An Integrative Model of Organizational Trust- Past Present and Future

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
A considerable amount of research has examined trust since our 1995 publication. We revisit some of the critical issues that we addressed and provide clarifications and extensions of the topics of levels of analysis, time, control systems, reciprocity, and measurement. We also recognize recent research in new areas of trust, such as affect, emotion, violation and repair, distrust, international and cross-cultural issues, and context-specific models, and we identify promising avenues for future research.
  EDITOR’S FORUM AN INTEGRATIVE MODEL OFORGANIZATIONAL TRUST: PAST, PRESENT,AND FUTURE F. DAVID SCHOORMANPurdue UniversityROGER C. MAYERThe University of AkronJAMES H. DAVISUniversity of Notre Dame A considerable amount of research has examined trust since our 1995 publication. Werevisit some of the critical issues that we addressed and provide clarifications andextensions of the topics of levels of analysis, time, control systems, reciprocity, andmeasurement. We also recognize recent research in new areas of trust, such as affect,emotion, violation and repair, distrust, international and cross-cultural issues, andcontext-specific models, and we identify promising avenues for future research. As we wrote our 1995 paper on trust (Mayer,Davis, & Schoorman, 1995), we were struck bythe relative scarcity of research in the main-stream management literature focusing directlyon trust. This led us to several bodies of litera-ture, including management, psychology, phi-losophy, and economics. We found that scholarsfrom diverse disciplines were presenting manyinsightful views and perspectives on trust butthat many of them seemed to talk past one an-other. Our goal was to integrate these perspec-tives into a single model.This work came to fruition at about the sametime as several other works on trust. Papers ontrust by Hosmer (1995) and McAllister (1995) werealso published in Academy of Managementjournals that year, followed the next year by abook edited by Kramer and Tyler (1996). The con-fluence of these works, fueled by practical con-cerns raised by now infamous government andcorporate scandals over the next decade, pro-duced a groundswell of interest in understand-ing this basic and ubiquitous construct.Since we were drawing perspectives frommultiple disciplines as inputs to the model, wewanted to provide a model that was generallyapplicable and would be used across multipledisciplines. We were gratified to find in a recentsearch that our paper has been cited over 1,100times (according to Google Scholar). In additionto management and general business, it hasbeen cited in such diverse areas as marketing,accounting, finance, economics, informationsystems, industrial engineering, political sci-ence, communication, ethics, law, psychology,sociology, health care, and agribusiness. Wewould like to use this opportunity to revisit someof the issues raised by our 1995 paper and re-view how the field has dealt with them. We willalso discuss the new concerns and opportunitiesfor future research on trust. CLARIFICATIONS AND EXTENSIONS OF THEMODEL OF TRUSTTrust As an Aspect of a Relationship One of the difficult conceptual decisions thatwe faced as we developed our definition of trustwas to break with the widely accepted ap-proach, to that point, that trust was dispositionaland “trait-like” and to argue that trust was anaspect of relationships. That meant it variedwithin person and across relationships. Withsome exceptions (e.g., Driscoll, 1978; C. L. Scott,1980), the dominant conceptual and operational   Academy of Management Review 2007, Vol. 32, No. 2, 344–354.Copyright of the Academy of Management, all rights reserved. Contents may notbe copied, emailed, posted to a listserv, or otherwise transmitted without thecopyright holder’s express written permission. Users may print, download, or emailarticles for individual use only. 344  definition of trust in the literature was Rotter’s(1967). We then went the next step and includedability as an antecedent of trust that allowed aparty’s trust to vary within a given trustee butacross domains. The dispositional aspects oftrust considered by Rotter are contained in theconstruct of propensity to trust in our model. Theliterature that has followed our model has notquestioned this decision and has accepted theview that trust is based in relationships. Application Across Levels of Analysis The importance of multilevel and cross-levelperspectives is gaining increasing attention inorganizational research. This has led to a callforexaminingtrustacrosslevelsoforganization-al analysis (e.g., Rousseau, Sitkin, Burt, & Cam-erer, 1998). There is a need to understand trustboth within and between organizations becausemethodological difficulties can arise in the ab-sence of a clear multilevel conceptual model(Currall & Inkpen, 2002; Mossholder & Bedeian,1983; Rousseau, 1985).We have heard from a number of scholars thatthe 1995 framework is fairly robust across levelsof analysis. A bit of history on the developmentof our theory may shed light on this issue. Earlydrafts of our paper developed our trust modelacross multiple levels of analysis. One of ourinitial goals was to develop a theory that wouldbeapplicableacrosslevelsofanalysis.Wewerecareful to develop constructs that would crosslevels of analysis, and we developed examplesof how cross-level applications of the modelwould work. Perhaps it was fortunate that earlyreviewers of our paper made the accurate obser-vation that the paper was very cumbersome(and long) because it developed the multilevelmodel. They recommended that we restrict ourpaper to a single level. The fact that our initialgoal was to develop a multilevel theory is prob-ably why the model works as well as it doesacross levels, but we do agree with those whoargue that one of the weaknesses in much of thecurrent trust research is that it is