28
http://hum.sagepub.com/ Human Relations http://hum.sagepub.com/content/53/10/1329 The online version of this article can be found at: 2000 53: 1329 Human Relations Tina Robbins, Timothy P. Summers and Janice L. Miller Intra- and inter-justice relationships: Assessing the direction Published by: http://www.sagepublications.com On behalf of: The Tavistock Institute can be found at: Human Relations Additional services and information for http://hum.sagepub.com/cgi/alerts Email Alerts: http://hum.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: What is This? - Oct 1, 2000 Version of Record >> at Jazan University on March 3, 2014 hum.sagepub.com Downloaded from at Jazan University on March 3, 2014 hum.sagepub.com Downloaded from

Human Relations 2000 Robbins 1329 55

Embed Size (px)

Citation preview

Page 1: Human Relations 2000 Robbins 1329 55

http://hum.sagepub.com/Human Relations

http://hum.sagepub.com/content/53/10/1329The online version of this article can be found at:

  2000 53: 1329Human Relations

Tina Robbins, Timothy P. Summers and Janice L. MillerIntra- and inter-justice relationships: Assessing the direction

  

Published by:

http://www.sagepublications.com

On behalf of: 

  The Tavistock Institute

can be found at:Human RelationsAdditional services and information for    

  http://hum.sagepub.com/cgi/alertsEmail Alerts:

 

http://hum.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- Oct 1, 2000Version of Record >>

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 2: Human Relations 2000 Robbins 1329 55

Intra- and inter-justice relationships:Assessing the directionTina L. Robbins, Timothy P. Summers and Janis L. Miller

A B S T R AC T This field study used structural equation modeling to investigate the

relationships among: (1) distributive and procedural justice; (2) justice

components and organizational commitment; and (3) justice com-

ponents and behaviors/behavioral intentions. The results suggest that,

over time, procedural justice judgments are likely to influence per-

ceptions of distributive justice, but not vice versa. In addition, the

results suggest that both distributive justice and procedural justice

have reciprocal relationships with commitment and turnover inten-

tions, although in some cases they are contingent on lagged effects.

Relationships between procedural justice and behaviors (i.e. compli-

ance, performance) were unidirectional, significant only in the justice-

to-behavior direction.

K E Y W O R D S distributive justice � employee turnover � organizationalcommitment � organizational justice � procedural justice

1 3 2 9

Human Relations

[0018-7267(200010)53:10]

Volume 53(10): 1329–1355: 014107

Copyright © 2000

The Tavistock Institute ®

SAGE Publications

London, Thousand Oaks CA,

New Delhi

03robbins(ds) 2/8/00 3:01 pm Page 1329

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 3: Human Relations 2000 Robbins 1329 55

Historically, the organizational justice literature has focused much of itsattention on distributive justice (Homans, 1961) or the perceived fairness ofthe amounts of compensation and other outcomes received. In the mid-1980s, theorists (Folger & Greenberg, 1985; Greenberg & Tyler, 1987; Lind& Tyler, 1988) turned to the fairness of the procedures used to determinethose outcomes, what is now known as procedural justice. Most recently,focus has been on the interrelationship among the different forms of justice(Sweeney & McFarlin, 1993) and their relationships to other variables(Brockner & Wiesenfeld, 1996).

Despite the volume of research, the nature of the conceptual or theor-etical link between the two aspects of justice remains in question. Althoughsome studies (Alexander & Ruderman, 1987; Folger & Konovsky, 1989;Greenberg, 1987; Sweeney & McFarlin, 1993) have tested intra-justicemodels, they have not examined the possibility of reciprocal effects, eitherwithin justice (i.e. distributive justice and procedural justice) or betweenjustice and the variables that have been associated with justice. The purposeof this research is to examine that possibility with the use of longitudinaldata. In doing so, we respond to the many requests (e.g. Folger & Konovsky,1989; Konovsky & Pugh, 1994; Korsgaard & Roberson, 1995; Sweeney &McFarlin, 1993) that future research capitalize on longitudinal data to bettertest causal directions regarding procedural and distributive justice in organiz-ational settings.

In addition, we believe that there may be a conceptually importantbenefit to the use of longitudinal data. This approach also permits a view ofrelationships that may develop only with the passage of time. Early researchof distributive justice (e.g. Cosier & Dalton, 1983; Vecchio, 1982) proposedinequity effects as a dynamic process, an accurate assessment of whichrequires a time parameter. Time-lagged effects are likely to occur becauselater events sometimes trigger reactions to prior levels of perceived inequity(Cosier & Dalton, 1983). For example, an annual salary adjustment (whenless than expected), hiring a new employee at a ‘high’ wage, or a sudden needfor money, each may make prior injustice more salient and consequently leadto a change in attitudes and behaviors. In addition, the concept of thresholdsof tolerance for injustice also necessitates capturing time-lagged effects as ‘thenotion of time and repeated inequity is central to understanding behavior insuch cases’ (Cosier & Dalton, 1983: 316). Cross-sectional studies of reac-tions to injustice cannot capture the effects of reaching this threshold overtime. Therefore a second purpose of this study is to address the possibilitythat some of the relationships examined may develop only with the passageof time. Although early equity theory research was criticized for ignoringtime-lagged effects, this trend has continued in recent justice research.

Human Relations 53(10)1 3 3 0

03robbins(ds) 2/8/00 3:01 pm Page 1330

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 4: Human Relations 2000 Robbins 1329 55

Furthermore, by using both cross-sectional and longitudinal data, we will beable to determine whether this is a viable concern and also identify whichrelationships are more susceptible to time-lagged effects. We begin in thefollowing section by clarifying the distinction between procedural and dis-tributive justice and then follow with a review of the literature and evidencesuggesting reciprocal relationships between these distinct justice componentsand in each one’s relationship with other organizational attitudes, intentionsand behaviors.

Distinguishing distributive and procedural justice

Because of its focus on outcome fairness, Adams’ (1963, 1965) equity theoryhas been commonly used as the basis for the concept of distributive justice(see review by Greenberg, 1990a). Following the equity theory research ofGoodman (1974), Ronen (1986) and Summers and DeNisi (1990), percep-tions of distributive justice arise when employees make comparisons of theirwork outcomes, given their inputs, against certain referent others. Implicit inmuch of the research on referents (Carrell & Dittrich, 1978; Goodman &Friedman, 1971; Weick & Nesset, 1968) is the assumption that the percep-tion of fairness may be based on more than one comparison. Though thereare many possible comparison ‘others’, research points to three key cat-egories, internal, external, and system referents, where internal refers to acomparison other within the organization, external refers to comparingoutside the organization, and system refers to an organization’s policies andsystems, e.g. promises made to prospective employees during the selectionprocess (Goodman, 1974; Summers & DeNisi, 1990).

While distributive justice focuses on the fairness of the outcomes, pro-cedural justice addresses the fairness of the procedures used to achieve thoseoutcomes (Greenberg, 1990a). While these two general forms of justice haveunique qualities, they also overlap, particularly in the area of performanceevaluation (Korsgaard & Roberson, 1995). While performance evaluation isa formal or structural procedure, the enactment of which is an aspect ofsupervisory treatment of employees, performance evaluation also impactsdirectly on outcomes such as merit pay raises in many organizations, and thusaffects distributive justice as well. The interpersonal treatment componentmay be responsible for the findings (Alexander & Ruderman, 1987; Folger& Konovsky, 1989; McFarlin & Sweeney, 1992; Tyler & Bies, 1989) ofunique effects of procedural justice over distributive justice in its relation-ships with other attitudes and behaviors. Greenberg (1990a) summarizedthat studies provide, ‘. . . compelling evidence that people consider the nature

Robbins et al. Intra- and inter-justice relationships 1 3 3 1

03robbins(ds) 2/8/00 3:01 pm Page 1331

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 5: Human Relations 2000 Robbins 1329 55

of their treatment by others as a determinant of fairness’ (p. 411). Proceduraljustice, like distributive justice, is therefore a multi-dimensional construct,consisting of perceptions about the fairness of formal procedures and inter-personal treatment.

In this study, we focus on interpersonal treatment, the aspect of pro-cedural justice that is most conceptually remote from distributive justice. Thiswill permit a clearer understanding of the intra-justice relationship, as wellas the unique roles of distributive and procedural justice in relationships withother variables. Past research has established many aspects of interpersonaltreatment. For example, interactional justice (Bies & Moag, 1986; Tyler &Bies, 1989) or enactment justice (Lind & Lissak, 1985) refers to behaviorduring the enactment of procedures (e.g. during a performance appraisalinterview), and these perceptions are likely to be associated with the personwith whom the employee is interacting. The opportunity to express feelingsduring performance evaluations (Dipboye & dePontbraind, 1981; Landy etal., 1978), and having two-way communication during the formal perform-ance evaluation (Greenberg, 1986), are also considered to be importantaspects of procedural justice in general, and interpersonal treatment in particular. Employees are also likely to base their perceptions of proceduralfairness on any informal behavior/procedures or ongoing actions of thesupervisor/manager, such as the provision of timely and informative feedback(Folger & Konovsky, 1989; Tyler & Bies, 1989), the identification of goalsto help eliminate weaknesses (Landy et al., 1978), and the general treatmentof others with civility and dignity (Bies, 1986; Karambayya & Brett, 1989;Tyler, 1989).

The intra-justice relationships

Based on the models and research showing procedural justice directly affect-ing distributive justice (Greenberg, 1987; Sweeney & McFarlin, 1993), weexpect procedural justice to have some causal influence on distributivejustice. Folger and Greenberg (1985) stated that, ‘. . . perceived fairness oforganizational outcomes is likely to be based, at least in part, on the pro-cesses through which they are determined’ (p. 142), because, ‘. . . the fairerthe procedures used to determine the outcomes, the more psychologicallyacceptable those outcomes are likely to be’ (p. 148). In other words, percep-tions of procedural fairness are what cause outcomes to be seen as fair(Deutsch, 1975; Leventhal, 1976).

However, the reverse causal direction has also been suggested by many(see review by Sweeney & McFarlin, 1993) who have argued that procedures

Human Relations 53(10)1 3 3 2

03robbins(ds) 2/8/00 3:01 pm Page 1332

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 6: Human Relations 2000 Robbins 1329 55

are seen as fair only when distributive justice exists. This perspective ‘. . . sub-ordinates procedures to distributions in that the “ends” drive the evaluationsof the “means” ’ (Sweeney & McFarlin, 1993: 25). Past research has resultedin mixed findings regarding the influence of outcome fairness on the per-ceived fairness of the procedures used to determine those outcomes (seereview by Greenberg, 1987). Much of the dispute resolution research(Conlon et al., 1989; Lind & Lissak, 1985) suggests that outcomes are usedto draw inferences about procedural justice, and that distributive justice islikely to play a role in this process. ‘Outcomes can themselves carry infor-mation about how they were decided and this information can in turn affectjudgments about the procedure’ (Conlon et al., 1989: 1096). Attributiontheory (e.g. Miller & Ross, 1975; Staw et al., 1983; Zuckerman, 1979) wouldpredict that, in reacting to unfair outcomes, people expect their behavior tolead to the desired outcomes, and, when it does not, they look to situational(e.g. procedures) explanations (Gooding & Kinicki, 1995).

H1: Distributive and procedural justice judgments are reciprocally related.

The relationships among justice components andorganizational commitment

Much of the recent justice literature has focused on examining the attitudinalcorrelates of both procedural and distributive justice in various contexts, ‘Inessence, the belief of researchers who support the value of organizationaljustice is that if employees believe that they are treated fairly, they will bemore likely to hold positive attitudes about their work, work outcomes, andtheir supervisors’ (Moorman, 1991: 845). Specifically, employees who believetheir contributions are highly (i.e. justly) regarded are likely to be more com-mitted to the organization (e.g. Buchanan, 1974; Pearce & Porter, 1986) andstudies provide evidence of this link (e.g. Alexander & Ruderman 1987,Folger & Konovsky, 1989). Also, recent experiments in field settings (e.g.Greenberg, 1994; Korsgaard & Roberson, 1995) provide support for the per-ceived justice to commitment relationship.

Most of the studies of justice and its relationship to attitudes and com-mitment have assumed that justice perceptions are the predictor variable withcausal influence. However, it is also possible that attitudes toward theorganization influence justice perceptions. Becoming committed requires theinternalization of the values of an organization, so once committed, anemployee can be expected to also accept the organization’s norms and prac-tices, believing them to be legitimate and fair. Eisenberger et al. (1986) found

Robbins et al. Intra- and inter-justice relationships 1 3 3 3

03robbins(ds) 2/8/00 3:01 pm Page 1333

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 7: Human Relations 2000 Robbins 1329 55

that perceived organizational support, which is likely to be positively corre-lated to commitment, may be used to judge the potential gain in benefits thatwould result from increased work effort (i.e. essentially a justice belief). Lindand Tyler (1988) also suggested the possibility of reciprocal relationships instating that, ‘. . . attitudes toward the group might affect procedural justicejudgments, just as procedural justice judgments might affect evaluations ofthe group’ (p. 232). Likewise, Folger and Konovsky (1989) suggested thatcommitment could be a predictor, as well as an outcome, in its relationshipwith justice perceptions.

H2: Organizational commitment and distributive justice judgments arereciprocally related.

H3: Organizational commitment and perceptions of procedural justice arereciprocally related.

Justice perceptions and turnover intentions

Turnover intentions have been the subject of much of the justice research (e.g.Alexander & Ruderman, 1987; Konovsky & Cropanzano, 1991) and theseintentions have been linked to both distributive and procedural justice.Mowday et al. (1982) suggest that turnover is likely to be very sensitive toviolations of both distributive justice and procedural justice. Employees whoperceive that the organization is supportive of them are expected to be morecommitted to fulfilling the role requirements of their jobs and less likely toseek jobs in other organizations (Eisenberger et al., 1990). Even though com-mitment and turnover intentions have been suggested to be causally relatedin past research (Mobley, 1982), each may still have direct influences onjustice. Perceived injustice is likely to lead employees to believe that their per-ceived employment contracts have been violated. Robinson et al. (1994) sug-gested that employees who perceive such a violation are less likely to want along-term relationship with the current employer. Dailey and Kirk (1992)found that justice perceptions appear to be stronger predictors of intent toquit than are core work attitudes that have long been associated with intentto leave the organization (i.e. commitment, satisfaction).

It is likely that the justice–turnover intentions relationship is also reci-procal in nature. Someone may justify their intentions by blaming the sourcesof injustice or by adopting negative attitudes about those sources. Employ-ees may, ‘rationalize their desire to quit by finding “evidence” which illus-trates how unfairly rewards are distributed and how “biased” theperformance evaluation system is’ (Dailey & Kirk, 1992: 314). As theysuggested, employees may attribute their intentions to leave an organization

Human Relations 53(10)1 3 3 4

03robbins(ds) 2/8/00 3:01 pm Page 1334

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

user
Surligner
Page 8: Human Relations 2000 Robbins 1329 55

to problems with the organization, as opposed to their own performance orother internal causes, thus externalizing the cause of their intentions. Theseattributions are likely to come in the form of perceptions of either unfairtreatment by supervisors or the awarding of unfair levels of compensation.

H4: Turnover intentions and distributive justice judgments are reciprocallyrelated.

H5: Turnover intentions and perceptions of procedural justice are recipro-cally related.

Justice perceptions and compliance/reprimand actions

Some previous research has also focused on the relationship between justiceand behaviors such as performance and compliance. Lind and Tyler (1988)concluded from their review that fair procedures in organizations lead togreater compliance with rules and decisions. Equity theory would suggestthat many types of compliance behaviors (e.g. performance, attendance, notstealing, following instructions) are obvious employee inputs that may beadjusted in response to employee perceptions of outcome injustice. Earlywork (Thibaut et al., 1974) also found that adherence to rules is affected notonly by perceptions of outcome justice but also by rule-maker-to-individualinteractions, which are likely to reflect interpersonal treatment. Morerecently (e.g. Greenberg, 1990b, 1993), pay inequity, as well as interpersonaltreatment (i.e. sensitivity) have been shown to influence employee theft. Per-ceived inequitable treatment, with respect to outcomes, procedures, and theenactment of those procedures, has also been identified as one of the primarymotivations for noncompliance in the form of sabotage behaviors (see reviewby Crino, 1994). Research in the legal arena has also shown that proceduraljustice has a significant impact on compliance with the law (Tyler, 1988).

Although most of the justice research has assumed otherwise, thereis evidence suggesting justice may play a criterion-variable role in itsrelationship to behaviors as well. Specifically, self-perception theory,which has been well supported (e.g. Abelson, 1972; Bem, 1972; Kiesler etal., 1969), suggests that attitudes are used, after the fact, to explain orjustify behaviors. Although compliance behavior may influence distribu-tive justice through self-perceptual processes, or indirectly through per-ceptions of procedural justice (as predicted in Hypothesis 1), there isstronger evidence for a reciprocal relationship between procedural justiceand these behaviors. Employees are likely to make rationalizations aboutprocedures (Lind & Tyler, 1988), referencing interpersonal treatment to‘justify’ noncompliance. Compliance (i.e. noncompliance) may also affect

Robbins et al. Intra- and inter-justice relationships 1 3 3 5

03robbins(ds) 2/8/00 3:01 pm Page 1335

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 9: Human Relations 2000 Robbins 1329 55

perceived procedural justice indirectly through perceptions of inter-personal treatment and procedural fairness associated with managerialand disciplinary actions. For example, Ball et al. (1994) found that apunishment event provides opportunities for perceptions of justice to bedeveloped or affected. On the other hand, those who have not been rep-rimanded are more likely to believe that the organization’s procedures arejust. Compliance/noncompliance is undertaken with respect to policiesand procedures.

H6: Compliance behaviors and/or reprimand actions are reciprocallyrelated to perceptions of procedural justice.

Justice perceptions and performance/performanceevaluations

Researchers have often explained the effects of justice on performance interms of equity theory and the exchange relationship (i.e. transactional, econ-omic). Employees are likely to adjust their effort, which in turn will affectperformance, in response to perceived injustice. Others have proposed thatprocedural justice, in particular, should lead to enhanced performance, andempirical evidence supports the relationship between procedural justice andperformance (see reviews by Konovsky & Cropanzano, 1991; Lind & Tyler,1988). Fasolo et al. (1990) suggested that perceived procedural justice leadsto an inference that the organization is supportive of employees, and that,within a social exchange, this support should result in employee reciproca-tion in the form of higher job performance.

However, procedural justice may serve as a criterion, as well as a pre-dictor variable in its relationship with performance. Self-perception theoryreferenced earlier would predict that employees use their attitudes about fair-ness to justify high or low performance. The performance-to-proceduraljustice relationship may also work indirectly through reactions to evaluationsof performance. High or low performance evaluations may cause employeesto evaluate organizations as either fair or unfair, respectively (Konovsky &Cropanzano, 1991). For example, the self-serving bias inherent in formingattributions would predict that favorable performance evaluations associatedwith higher performance will lead to higher perceived fairness of the pro-cedures used to recognize and reward one’s ability and effort (i.e. internalattributions). On the other hand, poor ratings are likely to be externalized topoor procedures and treatment.

Human Relations 53(10)1 3 3 6

03robbins(ds) 2/8/00 3:01 pm Page 1336

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 10: Human Relations 2000 Robbins 1329 55

H7: Performance and/or performance evaluations and perceptions of pro-cedural justice are reciprocally related.

Method

A textile products company located in the southeastern US agreed to par-ticipate in our research. Employees of two plants volunteered data on a widerange of organizational experience variables. A subset of those variables isconsidered in this paper. Data collection for all variables except performanceand reprimands (which were collected at Time 2 only) occurred at two timesin each plant, with a separation interval of 10 months. A 33 percent sam-pling of participants at Time 1 was selected, using a random number gener-ator, from among all hourly, nonexempt employees. A list of names wasgenerated and distributed, and 80 percent of those listed participated.Surveys were administered during the workers’ shifts, so the incentive to par-ticipate was high. Participants were asked to provide their names, only as ameans to permit collection of further data from company records. They wereassured by the researchers that their individual responses were to be usedonly for research purposes, and that the company would receive only asummary of the ‘average answers’. Useable responses at Time 1 wereobtained from 310 employees, of whom 297 provided names.

Time 2 data were collected from those who had provided their namesat Time 1, were still employed, attended work the day of data collection, vol-unteered to participate a second time, and provided their names a second time.Useable responses were collected from 180. Mean age was 37. The samplewas 59 percent male, 62 percent white and 37 percent African American.

The two plants utilized similar equipment, had similar jobs and similarworking conditions, and pursued similar management and personnel poli-cies. Furthermore, because there were no significant differences between theplants with respect to the variables under study, the demographic features ofthe two plants were similar, ‘location’ was not considered as a study variable,and preliminary results revealed the same pattern of findings for each plant,the data from the two plants were combined for this research.

Measures

Distributive justice (DJ)

The distributive justice measure was designed to measure internal, externaland system equity perceptions. For example, employees were asked tocompare their pay (i.e. cash wage) to ‘others doing the same type of work at

Robbins et al. Intra- and inter-justice relationships 1 3 3 7

03robbins(ds) 2/8/00 3:01 pm Page 1337

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 11: Human Relations 2000 Robbins 1329 55

your company’, ‘others in your job category at your company’, ‘others inyour job category outside your company’ and ‘others doing the same type ofwork at other companies’. The remaining items asked ‘How does your paycompare to the pay you should get according to company policy?’ and ‘youshould get under your arrangements with the company’ (system referent com-parisons). A seven-point scale that ranged from ‘much less than I deserve’ to‘much more than I deserve’ was used to answer all the distributive justiceitems. This measure, initially reported by Summers and DeNisi (1990), isbased on the work of Goodman (1974). It also proved reliable and predic-tive in the study conducted by Summers and Hendrix (1991).

Procedural justice (PJ)

Perceptions of procedural justice were also measured using six items that tapinformal procedures or treatment where the source of the injustice is on thepart of another individual (i.e. the supervisor) with whom the employee inter-acts. These items addressed aspects of interpersonal treatment identified inpast research. One item assessed the degree to which the employee felt thesupervisor (but not the performance evaluation system) evaluated his or herperformance fairly and another assessed the extent to which the employeehad a voice in the performance appraisal process. Two-way communicationand the opportunity to have one’s views considered (Greenberg, 1986); orthe opportunity to state one’s own side of issues (Dipboye & dePontbriand,1981) have been highlighted as important aspects of procedural justice. Twoadditional items measured the degree to which the supervisor provided feed-back and worked with the employee in developing performance-related plansand goals. The importance of timely and informative feedback to perceptionsof interactional and procedural justice has been well established (e.g.Gilliland, 1993; Tyler & Bies, 1989). Prior research (Dailey & Kirk, 1992;Dipboye & dePontbriand, 1981) found that planning and goal setting on thepart of the supervisor are important aspects of procedural justice. Theseactions are likely to suggest to employees that they have received adequatenotice of performance criteria (Folger & Konovsky, 1989). The remainingitems assessed general supervisory treatment and support of the employee, inorder to capture any other aspects of treatment that might be salient to theemployees.

Organizational commitment (CO)

Organizational commitment was measured using a six-item scale firstreported by DeCotiis and Summers (1987). Their measure demonstrated ahigh level of construct validity in research that supported commitment’s

Human Relations 53(10)1 3 3 8

03robbins(ds) 2/8/00 3:01 pm Page 1338

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 12: Human Relations 2000 Robbins 1329 55

hypothesized effect on motivation, job performance, desire to leave, andvoluntary turnover. Its coefficient alpha reliability was reported in their studyas .88. The measure has since been reported in studies modeling the role oforganizational commitment (e.g. Summers & Hendrix, 1991), and wasincluded in a recent meta-analysis of the commitment–turnover relationship(Cohen, 1993). The measure derives from a conceptual clarification of theconcept of commitment, in which commitment is viewed as, ‘The extent towhich an individual accepts and internalizes the goals and values of anorganization and views his or her organizational role in terms of its contri-bution to those goals and values, apart from any personal instrumentalitiesthat may attend his or her contribution’ (p. 448). This definition excludesthose concepts that should be treated as consequences of commitment, forexample, motivation (see DeCotiis & Summers, 1987). Three of the itemswere, ‘I am proud of the products and services my company provides to itscustomers’, ‘I feel as if my company’s problems are my problems’, and ‘I donot feel that what my company is trying to accomplish is worthwhile’. Theshort length of the measure was beneficial for the research.

Turnover intentions (TI)

Turnover intentions were measured with the three items suggested by Mobley(1977), who used his measure to establish the intervening role of intentionto leave between job satisfaction and voluntary turnover. One of the items,for example, assessed the extent of agreement with, ‘I am actively looking foranother job’.

Job performance/evaluations of performance (PERF)

For each participant, the company provided the mean of supervisory-ratedperformance across several dimensions, including knowledge of job, qualityof work, and quantity of work. The rating scale was from one, ‘inadequate’,to five, ‘superior’. Only the average rating was available, and the companyused that average (with equal weighting of each dimension) for personneldecision making. Employees who left the company between data collectionpoints or who had been employed less than one year had no current per-formance ratings on file. Performance data were thus obtained at Time 2 for239 of the 297 employees who provided their names at Time 1.

Compliance/reprimand action (CMPL)

The company also provided the number of reprimand actions pending foreach participant, which was also used to represent compliance in this study.

Robbins et al. Intra- and inter-justice relationships 1 3 3 9

03robbins(ds) 2/8/00 3:01 pm Page 1339

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 13: Human Relations 2000 Robbins 1329 55

This number ranged from zero to three, with three being the largest numberpermitted before firing. Reprimands remained active for a period of one year,and they were issued for noncompliance with published policies and pro-cedures. Specifically, of 65 pending reprimands, 58 percent were for viola-tions of a policy requiring issuance of a reprimand after three or moreunexcused absences. Thirty percent were for failing to follow job procedures,8 percent for violating safety procedures, and 4 percent for dishonesty, inthese cases, altering company records.

Results

As recommended by Anderson and Gerbing (1988) and Williams and James(1994), a two-step approach for path analysis with latent variables was fol-lowed utilizing the SAS System’s CALIS procedure and maximum likelihoodestimation. In the first step, confirmatory factor analysis was used to developa measurement model for each theoretical model that demonstrated anacceptable fit to the data. At this time, several minor, measurement-relatedmodifications were made. First, because of high correlations between pairsof indicators, the six indicators for both PJ and DJ were combined into threefactor scores.1 This method was described by Hair et al. (1992). Second, twoof the six indicators for the commitment variable were dropped because ofweak factor loadings. All of the remaining factor loadings were highly sig-nificant (p < .001) which suggests convergent validity for our factors. Dis-criminant validity was established by constraining, one at a time, each of thebetween-factor covariance terms to one. Nested chi-square tests revealedthat, in all cases, the better measurement model was the one in which the twovariables were viewed as distinct, but correlated, factors. Table 1 shows the

Human Relations 53(10)1 3 4 0

Table 1 Time 1 reliabilities and correlations between latent constructs

PJ DJ CO TI

PJ .80DJ .28 .79CO .61 .44 .64TI –.54 –.46 –.74 .73

Reliabilities are on the diagonal.All correlations significant at p < .05.

03robbins(ds) 2/8/00 3:01 pm Page 1340

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 14: Human Relations 2000 Robbins 1329 55

Time 1 zero-order correlations between latent variables, with the coefficientalpha reliability shown on the diagonal.

In the second step of our procedure, we estimated the theoreticalmodels of interest. Four fit indices for each of the five measurement modelsand their corresponding theoretical models appear in Table 2. We show chi-square, but because the chi-square statistic has been shown to be extremelysensitive to sample size effects (Bentler & Bonett, 1980), we include other fitindices, which suggest that all of our models were a good fit to the data. Inall but one case, both the GFI (Goodness of Fit Index) and the CFI (Com-parative Fit Index) were approximately .90, which is generally consideredindicative of a good fit (Bentler & Bonett, 1980; Sapienza & Korsgaard,1996). Also, our RMSR (root mean square residual) statistics are all below.09 and the (chi-square/d.f.) values of below 3.1 are considered indicative ofa good fit (Segars & Grover, 1993). Upon establishing acceptable model fit,individual parameter estimates were used as direct tests of our hypotheses.

Model development and analysis

Figures 1–5 identify the variables that are included in both the measurementand theoretical models including the standardized path coefficients that wereused to test the hypotheses in this study. Because we could not develop reci-procal models given only two waves of data, we developed separate uni-directional models to analyze each direction separately. A temporal dilemmais created when we try to analyze reciprocal relationships involving data col-lected at two periods within the model. This is because it makes no sense toconsider the lagged effect of a variable measured at Time 2, on a variablemeasured at Time 1. Significant relationships in both causal directions arerequired to provide evidence for the reciprocal relationships posited in thisstudy.

Because performance evaluations were done on the anniversary of thehire date and reprimands were cumulative (and remaining on the employee’srecord for one year), the behavioral data (i.e. performance and reprimands)collected at Time 2 actually occurred between Time 1 and Time 2. Since theyoccurred after Time 1 justice perceptions and before Time 2 justice percep-tions, they are referred to as Time 2 in Figure 3 and Time 1 in Figure 5. Giventhe high intercorrelation between commitment and turnover in this study,and the causal direction supported in past research (Mobley, 1982), we alsoincluded a path from commitment to turnover intentions, which was signifi-cant at (p < .01) in all models.

Figure 1 contains the model and standardized parameter estimates thatapply to the first hypothesis regarding the intra-justice reciprocal relationship.

Robbins et al. Intra- and inter-justice relationships 1 3 4 1

03robbins(ds) 2/8/00 3:01 pm Page 1341

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 15: Human Relations 2000 Robbins 1329 55

As shown in Figure 1, perceived procedural justice (PJ) influenced distributivejustice judgments (DJ) as illustrated by the significant (p < .05) path from pro-cedural justice to distributive justice. However, Hypothesis 1 was not sup-ported since the path from distributive justice to perceived procedural justicewas not significant. These results suggest that the intra-justice relationship isnot reciprocal, but unidirectional, in that procedural justice influences dis-tributive justice, but not vice versa.

Hypotheses 2 and 3 predicted that the relationships between distribu-tive justice and commitment, and procedural justice and commitment wouldbe reciprocal. The standardized path coefficients provide strong support fora causal direction from both procedural justice and distributive justice toorganizational commitment (CO). The cross-sectional model (containingTime 1 ‘independent’ and ‘dependent’ variables) shown in Figure 2 revealssignificant (p < .01) paths from procedural justice to commitment, and fromdistributive justice to commitment. The model of lagged effects from Time 1

Human Relations 53(10)1 3 4 2

Table 2 Model fit statistics

Model �2 d.f. GFI CFI RMSR

Figure 1Null 791 66Meas 130 48 .89 .89 .06Theoretical 135 49 .89 .88 .06

Figure 2Null 1158 78Meas 181 59 .91 .89 .06Theoretical 181 59 .91 .89 .06

Figure 3Null 745 105Meas 139 77 .91 .90 .06Theoretical 162 82 .90 .88 .07

Figure 4Null 1158 78Meas 181 59 .91 .89 .06Theoretical 181 60 .91 .89 .06

Figure 5Null 690 105Meas 131 77 .91 .91 .06Theoretical 176 82 .89 .84 .09

03robbins(ds) 2/8/00 3:01 pm Page 1342

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 16: Human Relations 2000 Robbins 1329 55

justice to Time 2 attitudes/intentions/behaviors presented in Figure 3 alsogenerated statistically significant coefficients from procedural justice (p < .01)and from distributive justice (p < .05) to commitment.

The other direction of influence, from commitment to proceduraljustice, was also consistently supported (p < .01) in both the cross-sectionalmodel, as shown in Figure 4, and in the lagged model, as shown in Figure5. In addition, while the lagged model provided support (p < .05) forcommitment-to-distributive justice effects, no support was found with the

Robbins et al. Intra- and inter-justice relationships 1 3 4 3

Figure 1 Lagged effects between procedural and distributive justice

Figure 2 Cross-sectional model of relationships of justice to attitudes and intentionsNotes: Path coefficients in brackets represent results of post-hoc interaction analysis (Figure 3also). The first coefficient is when sample is restricted to low values of the other justice variable.The second coefficient is when sample is restricted to low values of both justice variables.

03robbins(ds) 2/8/00 3:01 pm Page 1343

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 17: Human Relations 2000 Robbins 1329 55

cross-sectional model. In sum, these results support both Hypotheses 2 and3 in that the relationships between both procedural and distributive justicewith commitment are reciprocal, albeit only with respect to time-laggedeffects in the commitment-to-distributive justice direction.

Hypotheses 4 and 5 proposed reciprocal relationships between justiceand turnover intentions. The influence of distributive justice on turnoverintentions (TI), received moderate support given the significant (p < .05) path

Human Relations 53(10)1 3 4 4

Figure 3 Lagged model of relationships of justice to attitudes, intentions and behaviors

Figure 4 Cross-sectional model of relationships of attitudes and intentions to justice

03robbins(ds) 2/8/00 3:01 pm Page 1344

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 18: Human Relations 2000 Robbins 1329 55

in the cross-sectional model (shown in Figure 2), but only marginally (p < .10) significant lagged path, as shown in Figure 3. This same causal direc-tion with procedural justice (i.e. PJ-turnover intentions effect) also receivedmoderate support. The cross-sectional coefficient (shown in Figure 2) wasonly marginally significant (p < .10), but the lagged path shown in Figure 3was statistically significant (p < .05).

