JUSTICE, TRUST, AND TRUSTWORTHINESS: A LONGITUDINAL

JUSTICE, TRUST, AND TRUSTWORTHINESS: ALONGITUDINAL ANALYSIS INTEGRATING THREE

THEORETICAL PERSPECTIVES

JASON A. COLQUITTJESSICA B. RODELLUniversity of Georgia

Despite a significant amount of theoretical and empirical attention, the connectionbetween justice and trust remains poorly understood. Our study utilized Mayer, Davis,and Schoorman’s (1995) distinction between trustworthiness (the ability, benevolence,and integrity of a trustee) and trust (a willingness to be vulnerable to the trustee) toclarify that connection. More specifically, we drew on a theoretical integration ofsocial exchange theory, the relational model, and fairness heuristic theory to derivepredictions about the relationships among justice, trustworthiness, and trust, withsupervisors as the referent. A longitudinal field study stretching over two periodsshowed that informational justice was a significant predictor of subsequent trustperceptions, even when analyses controlled for prior levels of trust and trustworthi-ness. However, the relationship between justice and trustworthiness was shown to bereciprocal. Procedural and interpersonal justice were significant predictors of subse-quent levels of benevolence and integrity, with integrity predicting subsequent levels ofall four justice dimensions. We describe the theoretical implications of these results forfuture research in the justice and trust literatures.

Scholarly interest in the topics of organizationaljustice and trust has exploded in the past severalyears, as evidenced by the increased attention thesetopics have received in organizational journals.One likely reason for that increased attention is thatjustice perceptions and trust assessments are con-sistent predictors of employee attitudes and behav-iors. For example, meta-analytic results support arelationship between justice perceptions and keyorganizational outcomes, including organizationalcommitment, task performance, and citizenship be-havior (Cohen-Charash & Spector, 2001; Colquitt,Conlon, Wesson, Porter, & Ng, 2001). Meta-analyseshave also linked trust assessments to similar out-comes, such as task performance, citizenship be-havior, and counterproductive behavior (Colquitt,Scott, & LePine, 2007; Dirks & Ferrin, 2002).

Organizational justice refers to perceptions offairness in decision-making and resource allocationenvironments (Greenberg, 1987) and is commonlydivided into four dimensions. Distributive justicerefers to the fairness of decision outcomes, andindividuals judge it by determining whether theperceived ratio of outcomes to inputs matchesthose of a comparison other (Adams, 1965), or

whether resource allocations match appropriatenorms (Leventhal, 1976). Procedural justice refersto the perceived fairness of decision-making proce-dures (Leventhal, 1980; Thibaut & Walker, 1975).Procedures are evaluated by their level of consis-tency, bias suppression, accuracy, correctability,ethicality, and the degree to which they allow voiceand input (Leventhal, 1980; Thibaut & Walker,1975). Interactional justice refers to the fairness ofinterpersonal treatment during decision-makingprocedures (Bies & Moag, 1986). Interactional jus-tice can be further divided into interpersonal andinformational components. Interpersonal justice re-flects the degree of respect and propriety authorityfigures use when implementing procedures, andinformational justice reflects the degree of justifi-cation and truthfulness offered during procedures(Colquitt, 2001; Greenberg, 1993).

The definition of trust has been much debated inthe literature (Rousseau, Sitkin, Burt, & Camerer,1998). Some scholars have defined trust as confi-dent, positive expectations regarding a trustee’sconduct, motives, and intentions in situations en-tailing risk (e.g., Cook & Wall, 1980; Gabarro &Athos, 1976; Lewicki & Bunker, 1995). Those con-fident, positive expectations create a willingness toact on the basis of the words, actions, and decisionsof the trustee (Cook & Wall, 1980; McAllister,1995). Measures of this conceptualization of trustoperationalize those confident, positive expecta-

Editor’s note: The manuscript for this article was ac-cepted for publication during the term of AMJ’s previouseditor-in-chief, Duane Ireland.

� Academy of Management Journal2011, Vol. 54, No. 6, 1183–1206.http://dx.doi.org/10.5465/amj.2007.0572

1183

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.

tions by asking respondents whether trustees arecompetent, skilled, efficient, caring, loyal, princi-pled, and fair (Cook & Wall, 1980; Gabarro & Athos,1976; McAllister, 1995; Nyhan & Marlowe, 1997;Roberts & O’Reilly, 1974).

Mayer and colleagues’ (1995) model of trust of-fers a different definition. These authors definedtrust as the willingness to be vulnerable to the ac-tions of a trustee on the basis of the expectation thatthe trustee will perform a particular action, irre-spective of any monitoring or control mechanisms(see also Schoorman, Mayer, & Davis, 2007). Mea-sures of this conceptualization of trust operational-ize that willingness to be vulnerable by asking re-spondents about their willingness to allow a trusteeto have significant influence over their workinglives (Mayer & Davis, 1999; Mayer & Gavin, 2005;Schoorman et al., 2007). Importantly, Mayer andcolleagues (1995) argued that some of the sameconcepts that are typically assessed in the confi-dent, positive expectations approach predict trustas they define it. In changing the relevant timehorizon from one represented by future-orientedexpectations to one captured in current state as-sessments, Mayer and colleagues (1995) framed theconcepts as facets of trustworthiness—attributes orcharacteristics of a trustee that inspire trust. Abilityreflects concepts such as competence, skills, effi-ciency, and dedication. Benevolence reflects thesense that the trustee wants to “do good” to thetruster, with “doing good” including concepts suchas being caring and open. Integrity reflects an ad-

herence to a set of acceptable principles or a set ofshared values.

A number of theoretical perspectives include ar-guments for significant relationships among justiceand trust concepts. These include social exchangetheory (Blau, 1964), the relational model (Lind &Tyler, 1988; Tyler & Lind, 1992), and fairness heu-ristic theory (Lind, 2001; Van den Bos, 2001). Anarrative review by Lewicki, Wiethoff, and Tomlin-son noted that “the volume of both theoretical andempirical work over the last 15 years clearly pointsto a strong relationship between trust and justice”(2005: 253). Meta-analyses in both literatures havealso pointed to moderately to strongly positive cor-relations among justice and trust concepts (Cohen-Charash & Spector, 2001; Colquitt et al., 2001; Dirks& Ferrin, 2002). Nevertheless, Lewicki and col-leagues also noted that “in spite of the presumedconnections between justice and trust in theory andpractice, the precise association between these con-structs has not been fully elaborated” (2005: 248).The authors further argued that “continued effortsto tease apart the two constructs and understandthe nature of their connectedness will require agreat deal of careful conceptual thinking and lin-guistic precision, no less the challenge of actuallycalibrating these relationships through empiricalresearch” (Lewicki et al., 2005: 265).

Figure 1 summarizes the conventional wisdomon the justice-trust connection as it currently existsin the literature. In our view, that conventionalwisdom is deceptively uncontroversial, masking

FIGURE 1The Conventional Wisdom on the Justice-Trust Connection

1184 DecemberAcademy of Management Journal

four distinct questions. First, it may be that justiceis more highly related to some expectations thanothers, as expectations about competence or skillscould have different nomological networks thanexpectations about concern or values. Second, itmay be that estimates of the relationship betweenjustice and confident, positive expectations havebeen inflated because those expectations often al-lude to fairness. Appendix A cites the 13 studiesthat have integrated justice and trust using estab-lished measures of confident, positive expecta-tions. Four of the five measures explicitly containfairness content (Cook & Wall, 1980; Gabarro &Athos, 1976; Nyhan & Marlowe, 1997; Roberts &O’Reilly, 1974). Third, it remains unclear whetherjustice is even related to the willingness to be vul-nerable, because past research has rarely connectedjustice to that concept. Fourth, some theoreticalperspectives frame concepts such as concern, loy-alty, values, and principles as antecedents of jus-tice, not consequences of it (Lind & Tyler, 1988;Tyler, 1989, 1994; Tyler & Lind, 1992), implyingthat the causal order in Figure 1 is either incorrector oversimplified.

With those issues in mind, we sought in thisstudy to clarify the relationship between justiceand trust. A central premise of our study is thatMayer and colleagues’ (1995) distinction betweentrustworthiness and trust can bring increased con-ceptual and operational precision to our under-standing of the justice-trust connection. For thisreason, our study adopts Mayer and colleagues’willingness to be vulnerable definition for trust, asopposed to the confident, positive expectationsdefinition used by other scholars (Cook & Wall,1980; Gabarro & Athos, 1976; Lewicki & Bunker,1995). Conceptually speaking, Mayer and col-leagues’ ability, benevolence, and integrity dimen-sions reposition the concepts commonly assessedin the confident, positive expectations approachinto the three facets of trustworthiness. Figure 2reflects this repositioning, with concepts such ascompetence, concern, values, and so forth nowrepresented by the ability, benevolence, andintegrity boxes in the new diagram. That reposi-tioning allows for the possibility that the justice-trust relationship differs in nature from the justice-trustworthiness relationship. Operationallyspeaking, the scales used to measure Mayer andcolleagues’ (1995) concepts do not include fairnesscontent to the same degree, providing the linguisticprecision that Lewicki and colleagues (2005) calledfor in their discussion of the justice-trust connec-tion. Fairness is included only in Mayer and col-leagues’ integrity facet, not in the ability, benevo-lence, or trust measures.

In addition to drawing on Mayer and colleagues’(1995) model, the repositioning in Figure 2 repre-sents an integration of social exchange theory(Blau, 1964), the relational model (Lind & Tyler,1988; Tyler & Lind, 1992), and fairness heuristictheory (Lind, 2001; Van den Bos, 2001). Specifi-cally, we argue that social exchange theory suggeststhat justice acts as an antecedent of the willingnessto be vulnerable, reflecting a justice to trust causalflow. In contrast, the relational model and fairnessheuristic theory contain propositions more relevantto the relationship between justice and trustworthi-ness concepts. Moreover, their predictions forthose relationships are contradictory. The rela-tional model seems to predict a trustworthiness tojustice causal flow, whereas fairness heuristic the-ory seems to predict a justice to trustworthinessordering. These two theories can therefore be usedto predict that the relationships between justiceand trustworthiness concepts will be reciprocal ornonrecursive.

We believe that our examination of the structurein Figure 2 can remove some of the theoreticalmuddiness that currently plagues the literature onjustice and trust and that this effort will have im-portant theoretical and practical implications. The-oretically speaking, the muddiness prevents schol-ars from operationalizing variables in a manner thatis consistent with a given theory. For example, ourintegration will show that social exchange theoryfocuses specifically on the willingness to be vul-nerable, whereas the relational model focuses spe-cifically on the concepts reflected in trustworthi-ness (Blau, 1964; Organ, 1990; Tyler, 1989, 1994).Failing to account for these nuances results in in-appropriate applications of those theories and po-tentially inaccurate portrayals of causal order andendogeneity. That muddiness also results in unnec-essarily coarse theorizing, preventing scholars fromdrawing distinctions between constructs that mayhave differing relationships with justice, such asability versus benevolence and integrity.