limited to re-lationships at a single level of analysis, consid-ering either dyadic trust relationships withinorganizations or trust between organizations.Several authors have recognized differencesin trust for single referents at different hierar-chical levels within an organization (e.g., Cook& Wall, 1980; Driscoll, 1978; D. Scott, 1980). Recentresearch points out that trust should be exam-ined at both the macro and micro levels withinan organization (McEvily, Perrone, & Zaheer,2003; Zaheer, McEvily, & Perrone, 1998). Whilethe need to understand trust has been noted inareas of study both within and between organi-zations, methodological difficulties can arise inthe absence of a clear multilevel conceptualmodel (Mossholder & Bedeian, 1983; Rousseau,1985).Just as perceptions about an individual’s abil-ity, benevolence, and integrity will have an im-pact on how much trust the individual can gar-ner, these perceptions also affect the extent towhich an organization will be trusted. We de-fined each of these trustworthiness dimensionsso that it could be applied to interpersonal, in-tergroup, or interorganizational levels of analy-sis.At higher levels of analysis, such as betweenorganizations, viewing the trustee in terms ofability and integrity seems to be well accepted.At macro levels of analysis, however, benevo-lence has received little attention. We definedbenevolence as the extent to which a party isbelieved to want to do good for the trustingparty, aside from an egocentric profit motive.Does the other company hold the focal compa-ny’s best interests as highly important? Whilewe may be able to identify situations, such assole proprietorships, where the owners havestrong bonds that display significant benevo-lence toward one another, the more traditionalmode is probably one wherein each company ismotivated primarily by its own financial inter-ests. If this is indeed the norm, benevolence isnot likely to be the most important factor in thedevelopment of interorganizational trust. How-ever, acts of benevolence (e.g., allowing bench-marking) from a potential partner in a joint ven-ture would help to build trust.We contend that all three factors of ability,benevolence, and integrity can contribute totrust in a group or organization. Consider, forinstance, a supplier-buyer relationship. Thebuyer may believe that a supplier is able toprovide a quality product in a timely fashion.However, this only assures that the supplier could  perform. This does not mean that it  will perform, and, therefore, the supplier will notnecessarily be trusted. The perception that thesupplier has integrity suggests that it will fulfillagreements as promised. Yet even if there is an 2007 345 Schoorman, Mayer, and Davis  agreement, if the supplier’s ability to deliver isquestionable, it will not be trusted. If the sup-plier is perceived as benevolent, it will have astrong desire to serve this particular buyer’sneeds. If the supplier’s integrity is suspect be-cause, for instance, its track record with otherfirms is inconsistent with its stated policies,trust will again be lacking. As the perception ofeach of these factors increases, we would expectan increase in willingness to take a risk in therelationship.The trust of either the dominant coalition orthe management team is critical to understand-ing organizational trust, since it is this level oftrust that will govern the strategic actions of theorganization (Cyert & March, 1963; Simon, 1957).As with individuals, we propose that some or-ganizations develop greater propensities totrust than do others. For organizations, thesepropensities develop from geographic, industry,and economic histories. A series of previous in-teractions with other organizations that resultedin, for example, lawsuits or monetary losseswould lower an organization’s propensity totrust. Conversely, a series of such experiencesas mutual benchmarking with various organiza-tions that significantly improved the qualityprocesses for an organization would increase itstrust propensity.In summary, groups and organizations canboth garner trust from other parties and trustother parties. Our model was designed to under-stand the major factors that explain trust fromnot only the individual level but from the groupand organizational perspectives as well. The Time Dimension One of the issues explicit in our theory wasthat “time” would play an important role in themeaningfulness of the variables in the model.We noted that propensity, as a dispositionalquality, would be an important factor at the verybeginning of the relationship. We also notedthat judgments of ability and integrity wouldform relatively quickly in the course of the rela-tionship and that benevolence judgments wouldtake more time. Proposition 3 states that “theeffect of integrity on trust will be most salientearly in the relationship prior to the develop-ment of meaningful benevolence data” (1995:722); Proposition 4 states that “the effect of per-ceived benevolence on trust will increase overtime as the relationship between the parties de-velops” (1995: 722).Despite these assertions, in many empiricalstudies researchers have raised questions aboutthe high observed correlation between benevo-lence and integrity and have questioned the in-dependence of these variables. In a discussionof several empirical studies, Schoorman (2002)observed that the findings as a whole were com-pletely consistent with the model. Those studiesconducted in laboratory settings were morelikely to show a high correlation between be-nevolence and integrity because the relation-ships had not had time to develop any real dataabout benevolence. In field samples where theparties had longer relationships, benevolenceand integrity were more likely to be separablefactors. We continue to find this pattern to beconsistent in our research. We think it would beinteresting for future research to establish morespecifically the process and time frames inwhich each of the variables contributes to