not in the same boat: how status inconsistency affects ... · (d’aveni, 1996; trieschmann,...

r Academy of Management Journal2018, Vol. 61, No. 3, 1021–1049.https://doi.org/10.5465/amj.2015.0969

NOT IN THE SAME BOAT: HOW STATUS INCONSISTENCYAFFECTS RESEARCH PERFORMANCE IN

BUSINESS SCHOOLS

MICHAEL JENSENUniversity of Michigan

PENGFEI WANGBI Norwegian Business School

This study examines the consequences of status inconsistency for the performance ofmulti-unit organizations. Status inconsistency refers to the extent to which social actorsoccupy status positions accorded different amounts of social esteem and prestige indifferent social systems. Status inconsistency affects the performance of multi-unit or-ganizations by weakening the well-established positive effects of status on performancebecause it creates status ambiguity that makes external evaluations of the organizationand its individual units more difficult. Distinguishing between high-status and low-status multi-unit organizations, we argue further that status inconsistency is par-ticularly problematic for high-status multi-unit organizations, whereas low-statusmulti-unit organizations may actually benefit from status inconsistency. We test our ar-guments using a longitudinal sample comprised of 109 international business schools andtheir finance, accounting, marketing, management, and operations departments from 2002to 2013. Our study concludes with a discussion of the contributions our research makes tostatus theory and research, and the managerial relevance of our findings.

Status is often defined by the hierarchical positionsocial actors occupy in a particular social system(Gould, 2002; Jensen, Kim, & Kim, 2011), but mostsocial actors participate in multiple social systemssimultaneously. When social actors participate indifferent social systems, they may occupy unequalstatus positions in each social system, thus makingstatus inconsistency a concern for status theory(Benoit-Smullyan, 1944; Hughes, 1945). Status in-consistency shifts attention, accordingly, from howstatus affects social actors within a single socialsystem (Podolny, 1993) to the consequences of si-multaneously occupying unequal status positions indifferent social systems (Lenski, 1954). Status in-consistency poses a problem for social actors be-cause it creates ambiguity about their social identity,which raises questions about what to expect from

them and how to evaluate them (Smith, 2013; Zhang,2008). Despite the importance of status for organiza-tions (Jensen et al., 2011), status inconsistency inorganizations remains, with a few exceptions (Kang,2010; Zhao & Zhou, 2011), unexplored. The lack oforganizational status-inconsistency research is par-ticularly unfortunatewhen it comes to understandingthe importance of status for multi-unit organizations.Multi-unit organizations are both very common and,by virtue of operating in different industries or mar-kets simultaneously (Zuckerman, 1999), particularlysusceptible to status inconsistency, thus highlightingthe importance of extending status inconsistencytheory to multi-unit organizations.

Status inconsistency can have negative conse-quences for multi-unit organizations because occu-pying unequally ranked status positions in differentindustries or markets creates status ambiguity. Weargue specifically that status inconsistency affectsmulti-unit organizations primarily by weakening thecommonly observed positive effect of status on orga-nizational performance (Benjamin & Podolny, 1999;Malter, 2014; Pollock, Chen, Jackson, & Hambrick,2010). Focusing on how status inconsistencyweakensthe effect of status on organizational performance, weemphasize the interaction between status and status

We thank Bo Kyung Kim, Heeyon Kim, and seminarparticipants at the National University of Singapore,University of Lugano, University of Michigan, and theAcademy of Management conference for their helpfulcomments. Associate Editor Linus Dahlander and threeanonymous reviewers provided excellent guidancethroughout the review process. Please direct all corre-spondence to Michael Jensen.

1021

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

https://doi.org/10.5465/amj.2015.0969

inconsistency, rather than the main effect of statusinconsistency. We identify two complementarymechanisms that account for the negative inter-action between status, status inconsistency, andperformance. By creating status ambiguity, statusinconsistency makes status a less effective signal ofperceived quality (Podolny, 1993) and a less effectivetool for managing accountability pressures (Jensen,2006), both of which make status less useful forexternal audiences deciding with whom to formexchange relationships. Specifically, status incon-sistency creates ambiguity about the status of themulti-unit organization, which, in turn, reduces thevalue of the status of the multi-unit organizationand its units for each individual unit. We argue,moreover, that status ambiguity stemming fromstatus inconsistency is particularly problematic forhigh-status organizations and high-status units thattraditionally have benefitted most from their ad-vantageous status position.1

Our study makes several contributions to statusresearch. We contribute directly to status in-consistency research by shifting focus from single-unit organizations being ranked unequally ondifferent status dimensions to multi-unit organi-zations whose different units are ranked unequallyon the same status dimensions. Focusing onmulti-unit organizations and how status inconsistencyaffects the positive association between status andperformance contributes more broadly to statusresearch, as well by identifying an important lim-itation to status theory. Specifically, a core claim instatus theory is that status is most valuable whenproduct quality is uncertain because it allows de-cision makers to shift attention from unobservableproduct quality to observable status positions(Podolny, 1994). We add that even when status

positions are observable, status inconsistency cancreate uncertainty about how to interpret and assignmeaning to inconsistent status positions. Statusinconsistency, in other words, allows the problemof uncertainty about product quality to reappear inthe form of uncertainty about status position. Onepreviously unexplored implication for multi-unitorganizations is clear: “A potentially importantcondition for successfully occupying multiple po-sitions is that themultiple positions across differenthorizontal categories are within the same verticalstatus” (Jensen et al., 2011: 100). Our study begins toaddress this important aspect by theorizing andempirically testing how and when status inconsis-tency affects multi-unit organizations.

We also contribute to status theory by focusingon how involuntary associations affect status be-liefs. Social-networks-based research on status hasfocused mostly on voluntary relationships, suchas underwriting syndicates (Cowen, 2012; Jensen,2008; Podolny, 1993; Shipilov, Li, & Greve, 2011),strategic alliances (Ahuja, Polidoro, & Mitchell,2009; Gulati & Garguilo, 1999; Ozmel, Reuer, &Gulati, 2013), venture capital financing (Hsu, 2004;Petkova, Rindova, &Gupta, 2013; Pollock, Lee, Jin, &Lashley, 2015), and accounting firm certification(Greenwood & Suddaby, 2006; Jensen & Roy, 2008).Voluntary relationships provide the structural foun-dation of status hierarchies (Podolny, 1994) and af-fect how external audiences view a particular socialactor (Stuart, Hoang, & Hybels, 1999). We add thateven though involuntary associations may not im-ply as strong a form of interorganizational endorse-ments as voluntary associations, they still have thepotential to affect the status beliefs of external au-diences. By extending status theory to involuntaryassociations, our studynot onlyopensup for applyingstatus consideration to hitherto unexplored researchareas, such as status inconsistency as a source ofvertical diversification discount (Villalonga, 2004;Zuckerman, 2000), but also emphasizes the impor-tance of acknowledging that even involuntary asso-ciation never intended to have status consequencesmay nevertheless create status ambiguity amongexternal audiences.

We study the effect of status inconsistency onthe performance of multi-unit organizations in asample of 109 U.S. and non-U.S. business schoolsand their accounting, finance, management, mar-keting, and operations departments from 2002 to2013. Most business schools are multi-unit orga-nizations in which teaching and research activitiesare organized into separate academic departments

1 Status ambiguity is different from status noise orvagueness. Status ambiguity refers to situations inwhich thestatus of a multi-unit organization cannot be definitely re-solved but specific and distinct interpretations are possible,whereas status noise or vagueness precludes specific anddistinct interpretations. For example, a multi-unit organi-zation with a high- and a low-status unit could, dependingon the context, be equally meaningfully interpreted as ahigh- or a low-status multi-unit organization, whereas amulti-unit organizationwhoseunits operate inmarketswithpoorly defined status hierarchies could not be given a spe-cific and distinct interpretation. We focus on how statusambiguity induced by status inconsistency weakens the ef-fect of status, rather than the effect of status inconsistencyitself, because status inconsistency, compared to status, isa second-order concern for external audiences.

1022 JuneAcademy of Management Journal

(D’Aveni, 1996; Trieschmann, Dennis, Northcraft,& Nieme, 2000). We define business school per-formance in terms of research performance notbecause teaching performance is unimportant, butbecause research performance, as indicated bypublications in top academic journals, is more di-rectly comparable across business schools. More-over, because we focus on academic status, asindicated by the centrality of business schools anddepartments in PhD exchange networks (Bothner,Smith, & White, 2010; Burris, 2004), research per-formance is a particularly relevant performancemeasure. We begin our theory development by de-fining status inconsistency and theorizing whystatus inconsistency in general weakens the posi-tive association between status and organizationalperformance. To ground our hypothesis develop-ment in the business school context, we next dis-cuss why status is important for publication in topacademic journals and why status inconsistencyweakens the association between status and re-search performance. We conclude our study witha discussion of its contributions to status theory andresearch, as well as its managerial relevance.

STATUS INCONSISTENCY ANDORGANIZATIONAL PERFORMANCE

Status inconsistency refers to the extent to whichsocial actors occupy unequally ranked status po-sitions accorded different amounts of prestige indifferent social systems. Most research on statusinconsistency focuses on single-unit social actorsthat participate in social systems that are stratifiedaround different status dimensions.2 An individualoccupying unequally ranked educational and oc-cupational status positions (e.g., a PhD employedin a coffee shop) and a single-product organiza-tion occupying unequally ranked status positionsin different markets (e.g., a film studio that isacclaimed by film critics but unknown to regularaudiences) are examples of status-inconsistentsingle-unit social actors. Following Lenski (1954,

1956), most status inconsistency research has fo-cused on individuals and how status inconsistencyin, for example, income, occupation, and educationstatus hierarchies affects mental stress, social iso-lation, and political attitudes (see Stryker &Macke,1978; Whitt, 1983; Zhang, 2008). More recently,some research has examined the consequences ofstatus inconsistency for organizations and prod-ucts that have acquired multiple status indicators.Kang (2010) found that U.S. venture capital firmswith different achievement status (own achievedstatus) and reference status (peer comparison sta-tus) tend to deviate from modal investment strate-gies. In addition, Zhao and Zhou (2011) showedthat differences between wine critics and wineappellations in the evaluation of Californian pre-miumwines undermine the status claims of status-inconsistent wineries.

A different form of status inconsistency existswhen the units in a multi-unit social actor areranked unequally on the same status dimensionsby external audiences in their own social systems.A corporation with multiple product divisions anda university with multiple colleges, each of whichare ranked unequally in terms of perceived prod-uct and research quality, are examples of status-inconsistent multi-unit organizations. A vast amountof research has examined how horizontal diversifi-cation into different product categories affects theperformance of firms (see Benito-Osorio, Guerras-Martın, & Zuñiga-Vicente, 2012), and how bridgingdifferent horizontal product categories affects the per-formance of individual products (Hsu, 2006; Jensen,2010; Kim & Jensen, 2011; Smith, 2011). Despite thewidespread interest in horizontal product diversi-fication and product category combinations, the con-sequences of occupying unequally ranked statuspositions in different product categories for hori-zontally diversified organizations remain largelyunexplored. As noted above, however, successfulparticipation in different horizontal product cate-gories might require that organizations occupyequally ranked status positions in all the productcategories (Jensen et al., 2011). We shift attentionaccordingly from status-inconsistent single-unit or-ganizations to status-inconsistent multi-unit organi-zations to examine how status inconsistency affectsthe performance of each of the individual units ina multi-unit organization.

Status inconsistency in multi-unit organizationscan occur for different reasons. First, status in-consistency occurs when single- or multi-unit or-ganizations add newunits by diversifying into new

2 Kovacs and Liu (2016) used “status multiplicity” todescribe similar situations in which the status of a single-unit actor differs across different audiences.Weuse “statusinconsistency” because it implies unequally ranked statuspositions, whereas status multiplicity could imply beingranked similarly by different audiences (such as whena firm operating in three different markets is high status inall three markets). See also Smith (2013) on “evaluationdiscordance” in multiple audience contexts.

2018 1023Jensen and Wang

industries in which they come to occupy statuspositions that are different from the positions theyhold in their current industries. Commercial banksentering investment banking (Jensen, 2003) andaccounting firms entering management consulting(Greenwood & Suddaby, 2006) are examples ofdiversification that could result in status inconsis-tency. Second, status inconsistency emerges grad-ually in multi-unit organizations when the statusof their units changes at different rates becauseeach unit makes different investments in status-conferring activities. When the different businessunits in a corporation invest unequally in productquality, for example, or when the academic de-partments in a college invest unequally in research-productive faculty, status inconsistency emergesmore gradually. Finally, status inconsistency canhappen abruptly in a multi-unit organization if oneof its units experiences a sudden increase or de-crease in status due to positive or negative externalevaluations. If an academic department jumps instatus because one of its faculty wins a Nobel Prize,for example, or if a business unit is forced to recalla product due to real or perceived quality problems(Rhee & Haunschild, 2006), status inconsistencyhappens more abruptly.

Regardless of the origin of status inconsistency, itposes a potential problem for multi-unit organiza-tions because status provides a social identity thatcodifies what to expect from the occupants of thatstatus position (Jensen et al., 2011). When multi-unit organizations occupy inconsistent status po-sitions, their social identities are more difficult toascertain by external audiences. Identity ambiguityis a problem in horizontal diversification. Whenfirms diversify into different industries, for exam-ple, they make evaluations difficult for securityanalysts, which decreases analyst coverage and,therefore, the firms’ stock market price (Zuckerman,1999). Similar problems may occur when multi-unit organizations occupy different status positionsin different markets. A firm may leverage its statusin one market to build status in another (Jensen,2003), but persistent status inconsistency acrossbusiness units creates status ambiguity. We arguenext that the consequences of status inconsistencyfor multi-unit organizations and, more importantly,their individual units are likely negative, in partic-ular for high-status organizations and high-statusunits, because status ambiguity decreases the effec-tiveness of status as a signal of perceived quality andthe effectiveness of status in reducing accountabilitypressures.

STATUS INCONSISTENCY INBUSINESS SCHOOLS

Business schools are complex multi-unit organi-zations that serve a variety of audiences withdifferent products, such as student education, aca-demic research, and business consulting, each ofwhich may cause a particular business school to beranked differently compared to others (D’Aveni,1996). Most business schools are comprised of dif-ferent academic departments, such as accounting,finance, management, marketing, and operations,each of which may also be ranked unequally byits respective audience. Trieschmann et al. (2000)showed, for example, that different business schoolsnot only rank differently in education (MBA pro-grams) and research (scholarlypublications), but alsothat the different academic departments within aparticular business school often rank differently inresearch. A business schoolmay therefore experiencestatus inconsistency both because it ranks unequallyon different status dimensions (education and re-search) and because its departments rank unequallyon the same status dimension (research).We focus onthe latter form of status inconsistency and how statusinconsistency among academic departments withina business school affects their research performance.To theorize why status inconsistency may affect re-searchperformance, it is necessary to first understandwhy department and business school status are im-portant for research performance.

Status and Research Performance

Department status and business school status areimportant for research performance because recog-nition and resources accrue disproportionally atthe top of the status hierarchies. Merton (1968: 58)noted that academic recognition is shaped by theMatthew effect: “the accruing of greater incrementsof recognition for particular scientific contributionsto scientists of considerable repute and the with-holding of such recognition from scientists whohave not yet made their mark.” The Matthew effectis self-perpetuating: high-status academics getmorerecognition for their research, which increases theiraccess to resources that can be invested in new re-search, which gives them even more recognitionand resources (Bedeian, Cavazos, Hunt, & Jauch,2010; Burris, 2004; D’Aveni, 1996). Because theMatthew effect is rooted in the uncertainty associ-ated with evaluating research quality for externalaudiences (Pfeffer, 1993; Starbuck, 2009; Welch,


2014), we focus on how status affects externalevaluations of research quality, rather than internalresource allocations.3 Specifically, we examine theimplications of external evaluators, such as editorsand reviewers, that focus on evaluating the qualityof research, but may also be affected by social statusin their evaluations.

By functioning as a signal of perceived quality(Jensen, 2003; Podolny, 1993; Stuart et al., 1999),status provides unique information for editors andreviewers. To understand why status is an effectivesignal of perceived research quality, it is useful todecompose research quality. Ellison (2002) arguedthat academic papers vary on two different di-mensions,q and r, that reflect the importance of theirdifficult-to-revise main contribution (q) and theeasier-to-revise other aspects of quality, includingtheoretical framing, literature review, and empiri-cal analyses (r). Status is positively correlated withboth aspects of quality. Although important research(q-quality) happens in many different places, high-status business schools and high-status depart-ments are typically able to recruit the faculty behindimportant research from low-status schools and de-partments because they have greater access to re-sources, including financial and human capital, thatcan be used to revise (r-quality) manuscripts suc-cessfully. The status of a business school and adepartment cannot, however, be reduced to their fi-nancial and human capital because status tends to berobust to changes in financial and human capital, atleast in the short term. Moreover, status cannot beappropriated by individual faculty members but isavailable to everybody in the business school, thussignaling quality for everybody affiliated with thebusiness school and department, regardless of theirown human and financial capital.

Status not only functions as a signal of perceivedquality, but also reduces accountability pressures(Jensen, 2006). Accountability refers to the implicitor explicit expectation that social actors can becalled on at any time to explain and justify their de-cisions (Lerner & Tetlock, 1999; Scott & Lyman,1968). Editors are accountable for the academic in-tegrity and the status of their journals, as commonlyindicated by citation impact, inclusion in externalquality surveys, and centrality in cocitation networks(Baum, 2011; Onder & Tervio, 2015; Trieschmann

et al., 2000). Status reduces accountability pressuresbecause it is easier to defend having mistakenly ac-cepted a low-quality manuscript from a high-statusthan a low-status institution, and having mistakenlyrejected a high-quality manuscript from a low-statusthan a high-status institution. Because low-qualityresearch by high-status scientists diffuses faster thanlow-quality research by low-status scientists (Cole,1970; Judge, Cable, Colbert, & Rynes, 2007), mis-takenly accepting research from high-status institu-tions is less of an issue for editors who are worriedabout journal citations than are similarmistakeswithlow-status institutions. It is, in other words, moreeasily justifiable for editors to give, knowingly orunknowingly, the benefit of the doubt to facultyfrom high-status business schools and departmentsthan to faculty from low-status business schools anddepartments.4

In sum, status is important for research perfor-mance because status functions as a signal of per-ceived quality, thus helping editors make the righteditorial decisions, and because it reduces account-ability pressures, thus helping editors avoid beingblamed for the wrong editorial decisions. Becausebusiness school status and department status bothfunction as signals of perceived research quality andreduce accountability pressures, we expect positiveassociations between business school status anddepartment research performance and between de-partment status and department research perfor-mance. We argue next that status inconsistencyweakens the positive associations between businessschool status, department status, and departmentperformance bydecreasing the effectiveness of statusas a signal of perceived quality and the effectivenessof status in reducing accountability pressures.

Status Inconsistency and Research Performance

The effectiveness of a market signal depends onthe correlation between the observable signal and

3 We control statistically for resource allocations be-cause status is a source of power that affects access to re-search performance enhancing resources (Salancik &Pfeffer, 1974).

4 The idea that journal citations influence editors canappear offensive.We see no reason, however,why incentivesystems would not influence, knowingly or unknowingly,editors in the same way they influence other people. Weagree, therefore, with the editor of Administrative ScienceQuarterly that “because so much is at stake, the incentivesfor gaming the system are irresistible to many editors andauthors” (Davis, 2015: 182). See also Peters and Ceci (1982)on how editors and reviewers, who probably would havedenied being influenced by status, were clearly positivelyaffected by status in their editorial decision making.


unobservable quality (Connelly, Certo, Ireland, &Reutzel, 2011). Status inconsistency differentlyreduces the correlation between the observablestatus signal and unobservable quality for high-and low-status departments in a business school.Status inconsistency likely results in status defla-tion at the top of the hierarchy because it impliesnegative status transfers to low-status departments(Podolny, 1994), thus reducing the benefits of statusfor high-status departments. Status inconsistencylikely results in status inflation at the bottom of thehierarchy, on the other hand, because it implies pos-itive status transfers from high-status departments,thus increasing the benefits of status for low-statusdepartments. Alternatively, status inconsistency couldsimultaneously create status deflation at the top ofthe status hierarchy and status inflation at the bot-tom, thus diminishing the perceived differences be-tween high- and low-status departments. Whetherstatus inconsistency results in status deflation or statusinflation, the aggregate result, although not necessarilythe result for all departments, as argued below, is toweaken the positive association between status andperformance. Status inconsistency, in other words,creates doubt about the status of a business schoolthrough status ambiguity and, as a consequence, thestatus of its individual departments through status in-flation and deflation, which makes business schoolstatus and department status less useful signals of per-ceived quality.

The effectiveness of status as a mechanism to re-duce accountability depends on the extent to whichstatus identifies a clearly defined set of expectationson which to base external evaluations (Jensen et al.,2011). By creating status ambiguity, status incon-sistencymakes it difficult to evaluate research and tojustify these evaluations, which forces editors andreviewers to be more critical and form complex andintegrated views that are easier to defend (Tetlock,Skitka, & Boettger, 1989). Specifically, status in-consistency makes it easier for external audiencesand the editors themselves to question status-basedjustifications of editorial decisions. For high-statusdepartments, status inconsistency makes it harderjustify “accept mistakes” by invoking departmentstatus: status inconsistency should have deflateddepartment status. For low-status departments, sta-tus inconsistent makes it harder to justify “rejectmistakes” by invoking department status: status in-consistency should have inflated department status.As a result, status inconsistency diminishes theperceived differences between high- and low-statusdepartments, which makes status a less effective

mechanism for external evaluators to reduce ac-countability pressures. Putting the arguments to-gether, status inconsistency creates reasonabledoubt about the actual status of a business school andits different departments, as discussed above, thusensuring ex ante uncertainty about expected quality(signaling problem) and ex post uncertainty aboutdecision justifiability (accountability problem).

The status inconsistency argument suggests, inother words, that status inconsistency weakens thepositive association between status and researchperformance because it weakens the effectivenessof status as a signal of perceived quality and theeffectiveness of status in reducing accountabilitypressures. We therefore hypothesize that:

Hypothesis 1. Status inconsistency diminishes thepositive associations between (a) department sta-tus and department research performance and (b)business school status and department researchperformance.

The effects of status inconsistency may not be thesame for high- and low-status business schools.The advantages of status accrue disproportionally atthe top of the status hierarchy (Blau, 1994; Sørensen,1996), which implies that status is less important atthe bottom of the hierarchy. Because low-status ac-tors tend to benefit little from status hierarchy, theyhave less to lose from deviating from status-basednorms and expectations (Phillips & Zuckerman,2001). High-status actors, on the other hand, maydeviate in identity-confirming ways, such as en-couraging unconventional research, but not in anidentity-threatening way (Phillips & Zuckerman,2001). The implication is not that high-status orga-nizations always avoid identity-threatening affilia-tions with low-status organizations. High-statusorganizations may indeed value low-status affilia-tions if they can extract more resources from low-status organizations in their exchange relationships(Castellucci & Ertug, 2010; Cowen, 2012). The valuefor high-status organizations of extracting moreresources through low-status affiliations depends,however, on the affiliations being voluntary, withindependent organizations that bring external re-sources to the high-status organizations. When theaffiliations are involuntary affiliations within amulti-unit organization, the resource extraction ad-vantages of low-status affiliations are less relevant.However, the status threat to the high-status multi-unit organizations and their high-status units stillapplies because it reduces their ability to extractexternal resources.


In addition, status differences tend to diminishand become less salient at the bottom of the statushierarchy, which makes status a less useful mecha-nism to distinguish among low-status businessschools. Being in the first decile of a ranking is typi-cally seen as something very different from being inthe third decile, for example, whereas being in theseventh decile is hardly distinguishable from beingin the ninth decile. The differences between thedeciles of business schools and departments at thetop of the status hierarchy are simply more mean-ingful than the differences at the bottom of the hier-archy. Because status itself is more ambiguous at thebottom of the hierarchy, status inconsistency at thebottom of the hierarchy is less consequential than atthe top of the hierarchy. Status inconsistency is, inother words, more likely to pose a problem for high-status than low-status business schools.We thereforehypothesize that:

Hypothesis 2. Status inconsistency is more likely todiminish the positive associations between (a) de-partment status and department research perfor-mance and (b) business school status anddepartmentresearch performance for high-status businessschools than for low-status business schools.

The effects of status inconsistency for high- andlow-status departments aremore complex. For high-status departments, being in a status-inconsistentbusiness school implies the co-location of oneor more low-status departments, which may posea problem for the high-status departments becausetheir status may be deflated by the low-status de-partments. The mere association with low-statusdepartments, even involuntary associations, increasesthe risk of status deflation because external audiencestend to attributemeaning to simple spatial co-locations(Campbell, 1958; Kim & Jensen, 2011). Specifically,low-status departments may taint the high-status de-partments in the eyes of external audiences, whichmay result in the external audiences effectively de-flating the status of high-status departments in status-inconsistent business schools. Status inconsistencyreduces, in other words, the value of status forhigh-status departments because even involuntaryassociations with low-status departments can createambiguity about the actual status of the high-statusdepartments. For high-status departments, being instatus-consistent business schools implies the co-location of one or more other high-status depart-ments, thus avoiding the risk of status deflation andreducing status ambiguity. In addition, when thestatus of the business school and the status of its

departments are unambiguously high, the effective-ness of status in signaling perceived quality and re-ducing accountability pressures is likely strengthened(Zhao & Zhou, 2011).

For low-status departments, being in status-inconsistent business schools implies the co-location of one or more high-status departments.Unlike high-status departments, low-status de-partments benefit from status leakage to the extentthat being co-locatedwith high-status departmentsis viewed as an implicit endorsement of the low-status departments by the high-status departments(Stuart et al., 1999). Being co-located with high-status departments may, however, result in statusdisappointment if the audiences identify the low-status departments with their co-located high-status departments and inflate their expectationsto the low-status departments too much. The valueof high-status departments as endorsements de-pends ultimately on the low-status departmentsbeing able to actually meet the expectations ofhigh-status departments, which is more likelywhen the high-status departments associate vol-untarily with low-status departments after havingcarefully evaluated the low-status departments(Stuart et al., 1999). For low-status departments,being in status-consistent business schools impliesthe co-location of one or more other low-statusdepartments. When the status of the businessschool and the status of its departments areunambiguously low, it is difficult to furtherweaken their effectiveness in signaling perceivedquality and reducing accountability pressures,thus making status considerations generally lessrelevant.

In sum, because both status inflation and statusdisappointment affect low-status units, even if onedominates in a particular context, status inconsis-tency is less likely to reduce the benefits of businessschool and department status for low-status de-partments than for high-status departments.5 Wetherefore hypothesize that:

5 Our argument that status inconsistency is more prob-lematic for high-status departments is consistent with ouremphasis on external evaluations, not internal resourceallocations, in theorizing the effect of status inconsistencyon research performance. Had we emphasized internalresource allocations, the argument would instead havebeen that status inconsistency is more beneficial for high-status departments because status inconsistency givesthem a power advantage in securing resource allocations(Salancik & Pfeffer, 1974).


Hypothesis 3. Status inconsistency is more likely todiminish the positive associations between (a) de-partment status and department research perfor-mance and (b) business school status anddepartmentresearch performance for high-status departmentsthan for low-status departments.

Finally, status inconsistency is likely to weakenthe positive association between business schoolstatus and research performance more than itweakens the positive association between depart-ment status and research performance. Departmentstatus is a more proximate signal of quality andaccountability than business school status, whichmakes department status less susceptible to statusinconsistency than business school status. Prox-imity refers to the extent to which audiences arepart of a status-stratified social system and there-fore know who occupies what status position.Specifically, most academic journals specializeby academic field and most editors and reviewersare experienced scholars that know the statushierarchies of their own fields. Even if status in-consistency creates ambiguity about the statusof a business school, status inconsistency is lesslikely to affect how editors and reviewers viewdepartments in their own fields because they havemore detailed knowledge about the status hier-archies in their own fields. By distinguishingbetween the effects of status inconsistency ondepartment status and business school status, weargue implicitly that editors and reviewers do notknow and consider the status of all the departmentsin a business school equally well. We also narrowthe scope of our status inconsistency argumentby arguing that status inconsistency across de-partments affects proximate department statusconsiderations less when external evaluators areembedded in the same status hierarchies as thefocal department.

In sum, because status inconsistency is morelikely to create ambiguity about business schoolstatus than about department status, it ismore likelyto diminish the positive association between busi-ness school status and research performance thanthe positive association between department statusand research performance. We therefore hypothe-size that:

Hypothesis 4. Status inconsistency is more likely todiminish the positive association between businessschool status and department research performancethan the positive association between departmentstatus and department research performance.

METHODS

Business School Sample

The business school is a relatively young aca-demic institution. The first business school, EcoleSuperieure de Commerce de Paris, was founded in1819; the first U.S. business school, The WhartonSchool of Finance and Commerce, in 1881; andHarvard Business School offered the first Master ofBusiness Administration degree in 1908. Businessschools serve two purposes, teaching and research,whose relative importance has changed over time(Khurana & Spender, 2012; Simon, 1967). Whereasbusiness schools initially focused primarily onteaching and hardly paid attention to research, mid-century criticism of business school research forpoor quality led to an increased emphasis on re-search (Kaplan, 2014; Khurana, 2007). The leadingU.S. business schools responded aggressively toraise their academic status and shed the negativetrade-school image by requiring doctoral degreesfor tenured faculty appointments, rewarding pub-lications in peer-reviewed academic journals, andfavoring quantitative discipline-based researchmethods. To facilitate the transition to research-activeinstitutions, most business schools organized them-selves as independent schools or colleges andadopteda functional organizational structure with a deanoverseeing specialized academic departments suchas accounting, finance, management, marketing, andoperations in charge of teaching and research.

The business school has been a success: morethan 14,000 business schools enrolled thousands ofstudents in undergraduate, graduate, and executivedegree programs in 2014. We use a sample of 109research-active business schools drawn from theUniversity of Texas—Dallas (UTD) 2013 lists (theTop 100 North American Rankings and the Top 100Worldwide Rankings). We focus on research-activebusiness schools because the majority of the 14,000business schools are not involved in significant re-search activities but emphasize teaching only. Toensure comparability of business schools, our sam-ple is comprised mainly of U.S. business schools(86) and non-U.S. business schools from Canada(9), Europe (4), and Asia and Australia (9) thathave adopted the U.S. business school model andpublish in the top peer-reviewed management jour-nals (data availability, particularly in the Internetwebpage archive used to create the PhD exchangenetworks, eliminated some business schools on theUTD lists from our sample). We follow the 109business schools and their accounting, finance,


management, marketing, and operation departmentsfrom 2002 to 2013.6 Because business schools usedifferent names for their academic departments andsometimes group them differently, we combinedorganizational behavior, strategy, human resourcemanagement, entrepreneurship, and internationalbusiness into management, and management sci-ence, supply chain, logistics, and information sci-ence into operations. Appendix A lists the businessschools in our sample.

Dependent Variable

We define research performance in terms of aca-demic journal publications, and use theUTD researchcontribution score to measure journal publications.The annual UTD research contribution score countsand prorates, depending on the number of coauthors,the number of articles published in 24 leading aca-demic business journals.7 To measure departmentresearch performance in an efficient manner, weassigned the journals to individual departments;for example, The Accounting Review, the Journal ofAccounting and Economics, and the Journal of Ac-counting Review were assigned to the accounting de-partment (the journals, the UTD prorating method,and the journal assignmentsareprovided inAppendixB). This assignment approach could cause problemsbecause some journals could be assigned tomore thanone department since faculty from more departmentspublish regularly in these journals, such as operationsand management faculty publishing in ManagementScience, and some faculty could be in one departmentbut publish in journals assigned to other departments.To make sure our results were not affected by ambig-uous journal assignments, we switched journal as-signments and removed the most ambiguous journal(Management Science), finding that the results arerobust.Wealsohand-codeda subsampleof five top-20schools over three different years to match each

individual faculty to publications and departments,and found a reassuring 0.88 correlation between thedirect measure and the assigned measure of de-partment performance.

The UTD research contributions ranking uses amore comprehensive list of journals than the lists usedin other research on business school research perfor-mance (e.g., Trieschmann et al., 2000) but it never-theless ignores publications in lesser business journalsand publications in discipline-based journals. By ig-noring lesser business journals, the research perfor-mance of lower-status business schools is likelyunderestimated because faculty in these schools tendtopublish disproportionallymore in business journalsthat are unlikely to count in promotion and compen-sation decisions in higher-status business schools. Byignoring discipline-based journals, however, the re-search performance of higher-status business schoolsis likely to be underestimated because faculty in theseschools tend to publish disproportionally more indiscipline-based journals, such as psychology, soci-ology, and economics journals. Because our approachunderestimates the research performance of bothhigh- and low-status departments, we find it unlikelythe limited number of journals used by UTD favorseither high- or low-status departments.8 By focusingon department-level research performance, ratherthan individual-level research performance, our ap-proach is also unlikely to portray a highly produc-tive department as unproductive because departmentperformance depends on the performance of all thefaculties in a department, even if the performance ofa fewindividual facultymembers focusingexclusivelyon disciplinary journals may be misrepresented.9

Independent Variables

Status. The status of a business school refers to itsrank among other business schools. Focusing on re-search performance, the relevant ranking of business

6 We excluded economics departments because themajority of business schools in our sample do not haveeconomics departments and most business schools witheconomics departments have larger and academicallystronger economics departments outside the businessschool, including Stanford, MIT, and Chicago.

7 UTD prorates the number of publications by the num-ber of coauthors to avoid double-counting publications indepartments and business schoolswithmore collaborativeresearch and publication strategies. The results are robustto not prorating andusing the total number of publications.Weuse the annual proratednumber of publications but ourresults are robust to using a three-year moving average.

8 Although it is unlikely that our results are sub-stantively affected by the exclusion of disciplinary andlower-tier business journals, it is prudent to note this lim-itation of the UTD data.

9 We focus on departments rather than individuals bothbecause of our theoretical interest in multi-unit organiza-tions and because academic departments should ideallybe status inconsistent, comprised of individuals at dif-ferent stages of their careers, such as assistant professors,associate professors, and full professors, which makesdepartment-level status inconsistency less of a concernthan business-school-level status inconsistency.


schools for our study is based on the academic ac-complishments of the faculty and doctoral students atthe business schools. The academic status of a busi-ness school and its departments canbecapturedby theexchange (hiring and placement) of PhD students: theexchange of PhDs is “a symbolic act of mutual affir-mation” that confirms and reproduces status divisions(Burris, 2004: 244). Burris (2004) showed that cen-trality in the PhD exchange network explained 84%ofthevariance inexpert surveysofdepartmentalprestigein sociology (similarly in history andpolitical science)andClauset,Arbesman, andLarremore (2015) showedthat it is a better predictor of PhD placements thanaU.S.News&WorldReport rank in computer science,business, and history (see Amir and Knauff [2008] foreconomics). FollowingBurris (2004) andBothner et al.(2010), we create annual symmetric 109 3 109 PhDexchange matrices to calculate business school status,in which cells denote the total number of PhDs ex-changed between business schools i and j calculatedby adding the number of PhDs hired by businessschool i from business school j to the number hired bybusiness school j from business school i. We use sameapproach to calculate department status, except thatwe now use the department networks inwhich cells ijinclude the number of PhDs exchanged between de-partment z in school i and department z in school j. Toavoid random year-to-year fluctuations, we use five-year rolling hiring and placement data to create ourPhD exchange matrices: status in 2012 depends, inother words, on the hiring and placement of PhDs be-tween 2008 and 2012.

Following prior status research (Jensen, 2003;Podolny, 1993), we measure status using the standardnetwork centrality measure (Bonacich, 1987), accord-ing to which the status of a business school (academicdepartment) depends on the status of the schools(departments)withwhich it exchangesPhDstudents:10

sða,bÞ5a +‘

k5 0bk Rk1 1I

where a is a scaling factor that normalizes the mea-sure,b is aweighting factor (we followBenjamin andPodolny [1999] and set b at 0.995 of the reciprocal ofthe largest eigenvalue of R; see also Bonacich andRodan [2011] on choosing the b parameter) that re-flects how much the status of the focal school de-pends on the status of the schools with which itexchange PhD students, R is the 109 3 109 PhD ex-changematrix, and I is a columnvector of ones.11 The2012 status ranking of the 109 sampled businessschools is shown in Appendix A.

Status inconsistency. The status inconsistency ofa business school refers to the extent to which thestatus of the accounting, finance, management, mar-keting, and operations departments differ. We followLenski (1954) and measure status inconsistency as:

SI 5ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi+iðSi 2MÞ2

r

where Si is the status of department i and M is theaverage status of the different departments withina business school. Status inconsistency researchfaces the problem of separating the main effect ofstatus from the effect of status inconsistency, aproblem that is particularly severe in research fo-cusing on single-unit actors being ranked on dif-ferent status dimensions where the status of theactor is defined as the average of its status on each ofthe different dimensions (Jackson & Curtis, 1972;Treiman, 1966; Zhang, 2008). By focusing onmulti-department business schools, we partly avoid thisproblem because department status and status in-consistency are defined at two different measure-ment levels (department and business school),which ensures that status inconsistency is not alinear function of department status.

Figure 1 illustrates our status measures (businessschool status, department status, and status incon-sistency) for the top-10 business schools in 2012. Thesolid black line shows the business school status,the dashed gray line status inconsistency, and thestacked columns department status, calculated asdescribed above. Figure 1 shows, for example, thatHarvard Business School and University of Chicagohave relatively high status inconsistency due tothe dominance of their management and finance

10 Bonacich (1987: 1172–1173) noted that the s(a, b)measure is appropriate for symmetric and asymmetric re-lations but that s(a, b), according to the distinction betweencentrality and prestige in Knoke and Burt (1983), technicallymeasures prestige if the relations are asymmetric and cen-trality if they are symmetric.We followBurris (2004) andusesymmetric PhD exchange networks because, like Burris(2004: 252),wehaveno theoretical reasons to assigndifferentweights toplacingandhiringPhDs.Toensure that our resultsare robust, we follow Bothner et al. (2010: 976) and recalcu-late our status measure using the simpler eigenvector cen-trality measure (Bonacich, 1972). We also recalculate ourstatusmeasure using the s(a, b) measure on asymmetric PhDnetworks andpresent all the results inour robustness checks.

11 The correlation in the statusmeasure that results fromusing 0.995 and 0.75 is 0.99.


departments, whereas Stanford and Wharton haverelatively low status inconsistency.

Control Variables

We use a number of different control variables torule out alternative explanations for the relationshipbetween status, status inconsistency, and researchperformance. We control for prior research perfor-mance by including a one-year lagged dependentvariable (Prior Research Performance) because priorresearch performance is a strong predictor of bothstatus and future research performance and in-cluding a lagged dependent variable is a commonway of controlling for unobserved heterogeneity(Heckman & Borjas, 1980; Podolny, 1994).12 Wecontrol for department size by including the annualnumber of faculty in each department (Department

Size). Department size affects both the number ofarticles published and department status becausehigh-status departments tend to be favored whenopportunities to hirenew faculty are allocatedwithinbusiness schools. Specifically, research has shownthat the power of a department within a universityis highly correlated with national prestige, and thatthe power of a department increases the amount ofresources allocated to that department (Salancik &Pfeffer, 1974).13

We use a series of binary variables to control fordifferent qualitative aspects of a business school thatmay both be related to status and research perfor-mance, including the lack of a PhD program (No PhD

FIGURE 1Status Measures for Top 10 Business Schools in 2012

0

5

10

15

20

25

0

5

10

15

20

25

Stanford Wharton NYU HBS Chicago Northwestern MIT Columbia UC Berkley USC

Accounting Finance Operations Marketing Management School status Status inconsistency

Dep

artm

ent

Sta

tus

Sch

ool

Sta

tus

12 We also controlled for research performance in-consistency (main effects and interaction with status var-iables) and found that it was not significant and did notaffect our results.

13 The size of departments ranges from 0 to 121. Zeromeans that a department had no faculty in that year. In oursample, for example, Melbourne Business School had noor few faculty members in accounting for several years.Harvard Business School had over 100 faculty members inmanagement for many years. As noted earlier, however,the management department is a combination of organi-zational behavior, strategy, human resource management,entrepreneurship, and international business.


Program), whether the business school is part of aneliteuniversity (EliteUniversity) or apublicuniversity(Public University), and whether the business schoolisnamedafter adonor (DonorName).14Controlling forPhD program is important because having a PhDprogram is an indication that a business school ishighly committed to research and views the academicresearch community as an importance stakeholder (asopposed to only corporate recruiters of graduate andundergraduate students). Moreover, business schoolswithout PhD programs participate only partly in thestatus-granting PhD exchanges by virtue of hiring, butnot placing, PhD students. The size of the PhD pro-gram, as measured by the number of PhDs placed,provides an alternative indicator of status due to theskewed distribution of faculty production, with thetop 25% of schools producing almost 75% of faculty(see Clauset, Arbesman, & Larremore, 2015). The sizeof the PhDprogram is correlatedwith department sizeand research performance, however, thus suggestingthat we report, as a robustness check, the full model

with thepresenceof aPhDprogramsubstituted for thesize of the PhD program.

We control for the age of the business school(Business School Age), as well as for department (ac-counting as comparison group) and country (UnitedStates as comparison group). Some of the researchperformance observations (the dependent variable)weremissing, typicallybecauseof zeropublications inthe relevant journals, which led us to set the missingvariables to zero and enter controls for missing re-search performance and missing prior research per-formance (Missing Research Performance; MissingPriorResearchPerformance).15All themodels includefixed effects for year (not reported). Table 1 containssummary statistics and bivariate correlations.

Multicollinearity and Endogeneity

Multicollinearity and endogeneity are commonconcerns in nonexperimental research that couldinflate standard errors and bias coefficient estimates.Multicollinearity could be a concern in our studybecause department status, business school status,and status inconsistency are ultimately based on thesame PhD exchange networks, and are therefore

TABLE 1Descriptive Statistics and Bivariate Correlations (n 5 4,444)

Variables Mean SD Min. Max. 1 2 3 4 5

1. Research performance 1.66 1.96 0.00 15.412. Department Status 0.75 1.05 0.00 10.02 0.583. Business school (BS) status 4.05 4.40 0.02 23.30 0.59 0.784. BS status inconsistency 1.03 1.03 0.00 8.08 0.47 0.61 0.795. Prior research performance 1.61 1.93 0.00 15.41 0.71 0.59 0.60 0.476. Department size 22.28 14.58 0.00 121.00 0.36 0.39 0.30 0.27 0.357. No PhD program 0.05 0.22 0.00 1.00 20.10 20.12 20.16 20.14 20.108. Elite university 0.12 0.33 0.00 1.00 0.36 0.52 0.67 0.50 0.389. Public university 0.66 0.47 0.00 1.00 20.18 20.28 20.37 20.24 20.19

10. Business school donor name 0.62 0.49 0.00 1.00 0.16 0.11 0.15 0.02 0.1711. Business school age (ln) 4.18 0.50 1.61 4.89 0.13 0.16 0.21 0.15 0.1412. Finance 0.20 0.40 0.00 1.00 20.03 0.04 0.00 0.00 20.0213. Management 0.20 0.40 0.00 1.00 0.19 20.03 0.00 0.00 0.1614. Marketing 0.20 0.40 0.00 1.00 20.01 20.03 0.00 0.00 20.0115. Operations 0.20 0.40 0.00 1.00 0.02 0.02 0.00 0.00 0.0316. Canada 0.08 0.27 0.00 1.00 20.08 20.09 20.11 20.12 20.0817. Europe 0.04 0.20 0.00 1.00 0.02 0.03 0.05 0.14 0.0218. Asia & Australia 0.07 0.26 0.00 1.00 20.07 20.04 20.06 20.02 20.0719. Prior research performance missing 0.19 0.40 0.00 1.00 20.30 20.23 20.27 20.23 20.4120. Research performance missing 0.18 0.39 0.00 1.00 20.40 20.23 20.26 20.23 20.29

14 The following universities are defined as elite univer-sities: Columbia University, Cornell University, DartmouthCollege, Harvard University, Massachusetts Institute ofTechnology, Northwestern University, Stanford University,University of California at Berkeley, University of Californiaat Los Angeles, University of Chicago, University of Michi-gan at Ann Arbor, University of Pennsylvania, and YaleUniversity (Kacperczyk, 2013; Useem & Karabel, 1986).

15 Unnecessary controls may lead to estimate bias(Atinc, Simmering, & Kroll, 2012). Thus, in one of our ro-bustness checks, we removed the two dummyvariables formissing performance. The results were highly consistent.


relatively highly correlated. Endogeneity is almost al-ways a problem in status research because unobservedquality differences could account for both statusand research performance differences (Azoulay,Stuart, & Wang, 2014; Benjamin & Podolny, 1999;Podolny, 1994).

We address the multicollinearity concern byreporting results using (1) mean-centered statusvariables (Aiken & West, 1991; Guler & Guillen,2010), (2) orthogonalized main effects, and (3)orthogonalized main and interaction effects. First,mean-centering can facilitate the interpretationof coefficients on lower-order terms in the presenceof interactions, while results regarding interac-tion effects remain unchanged (Aguinis, Edwards, &Bradley, 2016). Second, following Ertug andCastellucci (2013), Hiatt, Sine, and Tolbert (2009),and Pollock and Rindova (2003), we orthogonalizeour status variables using the orthog command inStata 12 because of the high correlation betweenthese variables. By orthogonalizing first priorperformance, department status, business schoolstatus, and status inconsistency, and then also theinteractions between department status, businessschool status, and status inconsistency, we removethe effect of prior performance from the statusmeasures, the effect of department status on busi-ness school status, the effects of department statusand business school status on status inconsistency,

and the effects of the three main effects on the in-teraction effects.16 Although the magnitudes ofthe orthogonalized variables are hard to interpret

TABLE 1(Continued)

6 7 8 9 10 11 12 13 14 15 16 17 18 19

20.100.09 0.050.03 20.26 20.360.01 0.09 0.16 20.220.11 0.12 0.21 20.20 0.220.01 0.00 0.00 0.00 0.01 0.000.08 20.01 0.00 0.00 20.01 20.01 20.25

20.16 0.00 0.00 0.00 0.00 20.01 20.25 20.250.20 0.00 0.00 0.00 20.01 20.01 20.25 20.25 20.25

20.09 20.07 20.11 0.21 20.12 0.00 0.00 0.00 0.00 0.000.04 20.05 20.08 20.07 20.27 20.02 0.00 0.00 0.00 0.00 20.060.10 20.07 20.10 0.16 20.35 20.54 20.01 0.00 0.01 0.01 20.08 20.06

20.18 0.04 20.16 0.07 20.15 20.06 0.05 20.06 0.05 20.25 0.08 0.04 0.0320.17 0.03 20.15 0.07 20.12 20.04 0.05 20.06 0.04 20.24 0.07 0.03 0.01 0.39

Notes: Coefficients greater than .04 in absolute value are significant at p , .01.

16 Weorthogonalizedepartment-levelvariables first for thefollowing reasons. First, our unit of analysis is departmentandwemeasure performance accordingly at the departmentlevel. Second, we expect department status to have thestrongest effect on department performance. Third, editorsare closer to the department unit of analysis than to thebusiness school unit of analysis. Fourth, our status measureis rooted in departments because hiring decisions aredepartment-level decisions, which means that department-level networks aggregate to business-school-level networks,but not the other way around. Finally, if we orthogonalizeschool statusbeforedepartment status, the variance “shared”between the school and the departments would be “allo-cated” to the school. Becausebusiness schools resemble theirhigh-status departments more than their low-status de-partments (as shown in regression analyses, includingbusiness-school-level regressions not reported here), thevariance “left” for the departments will be the varianceunique to the low-status departments. Switching the or-thogonalization order should therefore switch the expectedsign of the department status 3 school status inconsistencyinteraction fromnegative to positive,which is similar to low-status departments in Model 18. Unreported reestimatedorthogonalized models with business school status orthogo-nalized before department status show that key interactionsindeed switch sign, thus confirming our approach to or-thogonalize department status before business school status.


(Cohen & Cohen, 1983) because of the way in whichorthogonalization transforms the original data, theuncertainty about the direction and statistical sig-nificance of the highly correlated variables is re-moved (Pollock & Rindova, 2003).

To address the endogeneity concern, it wouldbe ideal to use a randomized experiment by assign-ing the same papers to departments and businessschools with different levels of status and status in-consistency to see whether this affects how editorsevaluate papers. Unfortunately, it is not possible toconduct randomized experiments in our context, norcan we take advantage of idiosyncratic natural ex-periments, such as randomly missing author namespreviously used to identify status effects (Simcoe &Waguespack, 2011). We rely instead on a compre-hensive set of control variables, including prior re-search performance, department size, and presenceor size of PhDprogram, as discussed above, to reducethe likelihood that unobserved quality differencesbias our results.Moreover, we include prior researchperformance, the most direct measure of researchquality, in the set of hierarchically orthogonalizedvariables as the base variable, which means that thecommon variance of prior research performance,department status, business school status, and statusinconsistency is removed before regressing researchperformance on these variables. To the extent thatresearch (or faculty) quality is the common un-observed factor behind the effects of prior researchperformance and status on future research perfor-mance, orthogonalization effectively removes thiscommon unobserved factor, thus alleviating theendogeneity concern ( Denis & Sarin, 1999).

Statistical Analysis

Weuse amulti-levelmixed-effectsmodel to test ourhypotheses because academic departments are nestedwithin a business school and therefore are not in-dependent of each other (Rabe-Hesketh & Skrondal,2008). Specifically,wedefine a three-levelmodelwithrandom intercepts at the business school and aca-demic department levels and with country, depart-ment type, and year fixed effects (described above).The intraclass correlation tests (ICC) comparing thethree-level models with one-level ordinary linear re-gressionmodels are highly significant,which suggeststhat the three-level mixed-effects approach is appro-priate.We reestimated the three-levelmodels, treatingbusiness school and academic department as fixedeffects, and found similar results. To compare theoverall goodness of fit across models, we used the

Aikake Information Criterion (AIC) and the BayesianInformation Criterion (BIC), which allow for thecomparison of mixed models with different numbersof levels and predictors (Arregle, Miller, Hitt, &Beamish, 2013; Burnham & Anderson, 2004). In gen-eral, a decrease in AIC and BIC of more than 10 in-dicates a substantial improvement; a decrease ofmorethan 4 indicates a considerable improvement; anda decrease of less than 2 indicates no improvement inmodel fit (Burnham & Anderson, 2004: 270).

RESULTS

Table 2 presents the results of multi-level mixed-effects regression of departmental research perfor-mance. Models 1 to 5 contain mean-centered statusvariables; Model 6 and Model 7 contain orthogo-nalized main effects status variables; Model 8 con-tains orthogonalized main- and interaction effectsstatus variables; Model 9 contains orthogonalizedprior research performance variable and the main-and interaction effects status variables.

Model1containscontrol variablesonly. It shows, asexpected, that larger departments in elite universitieswith PhD programs that were more research pro-ductive in the past tend to also be more productive inthe future. The control variables also show that aca-demic departments differ in performance, with oper-ation departments being the most and accountingdepartments the least productive, which is probablya reflection of differences in research practices, withdepartments relying more on modeling research de-signs (operations and marketing) publishing morepapers (or differences in the journals covered by theUTD ranking system). There are no performance dif-ferences betweenU.S. and other schools, which is nota surprise because only highly research active non-U.S. schools are included in the sample. Finally, thecontrols for missing research performance observa-tions shows that missing research performance is as-sociatedwith lower researchperformancebutmissingprior research performancewith higher performance,which probably reflects simple mean reversion.

Model 2 includes the status variables. The effects ofdepartment status and business school status are con-sistently positive and significant, thus showing thatstatus is positively associated with department-levelresearch performance. The support for Hypothesis 1a,which suggested that status inconsistency is associatedwith a weakened effect of department status, is mixedin the unorthogonalized models. The department-status-by-status-inconsistency interaction variable isnegative, as suggested in Hypothesis 1a, but only


TABLE2

Multi-leve

lMixed

Effec

ts:D

epartm

entR

esea

rchPerform

ance

(n5

4,44

4;schoo

ls5

106)

Variables

Mod

el1

Mod

el2

Mod

el3

Mod

el4

Mod

el5

Mod

el6

Mod

el7

Mod

el8

Mod

el9

Sch

ools

tatus3

schoo

lstatus

inco

nsisten

cy20.02

**20.03

**20.04

**20.13

**20.13

**(0.00)

(0.01)

(0.01)

(0.03)

(0.02)

Dep

artm

ents

tatus3

schoo

lstatus

inco

nsisten

cy20.02

†0.03

20.03

**20.07

**20.06

**(0.01)

(0.02)

(0.01)

(0.02)

(0.02)

Sch

ools

tatusinco

nsisten

cy0.00

0.03

0.13

**0.13

**0.00

0.07

**20.02

20.02

(0.04)

(0.04)

(0.05)

(0.05)

(0.02)

(0.03)

(0.02)

(0.02)

Sch

ools

tatus

0.10

**0.10

**0.12

**0.13

**0.28

**0.30

**0.29

**0.28

**(0.02)

(0.02)

(0.02)

(0.02)

(0.04)

(0.04)

(0.04)

(0.03)

Dep

artm

ents

tatus

0.17

**0.21

**0.16

**0.11

**0.54

**0.52

**0.57

**0.42

**(0.04)

(0.05)

(0.04)

(0.05)

(0.05)

(0.05)

(0.05)

(0.03)

Prior

research

perform

ance

0.18

**0.19

**0.18

**0.18

**0.18

**0.19

**0.18

**0.18

**0.73

**(0.02)

(0.02)

(0.02)

(0.02)

(0.02)

(0.02)

(0.02)

(0.02)

(0.04)

Dep

artm

ents

ize

0.02

**0.02

**0.02

**0.01

**0.01

**0.02

**0.02

**0.01

**0.01

**(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

NoPhDprogram

20.81

**20.20

20.18

20.13

20.13

20.20

20.20

20.13

20.11

(0.31)

(0.21)

(0.21)

(0.20)

(0.20)

(0.21)

(0.21)

(0.20)

(0.20)

Elite

university

1.43

**0.30

†0.30

†0.29

†0.29

†0.30

†0.31

†0.28

†0.26

(0.23)

(0.18)

(0.18)

(0.17)

(0.17)

(0.18)

(0.18)

(0.17)

(0.17)

Public

university

20.23

0.04

0.04

0.03

0.03

0.04

0.03

0.04

0.03

(0.17)

(0.11)

(0.11)

(0.11)

(0.11)

(0.11)

(0.11)

(0.11)

(0.10)

Businessschoo

ldon

ornam

e0.12

0.15

†0.16

†0.20

**0.20

**0.15

†0.18

**0.20

**0.20

**(0.10)

(0.09)

(0.09)

(0.08)

(0.08)

(0.09)

(0.09)

(0.08)

(0.08)

Businessschoo

lage

(ln)

0.10

0.02

0.02

0.03

0.03

0.02

0.02

0.03

0.01

(0.15)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.09)

Finan

ce0.26

**0.25

**0.26

**0.26

**0.26

**0.25

**0.26

**0.26

**0.26

**(0.11)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

Operations

0.90

**0.91

**0.90

**0.92

**0.92

**0.91

**0.92

**0.92

**0.92

**(0.11)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

Marke

ting

0.41

**0.41

**0.41

**0.41

**0.42

**0.41

**0.41

**0.42

**0.41

**(0.11)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

Man

agem

ent

0.13

0.15

0.16

0.16

0.16

0.15

0.15

0.16

0.29

**(0.11)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

(0.10)

Can

ada

20.05

20.03

20.02

0.01

0.00

20.03

20.01

20.00

20.02

(0.26)

(0.17)

(0.17)

(0.16)

(0.16)

(0.17)

(0.17)

(0.16)

(0.16)


TABLE2

(Con

tinued

)

Variables

Mod

el1

Mod

el2

Mod

el3

Mod

el4

Mod

el5

Mod

el6

Mod

el7

Mod

el8

Mod

el9

Europe

0.26

0.10

0.09

0.03

0.03

0.10

0.09

0.04

0.05

(0.37)

(0.25)

(0.24)

(0.24)

(0.24)

(0.25)

(0.25)

(0.24)

(0.23)

Asia&Australia

20.16

20.19

20.19

20.16

20.16

20.19

20.18

20.16

20.18

(0.30)

(0.21)

(0.20)

(0.20)

(0.20)

(0.21)

(0.21)

(0.20)

(0.19)

Resea

rchperform

ance

missing

20.96

**20.96

**20.96

**20.95

**20.95

**20.96

**20.96

**20.95

**20.98

**(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

Prior

research

perform

ance

missing

0.11

**0.12

**0.12

**0.13

**0.13

**0.12

**0.12

**0.13

**0.12

**(0.06)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

(0.05)

Con

stan

t0.04

0.31

0.32

0.35

0.36

0.43

0.43

0.42

0.81

†

(0.64)

(0.46)

(0.45)

(0.44)

(0.44)

(0.46)

(0.46)

(0.44)

(0.43)

Interactiondifference

20.44

†20.10

**20.33

**20.32

**AIC

1434

214

252

1425

114

226

1422

514

252

1424

214

228

14,139

BIC

1453

414

463

1446

814

443

1444

914

463

1446

514

452

14,363

DAIC

290

21

226

227

210

224

DΒIC

271

5220

214

22

211

Notes:S

tandarderrors

inparen

theses;two-sided

tests.Mod

els6an

d7orthog

onalized

dep

artm

entstatus,schoo

lstatus,an

dstatusinco

nsisten

cy(SI);M

odel

8orthog

onalized

dep

artm

ents

tatus,schoo

lstatus,SI,an

dtheirinteractions;Mod

el9orthog

onalized

prior

perform

ance,d

epartm

ents

tatus,schoo

lstatus,SI,an

dtheirinteractions.

†p,

0.10

*p,

0.05

**p,

0.01


marginally significant inModel 3 (–0.02;p, 0.10) andnot significant (0.03; p . 0.10) in Model 5. Thedepartment-status-by-status-inconsistency interactionvariable is, however, negative and statistically signifi-cant in all the orthogonalized models (6–9), thus pro-viding more consistent support for Hypothesis 1awhen multicollinearity is removed. Model 4 includesthe interaction between business school status andstatus inconsistency and provides strong support forHypothesis 1b. The business-school-status-by-status-inconsistency interaction variable is negative and sig-nificant (–0.02; p, 0.001), thus suggesting that statusinconsistency diminishes the positive association be-tween business school status and department researchperformance. The support for Hypothesis 1b is con-firmed in the full Model 5 and all the orthogonalizedmodels.Thedecreases inAICandBICshowthatmodelfitness is also significantly improved.

Following Paolella and Durand (2016), we graphHypothesis 1b and 1b in Figure 2 based on the fullyorthogonalized Model 9 in Table 2. Figure 2A andFigure2Bshowthat themarginal effects ofdepartmentstatus and school status on department performancedecrease as business school status inconsistency in-creases, thus confirming a negative moderation effectof status inconsistency on both business school statusand department status. Figure 2A also shows that thepositive effect of department status becomes non-significant (p . 0.05) when status inconsistency ap-proaches four, which is approximately four standarddeviations above mean status inconsistency. In addi-tion, Figure 2B shows that the positive effect of schoolstatus becomes nonsignificant (p. 0.05) when statusinconsistency approaches two, which is approxi-mately two standard deviations above mean statusinconsistency. Finally, the negative effect of statusinconsistency on the positive association betweenstatus and research performance is substantively sig-nificant as well. A one standard deviation increase instatus inconsistency leads the effect of departmentstatus on research performance to decrease by 9%and a one standard deviation increase in status in-consistency would lead the effect of school status onresearch performance to decrease by almost 14%.17

We split the full (orthogonalized) sample and ana-lyze high- and low-status business schools and high-and low-status departments separately in Table 3.For comparison, Model 10 is identical to Model 9.Models 11 to 14 focus on the 50 highest- and 50lowest-status business schools and show that statusinconsistency weakens the positive effects of busi-ness school and department status on departmentresearch performance for high-status business schoolsbut not for low-status business schools. Hypothesis 2,which suggested that status inconsistency is morelikely to diminish the positive associations between(a) department status and department researchperformance and (b) business school status anddepartment research performance for high-statusschools than for low-status schools, is thereforesupported. Models 15 to 18 focus on the 50 highest-and 50 lowest-status academic departments andshow that status inconsistency decreases the posi-tive effects of business school and department statuson department research performance for high-statusdepartments but increases the positive effects forlow-status departments. Hypothesis 3, which sug-gested that status inconsistency is more likelyto diminish the positive associations between (a)department status and department research per-formance and (b) business school status and de-partment research performance for high-statusdepartments than for low-status departments, istherefore supported.

Although not an explicit test of the theoreticalmechanisms accounting for the status-inconsistencyeffects, the opposite status-inconsistency interactioneffects for high- and low-status departments inModel 16 and Model 18 suggest that the status in-consistency interaction effects are more likely causedby status deflation for high-status departments andstatus inflation for low-status departments, as ar-gued above, rather than resource concentration inhigh-status departments, which would have giventhe opposite results.

We graphHypothesis 2 in Figure 3 based onModel12 andModel 14. Figure 3A shows that the marginaleffect of department status on department perfor-mance for high-status business schools decreasesand becomes nonsignificant (p . 0.05) as status in-consistency increases above three (three standarddeviations above the mean). A one standard de-viation increase in status inconsistency leads theeffect of department status on research performanceto decrease by 11%. For low-status business schools,the marginal effects of department and businessschool status in Figure 3B and Figure 3D decrease as

17 Although we focus primarily on the research perfor-mance of individual departments, we repeat our analysesusing the research performance of business schools asthe dependent variable, conducting fixed-effects businessschool level. The results confirm our department-levelresults: status inconsistency diminishes the positive asso-ciation between business school status and businessschool research performance.


FIGURE 2Marginal Effects of Department Status and School Status on Department Performance

A. Marginal Effect of Department Status on Department PerformanceB. Marginal Effect of School Status on Department Performance

Mar

gin

al E

ffec

t of

Dep

artm

ent

Sta

tus

on D

epar

tmen

t P

erfo

rman

ceM

argi

nal

Eff

ect

of S

choo

l S

tatu

s on

Dep

artm

ent

Per

form

ance

A

B

–5 –4 –3 –2 –1 0 1 2 3 4 5 6

–5 –4 –3 –2 –1 0 1 2 3 4 5 6Status Inconsistency

Status Inconsistency

0

–0.5

–1

0.5

1

1.5

1.2

1

0.8

0.6

0.4

0.2

0

–0.2

–0.4

Notes: Bars indicate 95% confidence interval. Status inconsistency has negative values because it is orthogonalized.


TABLE3

Multi-leve

lMixed

Effec

ts:D

epartm

entR

esea

rchPerform

ance

(Subsam

ple)

BusinessSch

ool

Aca

dem

icDep

artm

ent

All

Top

50Bottom

50Top

50Bottom

50

Variables

Mod

el10

Mod

el11

Mod

el12

Mod

el13

Mod

el14

Mod

el15

Mod

el16

Mod

el17

Mod

el18

Sch

ools

tatus3

schoo

lstatusinco

nsisten

cy20.13

**20.19

**20.01

20.16

**0.35

**(0.02)

(0.04)

(0.17)

(0.04)

(0.08)

Dep

artm

ents

tatus3

schoo

lstatusinco

nsisten

cy20.06

**20.10

**20.15

20.09

**0.42

**(0.02)

(0.04)

(0.15)

(0.03)

(0.18)

Sch

ools

tatusinco

nsisten

cy20.02

0.02

0.02

20.05

20.13

20.03

20.02

0.01

0.40

**(0.02)

(0.03)

(0.03)

(0.04)

(0.16)

(0.03)

(0.03)

(0.04)

(0.12)

Sch

ools

tatus

0.28

**0.21

**0.29

**0.13

**0.12

0.21

**0.27

**0.44

**0.74

**(0.03)

(0.05)

(0.06)

(0.07)

(0.21)

(0.05)

(0.05)

(0.05)

(0.09)

Dep

artm

ents

tatus

0.42

**0.36

**0.43

**0.22

**0.07

0.39

**0.44

**0.47

**0.93

**(0.03)

(0.05)

(0.05)

(0.07)

(0.19)

(0.05)

(0.05)

(0.06)

(0.20)

Prior

research

perform

ance

0.73

**0.66

**0.76

**0.36

**0.28

†0.63

**0.70

**0.74

**0.99

**(0.04)

(0.06)

(0.06)

(0.07)

(0.15)

(0.05)

(0.06)

(0.06)

(0.11)

Dep

artm

ents

ize

0.01

**0.02

**0.02

**0.01

**0.01

**0.02

**0.02

**0.00

0.00

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

NoPhDprogram

20.11

20.48

20.27

20.06

20.06

20.58

20.43

0.04

20.01

(0.20)

(0.60)

(0.56)

(0.12)

(0.12)

(0.43)

(0.43)

(0.11)

(0.13)

Elite

university

0.26

0.13

0.04

0.52

**0.51

**0.22

0.14

0.39

**0.47

**(0.17)

(0.25)

(0.23)

(0.23)

(0.23)

(0.23)

(0.23)

(0.19)

(0.21)

Public

university

0.03

0.03

0.01

0.10

0.09

20.04

20.03

0.16

**0.17

**(0.10)

(0.20)

(0.19)

(0.09)

(0.09)

(0.18)

(0.17)

(0.08)

(0.08)

Businessschoo

ldon

ornam

e0.20

**0.22

0.39

**0.11

†0.12

†0.24

0.37

**0.08

0.08

(0.08)

(0.16)

(0.16)

(0.06)

(0.06)

(0.14)

(0.15)

(0.06)

(0.07)

Businessschoo

lage

(ln)

0.01

20.09

20.05

20.03

20.04

0.01

0.03

0.03

0.05

(0.09)

(0.25)

(0.24)

(0.06)

(0.06)

(0.19)

(0.19)

(0.06)

(0.07)

Finan

ce0.26

**0.46

**0.48

**0.04

0.04

0.41

**0.43

**0.09

0.09

(0.10)

(0.18)

(0.19)

(0.08)

(0.08)

(0.18)

(0.18)

(0.08)

(0.08)

Operations

0.92

**1.58

**1.63

**0.31

**0.31

**1.39

**1.47

**0.32

**0.33

**(0.10)

(0.19)

(0.19)

(0.08)

(0.08)

(0.18)

(0.19)

(0.08)

(0.08)

Marke

ting

0.41

**0.68

**0.70

**0.09

0.09

0.71

**0.73

**0.11

0.11

(0.10)

(0.18)

(0.18)

(0.08)

(0.08)

(0.18)

(0.18)

(0.08)

(0.08)

Man

agem

ent

0.29

**0.38

**0.40

**0.18

**0.18

**0.36

†0.40

**0.16

**0.17

**(0.10)

(0.19)

(0.19)

(0.08)

(0.08)

(0.18)

(0.18)

(0.08)

(0.08)

Can

ada

20.02

20.33

20.25

20.05

20.05

20.21

20.14

0.03

0.00

(0.16)

(0.32)

(0.30)

(0.11)

(0.11)

(0.28)

(0.28)

(0.10)

(0.11)


TABLE3

(Con

tinued

)

BusinessSch

ool

Aca

dem

icDep

artm

ent

All

Top

50Bottom

50Top

50Bottom

50

Variables

Mod

el10

Mod

el11

Mod

el12

Mod

el13

Mod

el14

Mod

el15

Mod

el16

Mod

el17

Mod

el18

Europe

0.05

20.14

20.05

0.34

0.30

20.32

†20.41

**(0.23)

(0.33)

(0.31)

(0.36)

(0.35)

(0.19)

(0.20)

Asia&Australia

20.18

20.41

20.29

20.06

20.06

20.32

20.17

0.15

0.10

(0.19)

(0.49)

(0.46)

(0.13)

(0.13)

(0.36)

(0.35)

(0.13)

(0.14)

Resea

rchperform

ance

missing

20.98

**21.22

**21.20

**20.91

**20.91

**21.29

**21.26

**20.90

**20.91

**(0.05)

(0.13)

(0.12)

(0.04)

(0.04)

(0.11)

(0.11)

(0.04)

(0.04)

Prior

research

perform

ance

missing

0.12

**0.19

0.19

0.00

0.00

0.11

0.13

0.05

0.03

(0.05)

(0.12)

(0.12)

(0.04)

(0.04)

(0.11)

(0.11)

(0.05)

(0.05)

Con

stan

t0.81

†1.05

0.70

1.01

**0.99

**0.62

0.44

1.09

**1.16

**(0.43)

(1.06)

(1.00)

(0.29)

(0.30)

(0.82)

(0.81)

(0.30)

(0.33)

Interactiondifference

20.32

**20.41

**2.12

20.40

**0.01

AIC

14,139

7,79

77,77

74,32

64,32

97,64

87,63

04,67

54,65

6BIC

14,363

7,98

57,97

44,50

54,51

97,83

57,82

84,85

94,85

1Observa

tion

s4,44

42,11

82,11

81,97

41,97

42,12

62,12

61,97

61,97

6Numbe

rof

grou

ps

106

4949

4949

7979

8282

Notes:S

tandarderrors

inparen

theses;t-sided

tests.

†p,

0.10

*p,

0.05

**p,

0.01


status inconsistency increases, but both fail tobecome significant. Figure 3C shows that the posi-tive effect of business school status becomes non-significant (p . 0.05) as status inconsistencyincreases above two (two standard deviations abovethe mean). A one standard deviation increase instatus inconsistency decreases the effect of schoolstatus on research performance by 15%. Finally, wegraph Hypothesis 3 in Figure 4 based on Model 16and Model 18. Figure 4A and Figure 4C show thatthe marginal effects of department status and busi-ness school status decrease as status inconsis-tency increases for high-status departments, witheffect sizes being similar to those in Figure 3A andFigure 3C. Figures 4B and 4D show, in contrast, that

the marginal effects of department status and schoolstatus increase as status inconsistency increases forlow-status departments. Specifically, a one standarddeviation increase in status inconsistency increasesthe effects of department status and business schoolstatus on research performance by roughly 8%.

Hypothesis 4 suggested that status inconsistencyis more likely associated with a weakened posi-tive effect of business school status on departmentresearch performance than a weakened positive ef-fect of department status. To test Hypothesis 4, wedivide the interaction effects by the main effects(b(department status by status inconsistency) / b(department status)

and b(school status by status inconsistency) / b (school status))and test whether the difference between them is

FIGURE 3Marginal Effects of Status on Department Research Performance: High- and Low-status Business Schools

–1

–0.5

0

0.5

1

1.5

–1

–1.5

–0.5

0

0.5

1

1.5

2

–5 –4 –3 –2 –1 0 1 2 3 4 5 6 –1 –0.5 0 0.5 1 1.5 2 2.5 3

–2

–1.5

–1

–0.5

0

0.5

1

–2

–1.5

–1

–0.5

0

0.5

1

1.5

Mar

gin

al E

ffec

t of

Dep

artm

ent

Sta

tus

on D

epar

tmen

t P

erfo

rman

ce

Mar

gin

al E

ffec

t of

Dep

artm

ent

Sta

tus

on D

epar

tmen

t P

erfo

rman

ce

–5 –4 –3 –2 –1 0 1 2 3 4 5 6Status Inconsistency

Status Inconsistency Status Inconsistency

–1 –0.5 0 0.5 1 1.5 2 2.5 3Status Inconsistency

Mar

gin

al E

ffec

t of

Sch

ool

Sta

tus

on

Dep

artm

ent

Per

form

ance

Mar

gin

al E

ffec

t of

Sch

ool

Sta

tus

on

Dep

artm

ent

Per

form

ance

A B

DC



statistically significant.We test thedifferencebetweenb(department status by status inconsistency) / b(department status)

and b(school status by status inconsistency) / b (school status)

rather than the difference between b(department status by

status inconsistency) and b(school status by status inconsistency)

because the importance of the status-inconsistencyinteraction effects depends not only on the size of theinteraction effects themselves, b(department status by status

inconsistency) andb(school status by status inconsistency), but alsoon the size of the main effects, b(department status) andb(department status). The size of the main effects needs tobe consideredbecause a large interactioneffect doesnotnecessarily mean that the interaction effect is sub-stantively important. If the main effect of businessschool status and department status is relatively large

compared to thestatus inconsistency interactioneffects,the status inconsistency interaction effects wouldhardly change the overall effect of business school anddepartment status.

Specifically, we used thenlcom command in Stata 13to testwhether themain-effects-scaled interaction effectsare significantly different from each other by testingwhether (b(school status by status inconsistency) / b (school status)) –(b(department status by status inconsistency) / b(department status))is different from zero. The significant InteractionDifference scores in Table 2 support Hypothesis 4:proportional to the main effects of business schoolstatus and department status, status inconsistency islikely to diminish the positive association betweenbusiness schools status on department research

FIGURE 4Marginal Effects of Status on Department Research Performance: High- and Low-status Academic Departments

–0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

–1.5

–1

–0.5

0

0.5

1

1.5

2

–0.5

0

0.5

1

1.5

2

2.5

3

3.5

–1

0

1

2

3

4

5

–5 –4 –3 –2 –1 0 1 2 3 4 5 6 –2 –1.5 –1 –0.5 0 0.5 1 1.5 2 2.5 3 43.5

–2 –1.5 –1 –0.5 0 0.5 1 1.5 2 2.5 3 43.5–5 –4 –3 –2 –1 0 1 2 3 4 5 6

Mar

gin

al E

ffec

t of

Dep

artm

ent

Sta

tus

on D

epar

tmen

t P

erfo

rman

ce


Status Inconsistency Status Inconsistency


Mar

gin

al E

ffec

t of

Sch

ool

Sta

tus

on

Dep

artm

ent

Per

form

ance

Mar

gin

al E

ffec

t of

Sch

ool

Sta

tus

on

Dep

artm

ent

Per

form

ance

A B

DC

Mar

gin

al E

ffec

t of

Dep

artm

ent

Sta

tus

on D

epar

tmen

t P

erfo

rman

ce



performance significantly more than it diminishesthe positive association between department statusand department research performance. Finally, thesignificant Interaction Difference scores inModel 12and Model 16 in Table 3 provide additional supportfor Hypothesis 4: status inconsistency is likely todiminish the positive association between businessschool status and department research performancesignificantly more than it diminishes the positiveassociation between department status and depart-ment research performance for high-status businessschools and high-status departments.18

DISCUSSION AND CONCLUSION

We argued in this study that status inconsistencyin multi-unit organizations diminishes the posi-tive association between status and performancebecause it weakens the effectiveness of status as asignal of perceived quality and as an accountability-reduction mechanism. Using business schools andtheir accounting, finance, management, marketing,and operations departments as the empirical setting,we argued that the positive association betweendepartment- and school-level status and department-level research performance depends on school-levelstatus consistency. Our statistical analyses providestrong support for our core arguments. Departmentstatus and business school status are positively asso-ciated with department- and school-level researchperformance, but status inconsistency diminishes thepositive associations between department status andschool status and department (and school) researchperformance. Status inconsistency is particularlyproblematic at the top of the status hierarchy: Statusinconsistency weakens the positive associations forhigh-status business schools and high-status de-partments, whereas it strengthens the positive associ-ations for low-status departments. Finally, statusinconsistency diminishes the positive association

between business school status and research per-formance more than it diminishes the positive as-sociation between department status and researchperformance.

In addition to the specific contributions discussedin the introduction, our study contributes morebroadly to status theory as well. By showing thatstatus inconsistency weakens the positive associa-tion between status and performance, we identifyan important constraint on self-perpetuating statusadvantages, thus providing a theoretical limit tothe cumulative advantage of status. Other limits tothe self-perpetuation of status advantages exist, suchas, for example, status homophily, status discrimi-nation, status deprivation, and status disruption,which increase the negative consequences of posi-tive status shifts (Jensen, 2008; Jensen & Kim, 2015;Podolny, 1994). Status inconsistency is particu-larly relevant for multi-unit organizations becauseit emphasizes the importance of coordinating status-influencing resource allocations among seeminglyindependent units within the same multi-unit orga-nization. Specifically, although status simplifiesexternal evaluations by functioning as a signal ofproduct quality that reduces uncertainty aboutproduct quality (Jensen &Roy, 2008; Podolny, 1993),status can also, as briefly discussed in the in-troduction, complicate external evaluations if statusinconsistency creates ambiguity about the actualstatus of a multi-unit organization. Indeed, rethink-ing status inconsistency as vertical unrelatedness,our study suggests that status inconsistency shouldbe integrated into product diversification theory tocomplement horizontal relatedness as a potentialconstraint on beneficial horizontal diversification(Markides, 1995; Zuckerman, 2000).

Our study also has important managerial impli-cations. The implications for business schoolsare straightforward. Whereas other research hasdocumented the negative effects on departmentresearch performance of an intradepartmentallow-productivity culture (Kim, Morse, & Zingales,2009), we document the negative effects of inter-departmental status inconsistency on researchperformance for high-status business schools. Sta-tus inconsistency in high-status business schoolstypically manifests in a particular form of statusinequality where a few departments rank signifi-cantly lower than the other highly ranked de-partments. Rather than increasing the status of thehigh-status departments even more, which canbe the consequence of winner-take-all incentivesystems such as, for example, allocating extra

18 We conduct a series of robustness checks. We sub-stitute the presence of PhD program for the size ofPhD program; use the eigenvector centrality measure(Bonacich, 1972) for status instead of the s(a, b) measure(Bonacich, 1987); use the s(a, b) measure (Bonacich, 1987)on asymmetric PhD placement networks instead of sym-metric PhD networks; reestimate Model 9 using U.S.schools only; exclude observations with zero publica-tion; use the total (not-prorated) number of publicationsas dependent variable; and employ a three-year movingwindow for research performance. All the robustnesschecks support our main results.


resources based on PhD student placements, ourstudy suggests that allocating extra resources tolower-status departments to move them up the sta-tus hierarchy could increase the research perfor-mance of all the departments in the business school.We obviously do not suggest that resources neces-sary for high-status departments to remain com-petitive relative to high-status departments in otherbusiness schools within their markets should bereallocated to low-status departments; only that itmight be more productive for everybody to useextra resources to increase the status of low-statusdepartments. The importance of status consis-tency generalizes straightforwardly to other types ofmulti-unit organizations that depend on externalaudiences.

Our study is not without limitations. We do notfully explore the scope conditions of our theoreticalarguments. Specifically, status inconsistency re-mains an issue as long as the units of a multi-unitorganization and their audiences are not com-pletely isolated from each other and have abso-lutely no knowledge about each other. The businessschool context meets this boundary condition be-cause most editors are senior faculty in research-active business schools that, over the years, havetypically acquired detailed knowledge about thestatus hierarchies in their own fields, some knowl-edge about the status hierarchies in neighboringfields, and rudimentary knowledge about the statushierarchies in other fields through their work on ten-ure committees and other interdepartmental func-tions.19 Status inconsistency is nevertheless moreimportant in contexts in which the audiences forthe different units are the same or pay closer atten-tion to all the units. The stock market is a contextwith likely stronger status inconsistency effects, forexample, because investors and analysts carefullyattend to all the business units of public firms andtheir market (Zuckerman, 1999). Our study provides,therefore, a conservative test of status inconsistencyand its effects on the positive association betweenstatus and performance, but it would nevertheless beuseful to empirically explore the scope conditionsmore comprehensively.

Finally,we cannot determinewhich one of the twotheoretical mechanisms, signaling quality and ac-countability reduction, behind the status inconsis-tency effects matter most. We are, however, lessconcerned with our inability to determine the rel-ative importance of the theoretical mechanismsbecause the primary purpose of our study is to es-tablish that status inconsistency matters for multi-unit organizations. Moreover, we focus exclusivelyon how external audiences respond to status in-consistency andnot onhow internal audiences, suchas the departments themselves or the businessschool leadership, respond. We can think of a num-ber of different types of responses, ranging frominterdepartmental power struggles to leadership-mandated closing-down of low-status departments,thatwouldmerit attention in future research.Despitethe aforementioned limitations, our study docu-ments the importance of status inconsistency formulti-unit organizations and draws attention to sta-tus inconsistency as an important constraint on self-perpetuating status advantages and an importantconstraint on horizontal diversification.

REFERENCES

Aguinis,H., Edwards, J. R., &Bradley,K. J. 2017. Improvingour understanding of moderation and mediation instrategic management research. Organizational Re-search Methods, 20: 665–685.

Ahuja, G., Polidoro, F., & Mitchell, W. 2009. Structuralhomophily or social asymmetry? The formation ofalliances by poorly embedded firms. Strategic Man-agement Journal, 30: 941–958.

Aiken, L. S., & West, S. G. 1991. Multiple regression:Testing and interpreting interactions. NewburyPark, CA: Sage.

Amir, R., & Knauff, M. 2008. Ranking economics de-partments worldwide on the basis of PhD placement.The Review of Economics and Statistics, 90: 185–190.

Arregle, J. L., Miller, T. L., Hitt, M. A., & Beamish, P. W.2013. Do regions matter? An integrated institutionaland semiglobalization perspective on the interna-tionalization of MNEs. Strategic Management Jour-nal, 34: 910–934.

Atinc, G., Simmering, M. J., & Kroll, M. J. 2012. Controlvariable use and reporting in macro and micro man-agement research.Organizational Research Methods,15: 57–74.

Azoulay, P., Stuart, T., &Wang, Y. 2014.Matthew: Effect orfable? Management Science, 60: 92–109.

Baum, J. A. 2011. European and North American ap-proaches to organizations and strategy research: An

19 Goldstein and Gigerenzer (2002) showed that com-plete knowledge is not necessary to generate accuraterankings when mere recognition is correlated with theranking criterion, which is typically the case in social do-mains because recognition is often correlated with keystatus dimensions such as quality, resources, wealth, andpower (see also Gigerenzer & Goldstein, 1996).


Atlantic divide? Not. Organization Science, 22:1663–1679.

Bedeian, A. G., Cavazos, D. E., Hunt, J. G., & Jauch, L. R.2010. Doctoral degree prestige and the academicmarketplace: A study of career mobility within themanagement discipline. Academy of ManagementLearning & Education, 9: 11–25.

Benito-Osorio, D., Guerras-Martın, L. A., & Zuñiga-Vicente,J. A. 2012. Four decades of research on product di-versification: A literature review. Management De-cision, 50: 325–344.

Benjamin, B. A., & Podolny, J.M. 1999. Status, quality, andsocial order in the California wine industry. Admin-istrative Science Quarterly, 44: 563–589.

Benoit-Smullyan, E. 1944. Status, status types, and statusinterrelations. American Sociological Review, 9:151–161.

Blau, P. M. 1994. Structural context of opportunities.Chicago, IL: University of Chicago Press.

Bonacich, P. 1972. Factoring and weighting approaches tostatus scores and clique identification. The Journal ofMathematical Sociology, 2: 113–120.

Bonacich, P. 1987. Power and centrality: A family of mea-sures.American Journal of Sociology, 92: 1170–1182.

Bonacich, P., & Rodan, S. 2011. Comment & response onchoosing the ‘b’. Journal of Social Structure: JOSS,12: 1–4.

Bothner,M.S., Smith, E.B., &White,H.C. 2010.Amodel ofrobust positions in social networks. American Jour-nal of Sociology, 116: 943–992.

Burnham, K. P., & Anderson, D. R. 2004. Multimodel in-ference understanding AIC and BIC inmodel selection.Sociological Methods & Research, 33: 261–304.

Burris, V. 2004. The academic caste system: Prestige hier-archies in PhD exchange networks. American Socio-logical Review, 69: 239–264.

Campbell, D. T. 1958. Common fate, similarity, and otherindices of the status of aggregates of person as socialentities. Behavioral Science, 3: 14–25.

Castellucci, F., & Ertug, G. 2010. What’s in it for them? Ad-vantages of higher-status partners in exchange relation-ships.Academy of Management Journal, 53: 149–166.

Clauset, A., Arbesman, S., & Larremore, D. B. 2015. Sys-tematic inequality and hierarchy in faculty hiringnetworks. Science Advances, 1: e1400005.

Cohen, J., & Cohen, P. 1983.Applied multiple regression/correlation analysis for the behavioral sciences.Hillsdale, NJ: Erlbaum.

Cole, S. 1970. Professional standing and the reception ofscientific discoveries. American Journal of Sociol-ogy, 76: 286–306.

Connelly, B. L., Certo, S. T., Ireland, R. D., & Reutzel, C. R.2011. Signaling theory: A review and assessment.Journal of Management, 37: 39–67.

Cowen, A. P. 2012. An expanded model of status dynam-ics: The effects of status transfer and interfirm co-ordination. Academy of Management Journal, 55:1169–1186.

D’Aveni, R.A. 1996.Amultiple-constituency, status-basedapproach to interorganizational mobility of facultyand input-output competition among top businessschools. Organization Science, 7: 166–189.

Davis, G. F. 2015. What is organizational research for?Administrative Science Quarterly, 60: 179–188.

Denis, D. J., & Sarin, A. 1999. Ownership and boardstructures in publicly traded corporations. Journal ofFinancial Economics, 52: 187–223.

Ellison, G. 2002. The slowdown of the economics pub-lishing process. Journal of Political Economy, 110:947–993.

Ertug, G., & Castellucci, F. 2013. Getting what you need:How reputation and status affect team performance,hiring, and salaries in the NBA. Academy of Man-agement Journal, 56: 407–431.

Gigerenzer, G., & Goldstein, D. G. 1996. Reasoning the fastand frugal way: Models of bounded rationality. Psy-chological Review, 103: 650–669.

Goldstein, D. G., & Gigerenzer, G. 2002. Models of ecolog-ical rationality: The recognition hypothesis. Psycho-logical Review, 109: 75–90.

Gould, R. V. 2002. The origins of status hierarchies: Aformal theory and empirical test. American Journalof Sociology, 107: 1143–1178.

Greenwood, R., & Suddaby, R. 2006. Institutional entre-preneurship in mature fields: The big five accountingfirms. Academy of Management Journal, 49: 27–48.

Gulati, R., &Garguilo,M. 1999.Wheredo interorganizationalnetworks come from? American Journal of Sociology,104: 1439–1493.

Guler, I., & Guillen, M. F. 2010. Home country networksand foreign expansion: Evidence from the venturecapital industry. Academy of Management Journal,53: 390–410.

Heckman, J. J., &Borjas,G. J. 1980.Doesunemployment causefuture unemployment? Definitions, questions and an-swers from a continuous time model of heterogeneityand state dependence. Economica, 47: 247–283.

Hiatt, S. R., Sine,W. D., & Tolbert, P. S. 2009. FromPabst toPepsi: The deinstitutionalization of social practicesand the creation of entrepreneurial opportunities.Administrative Science Quarterly, 54: 635–667.

Hsu, D. H. 2004. What do entrepreneurs pay for venture cap-ital affiliation? The Journal of Finance, 59: 1805–1844.


Hsu, G. 2006. Jacks of all trades and masters of none: Au-diences’ reactions to spanning genres in feature filmproduction. Administrative Science Quarterly, 51:420–450.

Hughes, E. C. 1945. Dilemmas and contradictions of status.American Journal of Sociology, 51: 353–359.

Jackson,E.F.,&Curtis,R.F.1972.Effectsof verticalmobilityand status inconsistency: A body of negative evidence.American Sociological Review, 37: 701–713.

Jensen, M. 2003. The role of network resources in marketentry: Commercial banks’ entry into investmentbanking, 1991–1997. Administrative Science Quar-terly, 48: 466–497.

Jensen, M. 2006. Should we stay or should we go? Ac-countability, status anxiety, and client defections.Administrative Science Quarterly, 51: 97–128.

Jensen, M. 2008. The use of relational discrimination tomanage market entry: When do social status andstructural holes work against you? Academy ofManagement Journal, 51: 723–743.

Jensen, M. 2010. Legitimizing illegitimacy: How creatingmarket identity legitimizes illegitimate products. Re-search in the Sociology of Organizations, 31: 39–80.

Jensen, M., & Kim, H. 2015. The real Oscar curse: Thenegative consequences of positive status shifts. Or-ganization Science, 26: 1–21.

Jensen, M., & Roy, A. 2008. Staging exchange partnerchoices: When do status and reputation matter?Academy of Management Journal, 51: 495–516.

Jensen, M., Kim, B. K., & Kim, H. 2011. The importanceof status in markets: A market identity perspective.In J. J. Pearce (Ed.), Status in management and or-ganizations: 87–117. Cambridge, U.K.: CambridgeUniversity Press.

Judge, T.A., Cable,D.M.,Colbert,A. E., &Rynes, S. L. 2007.What causes a management article to be cited—Article, author, or journal?Academy ofManagementJournal, 50: 491–506.

Kacperczyk, A. J. 2013. Social influence and entrepre-neurship: The effect of university peers on entrepre-neurial entry. Organization Science, 24: 664–683.

Kang, J. H. 2010. Status inconsistency and strategic be-havior: US venture capital investments, 1981 to 2002.Korean Journal of Sociology, 44: 1–25.

Kaplan, A. 2014. European management and Europeanbusiness schools: Insights from the history of businessschools. EuropeanManagement Journal, 32: 529–534.

Khurana, R. 2007. From higher aims to hired hands: Thesocial transformation of American business schoolsand the unfulfilled promise of management asa profession. Princeton, NJ: Princeton UniversityPress.

Khurana, R., & Spender, J. C. 2012. Herbert A. Simon onwhat ails business schools: More than “a problem inorganizational design”. Journal of ManagementStudies, 49: 619–639.

Kim, B. K., & Jensen, M. 2011. How product order affectsmarket identity repertoire ordering in the US operamarket. Administrative Science Quarterly, 56:238–256.

Kim, E. H., Morse, A., & Zingales, L. 2009. Are elite uni-versities losing their competitive edge? Journal ofFinancial Economics, 93: 353–381.

Knoke, D., & Burt, R. S. 1983. Prominence. In R. S. Burt &M. J. Minor (Eds.), Applied network analysis: Amethodological introduction. Beverly Hills, CA:Sage.

Kovacs, B., & Liu, C. 2016. Audience structure and statusmultiplicity. Social Networks, 44: 36–49.

Lenski, G. E. 1954. Status crystallization: A non-verticaldimension of social status. American SociologicalReview, 19: 405–413.

Lenski, G. E. 1956. Social participation and status crys-tallization. American Sociological Review, 21:458–464.

Lerner, J. S., & Tetlock, P. E. 1999. Accounting for the ef-fects of accountability. Psychological Bulletin, 125:255–275.

Malter, D. 2014. On the causality and cause of returnsto organizational status: Evidence from the GrandsCrus Classes of the Medoc. Administrative ScienceQuarterly, 52: 271–300.

Markides, C. C. 1995. Diversification, restructuring andeconomic performance. Strategic ManagementJournal, 16: 101–118.

Merton, R. K. 1968. The Matthew effect in science. Sci-ence, 159: 56–63.

Onder, A. S., & Tervio, M. 2015. Is economics a housedivided? Analysis of citation networks. EconomicInquiry, 53: 1491–1505.

Ozmel, U., Reuer, J. J., & Gulati, R. 2013. Signals acrossmultiple networks: How venture capital and alliancenetworks affect interorganizational collaboration.Academy of Management Journal, 56: 852–866.

Paolella, L., & Durand, R. 2016. Category spanning, eval-uation, and performance: Revised theory and test onthe corporate law market. Academy of ManagementJournal, 59: 330–351.

Peters, D. P., & Ceci, S. J. 1982. Peer-review practices ofpsychological journals: The fate of published articles,submitted again. Behavioral and Brain Sciences, 5:187–195.

Petkova,A. P., Rindova,V. P., &Gupta,A.K. 2013.Nonewsis bad news: Sensegiving activities, media attention,


and venture capital funding of new technology orga-nizations. Organization Science, 24: 865–888.

Pfeffer, J. 1993. Barriers to the advance of organizationalscience: Paradigm development as a dependent vari-able.AcademyofManagementReview, 18: 599–620.

Phillips, D. J., & Zuckerman, E. W. 2001. Middle-statusconformity: Theoretical restatement and empiricaldemonstration in two markets. American Journal ofSociology, 107: 379–429.

Podolny, J. M. 1993. A status-based model of marketcompetition. American Journal of Sociology, 99:829–872.

Podolny, J. M. 1994. Market uncertainty and the socialcharacter of economic exchange. AdministrativeScience Quarterly, 39: 458–483.

Pollock, T. G., Chen, G., Jackson, E. M., & Hambrick, D. C.2010. How much prestige is enough? Assessing thevalue of multiple types of high-status affiliates foryoung firms. Journal of Business Venturing, 25: 6–23.

Pollock, T. G., Lee, P. M., Jin, K., & Lashley, K. 2015.(Un)tangled: Exploring the asymmetric coevolutionof new venture capital firms’ reputation and status.Administrative Science Quarterly, 60: 482–517.

Pollock, T. G., & Rindova, V. P. 2003. Media legitimationeffects in the market for initial public offerings.Academy of Management Journal, 46: 631–642.

Rabe-Hesketh, S., & Skrondal, A. 2008. Multilevel andlongitudinal modeling using Stata. College Station,TX: STATA press.

Rhee, M., & Haunschild, P. R. 2006. The liability ofgood reputation: A study of product recalls in theUS automobile industry. Organization Science, 17:101–117.

Salancik, G. R., & Pfeffer, J. 1974. The bases and use ofpower in organizational decision making: The case ofa university. Administrative Science Quarterly, 25:453–473.

Scott, M. B., & Lyman, S. M. 1968. Accounts. AmericanSociological Review, 33: 46–62.

Shipilov, A. V., Li, S. X., & Greve, H. R. 2011. The princeand the pauper: Search and brokerage in the initiationof status-heterophilous ties. Organization Science,22: 1418–1434.

Simcoe, T. S., & Waguespack, D. M. 2011. Status, quality,and attention: What’s in a (missing) name? Manage-ment Science, 57: 274–290.

Simon, H. A. 1967. The business school: A problem in or-ganizational design. Journal ofManagementStudies,4: 1–16.

Smith, E. B. 2011. Identities as lenses: How organizationalidentity affects audiences’ evaluation of organizational

performance. Administrative Science Quarterly, 56:61–94.

Smith, E. B. 2013. From role conflict to evaluationdiscordance. How do conflicting performance eval-uations affect risk taking in multiple audiencecontexts? Research in organizational sciences. InD. Svyantek & K.Mahoney (Eds.),Receivedwisdom,kernels of truth, and boundary conditions in orga-nizational studies. 271–295 Charlotte, NC: Infor-mation Age Publishing.

Sørensen, A. B. 1996. The structural basis of social in-equality. American Journal of Sociology, 102: 1333–1365.

Starbuck, W. H. 2009. The constant causes of never-ending faddishness in the behavioral and socialsciences. Scandinavian Journal of Management,25: 108–116.

Stryker, S., & Macke, A. S. 1978. Status inconsistency androle conflict.Annual Review of Sociology, 84: 57–90.

Stuart, T. E., Hoang, H., & Hybels, R. C. 1999. In-terorganizational endorsements and the performanceof entrepreneurial ventures. Administrative ScienceQuarterly, 44: 315–349.

Tetlock, P. E., Skitka, L., & Boettger, R. 1989. Social andcognitive strategies for coping with accountability:Conformity, complexity, and bolstering. Journal ofPersonality and Social Psychology, 57: 632–640.

Treiman, D. J. 1966. Status discrepancy and prejudice.American Journal of Sociology, 72: 651–664.

Trieschmann, J. S., Dennis, A. R., Northcraft, G. B., &Nieme,A.W. 2000. Serving constituencies in businessschools: MBA program versus research performance.Academy of Management Journal, 43: 1130–1141.

Useem, M., & Karabel, J. 1986. Pathways to top corporatemanagement. American Sociological Review, 51:184–200.

Villalonga, B. 2004. Diversification discount or premium?New evidence from the business information trackingseries. The Journal of Finance, 59: 479–506.

Welch, I. 2014. Referee recommendations. Review of Fi-nancial Studies, 27: 2773–2804.

Whitt, H. P. 1983. Status inconsistency: A body of negativeevidence or a statistical artifact? Social Forces, 62:201–233.

Zhang, X. 2008. Status inconsistency revisited: An im-proved statistical model. European Sociological Re-view, 24: 155–168.

Zhao, W., & Zhou, X. 2011. Status inconsistency andproduct valuation in the California wine market. Or-ganization Science, 22: 1435–1448.


Zuckerman, E. W. 1999. The categorical imperative: Secu-rities analysts and the illegitimacy discount.AmericanJournal of Sociology, 104: 1398–1438.

Zuckerman, E. W. 2000. Focusing the corporate product:Securities analysts and de-diversification. Adminis-trative Science Quarterly, 45: 591–619.

Michael Jensen ([email protected]) is a professor ofstrategy at the Stephen M. Ross School of Business, Uni-versity ofMichigan, and an international research fellowat

the Oxford University Centre for Corporate Reputation. Hereceived his PhD in management and organizations fromNorthwestern University. His research focuses on status,reputation, and identity.

Pengfei Wang ([email protected]) is an assistantprofessor of strategy at the BI Norwegian BusinessSchool. He received his PhD in strategy from Rotter-dam School of Management, Erasmus University Rot-terdam. His research focuses on status, networks, andinnovation.

APPENDIX A: BUSINESS SCHOOL SAMPLE

TABLE A1Business Schools Ranked by Status (2012)

School Name Rank School Name Rank School Name Rank

Stanford Univ. 1 Indiana Univ. at Bloomington 38 Univ. of South Carolina 75Univ. of Pennsylvania 2 Western Univ. 39 Univ. of Virginia, Darden 76New York Univ. 3 Rutgers Business School 40 Univ. of Arizona 77Harvard Business School 4 City Univ. of Hong Kong 41 Rice Univ. 78Univ. of Chicago 5 HEC France 42 George Mason Univ. 79Northwestern Univ. 6 Arizona State Univ. 43 Univ. of Houston 80Massachusetts Institute of Technology 7 Univ. of Texas at Dallas 44 Virginia Tech 81Columbia Univ. 8 Georgia State Univ. 45 Univ. of Wisconsin at Milwaukee 82Univ. of California at Berkeley 9 Boston College 46 Iowa State Univ. 83Univ. of Southern California 10 Georgetown Univ. 47 Univ. of Kentucky 84Univ. of Michigan at Ann Arbor 11 Southern Methodist Univ. 48 Univ. of Miami 85Carnegie Mellon Univ. 12 Univ. of Oregon 49 Univ. of California at San Diego 86London Business School 13 Univ. of California at Irvine 50 State Univ. of New York at Buffalo 87Univ. of Toronto 14 Univ. of Georgia 51 Univ. of Arkansas at Fayetteville 88Univ. of Texas at Austin 15 Univ. of Alberta 52 Case Western Reserve Univ. 89Duke Univ. 16 Temple Univ. 53 Syracuse Univ. 90INSEAD 17 Univ. of Pittsburgh 54 Vanderbilt Univ. 91Univ. of North Carolina at Chapel Hill 18 Drexel Univ. 55 Univ. of Kansas 92HKUST 19 George Washington Univ. 56 Univ. of Missouri at Columbia 93Univ. of Minnesota at Twin Cities 20 Purdue Univ. 57 Univ. of Colorado 94Yale Univ. 21 Chinese Univ. of Hong Kong 58 Louisiana State Univ. 95Pennsylvania State Univ. at Univ. Park 22 Univ. of Wisconsin-Madison 59 Univ. of Tennessee at Knoxville 96Univ. of Illinois at Urbana-Champaign 23 Texas A&M Univ. 60 Northeastern Univ. 97Univ. of British Columbia 24 Nanyang Technological Univ. 61 Univ. of Calgary 98Boston Univ. 25 Brigham Young Univ. 62 Univ. of California at Davis 99Univ. of Maryland at College Park 26 Univ. of Hong Kong 63 Univ. of California at Riverside 100National Univ. of Singapore 27 Queen’s Univ. 64 Clemson Univ. 101Cornell Univ. 28 Singapore Management Univ. 65 Florida State Univ. 102Emory Univ. 29 Univ. of Rochester 66 Univ. of Virginia, Mclntire 103Univ. of Washington at Seattle 30 Univ. of Notre Dame 67 Simon Fraser Univ. 104Georgia Institute of Technology 31 Univ. of Connecticut 68 American Univ. 105Univ. of California at Los Angeles 32 City Univ. of New York, 69 Univ. of Texas at Arlington 106Ohio State Univ. 33 Univ. of Utah 70 Texas Christian Univ. 107Univ. of Florida 34 Michigan State Univ. 71 York Univ. 108McGill Univ. 35 Univ. of Iowa 72 Univ. of Melbourne 109Washington Univ. at St. Louis 36 Hong Kong Polytechnic Univ. 73Univ. of Navarra 37 Dartmouth College 74


mailto:[email protected]

mailto:[email protected]

APPENDIX B

UTD RESEARCH RANKING DATA AND JOURNAL ASSIGNMENTS

MEASUREMENT OFRESEARCH PERFORMANCES

“A single-authored paper results in the school of affili-ation being credited with a score of 1. If there are multipleauthors from different schools, each school gets a score ofp/n, wherep is the number of authors from the same school

and there are a total of n authors on the article. If an authorlists multiple affiliations, each of the schools that author isaffiliated with gets a corresponding scaled score. For ex-ample, if one of the n authors lists m affiliations, eachschool that author is affiliated with gets a score of 1/nm.”

Source: http://jindal.utdallas.edu/the-utd-top-100-business-school-research-ranking

TABLE B1Journal Assignments

Department UTD Journals Department UTD Journals

Accounting Accounting Review Marketing Journal of Consumer ResearchJournal of Accounting & Economics Journal of MarketingJournal of Accounting Research Journal of Marketing Research

Finance Journal of Finance Marketing ScienceJournal of Finance Economics Operations Management ScienceThe Review of Financial Studies Operation Research

Management Administrative Science Quarterly Journal of Operation ManagementAcademy of Management Journal MIS QuarterlyAcademy of Management Review Information Systems ResearchStrategic Management Journal Journal on ComputingJournal of International Business Studies Manufacturing and Service Operation ManagementOrganization Science Production and Operation Management


http://jindal.utdallas.edu/the-utd-top-100-business-school-research-ranking

http://jindal.utdallas.edu/the-utd-top-100-business-school-research-ranking

not in the same boat: how status inconsistency affects ... · (d’aveni, 1996; trieschmann,...

Documents