Practically speaking, this muddiness can lead tomanagerial implications that are incomplete or un-founded. For example, if employees have alreadyformed perceptions of trustworthiness, will justice-relevant actions on the part of managers continue topredict intentions to accept vulnerability? As an-other example, which should managers focus moreof their time on? Being distributively fair, proce-durally fair, interpersonally fair, or informationallyfair? Or is more gained by promoting a sense ofability or a sense of benevolence and integrity? Adeeper level of theoretical nuance is needed to beable to convey to managers the most promisingroute for building a willingness to be vulnerable in

2011 1185Colquitt and Rodell

their units. Attaining that level of nuance requiresthe simultaneous operationalization over time oftrustworthiness, justice, and willingness to bevulnerable.

THEORY AND HYPOTHESES

The sections to follow review the theoretical per-spectives relevant to the connections in Figure 2,along with representative empirical tests of thoseperspectives. We begin by focusing on the pathsinvolving trust before moving on to the relationshipbetween justice and trustworthiness. We then offer

the specific hypotheses that were tested in a two-period longitudinal field study that used supervi-sors as the referent for justice, trustworthiness, andtrust. George and Jones (2000) suggested that timeshould become a more central element in scholarlytheorizing and that the time lags used in longitudi-nal research should be based on sound conceptuallogic. In particular, they suggested that scholarsshould apply time concepts to the conceptualiza-tion of constructs (e.g., how are the past, present,and future represented in a construct, how stable isit, how are experiences with it aggregated or brack-eted in one’s mind?) and to the description of rela-

FIGURE 2A More Nuanced Model Utilizing a Mayer and Colleagues (1995) Lens


tionships (e.g., do perceived causes have immedi-ate or delayed effects, what time aggregations areappropriate for observing effects?). With these rec-ommendations in mind, we draw attention to therole of time in the theorizing that supports ourhypotheses.

Predicting Trust: Social Exchange Theory

Figure 2 casts both trustworthiness and justice aspredictors of trust. The relationship between jus-tice and trust has most often been described usingsocial exchange theory (Blau, 1964), which summa-rizes how the transactions in an exchange relation-ship evolve and change over time. Conceptualiza-tions of exchange theory have stressed that anumber of resources can be involved in exchanges.For example, Foa and Foa (1980) identified sixclasses of resources—love, status, information,money, goods, and services—that could be con-trasted on two dimensions: particularism and con-creteness. Particularism reflects how significant theprovider of a resource is to the meaning and valueof the resource; status and services are more partic-ular, and money and goods are more universal.Concreteness reflects whether a resource is overtlytangible or more symbolic; goods and services aremore concrete, and status and information are moresymbolic.

Blau’s (1964) conceptualization of exchange the-ory distinguishes between economic exchange re-lationships, which are based on strict, quid pro quoexchanges, and social exchange relationships,which are based on long-term and unspecified ex-changes of tangible and intangible obligations. Heprovided several examples of intangible resources,including acceptance, compliance, advice, assis-tance, and support, all of which are similar to Foaand Foa’s (1980) status and information. Consider acircumstance in which a supervisor provides assis-tance and support to an employee. The social ex-change theory argument is that these rewardingactions obligate the employee to reciprocate in kind(see also Gouldner, 1960). To discharge that obliga-tion, the employee could reciprocate with accep-tance and compliance, or even with his/her ownbrand of assistance and support.

Blau described the importance of trust in socialexchanges: “Since there is no way to assure anappropriate return for a favor, social exchange re-quires trusting others to discharge their obliga-tions” (1964: 94). In the example given above, nei-ther the supervisor nor the employee can beabsolutely certain that their actions are going to bemet with reciprocated benefits, because no formalcontract governs such events. Fortunately, Blau ar-

gued that “the gradual expansion of exchangetransactions promotes the trust necessary for them”(1964: 98). In this way, the assistance and supportthe supervisor provided promotes trust on the partof the employee, making the employee more will-ing to reciprocate. Blau’s (1964) description of trustseems quite similar to Mayer and colleagues’ (1995)willingness to be vulnerable, as it implies acceptingvulnerability on the basis of the expectation that anaction will occur irrespective of monitoring orsafeguards.

Blau’s (1964) arguments were not applied to jus-tice concepts until Organ’s (1990) analysis of mo-tives for engaging in citizenship behavior. Organsuggested that fair actions on the part of a supervi-sor could themselves comprise the sort of reward-ing actions that engender reciprocation, one form ofwhich would be citizenship behavior. That propo-sition suggests that equity, consistency, bias sup-pression, voice, respect, and justification are func-tionally similar to exchange currencies such asstatus, information, acceptance, compliance, ad-vice, assistance, and support (Blau, 1964; Foa &Foa, 1980). Along these lines, Konovsky and Pugh(1994) suggested that justice can increase the ac-ceptance of vulnerability because it signals howdecision making will proceed over the long term. Ifan authority has adhered to justice rules before, the“default view” is often that they will do so again(Brockner & Siegel, 1995).

In their review, Cropanzano and Mitchell (2005)argued that trust is a promising mechanism forexplaining the social exchange benefits of justiceand other related behaviors, though empirical testshave been sparse. In fact, the relationship betweenjustice and the willingness to be vulnerable form oftrust remains unclear, because past research hastended to rely on measures based on the confident,positive expectations form (e.g., Aryee, Budhwar, &Chen, 2002; Konovsky & Pugh, 1994). It also re-mains unclear whether justice predicts the willing-ness to be vulnerable when considered simultane-ously with trustworthiness, as in Figure 2. Colquittet al. (2007) suggested that some facets of trustwor-thiness could themselves be viewed as currenciescapable of fostering a social exchange. That asser-tion suggests that concern, loyalty, shared values,principles, and so forth are also functionally simi-lar to the exchange currencies of status, informa-tion, acceptance, compliance, advice, assistance,and support (Blau, 1964; Foa & Foa, 1980).

If justice and trustworthiness are viewed as sig-nals for employees to trust their supervisor, howlong does it take for these signals to impact thewillingness to accept vulnerability? Drawing abridge to the attitude-behavior literature (Azjen,


1991; Azjen & Fishbein, 1980), we suggest that trustreflects a behavioral intention to be vulnerable,whereas trustworthiness and justice reflect atti-tudes toward a supervisor. As with other behav-ioral intentions, employees’ intentions to be vul-nerable likely result from a deliberate processwherein they carefully consider relevant informa-tion (Azjen, 1991; Azjen & Fishbein, 1980), such astrustworthiness and trust. The controlled and cog-nitive nature of this process suggests that employ-ees need enough separation in time between super-visory actions and trust intentions to evaluate therelevant information. Although this length of timemay vary among individuals and situations, it isbounded by the opportunity to experience repeatedinteractions with supervisors that provide justiceand trustworthiness information (Cropanzano,Byrne, Bobocel, & Rupp, 2001; Mayer et al., 1995).

Moving beyond issues of time frame, how canone draw predictions about the unique effects ofjustice and trustworthiness on trust, given theirsimilar positioning in exchange theorizing? Oneway is to consider the specific concepts encom-passed by the justice and trustworthiness dimen-sions to gauge whether those concepts are uniquelyrelevant to the willingness to be vulnerable. To theextent that it even is an exchange currency, abilityseems to represent one that is much different fromthe other trustworthiness dimensions and the jus-tice dimensions. In Foa and Foa’s (1980) conceptu-alization of exchange resources, competence, skills,and efficiency seem more concrete, bordering onthe goods and services classes rather than the statusand information classes. This conceptualizationtherefore seems less relevant to the discretionaryexchanges at the core of Blau’s (1964) theorizing.Benevolence does seem to belong in the same classas the other justice dimensions, particularly inter-personal justice, but it encompasses a depth ofrelationship and loyalty that the justice formsdo not (Mayer et al., 1995; Mayer & Davis, 1999).Similarly, integrity encompasses concepts that layoutside of any of the four justice forms, includingvalue congruence, discreetness, reliability, andpromise fulfillment (Mayer et al., 1995; Mayer &Davis, 1999). We therefore hypothesized:

Hypothesis 1. Procedural, distributive, inter-personal, and informational justice at time 1are positively related to the willingness to bevulnerable form of trust at time 2.

Hypothesis 2. Ability, benevolence, and integ-rity at time 1 are positively related to the will-ingness to be vulnerable form of trust at time 2.

Predicting Justice: The Relational Model

As noted above, many of the trustworthiness andjustice dimensions are functionally similar and arelikely to be highly related to one another. We nowexplore whether those relationships are causal.One theory that is relevant in this regard is therelational model (Tyler, 1994; Tyler & Lind, 1992).Originally termed the “group-value model” (Lind &Tyler, 1988; Tyler, 1989), this model arose as ameans of explaining why voice improved reactionsto decision events even when it did not result inmore favorable outcomes. The basic assumption ofthe group-value model is that people value mem-berships in social groups because those groups pro-vide a source of self-validation. The model furthersuggests that individuals look to three judgments assignals of how much their groups value them (Ty-ler, 1989: 831): (1) trust, which involves “the beliefthat the intentions of a third party are benevolent,that they desire to treat people in a fair and reason-able way”; (2) neutrality, which involves whetheran authority in question “is an unbiased decisionmaker who is honest and who uses appropriatefactual information to make decisions”; and (3)standing, which involves “polite and respectfultreatment.”

In his initial test, Tyler (1989) focused on citizensof Chicago. He suggested that the city itself mightcomprise a meaningful group and that citizenscould look to their treatment by police or legalauthorities as signals of how much the city valuedthem. Tyler measured reactions to decision eventsusing several outcomes, including procedural anddistributive justice. He also gathered data on out-come favorability by asking individuals whetheractual decisions were in their favor. The studyshowed that all three group-value judgments weresignificant predictors of procedural and distribu-tive justice, even when outcome favorability wascontrolled for. Thus, citizens did seem to look totrust, neutrality, and standing when deciding howto react to city authorities. At first glance, theseresults seem in direct opposition to Hypothesis 1,as they imply a trust to justice causal order. How-ever, it must be noted that Tyler (1989) conceptu-alized trust as positive expectations about the be-nevolence and fairness of authority. In Mayer andcolleagues’ (1995) terms, that variable would belabeled a combination of benevolence and integ-rity. Thus, Tyler’s (1989) study actually supports atrustworthiness to justice prediction.

The group-value model was subsequently rela-beled the “relational model,” with the three group-value judgments now grouped under the “rela-tional judgments” heading (Tyler, 1994; Tyler &


Lind, 1992). One change that is particularly rele-vant to this discussion was a shift from the trustlabel to the trustworthiness label in descriptions ofthe relational judgments. Tyler (1994) conceptual-ized trustworthiness as positive perceptions aboutthe motives of authority, which would fall underMayer and colleagues’ (1995) benevolence facet.Indeed, Tyler (1994) himself utilized the benevo-lence label in one passage referring to the trustwor-thiness variable and asked employees about theirrecent experiences with a supervisor in getting helpon work-related problems, making resource alloca-tion decisions, or settling disputes. As in the earlierstudy, Tyler (1994) measured reactions to decisionevents using several outcomes, including proce-dural and distributive justice, and also gathereddata on outcome favorability. The results of thestudy again showed that trustworthiness, neutral-ity, and standing were significant predictors of pro-cedural and distributive justice, even given con-trols for outcome favorability.

Taken together, these studies suggest that indi-viduals look to three judgments—trustworthi-ness, neutrality, and standing—as signals of howmuch they are valued by their groups (Tyler,1989, 1994). They consider those issues most in-tently when reacting to decision-making eventsand, as a result, those three judgments help shapejustice perceptions. With respect to the role oftrustworthiness specifically, the perception thatauthorities have benevolence and integrity is be-lieved to create a lens through which people viewsubsequent decision-making actions over thelong term (Tyler, 1989, 1994). That rationale isconsistent with suggestions by trust scholars. Forexample, Robinson (1996) argued that believingauthorities to be trustworthy creates selectiveperception, reducing the likelihood that the au-thorities’ actions will be perceived as breachingagreed-upon codes of conduct. Similarly, Dirksand Ferrin (2001) noted that many of the actionsobserved in organizations are ambiguous to somedegree, meaning that perceived trustworthinesscan shape interpretations.

With regard to the time span involved in thisrelationship, it is important to note that manyscholars believe that justice judgments take time todevelop because they are based on an accumulationof experiences with a supervisor, such as resourceallocations, dispute resolutions, or other similarevents (Cropanzano et al., 2001). The reasoning thatunderlies the relational model suggests that evi-dence of trustworthiness—which also accumulatesover repeated interactions with the supervisor(Mayer et al., 1995)—is one piece of informationthat employees collect during these experiences.

We therefore advanced the predictions below forthe relationships among benevolence, integrity,and the justice dimensions. Note that we omittedability from our hypotheses because tests of therelational model have focused more on the benev-olence and integrity of authorities than on theircompetence (Lind, Tyler, & Huo, 1997; Tyler, 1989,1994; Tyler, Degoey, & Smith, 1996). Note also thatwe included all four justice dimensions in our hy-potheses. Although the relational model discussesreactions to decision events in relatively generalterms, Tyler’s (1989, 1994) tests of the model haveincluded procedural and distributive justice, omit-ting the interactional justice facets. Two reasonscould explain this omission. First, the justice liter-ature as a whole has tended to lump procedural andinteractional justice together. Second, elements ofinterpersonal justice are already represented in thestanding portion of the model. Nevertheless, thecore logic of the model would apply to interper-sonal and informational reactions as well. Wetherefore predicted:

Hypothesis 3a. Benevolence at time 1 is posi-tively related to procedural, distributive, inter-personal, and informational justice at time 2.

Hypothesis 3b. Integrity at time 1 is positivelyrelated to procedural, distributive, interper-sonal, and informational justice at time 2.

Predicting Trustworthiness: FairnessHeuristic Theory

One other theory contains propositions relevantto the justice-trustworthiness connection. Fairnessheuristic theory suggests that individuals in organ-izations face a “fundamental social dilemma”wherein cooperating with authorities can lead tobetter outcomes but can also increase the risk ofexploitation (Lind, 2001; Van den Bos, 2001). Ac-cording to the theory, in an effort to cope with thisdilemma, individuals use a “fairness heuristic” as ashortcut to determine whether or not to cooperate.The theory details two general phases of workinglife: a relatively short “judgmental phase” and alonger “use phase.” During the judgmental phase,individuals quickly arrive at a general justice judg-ment by assessing the available information. Thatinformation could be procedural, distributive, in-terpersonal, or informational in nature—the theo-ry’s argument is that the individuals will use what-ever information is encountered first and is mostinterpretable to create the judgment. In the usephase, they assume this general heuristic to be ac-curate, and it impacts daily attitudes and behaviors(e.g., prosocial behavior, acceptance of authority).


In talking about the importance of the funda-mental social dilemma to the development offairness heuristic theory, Van den Bos wrote,“These ideas led us to ask: Do people often havedirect information about an authority’s trustwor-thiness? We suggested that they frequentlydo not. . . . We suggested that in such situa-tions—in which information about the authori-ty’s trustworthiness is missing—people refer tothe fairness of the authority’s procedures to de-cide how to react” (2001: 73). This logic seems tosuggest that individuals will use data on justiceas a signal when they form impressions of trust-worthiness. However, the theory does not explic-itly make that prediction, instead focusing onuncertainty about trustworthiness as a moderatorof justice-outcome relationships. For example, inone experiment by Van den Bos, Wilke, and Lind(1998), participants were asked to imagine thatthey were applying for a grant, with the trustwor-thiness of the grant committee being manipulatedto be high, low, or unknown. Procedural justicewas shown to have stronger effects on satisfac-tion ratings when trustworthiness was unknownthan when it was either high or low.

The Van den Bos and colleagues (1998: 1453)experiments manipulated trustworthiness informa-tion in a global fashion (i.e., by telling participantsthat the authority was “not to be trusted at all” or is“to be trusted completely”). Trustworthinesswas not framed in more specific terms using infor-mation about ability, benevolence, or integrity.Moreover, the fact that justice and trustworthinesswere manipulated so that they were orthogonal toone another prevented any sort of investigation ofthe justice-trustworthiness connection. However,discussions of fairness heuristic theory do make thecase that justice perceptions can be formed morequickly and easily than trustworthiness judgments(Lind, 2001; Van den Bos, 2001). This is becausejustice can be gauged in terms of adherence to rulessuch as consistency, accuracy, correctability, re-spect, and justification—rules that are frequentlyencountered and often clear in their interpretation(Bies & Moag, 1986; Leventhal, 1980). In contrast,Mayer and colleagues (1995) suggested that trust-worthiness is often difficult to gauge, particularlyin the case of benevolence. Such differences sug-gest that individuals will form justice perceptionsbefore they form trustworthiness perceptions andbe relatively more certain about the former, whichcomprise data for the assessment of trustworthi-ness. This line of reasoning further suggests that, ifjustice perceptions are to inform trustworthinessimpressions, employees need sufficient opportu-

nity to experience justice-laden events. We there-fore predicted:

Hypothesis 4a. Procedural, distributive, inter-personal, and informational justice at time 1are positively related to benevolence at time 2.

Hypothesis 4b. Procedural, distributive, inter-personal, and informational justice at time 1are positively related to integrity at time 2.

METHODS

Participants

Participants in this study were registered alumnifrom a large southeastern university. They wereemployed in a variety of different industries, themost common being retail, law, and hospitality.On average, participants were 44 years old(s.d. � 7.9 years), and 32 percent were female.Their average tenure with their current employerwas 11.2 years (s.d. � 7.5 years), a value that isconsistent with those in past research integratingjustice with trust. Of the 13 studies cited in Appen-dix A, 5 had average respondent tenure levels of10 years or more (Aryee et al., 2002; Kernan &Hanges, 2002; Kim & Mauborgne, 1993; Konovsky& Pugh, 1994; Korsgaard, Sapienza, and Schweiger,2002). That correspondence is helpful to supportthe notion that any differences between our resultsand extant research are a function of our decompo-sition of trustworthiness and trust and our use oflongitudinal methods (as opposed to being a func-tion of differences in tenure levels).

Procedure

The present study used a longitudinal designwith two periods. George and Jones (2000) notedthat most of the relationships studied in organiza-tional behavior are not instantaneous, with changesin one construct instantly accompanied by changesin another. Instead, “some level of time aggregationis necessarily involved in most, if not all, theoriz-ing” (George & Jones, 2000: 670). As noted in thedevelopment of our hypotheses, the key question intesting our predictions concerns the time lag be-tween the presence of causes and the demonstra-tion of effects. Our goal was to select a time lag thatwas appropriate given the theoretical parametersreviewed in our introduction, which meant meet-ing four conditions: (1) that employees could haveinteractions with their supervisors that could sup-ply relevant data on trustworthiness, (2) that em-ployees could experience the sorts of decisionevents that can be used to gauge justice, (3) that


employees could experience particular choices andcontexts in which a willingness to be vulnerable totheir supervisor could be considered, and (4) thatemployees would have enough time to engage in acontrolled, cognitive translation of attitudes (e.g.,justice, trustworthiness) into behavioral intentions(e.g., trust). We felt that these parameters pointed toa lag of months, as opposed to seconds, minutes,weeks, or years. Within that broad range, we spe-cifically chose four months because it fulfilled theparameters above and fell within the boundariesused in prior multiwave studies of justice and trust,which have used lags from six weeks to ten months(e.g., Frazier, Johnson, Gavin, Gotty, & Snow, 2010;Kernan & Hanges, 2002).

Participants were recruited by mail through aletter describing the longitudinal nature of thestudy, which was described as focusing on “atti-tudes toward supervisors.” This mailing also in-cluded a consent form and the first survey. Eighthundred thirty participants were sent the initialrequest letter and survey. One hundred sixty-nineof these requests were undeliverable, resulting inan initial pool of 661 individuals. Of these, 256people completed the first survey, for a time 1response rate of 39 percent. Four months after com-pleting the first survey, participants were mailedthe second survey. Two hundred nine individualsreturned the second survey, resulting in a time 2response rate of 82 percent. Two weeks after mail-ing each survey, we sent reminder postcards toparticipants who had not yet returned their com-pleted survey. Of the final 209 participants, 7 hadincomplete data and were removed from the study,resulting in a final sample of 202 participants (andan overall response rate of 31 percent). Data collec-tion took place from January 2006 to January 2007,and individuals received a $17 honorarium fortheir complete participation. Specifically, we at-tached $2 to the first survey and provided the re-mainder of the payment as the surveys were com-pleted and returned. Participants were also mailedreminder postcards to ensure the return of theircompleted surveys.

Measures

The study’s framing and instructions utilizedparticipants’ immediate supervisors as the refer-ents for their responses, rather than a more generalauthority such as top management or their overallorganization. Although justice researchers have tra-ditionally referenced procedural and distributivejustice to organization and interpersonal and infor-mational justice to supervisor, scholars now recog-nize that each of the justice dimensions can ema-

nate from either organizational or supervisorysources (e.g., Blader & Tyler, 2003; Rupp & Cropan-zano, 2002). From this perspective, adherence tojustice rules can be a function of a supervisor’spersonal leadership style as much as of any formalsystem or set of policies. To guard against itempriming or item context effects on the survey, wevaried the order of the trustworthiness, justice, andtrust scales (Podsakoff, MacKenzie, Lee, & Podsa-koff, 2003). Specifically, some versions of the sur-vey began with the trustworthiness scales, othersbegan with the justice scales, and still others beganwith the trust scale. The order of the remaining twosets of scales was then varied within each of thoseversions. All measures utilized previously pub-lished scales. Table 1 contains a complete listof items.

Trustworthiness. The trustworthiness facetswere measured using scales designed by Mayer andDavis (1999) to assess the ability, benevolence, andintegrity of a supervisor. Individuals used a scaleranging from 1, “strongly disagree,” to 5, “stronglyagree,” to assess the extent to which they agreedwith the statements in each scale. Given concernsabout the inclusion of fairness content in trustwor-thiness and trust scales, we omitted two itemswhen creating the integrity scale: “My supervisorhas a strong sense of justice” and “My supervisortries hard to be fair in dealing with others.” It isimportant to note, however, that the tests of ourhypotheses did not differ when our model testingincluded versus excluded these items. Coefficientalpha values for times 1 and 2, respectively, were:ability, .96 and .96; benevolence, .94 and .95; integ-rity, .87 and .85.

Justice. We measured participant perceptionsof procedural justice, distributive justice, inter-personal justice, and informational justice withthe scale developed and validated by Colquitt(2001). Extent of agreement was assessed on ascale ranging from 1, “to a very small extent,” to5, “to a very large extent.” For procedural justice(� � .86, time 1, and .90, time 2), participantswere told to “refer to the procedures your imme-diate supervisor uses to make decisions aboutpay, rewards, evaluations, promotions, assign-ments, etc.” when assessing their agreement witheach of seven items. These seven items assessedadherence to the rules outlined in Leventhal(1980) and Thibaut and Walker (1975). For dis-tributive justice (� � .97, time 1, and .98, time 2),participants were instructed to “refer to the out-comes you receive from your job, such as pay,rewards, evaluations, promotions, assignments,etc.” when assessing their agreement with fouritems. This four-item scale assessed adherence to


TABLE 1Factor Loadings and Equivalence in Measurement Model

ItemsTime 1Loading

Time 2Loading

EqualityConstraint

Ability1. My supervisor is very capable of performing his/her job. 0.95 0.97 0.122. My supervisor is known to be successful at the things s/he tries to do. 0.94 0.92 0.093. My supervisor has a lot of knowledge about the work to be done. 0.87 0.89 0.054. I feel very confident about my supervisor’s skills. 0.97 0.98 0.005. My supervisor has specific capabilities that can increase our

performance.0.88 0.91 0.14

6. My supervisor is well qualified. 0.98 0.99 0.02Benevolence1. My supervisor is very concerned about my welfare. 0.96 0.96 0.002. My needs and desires are very important to my supervisor. 0.93 0.97 0.413. My supervisor would not knowingly do anything to hurt me. 0.84 0.84 0.014. My supervisor really looks out for what is important to me. 0.96 0.97 0.035. My supervisor will go out of his/her way to help me. 0.90 0.93 0.17Integrity1. My supervisor has a strong sense of justice. Omitted from analysis2. I never have to wonder whether my supervisor will stick to his/her

word.0.84 0.87 0.20

3. My supervisor tries hard to be fair in dealing with others. Omitted from analysis4. My supervisor’s actions and behaviors are not very consistent. 0.63 0.54 0.825. I like my supervisor’s values. 0.94 0.95 0.046. Sound principles seem to guide my supervisor’s behaviors. 0.96 0.97 0.02Procedural justice1. Are you able to express your views during those procedures? 0.71 0.66 0.362. Can you influence the decisions arrived at by those procedures? 0.70 0.73 0.073. Are those procedures applied consistently? 0.74 0.86 2.404. Are those procedures free of bias? 0.81 0.87 0.675. Are those procedures based on accurate information? 0.79 0.91 2.296. Are you able to appeal the decisions arrived at by those procedures? 0.60 0.73 2.147. Do those procedures uphold ethical and moral standards? 0.82 0.88 0.56Distributive justice1. Do those outcomes reflect the effort you have put into your work? 0.95 0.95 0.012. Are those outcomes appropriate for the work you have completed? 0.93 0.99 0.893. Do those outcomes reflect what you have contributed to your work? 1.00 1.00 0.014. Are those outcomes justified, given your performance? 0.97 0.99 0.11Interpersonal justice1. Has your supervisor treated you in a polite manner? 0.95 0.94 0.042. Has your supervisor treated you with dignity? 0.99 0.99 0.003. Has your supervisor treated you with respect? 0.99 0.99 0.014. Has your supervisor refrained from improper remarks or comments? 0.82 0.83 0.02Informational justice1. Has your supervisor been candid when communicating with you? 0.82 0.87 0.452. Has your supervisor explained decision-making procedures

thoroughly?0.90 0.93 0.23

3. Were your supervisor’s explanations regarding procedures reasonable? 0.94 0.95 0.014. Has your supervisor communicated details in a timely manner? 0.77 0.85 1.195. Has your supervisor tailored communications to meet individuals’

needs?0.83 0.86 0.09

Trust1. I would be comfortable giving my supervisor a task or problem that

was critical to me, even if I could not monitor his/her actions.0.79 0.81 0.07

2. If someone questioned my supervisor’s motives, I would give mysupervisor the benefit of the doubt.

0.88 0.90 0.06

3. I would be willing to let my supervisor have complete control over myfuture in this company.

0.69 0.78 1.37

4. I really wish I had a good way to keep an eye on my supervisor. 0.73 0.65 0.895. If I had my way, I wouldn’t let my supervisor have any influence over

issues that are important to me.0.75 0.69 0.48


an equity rule (Adams, 1965; Leventhal, 1976).For interpersonal and informational justice, par-ticipants were simply asked to refer to “theirimmediate supervisor” when assessing agree-ment with the respective items. The items mea-suring interpersonal justice (� � .93, time 1, and.94, time 2) referred to the interpersonal treat-ment participants received from their supervisorswhile procedures were enacted. These four itemsassessed adherence to the respect and proprietyrules outlined in Bies and Moag (1986). For in-formational justice (� � .90, time 1, and .93, time2), participants were asked about their supervi-sors’ communications and explanations for deci-sion making. The five items assessed adherenceto the justification and truthfulness rules out-lined in Bies and Moag (1986).

Trust. We assessed trust with the five-item mea-sure developed by Mayer and Gavin (2005), anupdate of Mayer and Davis’s (1999) measure. Ex-tent of agreement was assessed on scale rangingfrom 1, “strongly disagree,” to 5, “strongly agree.”These items were designed to capture participants’willingness to be vulnerable to their supervisor.The coefficient alpha for the trust scale was .82 attime 1 and .84 at time 2.

Data Analysis

The first step in analyzing our data was examin-ing the adequacy of our measurement model. Asshown in Table 2, our hypothesized eight-factormodel provided a good fit to the data at both time 1(�2 � 1,163.16; �2/df � 1.63; CFI � .99; IFI � .99;SRMR � .047; RMSEA � .050) and time 2(�2 � 1,130.45; �2/df � 1.55; CFI � .99; IFI � .99;SRMR � .041; RMSEA � .052). Good model fit istypically inferred when the �2/df ratio falls below

3, when CFI and IFI rise above .90, and when SRMRand RMSEA fall near .05 (Kline, 2005). However,given the conceptual commonalities among manyof our constructs, we investigated alternative mea-surement models. These included a model thatcombined all three trustworthiness facets into onefactor, a model that combined trustworthiness andtrust into one factor, a model that combined inter-personal and informational justice into an “interac-tional justice” factor, a model that combined allthree forms of “process justice” into one factor, anda model that combined all four justice dimensionsinto a single “organizational justice” factor. All ofthese models resulted in a significant decrement infit as judged by a chi-square difference test, sohypothesis testing proceeded with the eight-factormodel. Table 1 provides the factor loadings for theeight scales. All factor loadings were statisticallysignificant, with time 1 and time 2 averages asfollows: ability (.93 and .94), benevolence (.92 and.93), integrity (.84 and .83), procedural justice (.74and .80), distributive justice (.96 and .98), interper-sonal justice (.94 and .94), informational justice(.85 and .89), and trust (.77 and .77).

Our hypotheses were tested using cross-laggedstructural equation modeling. Farrell (1994) pro-vided an extensive review of the steps involved inthis approach to analyzing longitudinal data (seeEpitropaki and Martin [2005] for one application inthe organizational behavior literature). One of thefirst steps in the analysis is demonstrating equiva-lence in the measurement model across time peri-ods. To test for equivalence, we introduced anequality constraint to the eight-factor model on anitem-by-item basis. Table 1 provides the chi-squarevalue for that constraint; values above 3.84 indicatethat constraining the loadings to be equal resultedin a significant decrement in fit (Kline, 2005). None

TABLE 2Comparison of Alternative Measurement Models

Model �2 df �2/df CFI IFI SRMR RMSEA

Time 1Eight-factor model 1,163.16 712 1.63 .99 .99 .047 .050 (.044, .056)Combined “trustworthiness” factor 1,915.24 725 2.64 .97 .97 .062 .115 (.110, .120)Combined “trustworthiness and trust” factor 1,974.22 730 2.70 .97 .97 .063 .116 (.111, .121)Combined “interactional justice” factor 1,468.52 719 2.04 .98 .98 .061 .082 (.077, .086)Combined “process justice” factor 1,704.31 725 2.35 .98 .98 .061 .096 (.091, .101)Combined “organizational justice” factor 2,226.81 730 3.05 .96 .96 .071 .121 (.117, .126)Time 2Eight-factor model 1,103.45 712 1.55 .99 .99 .041 .052 (.045, .058)Combined “trustworthiness” factor 1,845.68 725 2.55 .98 .98 .056 .112 (.107, .116)Combined “trustworthiness and trust” factor 1,870.15 730 2.56 .98 .98 .056 .111 (.106, .116)Combined “interactional justice” factor 1,338.95 719 1.86 .99 .99 .045 .076 (.071, .082)Combined “process justice” factor 1,498.57 725 2.07 .98 .98 .048 .087 (.082, .092)Combined “organizational justice” factor 2,184.22 730 2.99 .97 .97 .060 .117 (.113, .122)


of the chi-square values were significant, suggest-ing a fairly robust factor structure. Error covari-ances were then modeled between all of the time 1and time 2 indicators, because repeated measuresof the same variable often result in correlatederrors.

Within the structural model, each time 2 variablewas represented by its time 1 counterpart and theparticular variables included in our hypotheses.We also modeled ability effects even when nothypothesized, to make our results as informative aspossible. Following recommendations for cross-lagged structural equation modeling, we modeledno cause-effect relationships between variablesmeasured at the same time (Epitropaki & Martin,2005; Farrell, 1994). However, so-called synchro-nous correlations were modeled among the time 1and time 2 variables to capture within-time covari-ation (Epitropaki & Martin, 2005; Farrell, 1994).Time 1 synchronous correlations were modeled byallowing the exogenous latent variables to covary,and time 2 synchronous correlations were modeledby allowing the disturbance terms for the endoge-nous latent variables to covary. The combination ofall of these paths resulted in a good fit to the data(�2 � 3,986.13; �2/df � 1.35; CFI � .99; IFI � .99;SRMR � .050; RMSEA � .042).

RESULTS

Table 3 shows the means, standard deviations,and zero-order correlations among the time 1 andtime 2 variables. Table 4 summarizes the structuralpath coefficients used to test our hypotheses. Fig-

ures 3A–3C provide visual representations of thoseanalyses, with trust, justice, and trustworthiness asthe respective outcomes. It is important to note thatit is these three figures, in combination, that repre-sent the structural model being tested. In the inter-est of clarity, the figures omit factor loadings, errorcovariances between time 1 and time 2 indicators,covariances among latent exogenous variables, co-variances among time 2 disturbance terms, andnonsignificant paths.

Hypothesis 1 predicts that procedural, distrib-utive, interpersonal, and informational justicemeasured at time 1 will be positively related totrust measured at time 2. Hypothesis 2 predictsthat ability, benevolence, and integrity measuredat time 1 will be positively related to trust mea-sured at time 2. The portion of our structuralmodel relevant to these hypotheses regressedtime 2 trust onto its time 1 counterpart, alongwith the time 1 justice and trustworthiness di-mensions. As shown in Table 4 and Figure 3A,both hypotheses were partially supported. Whenthe analysis controlled for trust’s stability coeffi-cient (� � .35), informational justice (� � .18),benevolence (� � .19), and integrity (� � .21) hadsignificant unique relationships with trust attime 2. These results support the notion that jus-tice leads to trust, even when both trustworthi-ness and prior trust levels are controlled for.They also suggest that informational justice hasspecial relevance to the willingness to bevulnerable.

Hypothesis 3 predicts that benevolence and in-tegrity assessed at time 1 will be positively re-

TABLE 3Means, Standard Deviations, and Zero-Order Correlationsa

Variable Mean s.d. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time 11. Ability 4.12 0.882. Benevolence 3.76 0.91 .673. Integrity 3.77 0.89 .66 .784. Procedural justice 3.47 0.76 .49 .59 .585. Distributive justice 3.56 1.08 .43 .53 .46 .616. Interpersonal justice 4.34 0.83 .47 .68 .64 .51 .587. Informational justice 3.70 0.85 .65 .76 .73 .65 .51 .638. Trust 3.70 0.84 .70 .70 .73 .65 .43 .57 .69

Time 29. Ability 4.12 0.84 .75 .57 .59 .46 .46 .52 .61 .64

10. Benevolence 3.70 0.92 .53 .77 .69 .56 .54 .64 .66 .64 .6811. Integrity 3.74 0.86 .55 .65 .74 .52 .45 .59 .63 .67 .70 .8112. Procedural justice 3.46 0.80 .44 .58 .60 .70 .53 .51 .58 .56 .57 .75 .7213. Distributive justice 3.56 1.04 .31 .46 .43 .51 .67 .48 .42 .36 .41 .57 .54 .6514. Interpersonal justice 4.32 0.83 .44 .59 .61 .45 .49 .79 .55 .51 .60 .70 .75 .63 .5315. Informational justice 3.65 0.87 .54 .66 .69 .54 .48 .60 .71 .61 .66 .79 .79 .79 .58 .7216. Trust 3.64 0.81 .55 .66 .70 .52 .46 .54 .63 .71 .71 .80 .84 .72 .54 .67 .75

a n � 202. All correlations are significant at p � .01.


lated to procedural, distributive, interpersonal,and informational justice at time 2. The portionof our structural model relevant to these hypoth-eses regressed the time 2 justice dimensions ontotheir time 1 counterparts, along with the time 1trustworthiness dimensions. As shown in Table 4and Figure 3B, the hypothesis was partially sup-ported. When we controlled for their stabilitycoefficients (which ranged from .48 to .66), integ-rity had significant, unique relationships withprocedural justice (� � .25), distributive justice(� � .30), interpersonal justice (� � .25), andinformational justice (� � .23). These results sup-port the notion that the integrity dimension oftrustworthiness leads to justice, though the be-nevolence dimension does not. Effects for abilitywere modeled on an exploratory basis; abilityhad a significant, negative relationship with dis-tributive justice (� � �.15).

Hypothesis 4 predicts that procedural, distribu-tive, interpersonal, and informational justice mea-sured at time 1 will be positively related to benev-olence and integrity at time 2. The portion of ourstructural model relevant to these hypotheses re-gressed the time 2 trustworthiness dimensions ontotheir time 1 counterparts, along with the time 1justice dimensions. As shown in Figure 3C, thehypothesis was partially supported. Interpersonaljustice (� � .12) had a significant, unique relation-ship with benevolence when the analysis con-trolled for the benevolence stability coefficient(� � .66). Procedural justice (� � .15) and interper-sonal justice (� � .14) had significant, unique rela-tionships with integrity when the analysis con-trolled for the integrity stability coefficient(� � .48). Effects for ability were modeled and werenonsignificant, as expected.

DISCUSSION

In the present study, we attempted to move be-yond the conventional wisdom that justice leads totrust by bringing more nuanced theorizing andmore rigorous methods to the study of this impor-tant connection. Utilizing Mayer and colleagues’(1995) distinction between trustworthiness andtrust allowed us to look separately at the relation-ship between justice and the willingness to be vul-nerable and the relationship between justice andtrustworthiness. As a result, several key findingsemerged. First, with regard to the relationship be-tween justice and trust, informational justice ap-pears to hold a unique relevance to the willingnessto be vulnerable. Of the four justice dimensions, itwas the only one that predicted trust when trust-worthiness was considered simultaneously. Sec-ond, with regard to the relationship between justiceand trustworthiness, there appears to be a recipro-cal relationship between justice and benevolenceand integrity. Managers who adhered to proceduraland interpersonal justice rules were perceived to bemore trustworthy, and vice versa. Third, the con-nection between trustworthiness and justice isweaker when trustworthiness is operationalized inability terms. The competence, skills, and effi-ciency of a manager appear to have little bearing onwhether that manager adheres to justice rules.

Justice and Trust: Informational JusticeStands Out

One of the central theoretical questions exam-ined in our study concerns whether justice reallydoes lead to trust, as the conventional wisdom de-picted in Figure 1 suggests. More specifically,when trust is expressed as the willingness to be

TABLE 4Results of Structural Equation Modelinga

Time 2 Outcomes

Time 1 Predictors Ability Benevolence IntegrityProcedural

JusticeDistributive

JusticeInterpersonal

JusticeInformational

Justice Trust

Ability .66* — — �.06 �.15* .00 �.03 �.09Benevolence — .66* — .09 �.07 �.05 .13 .19*Integrity — — .48* .25* .30* .25* .23* .21*Procedural justice �.07 .11 .15* .55* — — — .03Distributive justice .05 .03 �.06 — .66* — — .04Interpersonal justice .10 .12* .14* — — .66* — �.05Informational justice .14 �.04 .13 — — — .48* .18*Trust — — — — — — — .35*

a Contents are standardized path coefficients. Dashes (—) represent paths fixed to zero within the structural equation model.* p � .05, one-tailed.


vulnerable, with justice concepts and trustworthi-ness concepts removed from its operationalization,does the relationship with justice still hold? Thenotion that fair decision making fosters a willing-ness to accept vulnerability has been a cornerstoneof the application of social exchange theory in thejustice literature (Konovsky & Pugh, 1994; Organ,

1990). Social exchanges cannot develop without awillingness to trust that an exchange partner willeventually discharge his or her obligations (Blau,1964), and justice could provide a signal that fairprinciples will govern exchanges with that partner(Brockner & Siegel, 1995; Konovsky & Pugh, 1994;Organ, 1990). Our results showed that informa-

FIGURE 3APortion of Structural Model Relevant to Hypotheses 1–2a

a The structural model tested represents a combination of Figures 3A, 3B, and 3C. Path coefficients are standardized. The followingpaths are omitted for the purposes of clarity: factor loadings, error covariances between time 1 and time 2 indicators, covariances amongexogenous latent variables, covariances among time 2 disturbance terms, and nonsignificant paths.


tional justice was the only justice dimension topredict the willingness to be vulnerable: employeeswere more trusting of supervisors who explainedand justified decision making in an honest andtruthful manner. That informational effect was sig-nificant even when we controlled for prior levels of

the willingness to be vulnerable and for prior per-ceptions of trustworthiness.

Why does informational justice seem to hold aunique relevance to the willingness to be vulnera-ble, particularly when meta-analyses have also sup-ported the importance of the procedural and dis-

FIGURE 3BPortion of Structural Model Relevant to Hypothesis 3a

a The structural model tested represents a combination of Figures 3A, 3B, and 3C. Path coefficients are standardized. The followingpaths are omitted for the purposes of clarity: factor loadings, error covariances between time 1 and time 2 indicators, covariances amongexogenous latent variables, covariances among time 2 disturbance terms, and nonsignificant paths.


tributive facets to other kinds of trust measures(Cohen-Charash & Spector, 2001; Colquitt et al.,2001; Dirks & Ferrin, 2002)? Two issues may offeranswers to that question. First, it may be that pastresearch has overstated the importance of the otherjustice facets by including justice concepts in trustmeasures. Second, it may be that the other justice

dimensions are more functionally similar to benev-olence and integrity—inspiring a willingness to bevulnerable for the very same reasons as those trust-worthiness dimensions. To borrow a distinctionfrom McAllister’s (1995) work, it may be that infor-mational justice captures affect-based reasons totrust, in addition to the more cognition-based rea-

FIGURE 3CPortion of Structural Model Relevant to Hypothesis 4a

a The structural model tested represents a combination of Figures 3A, 3B, and 3C. Path coefficients are standardized. The followingpaths are omitted for the purposes of clarity: factor loadings, error covariances between time 1 and time 2 indicators, covariances amongexogenous latent variables, covariances among time 2 disturbance terms, and non-significant paths.


sons reflected by trustworthiness. McAllister(1995) described affect-based trust as trust based onemotional investments and ties. In large part, hismeasure of affect-based trust focuses on the mutualsharing of ideas and feelings and the sense thatindividuals can talk freely to one another. In thevalidation of his justice measure, Colquitt (2001)suggested that informational justice conveys asense of inclusion, with being “in the know” sig-naling being “in the in-group.” Thus, managerswho share honest and truthful justifications for keyevents may foster emotional ties with subordi-nates—ties that create their own unique reasons fora willingness to be vulnerable.

These results have important theoretical implica-tions for social exchange theory, particularly as it isapplied in the justice literature. Applications of thetheory have tended to stress procedural justice asthe exchange currency capable of breeding recipro-cation in the form of citizenship and related behav-iors (Konovsky & Pugh, 1994; Organ, 1990). It maybe that informational justice is the more criticalexchange currency. After all, of the four justicedimensions, informational justice best captures theinformation and status resources described in Foaand Foa’s (1980) conceptualization of exchangetheory. Because such resources are highly particu-laristic and symbolic, they may be uniquely capa-ble of fostering the trust levels needed to breed asocial exchange.

Trustworthiness and Justice: A ReciprocalRelationship with Benevolence and Integrity

Applying the Mayer and colleagues (1995) lensrather than the confident, positive expectationsview of trust allowed us to examine, for the firsttime, the relationships among the justice dimen-sions and the ability, benevolence, and integrity ofsupervisors. For benevolence and integrity, our re-sults suggest that the relationship between thosetrustworthiness forms and justice is reciprocal,with both influencing one another, as shown inFigure 2. On the one hand, our results suggest thatintegrity predicts subsequent perceptions of allfour justice dimensions, even when their prior lev-els are controlled for. Managers who stuck to theirword and whose actions were guided by soundvalues and principles were perceived to be fairer.On the other hand, justice perceptions also pre-dicted subsequent perceptions of benevolence andintegrity, even when prior assessments were con-trolled for. Interpersonal justice had a positive ef-fect on perceived benevolence, suggesting that ad-hering to rules such as respect and proprietysignaled that authorities were concerned about the

welfare of employees. Procedural and interpersonaljustice had positive effects on perceived integrity,suggesting that adherence to those rules signaledthat authorities used sound values and principlesto guide their actions.

Taken together, these results have important the-oretical implications for the relational model andfor fairness heuristic theory. With respect to theformer, the relational model suggests that somecombination of benevolence and integrity—whether labeled “benevolence,” “trustworthiness,”or merely “trust” in formal specifications of themodel—acts as an antecedent of justice perceptions(Tyler, 1989, 1994; Tyler & Lind, 1992). Our resultssuggest that this causal flow is an oversimplifica-tion, in that justice guides views on trustworthinessas much as trustworthiness guides views on justice.With respect to the latter, fairness heuristic theoryimplies that justice perceptions guide formation oftrustworthiness perceptions, because informationon justice is encountered earlier (and is more inter-pretable) than information on trustworthiness(Lind, 2001; Van den Bos, 2001; Van den Bos et al.,1998). Once again, that causal flow seems to be anoversimplification. It may be that the perceptionthat an authority has integrity creates a sort of se-lective perception through which people view jus-tice-relevant actions (Dirks & Ferrin, 2001; Robin-son, 1996; Tyler, 1989, 1994).

In our opinion, future work in this area wouldbenefit from a theoretical compromise between therelational model and fairness heuristic theory, oneacknowledging a sound conceptual logic for a trust-worthiness to justice causal order and a justice totrustworthiness causal order. That sort of compro-mise would offer important new insights for themodeling of justice and trustworthiness in futurestudies. Specifically, it would suggest that scholarsshould either model nonrecursive paths betweenjustice and trustworthiness, or simply model themas correlates that occupy the same functional posi-tion in a causal system. If they are exogenous in acausal system, such as in cases in which they pre-dict reciprocative attitudes or behaviors, theywould be connected by a modeled covariance (i.e.,a curved arrow). If they are endogenous in a causalsystem, such as in cases in which they mediateorganizational or supervisory effects, they wouldeither be connected by nonrecursive paths or by adisturbance covariance (neither of which affordscausal primacy). This compromise matches Le-wicki and colleagues’ (2005) suggestion that justiceand trustworthiness can be described as correlatesthat “codevelop.”

An alternative possibility is that justice and trust-worthiness do not codevelop, but rather that justice


is a component of trustworthiness. This possibilityseems most applicable to the relationship betweenjustice and integrity, given that Mayer and Davis’s(1999) integrity scale includes two items that usethe words “justice” or “fairness.” We believe that atleast two conceptual reasons exist for maintainingthe separation of justice and integrity as constructs.First, Mayer et al. (1995) described integrity as an“attribute” or “characteristic” of a trustee. Integrityis something that a person has—just as “trustwor-thy” is something that a person is. In other words,integrity is a personal quality that exists acrosssituations. In contrast, justice reflects somethingthat a person does—more specifically, a set of rulesthat a person adheres to in the context of decision-making and resource allocation events. Second,Mayer and colleagues’ (1995) conceptualizationof integrity includes a number of concepts thatdo not overlap substantially with the rules repre-sented by procedural, distributive, interpersonal,or informational justice. Those concepts includevalue congruence, discreetness, reliability, andpromise fulfillment. Similarly, the rules used tooperationalize justice include a number of con-cepts that do not overlap substantially with in-tegrity. Those concepts include voice, input, ac-curacy, correctability, equity, and justification(Adams, 1965; Bies & Moag, 1986; Leventhal,1980; Thibaut & Walker, 1975).

Of course, even if justice and integrity can beseparated conceptually, the question of how to han-dle the inclusion of fairness content in Mayer andDavis’s (1999) scale remains. In a general sense,measures are created to serve as content-valid op-erationalizations of the constructs within a givenliterature (Schwab, 1980; Stone-Romero, 1994).Aside from working to establish discriminant orpredictive validity, researchers have paid relativelylittle attention to construct definitions or scaleitems in other literatures. This somewhat insularfocus does not become problematic until a givenliterature becomes integrated with another withsimilar construct content. For example, Mowday,Steers, and Porter (1979) defined organizationalcommitment as the strength of an individual’sidentification with and involvement in an organi-zation, characterizing involvement as a willingnessto exert extra effort on behalf of his/her organiza-tion. Mowday and colleagues’ (1979) organiza-tional commitment measure included indicators ofinvolvement such as “I am willing to put in a greatdeal of effort beyond what is expected” and “I talkup this organization to my friends.” Although thatconstruct definition and item content functionedeffectively in the organizational commitment liter-ature, it became problematic when the construct

was integrated with research on organizational cit-izenship behavior, which is defined as employees’discretionary effort that benefits their organization(Organ, 1990), and is measured with items verysimilar to the Mowday and colleagues’ (1979) ex-amples (e.g., Lee & Allen, 2002).

In the same way that the definitions and scalesused for organizational commitment and organiza-tional citizenship behavior evolved over time tominimize the area of overlap, it may be that thedefinitions and scales used for justice and integrityneed to evolve as well. That evolution could occurthrough an emphasis on the construct content thatdoes not overlap between the two constructs, ac-companied by new scales being validated to focuson that nonoverlapping content. Alternatively, thatevolution could occur by scholars’ ceding concep-tual space to one construct or the other on the basisof how central the content is to the constructs. Forexample, it may be that “fairness” is more central tojustice, meaning that it would be included in ajustice scale but not an integrity scale, but that“ethicality” is more central to integrity, meaningthat it would be included in an integrity scale butnot a justice scale. Of course, that sort of integrativeconstruct validation work has to occur for scholarsto begin using such scales. Until those debates andvalidation efforts have been completed, the bestsolution likely remains the omission of items withoverlapping content from scales such as Colquitt’s(2001) or Mayer and Davis’s (1999), as long as theadapted measure continues to tap the core of therelevant construct.

Trustworthiness and Justice: Less Importancefor Ability

Although our results support a nonrecursive re-lationship between justice and both benevolenceand integrity, the same cannot be said for ability.Ability was not a significant predictor of subse-quent justice levels, nor were the justice dimen-sions shown to be predictors of subsequent ability.The one exception concerned distributive justice,as managers who were perceived as more compe-tent and efficient were viewed as less equitable intheir outcome allocations. That result is surprising,as one would expect high levels of ability to trans-late into a more accurate perception of employeeinputs.

We can envision three potential reasons for anegative relationship between ability and distrib-utive justice. First, it may be that more able man-agers create more outcome differentiation in theirunits, because they have a clearer sense of therelevant “compensable factors” that are used in


formal and informal job evaluation efforts (Milk-ovich & Newman, 2002). Although that increaseddifferentiation may seem equitable to able man-agers, self-serving biases may create perceptionsof inequity among the rank and file. Second, itmay be that equity expectations are simply higherfor more able managers, with employees “raisingthe bar” for what constitutes a balanced outcome-input ratio. Third, it may be that able managersearn particularly high salaries precisely becauseof their high competence. Research suggests thatemployees may use upward comparisons whenconsidering equity levels (Scholl, Cooper, &McKenna, 1987), raising the possibility that theywill view their managers’ salaries as too high,even given their ability levels.

Aside from that result, however, the relation-ship between trustworthiness and justice wasdue primarily to benevolence and integrity. Therelative importance of those facets flows out ofboth the relational model and fairness heuristictheory, as neither theory discusses the abilitycomponent of trustworthiness to a significant de-gree. It should be noted, however, that many ofthe trust measures in our appendix included abil-ity concepts, likely driving down relationshipswith justice. Our results therefore offer anotherbenefit of applying the more specific Mayer andcolleagues (1995) lens to the study of justice,trust, and trustworthiness.

Limitations and Suggestions for Future Research

This study has several limitations that should benoted. First, although we used a longitudinal de-sign in which prior levels of the key variables werecontrolled, our data are still ultimately correla-tional. The only way to draw definitive causal con-clusions is to utilize experimental methods inwhich random assignment can be used to eliminatealternative explanations. Second, our study wasconducted using veteran employees, not organiza-tional newcomers. Although this sample made ourstudy consistent with those cited in Appendix A, itshould be noted that fairness heuristic theory’spropositions are centered on newcomers. Third, allof our measures are taken from a common source,raising the possibility of correlation inflation due tocommon method variance. Once again, however,this feature makes our study more comparable tovirtually all of the studies on justice and trust, asthat linkage is commonly examined using same-source data. Our longitudinal design did allow usto separate measurements in time, removing tran-sient sources of method variance such as “stateaffect” (Podsakoff et al., 2003). Fourth, as with

many studies in organizational behavior, our theo-rizing addressed the role of time in the relationshipbetween the constructs on a fairly imprecise basis.Thus, we reiterate the call of George and Jones(2000) to explicitly address time in one’s theoriz-ing, given that a number of different lags may bemore appropriate for relationships among justice,trustworthiness, and trust.

One last limitation points to an important avenuefor future research. Specifically, we neglected toinclude job performance in our study, an omissionpreventing us from assessing the relative perfor-mance effects of trustworthiness, justice, and trust.We decided to omit performance because the focusof our study was on bringing more nuance to thejustice-trust connection, and because justice, trust,and performance variables have been included infour separate meta-analyses (Cohen-Charash &Spector, 2001; Colquitt et al., 2001, 2007; Dirks &Ferrin, 2002), two of which coded the variables atthe level of specificity utilized in our study(Colquitt et al., 2001, 2007). Those results suggestthat all three trustworthiness dimensions, two ofthe four justice dimensions, and the willingness tobe vulnerable have significant relationships withjob performance. The willingness to be vulnerableeffect tends to be stronger than the others, which isconsistent with the view that trust lies “down-stream” from trustworthiness and justice, making ita more proximal predictor of performance. Still,few of the studies included in those meta-analyticreviews looked at relationships over time, and vir-tually none utilized the sort of panel data used inour study. We therefore recommend that scholarscollect performance data alongside justice andtrust data over time, to continue to tease out theperformance implications of these concepts.

As a second suggestion for future research, wewould call for more research integrating justice andtrust using a Mayer and colleagues (1995) lens. Thisstudy is insufficient for balancing the 13 studiesincluded in our appendix that use a confident,positive expectations framework, not to mentionthe other studies that use ad hoc or direct measuresof trust. For conducting such research, we wouldoffer the following suggestions to maximize inter-nal and external validity. With respect to internalvalidity, we continue to suggest that scholars trimany items that could create overlapping contentcorrelations, at least until new scales are validated,so long as doing so does not hinder content validityto an unacceptable degree. In addition, laboratorystudies could be used to examine the potentiallyreciprocal relationship between justice and trust-worthiness. Doing this would involve manipulat-ing information about justice and measuring per-


ceptions of trustworthiness, and manipulatinginformation about trustworthiness and measuringperceptions of justice. The laboratory could alsoprovide a useful venue for examining the cognitivemechanisms that underlie such justice-trustworthi-ness relationships. With respect to external valid-ity, it would be beneficial to explore other relation-ship referents in work integrating justice and trust.Meta-analytic reviews have suggested that manytrust and trustworthiness relationships are robustto use of different kinds of trust referents (Colquittet al., 2007; Dirks & Ferrin, 2002). Future workshould explore whether the relationships observedhere generalize when “top management” or “theoverall organization” is used as the referent for thejustice, trust, and trustworthiness measures.

Practical Implications

Our results have important practical implica-tions for managers. For example, they show thatoffering honest justifications and explanationscontinues to matter even after participants haveformed impressions about their trust in their su-pervisor and their perception of the supervisor’strustworthiness. Fortunately for managers, stud-ies have shown that supervisors can be trained toadhere to justice principles (Skarlicki & Latham,2005). We suspect that many organizations viewsuch training programs as relevant only to newsupervisors, so that those supervisors can “get offon the right foot.” The fact that informationaljustice continues to matter shows that such train-ing could benefit established supervisors as well.Our results for benevolence and integrity alsohave important practical implications. Beingconcerned about the welfare of employees, stick-ing to one’s word, and using sound values andprinciples to guide actions predicted subsequenttrust levels even after trust and justice percep-tions had been formed. It seems likely that super-visors would benefit from the incorporation ofbenevolence and integrity content into the devel-opment and assessment efforts in which they par-ticipate. More frequent feedback from employeeson managers’ benevolence and integrity couldhelp managers nurture and maintain their levelsof perceived trustworthiness.

Conclusions

For all the theoretical and empirical work in-tegrating justice and trust, the connection be-tween the two remains poorly understood (Le-wicki et al., 2005). Indeed, the literature seems tobe marked by the deceptively uncontroversial no-

tion that justice leads to trust (Cohen-Charash &Spector, 2001; Colquitt et al., 2001; Dirks & Fer-rin, 2002). We argued that this conventional wis-dom is the result of coarse theorizing that hashindered the accumulation of valid knowledgeabout the justice-trust connection. Mayer andcolleagues’ (1995) model is capable of bringingmore nuance and operational precision to thisarea of inquiry. Our hope is that future researchwill continue to assess trustworthiness, justice,and trust in combination when testing and apply-ing social exchange theory, the relational model,and fairness heuristic theory. Only then can re-search yield valid prescriptions for how manag-ers can best foster a willingness to be vulnerableon the part of their employees.

REFERENCES

Adams, J. S. 1965. Inequity in social exchange. In L.Berkowitz (Ed.), Advances in experimental socialpsychology, vol. 2: 267–299. New York: AcademicPress.

Ambrose, M. L., & Schminke, M. 2003. Organizationstructure as a moderator of the relationship be-tween procedural justice, interactional justice,perceived organizational support, and supervisorytrust. Journal of Applied Psychology, 88: 295–305.

Aryee, S., Budhwar, P. S., & Chen, Z. X. 2002. Trust as amediator of the relationship between organizationaljustice and work outcomes: Test of a social exchangemodel. Journal of Organizational Behavior, 23:267–285.

Azjen, I. 1991. The theory of planned behavior. Organi-zational Behavior and Human Decision Processes,50: 179–211.

Azjen, I., & Fishbein, M. 1980. Understanding attitudesand predicting social behavior. Englewood Cliffs,NJ: Prentice-Hall.

Begley, T. M., Lee, C., & Hui, C. 2006. Organizationalleadership as a moderator of the relationship be-tween justice perceptions and work-related reac-tions. Journal of Organizational Behavior, 27: 705–721.

Bies, R. J., & Moag, J. F. 1986. Interactional justice: Com-munication criteria of fairness. In R. J. Lewicki, B. H.Sheppard & M. H. Bazerman (Eds.), Research onnegotiations in organizations, vol. 1: 43–55. Green-wich, CT: JAI.

Blader, S. L., & Tyler, T. R. 2003. What constitutes fair-ness in work settings? A four-component model ofprocedural justice. Human Resource ManagementReview, 13: 107–126.

Blau, P. M. 1964. Exchange and power in social life.New York: Wiley.


Brockner, J., & Siegel, P. 1995. Understanding the inter-action between procedural and distributive justice:The role of trust. In R. M. Kramer & T. R. Tyler (Eds.),Trust in organizations: Frontiers of theory and re-search: 390–413. Thousand Oaks, CA: Sage.

Cohen-Charash, Y., & Spector, P. E. 2001. The role ofjustice in organizations: A meta-analysis. Organiza-tional Behavior and Human Decision Processes,86: 278–321.

Colquitt, J. A. 2001. On the dimensionality of organiza-tional justice: A construct validation of a measure.Journal of Applied Psychology, 86: 386–400.

Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter,C. O. L. H., & Ng, K. Y. 2001. Justice at the millen-nium: A meta-analytic review of 25 years of organi-zational justice research. Journal of Applied Psy-chology, 86: 425–445.

Colquitt, J. A., Scott, B. A., & LePine, J. A. 2007. Trust,trustworthiness, and trust propensity: A meta-ana-lytic test of their unique relationships with risk tak-ing and job performance. Journal of Applied Psy-chology, 92: 909–927.

Cook, J., & Wall, T. D. 1980. New work attitude measuresof trust, organizational commitment, and personalneed non-fulfillment. Journal of Occupational Psy-chology, 53: 39–52.

Cropanzano, R., Byrne, Z. S., Bobocel, D. R., & Rupp,D. E. 2001. Moral virtues, fairness heuristics, so-cial entities, and other denizens of organizationaljustice. Journal of Vocational Behavior, 58: 164–209.

Cropanzano, R., & Mitchell, M. S. 2005. Social exchangetheory: An interdisciplinary review. Journal ofManagement, 31: 874–900.

Cropanzano, R., Prehar, C., & Chen, P. Y. 2002. Usingsocial exchange theory to distinguish proceduralfrom interactional justice. Group and OrganizationManagement, 27: 324–351.

Dirks, K. T., & Ferrin, D. L. 2001. The role of trust inorganizational settings. Organization Science, 12:450–467.

Dirks, K. T., & Ferrin, D. L. 2002. Trust in leadership:Meta-analytic findings and implications for researchand practice. Journal of Applied Psychology, 87:611–628.

Epitropaki, O., & Martin, R. 2005. From ideal to real: Alongitudinal study of the role of implicit leadershiptheories on leader-member exchanges and employeeoutcomes. Journal of Applied Psychology, 90: 659–676.

Farrell, A. D. 1994. Structural equation modeling withlongitudinal data: Strategies for examining groupdifferences and reciprocal relationships. Journal

of Consulting and Clinical Psychology, 62: 477–487.

Foa, U. G., & Foa, E. B. 1980. Resource theory: Interper-sonal behavior as exchange. In K. J. Gergen, M. S.Greenberg & R. H. Willis (Eds.), Social exchange:Advances in theory and research: 77–94. NewYork: Plenum.

Folger, R., & Konovsky, M. A. 1989. Effects of proceduraland distributive justice on reactions to pay raisedecisions. Academy of Management Journal, 32:115–130.

Frazier, M. L., Johnson, P. D., Gavin, M., Gotty, J., &Snow, D. B. 2010. Organizational justice, trustwor-thiness, and trust: A multifoci examination. Groupand Organization Management, 35: 39–76.

Gabarro, J. J., & Athos, J. 1976. Interpersonal relationsand communications. Englewood Cliffs, NJ: Pren-tice-Hall.

George, J. M., & Jones, G. R. 2000. The role of time intheory and theory building. Journal of Manage-ment, 26: 657–684.

Gouldner, A. W. 1960. The norm of reciprocity: A pre-liminary statement. American Sociological Review,25: 161–178.

Greenberg, J. 1987. A taxonomy of organizational jus-tice theories. Academy of Management Review,12: 9–22.

Greenberg, J. 1993. The social side of fairness: Interper-sonal and informational classes of organizationaljustice. In R. Cropanzano (Ed.), Justice in the work-place: Approaching fairness in human resourcemanagement: 79–103. Hillsdale, NJ: Erlbaum.

Kernan, M. C., & Hanges, P. J. 2002. Survivor reactionsto reorganization: Antecedents and consequencesof procedural, interpersonal, and informationaljustice. Journal of Applied Psychology, 87: 916–928.

Kim, W. C., & Mauborgne, R. A. 1993. Procedural justice,attitudes, and subsidiary top management compli-ance with multinationals’ corporate strategic deci-sions. Academy of Management Journal, 36: 502–526.

Kline, R. B. 2005. Principles and practice of structuralequation modeling. New York: Guilford.

Konovsky, M. A., & Pugh, S. D. 1994. Citizenship behav-ior and social exchange. Academy of ManagementJournal, 37: 656–669.

Korsgaard, M. A., Sapienza, H. J., & Schweiger, D. M. 2002.Beaten before begun: The role of procedural justice inplanning change. Journal of Management, 28: 497–516.

Korsgaard, M. A., Schweiger, D. M., & Sapienza, H. J.1995. Building commitment, attachment, and trust


in strategic decision-making teams: The role of pro-cedural justice. Academy of Management Journal,38: 60–84.

Lee, K., & Allen, N. J. 2002. Organizational citizenshipbehavior and workplace deviance: The role of affectand cognitions. Journal of Applied Psychology, 87:131–142.

Leventhal, G. S. 1976. The distribution of rewards andresources in groups and organizations. In L. Berkow-itz & W. Walster (Eds.), Advances in experimentalsocial psychology, vol. 9: 91–131. New York: Aca-demic Press.

Leventhal, G. S. 1980. What should be done with equitytheory? New approaches to the study of fairness insocial relationships. In K. Gergen, M. Greenberg & R.Willis (Eds.), Social exchange: Advances in theoryand research: 27–55. New York: Plenum.

Lewicki, R. J., & Bunker, B. B. 1995. Trust in relation-ships: A model of development and decline. In B. B.Banker & J. Z. Rubin (Eds.), Conflict, cooperation,and justice: 133–173. San Francisco: Jossey-Bass.

Lewicki, R. J., Wiethoff, C., & Tomlinson, E. C. 2005.What is the role of trust in organizational justice?In J. Greenberg & J. A. Colquitt (Eds.), Handbook oforganizational justice: 247–270. Mahwah, NJ: Erl-baum.

Lind, E. A. 2001. Fairness heuristic theory: Justice judg-ments as pivotal cognitions in organizational rela-tions. In J. Greenberg & R. Cropanzano (Eds.), Ad-vances in organizational justice: 56–88. Stanford,CA: Stanford University Press.

Lind, E. A., & Tyler, T. R. 1988. The social psychology ofprocedural justice. New York: Plenum.

Lind, E. A., Tyler, T. R., & Huo, Y. J. 1997. Proceduralcontext and culture: Variation in the antecedents ofprocedural justice judgments. Journal of Personal-ity and Social Psychology, 73: 767–780.

Mayer, R. C., & Davis, J. H. 1999. The effect of the per-formance appraisal system on trust for management:A field quasi-experiment. Journal of Applied Psy-chology, 84: 123–136.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. 1995. Anintegrative model of organizational trust. Academyof Management Review, 20: 709–734.

Mayer, R. C., & Gavin, M. B. 2005. Trust in managementand performance: Who minds the shop while theemployees watch the boss? Academy of Manage-ment Journal, 48: 874–888.

McAllister, D. J. 1995. Affect- and cognition-based trustas foundations for interpersonal cooperation in or-ganizations. Academy of Management Journal, 38:24–59.

Milkovich, G. T., & Newman, J. M. 2002. Compensation.New York: Irwin/McGraw-Hill.

Mowday, R. T., Steers, R. M., & Porter, L. W. 1979. Themeasurement of organizational commitment. Jour-nal of Vocational Behavior, 14: 224–247.

Nyhan, R. C., & Marlowe, H. A. 1997. Development andpsychometric properties of the organizational trustinventory. Education Review, 21: 614–635.

Organ, D. W. 1990. The motivational basis of organiza-tional citizenship behavior. In L. L. Cummings &B. M. Staw (Eds.), Research in organizational be-havior, vol. 12: 43–72. Greenwich, CT: JAI.

Pillai, R., Schriesheim, C. A., & Williams, E. S. 1999.Fairness perceptions and trust as mediators for trans-formational and transactional leadership: A two-sample study. Journal of Management, 25: 897–933.

Pillai, R., Williams, E. S., & Tan, J. J. 2001. Are the scalestipped in favor of procedural or distributive justice?An investigation of the U.S., India, Germany, andHong Kong (China). International Journal of Con-flict Management, 12: 312–332.

Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsa-koff, N. P. 2003. Common method biases in behav-ioral research: A critical review of the literature andrecommended remedies. Journal of Applied Psy-chology, 88: 879–903.

Roberts, K. H., & O’Reilly, C. A. 1974. Measuring organ-izational communication. Journal of Applied Psy-chology, 59: 321–326.

Robinson, S. L. 1996. Trust and breach of the psycholog-ical contract. Administrative Science Quarterly,41: 574–599.

Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C.1998. Not so different after all: A cross-disciplineview of trust. Academy of Management Review, 23:393–404.

Rupp, D. R., & Cropanzano, R. 2002. The mediating ef-fects of social exchange relationships in predictingworkplace outcomes from multifoci organizationaljustice. Organizational Behavior and Human Deci-sion Processes, 89: 925–946.

Scholl, R. W., Cooper, E. A., & McKenna, J. F. 1987.Referent selection in determining equity percep-tions: Differential effects on behavioral and attitu-dinal outcomes. Personnel Psychology, 40: 113–124.

Schoorman, F. D., Mayer, R. C., & Davis, J. H. 2007. Anintegrative model of organizational trust: Present,past, and future. Academy of Management Review,32: 344–354.

Schwab, D. P. 1980. Construct validity in organizationalbehavior. In L. L. Cummings & B. M. Staw (Eds.),Research in organizational behavior, vol. 2: 3–43.Greenwich, CT: JAI.

Skarlicki, D. P., & Latham, G. P. 2005. How can trainingbe used to foster organizational justice? In J.Greenberg & J. A. Colquitt (Eds.), Handbook of


organizational justice: 499–522. Mahwah, NJ: Erl-baum.

Stinglhamber, F., De Cremer, D., & Mercken, L. 2006.Perceived support as a mediator of the relationshipbetween justice and trust. Group and OrganizationManagement, 31: 442–468.

Stone-Romero, E. F. 1994. Construct validity issues inorganizational behavior research. In J. Greenberg(Ed.), Organizational behavior: The state of thescience: 155–179. Hillsdale, NJ: Erlbaum.

Thibaut, J., & Walker, L. 1975. Procedural justice: Apsychological analysis. Hillsdale, NJ: Erlbaum.

Tyler, T. R. 1989. The psychology of procedural justice:A test of the group-value model. Journal of Person-ality and Social Psychology, 57: 830–838.

Tyler, T. R. 1994. Psychological models of the justicemotive: Antecedents of distributive justice and pro-cedural justice. Journal of Personality and SocialPsychology, 67: 850–863.

Tyler, T. R., Degoey, P., & Smith, H. 1996. Understandingwhy the justice of group procedures matters: A testof the psychological dynamics of the group valuemodel. Journal of Personality and Social Psychol-ogy, 70: 913–930.

Tyler, T. R., & Lind, E. A. 1992. A relational model ofauthority in groups. In M. P. Zanna (Ed.), Advancesin experimental social psychology, vol. 25: 115–191. San Diego: Academic Press.

Van den Bos, K. 2001. Fairness heuristic theory: Assessingthe information to which people are reacting has apivotal role in understanding organizational justice. InS. Gilliland, D. Steiner & D. Skarlicki (Eds.), Theoret-ical and cultural perspectives on organizationaljustice: 63–84. Greenwich, CT: Information Age.

Van den Bos, K., Wilke, H. A. M., & Lind, E. A. 1998.When do we need procedural fairness? The role oftrust in authority. Journal of Personality and SocialPsychology, 75: 1449–1458.

Jason A. Colquitt ([email protected]) is the William HarryWillson Distinguished Chair in the Department of Man-agement at the University of Georgia’s Terry College ofBusiness. He received his Ph.D. from Michigan StateUniversity’s Eli Broad Graduate School of Management.His research interests include organizational justice,trust, team effectiveness, and personality influences ontask and learning performance.

Jessica B. Rodell ([email protected]) is an assistant profes-sor in the Department of Management at the University ofGeorgia’s Terry College of Business. She earned a Ph.D.and an MBA from the University of Florida’s WarringtonCollege of Business Administration. Her research inter-ests include volunteering, organizational justice, andemotions.

This article continues on the next page with Appendix A.


APPENDIX ATrust Scales Used in Past Research on Justice and Trust

Developers ofMeasure Items Studies Using Measure

Roberts and O’Reilly(1974)

1. How free do you feel to discuss with your immediate supervisor theproblems and difficulties in your job without jeopardizing yourposition or having it held against you?

2. Immediate supervisors at times must make decisions, which seem tobe against the interest of subordinates. When this happens to you as asubordinate, how much trust do you have that your immediatesupervisor’s decision was justified by other considerations?

3. To what extent do you have trust and confidence in your immediatesupervisor regarding his general fairness?

• Folger and Konovsky (1989)• Kim and Mauborgne (1993)• Konovsky and Pugh (1994)• Korsgaard and Schweiger

(1995)• Korsgaard (2002)• Cropanzano, Prehar, and Chen

(2002)

Gabarro and Athos(1976)

1. I believe my supervisor has high integrity2. I can expect my supervisor to treat me in a consistent and predictable

fashion.3. My supervisor is not always honest and truthful.4. In general, I believe my supervisor’s motives and intentions are good.5. I don’t think my supervisor treats me fairly.6. My supervisor is open and upfront with me.7. I am not sure I fully trust my supervisor.

• Aryee, Budhwar, and Chen(2002)

• Begley, Lee, and Hui (2006)

Cook and Wall(1980)

1. My supervisor is sincere in his/her attempts to meet the workers’point of view.

2. Our firm has a poor future unless it can attract a better supervisor.3. My supervisor can be trusted to make sensible decisions for the firm’s

future.4. My supervisor at work seems to do an efficient job.5. I feel quite confident that my supervisor will always try to treat me

fairly.6. My supervisor would be quite prepared to gain advantage by

deceiving the workers.

• Kernan and Hanges (2002)

McAllister (1995) 1. We have a sharing relationship. We can both freely share our ideas,feeling, and hopes.

2. I can talk freely to my supervisor about difficulties I am having atwork and know that (s)he will want to listen.

3. We would both feel a sense of loss if one of us transferred and wecould no longer work together.

4. If I shared my problems with my supervisor, I know (s)he wouldrespond constructively and caringly.

5. I would have to say that we have both made considerable emotionalinvestments in our working relationship.

6. My supervisor approaches his/her job with professionalism anddedication.

7. Given my supervisor’s track record, I see no reason to doubt his/hercompetence and preparation for the job.

8. I can rely on my supervisor not to make my job more difficult bycareless work.

9. Most people, even those who aren’t close friends with my supervisor,trust and respect him/her as a coworker.

10. Other work associates of mine who must interact with my supervisorconsider him/her to be trustworthy.

11. If people knew more about my supervisor and his/her background,they would be more concerned and monitor his/her performance moreclosely.

• Ambrose and Schminke (2003)• Stinglhamber, De Cremer, and

Mercken (2006)

Nyhan and Marlowe(1997)

1. My level of confidence that my supervisor is technically competent atthe critical elements of his/her job is ____.

2. My level of confidence that my supervisor will make well thought outdecisions about his/her job is ____.

3. My level of confidence that my supervisor will follow through onassignments is ____.

4. My level of confidence that my supervisor has an acceptable level ofunderstanding of his/her job is ____.

5. My level of confidence that my supervisor will be able to do his orher job in an acceptable manner is ____.

6. When my supervisor tells me something, my level of confidence that Ican rely on what they tell me is ____.

7. My confidence in my supervisor to do the job without other problemsis ____.

8. My level of confidence that my supervisor will think through what heor she is doing on the job is ____.

9. My level of confidence that my supervisor will treat me fairly is ____.10. The level of trust between supervisors and workers in this organiza-

tion is ____.11. The level of trust between my supervisor and me is ____.12. The degree to which I can depend on my supervisor is ____.

• Pillai, Schriesheim, andWilliams (1999)

• Pillai, Williams, and Tan(2001)

• Aryee, Budhwar, and Chen(2002)


Documents

JUSTICE, TRUST, AND TRUSTWORTHINESS: A LONGITUDINAL