trust. Trust, Risk, and Control Systems In our model we argued that trust would leadto risk taking in a relationship (see Proposition5). Perceived risk moderates the relationship be-tween trust and risk taking in our model. Trust isthe “willingness to take risk,” and the level oftrust is an indication of the amount of risk thatone is willing to take. Clearly, control systemsare an alternate mechanism for dealing withrisk in relationships. Recently, several scholarshave speculated about the relationship betweentrust and control systems in dealing with risk(McEvily et al., 2003; Sitkin & George, 2005).Our views on this issue are developed furtherin Davis, Schoorman, and Donaldson (1997), inwhich we argue that one of the major distinc-tions between agency theory and stewardshiptheory is the use of trust versus control systemsto manage risk. However, we do not see thesemechanisms as being mutually exclusive. Onthe contrary, when the risk in a situation isgreater than the trust (and, thus, the willingnessto take risk), a control system can bridge thedifference by lowering the perceived risk to alevel that can be managed by trust. For exam-ple, in an organization that has a culture of“open book management” and transparency innumbers (a control system), the levels of per-ceived risk may be lower. There is a greater 346 April  Academy of Management Review  opportunity to empower employees by trustingthem to manage larger budgets and for employ-ees to trust the supervisor that performance-based compensation is fair.However, there is an important caveat thatmust be noted. If there is a very strong system ofcontrols in an organization, it will inhibit thedevelopment of trust. Not only will there be fewsituations where there is any remaining per-ceived risk but trustworthy actions will be at-tributed to the existence of the control systemrather than to the trustee (cf. Strickland, 1958).Thus, a trustee’s actions that should be inter-preted as driven by benevolence or by integritymay be viewed simply as responses to the con-trol systems. The use of control systems is howagency theory proposes dealing with risk man-agement, and this does not foster the develop-ment of trust. The Reciprocity of Trust One of the limitations of our model that wenoted in the conclusion of our 1995 paper wasthat our conceptualization was unidirectional.We did not explore the reciprocity in trustingrelationships. This is a particularly salient issuein the area of leader-subordinate relationships,since the dominant view among leadership the-orists is that leader-member exchange (LMX) ismutual and reciprocal (Graen & Uhl-Bien, 1995;Liden, Wayne, & Stillwell, 1993). In an extensionof our model, Brower, Schoorman, and Tan (2000)argued that, unlike LMX, trust is not necessarilymutual and is not reciprocal. One of the impli-cations of this argument is that, in a relation-ship, A can trust B, but B may not trust A. This iscompletely consistent with the approach to trustand trust formation that we presented in ourmodel but is inconsistent with the views in theleadership literature. Empirical studies that ex-amine this reciprocal linkage of how one party’strust affects the other party’s trust in return,rather than assuming them to be equivalent, arerare (Serva, Fuller, & Mayer, 2005). We feel thatthis presents a fruitful area for future research. Measurement of Trust As “Willingness to BeVulnerable” While the theory of trust described in the 1995paper has been very influential in the develop-ment of trust research, the measurement of trusthas been a different story. We defined trust as awillingness to be vulnerable to another party.As such, suitable measurement of the constructnecessitates that questions be asked that assessthe extent to which a trustor is willing to volun-tarily take risks at the hands of the trustee.We developed a short, four-item measure,with each of the items tapping into how willingthe trustor was to be vulnerable to the trustee.We (Schoorman, Mayer, & Davis, 1996a) foundthat veterinary doctors took bigger risks withthose employees they trusted more. The impactof trust went beyond that explained by the abil-ity, benevolence, and integrity of the trustee.Despite the measure’s brevity, we found its in-ternal consistency strong (Cronbach’s alpha   .82).We (Davis, Schoorman, Mayer, & Tan, 2000)used the same measure in a restaurant settingto measure the trust that employees had in theirgeneral manager and found an alpha of .62. Thismeasure of trust in the leader significantly pre-dicted subsequent sales, profits, and employeeturnover in the restaurants (Davis, Mayer, &Schoorman, 1995; Davis et al., 2000). Based onthese results, we concluded that if trust in thegeneral manager could be developed and sus-tained, it would be a significant competitive ad-vantage to the firm, and the framework, includ-ing ability, benevolence, and integrity, meritedfurther consideration as an approach to build-ing trust in management.Using this same measure, a quasi-experiment(Mayer & Davis, 1999) showed that trust in topmanagement was significantly improved byidentifying and replacing an invalid appraisalsystem. While the alpha in this study was lower(i.e., .59 and .60 in two waves of data), the test-retest reliability was quite strong, at .75 overfive months and .66 over nine months. Further-more, the quasi-experimental results were sig-nificant, even though the sizes of the groupsbeing compared were modest (i.e., twenty-twoand fifty-seven). In additional analyses wefound an average interitem correlation of  r  .32.For each item we calculated a correlation be-tween it and a composite of the other threeitems. These correlations ranged from .23 to .49,with an average of .38.These results fall well within Kline’s (1986)description of a measure of a complex constructthat has maximum validity. Kline noted thatsuch a measure would only have low internal 2007 347 Schoorman, Mayer, and Davis
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks