicklyunderstand subtle nuances of information andshare that understanding with others (Tushman& Scanlan, 1981). Furthermore, a more robustexposure to different experiences also increasesstars’ transactive memory, honing their abilityto locate specific information (Wegner, 1986).7

In addition to efforts to increase long-termmemory, organizations can also optimize howstars allocate their attention capacities to im-prove information outcomes (Kanfer & Acker-man, 1989). By designating specific times tocheck email, voicemail, and texts, organizationsmay reduce the cognitive processing load forstars. These steps standardize processing re-quirements and reduce interruptions. For in-stance, star employees may set aside a specifictime to check email each day, ignoring it for therest of the day. In this way stars can focus oninformation requests and send informationwhen they have time to fully process the infor-mation, replying in an efficient, effectivemanner.

In addition to capability, motivation may af-fect both information processing (LePine,Colquitt, & Erez, 2000; Sackett, Gruys, & Elling-son, 1998; Witt & Burke, 2002) and knowledgesharing performance (Szulanski, 2000). This

6 One strategy for dealing with information overload is tosimply ignore much of the information one receives. Whileignoring information flow may directly benefit the star byreducing his or her cognitive burden, it will likely underminecooperation and teamwork in the organization. Groysberg,Lee, and Abrahams note the potential problems of this strat-egy: “To get the best of your top performers, maintain a“no-jerks” policy: Stars who don’t play well with others won’tbenefit you in the long run” (2009). Thus, we assume thatencouraging stars to ignore the problem is not a viablesolution for the stars or the organization.

7 Although schemas may help stars speed decision mak-ing, they may not automatically improve decision outcomes.For instance, Malhotra notes, “When presented with ‘toomuch’ information, consumers may become confused, so thatthey are unable to effectively and efficiently process theinformation, and/or they may adopt some heuristic process-ing. While consumers may employ heuristics to limit theintake of information, these heuristics may often involve atradeoff between simplifying and optimizing. As the work byWright (1975: 62) suggests, ‘simplifying and optimizing arelikely to be antagonistic goals.’ Hence, in the context ofdecision making, it is entirely possible for a consumer toadopt a choice heuristic that may limit cognitive strain butthat may not lead to the ‘best’ or even to a satisfactorychoice” (1984: 438).


proves to be a serious challenge. As Lorsch andTierney note:

Employing stars is necessary but insufficient.They must also be aligned; that is, they mustbehave in ways that move the firm toward itsgoals. . . . Unfortunately, such behavior is usuallyan unnatural act. This is particularly true in PSFs[professional service firms], where the profes-sionals’ natural independence is compounded bythe inherently decentralized nature of the work(2002: 26).

In a network simulation, Tang, Xi, and Ma(2006) showed that the star actors with the mostconnected nodes who were highly motivated toshare information (and not overwhelmed) werenearly twice as effective in facilitating informa-tion flow than were stars with average informa-tion-sharing aspirations. Thus, by increasing astar’s motivation, an organization may inducehim or her to dedicate more attention to infor-mation processing (Hwang, Kettinger, & Yi, 2010;Kanfer & Ackerman, 1989), dramatically chang-ing information flow in the organization.

Organizations can increase motivation by rec-ognizing and rewarding stars for their efforts. Astar’s efforts to effectively process and shareinformation are often inadequately rewarded.For example, Perrow (1999) recounted an inter-view with a star engineer. The engineer wasseen by coworkers as highly visible and produc-tive, and, as a result, many people went to thisengineer for help.

At one point, the engineer approached the soft-ware manager and told him that he was havingtrouble balancing all the demands for his helpand completing his own deliverables. Accordingto the engineer, he was told, ‘Do your own workfirst, and then, if you want to help others, that isyour choice, but do it on your own time’ (Perrow,1999: 69).

In this context the star employee was actuallydiscouraged from sharing information withother employees, sending a negative messageto employees and decreasing the star’s motiva-tion to help others.

Another way to increase a star’s motivation isto turn over more control to the star. Scholarsargue that top performers who are included inmaking strategic decisions pay more attentionto the information they receive from others (Mor-ris, Alvarez, Barney, & Molloy, 2010). This specif-ically occurs for knowledge workers in profes-sional service industries. Furthermore,

Gargiulo, Ertug, and Galunic (2009) have foundthat employees can more effectively process in-formation when they are free from the control ofnetwork ties and the compulsion to volunteerinformation. In other words, when stars canchoose whether they respond to employees, theymay be more effective than their peers at pro-cessing large amounts of information. This ar-gument is supported by recent research demon-strating that people who feel a diminishedsense of power and control have significantlyimpaired information processing and decision-making faculties (Smith, Jostmann, Galinsky, &Van Dijk, 2008). According to Hallowell (2011),when stars feel as though they have less controlover their networks, their ability to make deci-sions will likely decrease, as will their ability toprioritize information, plan, organize, and im-plement new ideas. As a result, when stars arerequired to share more information with others,rather than being able to choose when theyshare information, they have a decreased abilityto cope with high information loads.

In light of these methods for improving staremployees’ long-term memory, information allo-cation skills, and motivation to share informa-tion, we propose the following.

Proposition 7: Increasing a star’s infor-mation processing capabilities (in-creasing working memory, buildingefficient attention allocation capabil-ities, and increasing his or her motiva-tion) will increase his or her ability tomanage information and preventoverload.

Organizational Conditions

In addition to individual factors that increasea star’s ability to manage high informationloads, HR professionals can develop organiza-tional processes and systems to help stars wardoff overload. For example, efficient search pro-cesses can dramatically reduce the number ofqueries sent to stars. When the stars’ colleaguesknow where to find information in an organiza-tion, they are much less likely to engage incostly and ineffective search activities (Walsh &Ungson, 1991). These searches, in which actorsindiscriminately query others for help, arecalled “greedy” searches (Huberman & Adamic,2004). HR activities that focus on facilitating ef-


ficient searches rather than greedy ones willdramatically reduce the search burden fallingon stars.

However, these efforts are likely to be partic-ularly challenging, since the employees en-gaged in greedy searches are likely to find suchsearches personally efficient (Adamic, Lukose,Puniyani, & Huberman, 2001). Moreover, in affil-iatory networks indiscriminate searches “natu-rally gravitate towards the high degree nodes”(Adamic et al., 2001: 5). We do not suggest herethat stars should not be involved in employeesearches; efficient search does not mean thatactors should not engage stars but, rather, thatactors should engage the right stars. Efficientsearches may also cause a reduction in the over-all network traffic by “intentionally choosinghigh degree nodes” (Adamic et al., 2001: 5) ifthese nodes are the right nodes. By increasingemployees’ transactive memory, employers canincrease effective searches, which allows actorsto access information directly and efficiently(Wegner, 1986). Developing transactive memorymerely requires knowing what all individualsknow across the organization.

Efficient search can also be encouraged byrequiring those searching for information to paysearch costs. For instance, within the WorldBank Group, project leaders have reduced stars’processing burdens by requiring project teamsand managers from other offices to pay for stars’time. Thus, the “costs” of processing more infor-mation are calculated into the organization andinto the employee’s work schedule (Morris etal., 2010).

In addition to promoting efficient search, HRprofessionals may also employ information fil-tering mechanisms and information technolo-gies (Bawden, 2001; Edmunds & Morris, 2000),which can help stars manage their informationburdens. Grant (1996) has argued that process-ing information for applications requires orga-nizational processes and information systemsthat enable an individual to actually use theinformation coming to him or her. These systemscodify and simplify information input, capturingknowledge in a storage system that both pre-serves the information and shifts the burden ofsharing it from the stars to the information sys-tems themselves. For example, organizationsmay capture a star’s valuable information inbrief “lessons learned” or some other sort oftemplate that allows users to directly access

and apply the star’s knowledge in a comprehen-sive and readily digestible format (Morris &Oldroyd, 2009). In addition, information systemscan work to eliminate fluctuations in the infor-mation flow. For instance, companies can de-crease information overload (e.g., Snell, Youndt,& Wright, 1996) with the use of specific processesand systems consisting of set routines or guide-lines about how information should be receivedand disseminated (Hall, 1992; Itami, 1987; Subra-maniam & Youndt, 2005; Walsh & Ungson, 1991).These templates can aid in overcoming the com-plexities and strains of processing information.

Through information systems and processes,knowledge often becomes decontextualized andarticulated in databases and other codified sys-tems; this allows employees to more easily un-derstand which information is helpful in whichcontexts. In this regard, technology provides em-ployees with an appropriate structural mecha-nism to receive and share information (Brock-bank & Ulrich, 2005; Davenport & Prusak, 1998).For example, Morris et al. (2009) found that cod-ifying knowledge and embedding it into exist-ing operations allows organizations to capture,roll out, maintain, promote, and distribute infor-mation to others in the organization. In additionto codifying information, organizations such asMcKinsey and Company have transitioned someof their star employees to full-time knowledge-sharing positions. In these positions the staremployees’ incentives for information process-ing and sharing are cleanly aligned with theirroles in the organization (Rasiel & Friga, 2002).

HR leaders can further foster trust and mean-ingful relationships with others within the orga-nization. By doing so they may influence stars’ability to effectively process information (Groys-berg & Lee, 2008). Quality social relations withother stars can actually decrease informationoverload by altering the cognitive processes ofstars. Psychiatrists argue that, aside from indi-viduals’ actual information processing abilities,feelings of information overload also have aneurological basis (Hallowell, 2011). When starswork in environments where trust and respectfor one another proliferate, the deep centers ofthe brain send messages through the pleasurecenter to the area that assigns resources to thefrontal lobes. Even under extreme forms of infor-mation overload, this sense of human connec-tion improves the functioning of top-level exec-utives (Hallowell, 2005). In contrast, those who


work in physical isolation feel more stress frominformation requests (Hallowell, 2005). As a re-sult, we propose the following.

Proposition 8: Organizational pro-cesses and systems (effective search,implementing information technolo-gies, specifying information roles, andfostering trust) will increase a star’sability to prevent information overload.

Structural Conditions

Not only do individual and organizationalconditions influence the likelihood that a starwill experience overload, but HR professionalsmay also affect the network’s structural proper-ties to help stars manage their information bur-dens. One method is to provide support staff tostars within networks; support staff can monitorincoming requests and information solicita-tions, acting as information gatekeepers. Gate-keepers make initial diagnoses with respect tothe urgency or utility of the information and thendecide whether the information should be givento a star employee (Shumsky & Pinker, 2003). If agatekeeper can adequately process and dissem-inate the information offered or requested, thereis no need to send it on to the star. The gate-keeper can also prioritize information so the starwill know which information to address first.Gatekeepers further act as quality control mech-anisms, ensuring that only valuable informationreaches the star. They can also ensure that po-tentially harmful or misrepresentative informa-tion is not presented to the star. As a result,network filters may influence the informationload for stars.

Another way to help manage a star’s informa-tion burden is by narrowing the breadth of thestar’s networks. Recent work in information sci-ence has shown that stars’ ability to manageinformation dramatically increases when theseindividuals are focused primarily on dense re-ciprocal interactions. For instance, Adamic andAdar (2002) found that when employees in an HPlab were encouraged to interact only with peo-ple known to reciprocate knowledge sharing, alinear, rather than power-law, distribution of in-formation flow emerged. By taking this researchinto account, HR professionals can help focusthe information flow of stars to core, reciprocalcommunications.

Increasing the density of a star’s network mayalso play a role in managing the side effects ofhis or her information flow. For example, schol-ars suggest that stars who operate in moreclosed networks within a given organization, orin networks where employees are robustly con-nected to one another rather than just throughthe stars or central nodes, will have reducedinformation burdens. Noting that informationquality deteriorates as it moves through only afew central figures, Baker (1984) argued thatmarkets with more dense networks, which aresimilar to random networks in which everyoneis communicating freely and rapidly, result indecreased information burdens for the centralfew. This improves organizational informationsharing on the whole.

HR professionals may also manage the effectsof information overload on stars by focusing thebenefits of the stars’ social capital on statusoutcomes rather than on information outcomes.In this respect the quality of a star’s informationprocessing becomes less important to the suc-cess of the organization as a whole. For exam-ple, Groysberg et al.’s (2008) study of star stockanalysts demonstrated how star employees lev-eraged their status and shared standard stockopinions with all contacts, reducing their needto share unique and customized information.Standardized information may be highly valuedby such a network; as a result, status may en-able stars to increase their ability to overcomeinformation overload. Hence, we propose thefollowing.

Proposition 9: Shaping network condi-tions (increasing network filters, net-work density, and the value of status)will increase a star’s ability to preventinformation overload.

IMPLICATIONS

A few key theoretical implications haveemerged from our more comprehensive theory ofinformation overload and the high levels of so-cial capital for star employees. First, we devel-oped a theory that links star employees andcognitive constraints, suggesting that informa-tion overload is a potential reason why somestars fall. More specifically, we have arguedthat stars are likely to possess exponentiallyhigh levels of social capital, resulting in large


volumes of information flow. Because they haveexponentially higher volumes than their peers,stars are more likely to be overloaded with in-formation. When stars are in a state of overload,their decision quality declines and their abilityto share information grinds to a halt, cripplingthe performance of both the stars and the orga-nizations in which they are embedded. In otherwords, stars’ abundant social capital may, if notcarefully managed, cause them to fall.

Second, even though many studies note thehigh mobility of star employees (e.g., Groysberg,2010), few have explored the intrinsic reasonswhy stars leave their organizations. While theirhigh visibility makes stars easy targets of com-petitive hiring practices, our theory suggeststhat star employees’ turnover may actually bepartially caused by the stars’ desire to avoidinformation overload. In other words, instead ofdealing with information overload, stars mayopt to become “shooting stars”—joining forceswith competitors.

Third, the implications for HR professionalswho seek to increase organizational knowledgesharing may be even more far reaching. Fordecades, organizations have sought to increasesocial capital by fostering knowledge, coordina-tion, collaboration, and information flow (e.g.,Davenport & Prusak, 1998; Szulanski, 1996;Thompson, 1967). The theory we have developedhere places an important caveat on these efforts,highlighting how, because of stars’ propensityfor overload, HR professionals may be betterserved by focusing on alternative strategies. Wesuggest that these strategies should shift frombuilding links to increasing stars’ informationprocessing capacity, aligning organizationalprocesses and systems to manage stars’ infor-mation processing burdens, using technologicaland human gatekeepers to guard stars’ time andattention, and shaping information networks toease stars’ responsibilities. These actions maymore effectively facilitate information flow inthe organizations. Moreover, we suggest a refo-cusing of managerial attention on stars, who arethe key sources of information in organizations.Without them, whole organizational knowledgesharing will likely fail.

Fourth, our theory suggests that individualstars may benefit by preserving the value ofrobust social capital. Efforts to increase long-term memory in the form of schemas, experi-ence, and other strategies can increase stars’

ability to process information. In addition, HRprofessionals can emphasize aspects of thestars’ social capital (such as status) that do notimpose a cognitive information processing bur-den. Because the value of status is not con-strained by cognition, stars may be betterserved by leveraging their status than by lever-aging their information. Finally, stars may moveaway from managing structural redundanciesin their network—a point emphasized and in-vestigated previously—and begin to managethe processes by which they gain information.Our theoretical model representing these asso-ciations is summarized in Figure 6.

Our theory suggests several avenues of futureresearch. For instance, future work could ex-plore how stars’ efforts to build social capitalmay result in different types of social capital,such as social capital that optimizes status orsocial capital that optimizes knowledge flow.8

This research could investigate the performanceeffects of firms that differentiate between situa-tions in which status or reputation is paramount.It could also investigate situations in which in-formation processing is more important, seekingto deleverage the information flow associatedwith their stars’ network position. And it couldexplore the synergies or trade-offs that exist be-tween the two types of social capital, along withtheir respective advantages.

Future research could also examine specificHR practices that are tied to hiring employeeswith greater information processing capacities.In addition, researchers could examine prac-tices involving training employees to deal moreeffectively with information overload and tostructure work practices so as to reduce infor-mation burdens for star employees. In this arti-

8 Prior research has identified an additional limitation tosocial networks—namely, the information redundancy thatis primarily due to structural constraints. Specifically, actorswho receive redundant information will yield less valuefrom their social capital than actors who have more uniqueinformation flows (Burt, 1997; Granovetter, 1995; Uzzi, 1997).However, since stars’ performance is primarily viewed asthe ability to utilize knowledge (rather than control knowl-edge), and since stars reside in the center (rather than thejunction) of networks (Berman, Down, & Hill, 2002; Groysberget al., 2008), we believe that this type of structural constraintis less central to our discussion of stars and how they main-tain their information advantage. Still, while informationoverload differs from our structural constraints, they mayboth affect a given star, leading to a type of dual constraint.


cle we have discussed general strategies thatorganizations can take to help employees man-age information overload. We also suggest thatsome stars may be better suited than others todeal with exponentially high levels of socialcapital. Future measures could examine the ex-tent to which individuals possess specific capa-bilities that allow them to deal with high levelsof information flow.

Empirical research may help us understandspecific HR practices and systems that manag-ers can incorporate to reduce the informationburden or increase the information processingcapacity of star employees. For example, Dessand Shaw (2001) originally proposed a theoreti-cal link between turnover, social capital, andorganizational performance. Shaw and col-leagues (2005) later tested and extended thattheory by examining social capital on the turn-over-performance relationship among thirty-eight restaurant chains. In a similar vein, thisresearch could be extended to include specific

HR practices and explanations of how they re-late to stars’ ability to deal with uncommonlyelevated levels of social capital.

Further research may also address how thetheory we have delineated can, at the outset,differentiate star employees from their averagecolleagues (e.g., Hausknecht et al., 2009). Socialnetwork analyses could be conducted on entireemployee populations across multiple organiza-tions; these analyses could measure the extentto which star employees are interconnected. Fu-ture network analyses could also consider ex-amining stars’ level of information inflow andoutflow. Psychological measures could then beused to examine the star employees’ feelings ofbeing overwhelmed. This can be a difficult task,since much of the research on work overloadand “time famine” is qualitative (e.g., Berg,Grant, & Johnson, 2010; Perlow, 1999). For re-searchers to understand the network and orga-nizational factors involved, future studiesshould span organizations to ensure a sample

FIGURE 6Theoretical Model

Turnover

Organizational

performance

Individual performance

Outcomes

+

+

Affiliatory

social capitalStar employee Information

flow

The upward spiral

+

Individual

capabilities

Organizational

processes and

norms

Network

properties

HR management tactics

+


variance. Then performance measurementscould consider both individual and group per-formance appraisals and turnover.

Finally, it would be interesting to explore howthe incessant overload of stars impacts variousaspects of their performance. Does overloadcause a decrease generally across all aspects ofperformance, or does it limit only certain kindsof performance? We posit that status will likelybe unaffected by overload, but these differencescould be fruitfully examined in future empiricalresearch.

This article has several important limitationsand boundary conditions. First, we have focusedon the effects of robust social capital involvingan “average” star employee. However, impor-tant differences likely exist in the characteris-tics of star employees, particularly regardingtheir ability to manage information flow. Whilewe have addressed such differences when dis-cussing the management of stars and informa-tion overload, other factors may be beneficial tothe star but harmful to the organization. Forinstance, some stars may be highly narcissisticand unwilling to share information. As a result,these employees may act as information blackholes, into which vast information is pouredwith nothing passed on to colleagues. In thesecases the information burden that the star em-ployee bears is reduced only by reducing theincoming information processed. In cases suchas this, it is possible to alleviate the side effectsof social capital while heightening the negativeeffects on the organization. Future researchcould expound upon this important nexus be-tween social capital, information overload, andactor motivation.

Similarly, some stars may manage excessivelevels of information flow by simply ignoringthem. This loss of information may not be due tonarcissism but, instead, may simply be a resultof a lack of priorities. One factor that allowsemployees to become stars is their dedication toa specific vision or line of work; star employees,then, may become so absorbed in this work thatthey lose track of requests for information orforget to ask others for information when theyneed it. Owing to their singular vision, they mayfail to share information with others along theway and so may inadvertently ease their owninformation burdens.

CONCLUSION

In sum, this article contributes to existing dis-cussions of star employees and how organiza-tions manage them by drawing on the socialcapital, information processing, and HR man-agement literature. We identify a unique theo-retical link between stars, affiliatory networkeffects, and information overload, calling atten-tion to practices that might result in a subse-quent decline in the job performance of star em-ployees. Exploring these unique associations,we present a new understanding of how theinformation processing constraints of stars mayinfluence information flow and the stars’performance.

We also refocus scholarly attention from staremployees’ structural advantages to the poten-tially burdensome cognitive constraints of so-cial capital, which underscores the situationalmechanism of information flow for star employ-ees. In so doing we attempt to develop amidlevel theory by using an internal analysis ofsystem behavior (Coleman, 1990). We explorehow key HR strategies targeted at the individ-ual, organizational, and network structure-widelevels may determine whether stars experienceoverload within and across organizations. Be-cause of the curvilinear relationship betweeninformation overload and performance, starsmust carefully manage the information flow re-sulting from their social capital, rather thanmerely focus on increasing their social capital.Thus, the development of this new theory of in-formation overload for stars highlights the im-plicit tension between the stars’ preferred socialstructures—the ones that grant information re-sources—and their ability to utilize these re-sources for individual and organizationalbenefit.

REFERENCES

Ackerman, P. L. 1986. Individual differences in informationprocessing: An investigation of intellectual abilities andtask performance during practice. Intelligence, 10: 109–139.

Adamic, L. A., & Adar, E. 2002. How to search a social net-work. Social Networks, 27: 187–203.

Adamic, L. A., Lukose, R. M., Puniyani, A. R., & Huberman,B. A. 2001. Search in power-law networks. Physical Re-view E, 64: 1–8.


Adler, P. S., & Kwon, S. W. 2002. Social capital: Prospects fora new concept. Academy of Management Review, 27:17–40.

Allen, T. J., & Cohen, S. I. 1969. Information flow in researchand development laboratories. Administrative ScienceQuarterly, 14: 12–19.

Allen, T. J., & Katz, R. 1985. Project performance and the locusof influence in the R&D matrix. Academy of Manage-ment Journal, 28: 67–87.

Allison, P. D., Long, J. S., & Krauze, T. K. 1982. Cumulativeadvantage and inequality in science. American Socio-logical Review, 47: 615–625.

Arenas, A., Díaz-Guilera, A., & Guimera, R. 2001. Communi-cation in networks with hierarchical branching. Physi-cal Review Letters, 86: 3196–3199.

Bacharach, S. B., & Lawler, E. 1980. Power and politics inorganizations. San Francisco: Jossey-Bass.

Baddeley, A. D. 1986. Working memory. New York: OxfordUniversity Press.

Baeza-Yates, R., Boldi, P., & Castillo, C. 2006. GeneralizingPageRank: Damping functions for link-based rankingalgorithms. Seattle: ACM Press.

Baker, W. E. 1984. The social structure of a national securitiesmarket. American Journal of Sociology, 89: 775–811.

Baker, W. E. 1990. Market networks and corporate behavior.American Journal of Sociology, 96: 589–625.

Barabasi, A. L., & Crandall, R. E. 2003. Linked: The newscience of networks. American Journal of Physics, 71:409–410.

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

Barney, J. B., & Hansen, M. H. 1994. Trustworthiness as asource of competitive advantage. Strategic Manage-ment Journal, 15: 175–190.

Barney, J. B., & Wright, P. M. 1998. On becoming a strategicpartner: The role of human resources in gaining com-petitive advantage. Human Resource Management, 37:31–46.

Bawden, D. 2001. Information overload. Library & Informa-tion Briefings, 92: 1–15.

Becker, B. E., & Huselid, M. A. 2006. Strategic human resourcemanagement: Where do we go from here? Journal ofManagement, 32: 898–925.

Berg, J. M., Grant, A. M., & Johnson, V. 2010. When callings arecalling: Crafting work and leisure in pursuit of unan-swered occupational callings. Organization Science, 21:973–994.

Berman, S. L., Down, J., & Hill, C. W. L. 2002. Tacit knowledgeas a source of competitive advantage in the NationalBasketball Association. Academy of Management Jour-nal, 45: 13–31.

Boone, C., van Olffen, W., & van Witteloostuijn, A. 2005.Personality composition, leadership, information acqui-sition, and informed decision making in teams: A simu-lation game study of the locus-of-control. Academy ofManagement Journal, 47: 633–656.

Bourdieu, P. 1986. The forms of capital. In J. G. Richardson(Ed.), Handbook of theory and research for the sociologyof education: 241–258. New York: Greenwood.

Brass, D. J. 1984. Being in the right place: A structural anal-ysis of individual influence in an organization. Admin-istrative Science Quarterly, 29: 327–343.

Brass, D. J. 1985. Men’s and women’s networks: A study ofinteraction patterns and influence in an organization.Academy of Management Journal, 28: 327–343.

Brockbank, W., & Ulrich, D. 2005. Higher knowledge forhigher aspirations. Human Resource Management, 44:489–504.

Burkhardt, M. E., & Brass, D. J. 1990. Changing patterns orpatterns of change: The effects of a change in technol-ogy on social network structure and power. Administra-tive Science Quarterly, 35: 104–127.

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

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

Buskens, V., & van der Rijt, A. 2008. Dynamics of networks ifeveryone strives for structural holes. American Journalof Sociology, 114: 371–407.

Carpenter, M. A., & Fredrickson, J. W. 2001. Top managementteams, global strategic posture, and the moderating roleof uncertainty. Academy of Management Journal, 44:533–545.

Cellier, J. M., & Eyrolle, H. 1992. Interference betweenswitched tasks. Ergonomics, 35(1): 25–36.

Chewning, E. C., Jr., & Harrell, A. M. 1990. The effect ofinformation load on decision makers’ cue utilization lev-els and decision quality in a financial distress decisiontask. Accounting, Organizations and Society, 15: 527–542.

Cohen, R., Havlin, S., & ben-Avraham, D. 2002. Structuralproperties of scale-free networks. Berlin: Wiley.

Cole, J. R., & Cole, S. 1973. Social stratification in science.Chicago: University of Chicago Press.

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

Conner, K. R., & Prahalad, C. K. 1996. A resource-based the-ory of the firm: Knowledge versus opportunism. Organi-zation Science, 7: 477.

Connolly, T. 1977. Information processing and decision mak-ing in organizations. In B. M. Staw & G. R. Salancik(Eds.), New directions in organizational behavior: 205–234. Chicago: St. Clair Press.

Cross, R., & Parker, A. 2004. The hidden power of socialnetworks: Understanding how work really gets done inorganizations. Boston: Harvard Business School Press.

Cummings, J. N., & Cross, R. 2003. Structural properties ofwork groups and their consequences for performance.Social Networks, 25: 197–210.

Cutrell, E., Czerwinski, M., & Horvitz, E. 2001. Notification,disruption, and memory: Effects of messaging interrup-tions on memory and performance. In M. Hirose (Ed.),


Human-Computer Interaction INTERACT ‘01: 263–269. To-kyo: IOS Press (for IFIP).

Daft, R. L., & Huber, G. P. 1987. How organizations learn: Acommunication framework. Research in the Sociology ofOrganizations, 5: 1–36.

Davenport, T., & Prusak, L. 1998. Working knowledge: Howorganizations manage what they know. Boston: HarvardBusiness School Press.

DeLong, T. J., Gabarro, J. J., & Lees, R. J. 2007. When profes-sionals have to lead: A new model for high performance.Boston: Harvard Business School Press.

Dess, G. G., & Shaw, J. D. 2001. Voluntary turnover, socialcapital, and organizational performance. Academy ofManagement Review, 26: 446–456.

Dodds, P. S., Muhamad, R., & Watts, D. J. 2003. An experimen-tal study of search in global social networks. Science,301: 827–829.

Ebel, H., Mielsch, L. I., & Bornholdt, S. 2002. Scale-free topol-ogy of email networks. Physical Review E, 66(3) 35103.

Eccles, R. F., & Crane, D. B. 1988. Doing deals. Boston: Har-vard Business School Press.

Edmunds, A., & Morris, A. 2000. The problem of informationoverload in business organisations: A review of the lit-erature. International Journal or Information Manage-ment, 20: 17–28.

Emerson, R. M. 1962. Power-dependence relations. AmericanSociological Review, 27: 31–41.

Eppler, M. J., & Mengis, J. 2004. The concept of informationoverload: A review of literature from organization sci-ence, accounting, marketing, MIS, and related disci-plines. Information Society, 20: 325–344.

Ernst, H., Leptein, C., & Vitt, J. 2000. Inventors are not alike:The distribution of patenting output among industrialR&D personnel. IEEE Transactions on Engineering Man-agement, 47: 184–199.

Foa, U. G., & Foa, E. B. 1974. Societal structures of the mind.Springfield, IL: Charles C. Thomas.

Gardner, T. M. 2005. Interfirm competition for human re-sources: Evidence from the software industry. Academyof Management Journal, 48: 237–256.

Gargiulo, M., & Benassi, M. 2000. Trapped in your own net?Network cohesion, structural holes, and the adaptationof social capital. Organization Science, 11: 183–196.

Gargiulo, M., Ertug, G., & Galunic, C. 2009. The two faces ofcontrol: Network closure and individual performanceamong knowledge workers. Administrative ScienceQuarterly, 54: 299–333.

Granovetter, M. 1995. Getting a job: A study of contacts andcareers (2nd ed.). Chicago: University of Chicago Press.

Grant, R. M. 1996. Toward a knowledge-based theory of thefirm. Strategic Management Journal, 17: 109–122.

Greenwood, R., Hinings, C. R., & Brown, J. 1990. “P2-Form”strategic management: Corporate practices in profes-sional partnerships. Academy of Management Journal,33: 722–755.

Greenwood, R., Oliver, C., Sahlin, K., & Suddaby, R. 2008.Sage handbook of organizational institutionalism. Lon-don: Sage.

Grove, A. S. 1983. High output management. New York: Ran-dom House.

Groysberg, B. 2010. Chasing stars: The myth of talent and theportability of performance. Princeton, NJ: Princeton Uni-versity Press.

Groysberg, B., & Lee, L. E. 2008. The effect of colleaguequality on top performance: The case of security ana-lysts. Journal of Organizational Behavior, 29: 1123–1144.

Groysberg, B., Lee, L. E., & Abrahams, R. 2009. The myth of thelone star: Why one top performer may not shine asbrightly as you hope. Wall Street Journal, March 23:http://online.wsj.com/article/SB123740508141175095.html.

Groysberg, B., Lee, L. E., & Nanda, A. 2008. Can they take itwith them? The portability of star knowledge workers’performance. Management Science, 54: 1213–1230.

Guimera, R., Díaz-Guilera, A., Vega-Redondo, F., Cabrales,A., & Arenas, A. 2002. Optimal network topologies forlocal search with congestion. Physical Review Letters,89: 248701.

Hall, S. S. 1992. How technique is changing science. Science,257: 344–349.

Hallowell, E. M. 2005. Overloaded circuits: Why smart peopleunderperform. Harvard Business Review, 83(1): 1–9.

Hallowell, E. M. 2011. Shine: Using brain science to get thebest from your people. Boston: Harvard Business ReviewPress.

Hansen, M. T., & Haas, M. R. 2001. Competing for attention inknowledge markets: Electronic document disseminationin a management consulting company. AdministrativeScience Quarterly, 46: 1–28.

Hausknecht, J. P., Rodda, J., & Howard, M. J. 2009. Targetedemployee retention: Performance-based and job-relateddifferences in reported reasons for staying. Human Re-source Management, 48: 269–288.

Herbig, P. A., & Kramer, H. 1994. The effect of informationoverload on the innovation choice process. Journal ofConsumer Marketing, 11: 45–54.

Hodge, M. H., & Reid, L. S. 1971. The influence of similaritybetween relevant and irrelevant information upon acomplex identification task. Perception and Psychophys-ics, 10: 193–196.

Huberman, B. A., & Adamic, L. A. 2004. Information dynamicsin the networked world. Lecture Notes in Physics, 650:371–398.

Huselid, M. A. 1995. The impact of human resource manage-ment practices on turnover, productivity, and corporatefinancial performance. Academy of Management Jour-nal, 38: 635–672.

Huselid, M. A., Beatty, R. W., & Becker, B. E. 2005. The work-force scorecard: Managing human capital to executestrategy. Boston: Harvard Business School Press.

Hwang, Y., Kettinger, W. J., & Yi, M. Y. 2010. Understandinginformation behavior and the relationship to job perfor-


mance. Communications of the Association for Informa-tion Systems, 27(8): 113–128.

Itami, H. 1987. Mobilizing invisible assets. Cambridge, MA:Harvard University Press.

Jacoby, J. 1977. Information load and decision quality: Somecontested issues. Journal of Marketing Research, 14:569–573.

Jeong, H., Tombor, B., Albert, R., Oltvai, Z. N., & Barabasi, A. L.2000. The large scale organization of metabolic net-works. Nature, 47: 651–654.

Jett, Q. R., & George, J. M. 2003. Work interrupted: A closerlook at the role of interruptions in organizational life.Academy of Management Review, 28: 494–507.

Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. 2003. Theexpertise reversal effect. Educational Psychologist, 38:23–31.

Kanfer, R., & Ackerman, P. L. 1989. Motivation and cognitiveabilities: An integrative/aptitude-treatment interactionapproach to skill acquisition. Journal of Applied Psy-chology, 74: 657–690.

Kang, S. C., Morris, S. S., & Snell, S. A. 2007. Relationalarchetypes, organizational learning, and value creation:Extending the human resource architecture. Academy ofManagement Review, 32: 236–256.

Katz, R., & Tushman, M. L. 1983. A longitudinal study of theeffects of boundary spanning supervision on turnoverand promotion in research and development. Academyof Management Journal, 26: 437–456.

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

Kirmeyer, S. L. 1988. Coping with competing demands: Inter-ruption with the type-A pattern. Journal of Applied Psy-chology, 73: 621–629.

Kostova, T., & Roth, K. 2003. Social capital in multinationalcorporations and a micro-macro model of its formation.Academy of Management Review, 28: 297–317.

Lakoff, G. 1987. Fire, women, and dangerous things. Chicago:University of Chicago Press.

Lazear, E. P. 1986. Raids and offer matching. Research inLabor Economics, 8: 141–165.

Lechner, C., Frankenberger, K., & Floyd, S. W. 2010. Taskcontingencies in the curvilinear relationships betweenintergroup networks and initiative performance. Acad-emy of Management Journal, 53: 865–889.

Lee, T. H., Gerhart, B., Weller, L., & Trevor, C. O. 2008. Under-standing voluntary turnover: Path-specific job satisfac-tion effects and the importance of unsolicited job offers.Academy of Management Journal, 51: 651–671.

Lepak, D. P., & Snell, S. A. 1999. The human resource archi-tecture: Toward a theory of human capital allocationand development. Academy of Management Review, 24:31–48.

Lepak, D. P., & Snell, S. A. 2002. Examining the human re-source architecture: The relationships among humancapital, employment, and human resource configura-tions. Journal of Management, 28: 517–543.

Lepak, D. P., Takeuchi, R., & Snell, S. A. 2003. Employmentflexibility and firm performance: Examining the interac-tion effects of employment mode, environmental dyna-mism, and technological intensity. Journal of Manage-ment, 29: 681–703.

LePine, J. A., Colquitt, J. A., & Erez, A. 2000. Adaptability tochanging task contexts: Effects of general cognitiveability, conscientiousness, and openness to experience.Personnel Psychology, 53: 563–593.

Locke, E. A. 1965. The relationship of task success to taskliking and satisfaction. Journal of Applied Psychology,49: 379–385.

Lorsch, J. W., & Tierney, T. J. 2002. Aligning the stars: How tosucceed when professionals drive results. Boston: Har-vard Business School Press.

Malhotra, N. K. 1984. Reflections on the information overloadparadigm in consumer decision making. Journal of Con-sumer Research, 10: 436–440.

Marx, M., Strumsky, D., & Fleming, L. 2009. Mobility, skills,and the Michigan non-compete experiment. Manage-ment Science, 55: 875–889.

McGregor, J. 2010. How to keep your star employees. Fortune,October 27: http://money.cnn.com/2010/10/22/pf/jobs/employee_retention.fortune/index.htm.

Meier, R. L. 1963. Communications overload: Proposals fromthe study of a university library. Administrative ScienceQuarterly, 7: 521–544.

Miller, G. A. 1956. The magical number seven, plus or minustwo: Some limits on our capacity for processing infor-mation. Psychological Review, 63: 81–91.

Minsky, M. A. 1975. A framework for the representation ofknowledge: The psychology of computer vision. NewYork: McGraw-Hill.

Morris, S., Alvarez, S., Barney, J., & Molloy, J. 2010. Employeeinvestments in firm-specific human capital and firmlevel competitive advantage. Working paper, FisherCollege of Business, The Ohio State University, Colum-bus.

Morris, S., & Oldroyd, J. 2009. To boost knowledge transfer,tell me a story. Harvard Business Review, 87(5): 23.

Morris, S., Wright, P., Trevor, J., Stiles, P., Stahl, G., Snell, S.,Paauwe, J., & Farndale, E. 2009. Global challenges toreplicating HR: The role of people, processes, and sys-tems. Human Resource Management, 48: 973–995.

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

Narin, F. 1993. Technology indicators and corporate strategy.Review of Business, 14(3): 19–23.

Narin, F., & Breitzman, A. 1995. Inventive productivity. Re-search Policy, 24: 507–519.

Newman, M. E. J. 2002. Assortative mixing in networks. Phys-ical Review Letters, 89: 208701.

Newman, M. E. J., & Park, J. 2003. Why social networks aredifferent from other types of networks. Physical ReviewLetters, 68: 036122.


Noe, R. A. 1988. An investigation of the determinants ofsuccessful assigned mentoring relationships. PersonnelPsychology, 41: 457–479.

Oldham, G. T., Kulik, C. T., & Stepina, L. P. 1991. Physicalenvironments and employee reactions: Effects of stimu-lus-screening skills and job complexity. Academy ofManagement Journal, 34: 929–938.

O’Reilly, C. A. 1980. Individuals and information overload inorganizations: Is more necessarily better? Academy ofManagement Journal, 23: 684–696.

Oskamp, S. 1965. Overconfidence in case-study judgments.Journal of Consulting Psychology, 29: 261–265.

Perlow, L. A. 1999. The time famine: Toward a sociology ofwork time. Administrative Science Quarterly, 44: 57–81.

Perrow, C. 1999. Organizing to reduce the vulnerabilities ofcomplexity. Journal of Contingencies and Crisis Man-agement, 7(3): 150–155.

Phillips-Jones, L. 1983. Establishing a formalized mentoringprogram. Training & Development Journal, 37(2): 38–42.

Podolny, J. M., & Baron, J. N. 1997. Resources and relation-ships: Social networks and mobility in the workplace.American Sociological Review, 62: 673–693.

Rasiel, E. M., & Friga, P. N. 2002. The McKinsey mind: Under-standing and implementing the problem-solving toolsand management techniques of the world’s top strategicconsulting firm. New York: McGraw-Hill.

Rudolph, J. W., & Repenning, N. P. 2002. Disaster dynamics:Understanding the role of quantity in organizationalcollapse. Administrative Science Quarterly, 47: 1–30.

Rumelhart, D. E. 1975. Notes on a schema for stories. In D. G.Bobrow & A. Collins (Eds.), Representation and under-standing: Studies in cognitive science: 185–210. NewYork: Academic Press.

Ryall, M. D., & Sorenson, O. 2007. Brokers and competitiveadvantage. Management Science, 53: 566–583.

Sackett, P. R., Gruys, M. L., & Ellingson, J. E. 1998. Ability-personality interactions when predicting job perfor-mance. Journal of Applied Psychology, 83: 545–556.

Schank, R. C., & Abelson, R. P. 1977. Scripts, plans, goals, andunderstanding. Hillsdale, NJ: Lawrence Erlbaum Associ-ates.

Schick, A. G., Gorden, L. A., & Haka, S. 1990. Informationoverload: A temporal approach. Accounting, Organiza-tions and Society, 15: 199–220.

Schwab, D. P. 1991. Contextual variables in employee per-formance-turnover relationships. Academy of Manage-ment Journal, 34: 966–975.

Seibert, S. E., Kraimer, M. L., & Liden, R. C. 2001. A socialcapital theory of career success. Academy of Manage-ment Journal, 44: 219–237.

Shaw, J. D., Duffy, M. K., Johnson, J. L., & Lockhart, D. E. 2005.Turnover, social capital losses, and performance. Acad-emy of Management Journal, 48: 594–606.

Shumsky, R. A., & Pinker, E. J. 2003. Gatekeepers and refer-rals in services. Management Science, 49: 839–856.

Smith, P. K., Jostmann, N. B., Galinsky, A. D., & Van Dijk,W. W. 2008. Lacking power impairs executive functions.Psychological Science, 19: 351–398.

Snell, S. A., Youndt, M. A., & Wright, P. M. 1996. Establishinga framework for research in strategic human resourcemanagement: Merging resource theory and organiza-tional learning. Human Resource Management, 14: 61–90.

Speier, C., Valacich, J. S., & Vessey, I. 1999. The influence oftask interruption on individual decision making: An in-formation overload perspective. Decision Sciences, 30:337–360.

Spence, M. 1973. Job market signaling. Quarterly Journal ofEconomics, 87: 355–374.

Sproull, L., & Kiesler, S. 1991. Computers, networks, andwork. Scientific American, 265: 116–123.

Sternberg, R. J. 1977. Component processes in analogicalreasoning. Psychological Review, 84: 353–378.

Streufert, S. C. 1973. Effects of information relevance on de-cision making in complex environments. Memory andCognition, 1: 224–228.

Subramaniam, M., & Youndt, M. A. 2005. The influence ofintellectual capital on the types of innovative capabili-ties. Academy of Management Journal, 48: 450–463.

Sweller, J., van Merrienboer, J. J. G., & Paas, F. G. W. C. 1998.Cognitive architecture and instructional design. Educa-tional Psychology Review, 10: 251–296.

Szulanski, G. 1996. Exploring internal stickiness: Impedi-ments to the transfer of best practice within the firm.Strategic Management Journal, 17(Winter Special Issue):27–43.

Szulanksi, G. 2000. Appropriability and the challenge ofscope. In G. Dosi, R. R. Nelson, & S. G. Winter (Eds.), Thenature and dynamics of organizational capabilities: 69–98. Oxford: Oxford University Press.

Tang, F., Xi, Y., & Ma, J. 2006. Estimating the effect of orga-nizational structural on knowledge transfer: A neuralnetwork approach. Expert Systems with Applications, 30:796–800.

Thompson, J. D. 1967. Organizations in action: Social sciencebases of administrative theory. New York: McGraw-Hill.

Tichy, N. M., & Tushman, M. L. 1979. Social network analysisfor organizations. Academy of Management Review, 4:507–519.

Trevor, C. O. 2001. Interactions among actual ease-of-movement determinants and job satisfaction in the pre-diction of voluntary turnover. Academy of ManagementJournal, 44: 621–638.

Trevor, C. O., Hausknecht, J. P., & Howard, M. J. 2007. Whyhigh and low performers leave and what they find else-where: Job performance effects on employment transi-tions. CAHRS Working Paper Series No. 07–11, CornellUniversity, Ithaca, NY.

Trevor, C. O., & Nyberg, A. J. 2008. Keeping your headcountwhen all about you are losing theirs: Downsizing, vol-untary turnover rates, and the moderating role of HRpractices. Academy of Management Journal, 51: 259–276.


Tushman, M. L., & Nadler, D. A. 1978. Information processingas an integrating concept in organizational design.Academy of Management Review, 3: 613–624.

Tushman, M. L., & Scanlan, T. J. 1981. Boundary spanningindividuals: Their role in information transfer and theirantecedents. Academy of Management Journal, 24: 289–305.

Uzzi, B. 1997. Social structure and competition in interfirmnetworks: The paradox of embeddedness. Administra-tive Science Quarterly, 42: 35–67.

Van Gerven, P. W. M., Paas, F., Tuovinen, J. E., & Tabbers, H.2003. Cognitive load measurement as a means to ad-vance cognitive load theory. Educational Psychologist,38: 63–71.

Wadhwa, A., & Kotha, S. 2006. Knowledge creation throughexternal venturing evidence from the telecommunica-tion equipment manufacturing industry. Academy ofManagement Journal, 49: 819–835.

Walsh, J. P., & Ungson, G. R. 1991. Organizational memory.Academy of Management Review, 16: 57–91.

Watts, D. J. 2004. The new science of networks. Annual Re-view of Sociology, 30: 243–270.

Watts, D. J., Dodds, P. S., & Newman, M. E. J. 2002. Identity andsearch in social networks. Science, 296: 1302–1305.

Wegner, D. M. 1986. Transactive memory: A contemporaryanalysis of the group mind. In B. Mullen & G. R. Goeth-

als (Eds.), Theories of group behavior: 185–205. New York:Springer–Verlag.

Whittaker, S., Swanson, J., Kucan, J., & Sidner, C. 1997. Man-aging lightweight interactions in the desktop. ACMTransactions on Computer-Human Interaction, 4: 137–168.

Witt, L. A., & Burke, L. A. 2002. Moderators of the openness toexperience-performance relationship. Journal of Mana-gerial Psychology, 17: 712–721.

Wright, P. 1975. Consumer choice strategies: Simplifying vs.optimizing. Journal of Marketing Research, 12: 60–67.

Wright, P. M., Dunford, B. B., & Snell, S. A. 2001. Humanresources and the resource based view of the firm. Jour-nal of Management, 27: 701–721.

Xiao, Z., & Tsui, A. S. 2007. When brokers may not work: Thecultural contingency of social capital in Chinese high-tech firms. Administrative Science Quarterly, 52: 1–31.

Zucker, L. G., Darby, M. R., & Armstrong, J. 1998. Geographi-cally localized knowledge: Spillovers or markets? Eco-nomic Inquiry, 26: 65–86.

Zuckerman, H. 1977. Scientific elite: Nobel laureates in theUnited States. New York: Free Press.

Zuckerman, H., & Merton, R. K. 1972. Age, aging, and agestructure in science. In M. W. Riley, M. Johnson, & A.Foner (Eds.), Aging and society: A sociology of age strat-ification, vol. 3: 292–356. New York: Russell Sage Foun-dation.

James B. Oldroyd ([email protected]) is an assistant professor of managementand human resources at the Fisher College of Business, The Ohio State University. Heearned his Ph.D. at the Kellogg School of Management, Northwestern University. Hisresearch focuses on the performance effects of social capital in international organi-zations, including some of the darker sides of social capital, such as informationoverload, brokering negative affect, and social sanctions.

Shad S. Morris ([email protected]) is an assistant professor of management andhuman resources at the Fisher College of Business, The Ohio State University. Heearned his Ph.D. at Cornell University. He conducts research in the area of strategichuman resource management, particularly focusing on how international organiza-tions create value through human and social capital.


Copyright of Academy of Management Review is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

catching falling stars: a human resource...

Documents