The alternative causal direction in the reciprocal relationship predictedby Hypothesis 4 (the influence of turnover intentions on distributive justice)was supported, with significant paths (p < .05) in both the cross-sectionalmodel, as shown in Figure 4, as well as in the lagged-effects model (Figure5). Attributional effects from turnover intentions to procedural justice (relevant to H5), however, received mixed support. The path from turnoverintentions to PJ in the cross-sectional model was only marginally significant.However, this same path in the lagged effects model was significant (p < .01).In sum, these results support both Hypotheses 4 and 5 in that the relation-ships between justice and turnover intentions are reciprocal, at least con-ditionally. More specifically, the relationship between PJ and turnoverappears to develop and strengthen over time.

Hypothesis 6 posited a reciprocal relationship between proceduraljustice and compliance behavior (CMPL) as measured by the number of rep-rimands the employee received between Time 1 and Time 2. Performance was

Robbins et al. Intra- and inter-justice relationships 1 3 4 5

Figure 5 Lagged model of relationships of attitudes, intentions and behaviors to justice

03robbins(ds) 2/8/00 3:01 pm Page 1345

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 19: Human Relations 2000 Robbins 1329 55

rated some time between Time 1 and Time 2, and was therefore treated tem-porally the same as compliance. As shown in Figure 3, procedural justice sig-nificantly (p < .05) influenced compliance behaviors, however, the reversecausal direction was not supported as evidenced by the nonsignificant pathfrom compliance (reprimands) to procedural justice in Figure 5. These resultslend no support for Hypothesis 6 because the relationship between pro-cedural justice and compliance behavior/reprimand actions appears to beonly unidirectional.

The final hypothesis predicted a reciprocal relationship between pro-cedural justice and performance (PERF). As shown in Figure 3, the influenceof procedural justice on performance was significant (p < .05). However, theresults failed to lend support for Hypothesis 7 since the reverse causal direc-tion in this relationship was not supported. As shown in Figure 5, the pathfrom performance evaluations to procedural justice judgments was not significant. As with compliance behaviors, these results suggest that thisrelationship is limited to justice-to-behavioral effects only.2

Discussion

The relationship among justice components, attitudes, intentions, andbehaviors does appear to be more complex than prior research suggests. Ingeneral, the results supported the hypothesized reciprocal relationships ofjustice to attitudes and intentions. These results suggest that not only areattitudes and intentions important consequences of justice perceptions asconsidered in most of the justice literature, but justice perceptions areimportant outcomes of these attitudes as well. Researchers and managersconcerned with perceptions of justice on the job should concentrate not onlyon objective predictors of such perceptions but also on previously held atti-tudes and intentions.

The results of this study should be considered and integrated withfuture research on commitment and turnover. Some of the variables identi-fied in prior research may serve as intervening links in the relationships foundhere and justice may serve as an intervening role in some prior relationshipsidentified. For example, trust, climate, satisfaction, as well as perceptions oforganizational dependability and personal importance to the organization,have been identified as antecedents to commitment in prior models (e.g.DeCotiis & Summers, 1987; Steers, 1977). These same variables may also beconsequences of justice and therefore serve as intervening links in thejustice–commitment relationship. In addition, justice, as a criterion in itsrelationship with commitment, may intervene in the relationship between

Human Relations 53(10)1 3 4 6

03robbins(ds) 2/8/00 3:01 pm Page 1346

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 20: Human Relations 2000 Robbins 1329 55

commitment and consequences of it (see for example, Mayer & Schoorman,1992). For example, performance and turnover may be direct consequencesof justice and only indirect effects of commitment. In sum, the reciprocalrelationship between justice and commitment suggests that justice be con-sidered as a potential predictor and criterion in models of commitment.

In addition, some of the causal relationships identified in prior researchon turnover (e.g. Williams & Livingstone, 1994) may be found to be spuri-ous when justice variables are included. A recent study (Martin & Bennett,1996) found that satisfaction and commitment were causally independentwhen justice judgments were taken into consideration. Similarly, since justiceinfluences both performance and turnover, studies that exclude it may erro-neously lead to the finding of a causal connection between the two.

Many of the relationships to justice that were found in this study weredependent on a time lapse even though the common methods between someof the variables could have introduced a bias in favor of more significantcross-sectional findings. For example, the reciprocal relationship betweenprocedural justice and turnover intentions was, in both causal directions,dependent on a time lag. This time-lagged effect of procedural justice onturnover intentions may exemplify a reaction or relationship in which anindividual has a relatively high threshold for injustice, but, once reached, mayexperience ‘the proverbial straw that broke the camel’s back’ (Cosier &Dalton, 1983: 316). Employees may expect interpersonal treatment to varyby supervisor, and most employees will have multiple supervisors during theirtenure in an organization. This belief may, in turn, lead to higher thresholdsfor procedural or treatment injustice, as opposed to distributive injustice.However, what one may expect to be only short-term treatment may, in fact,continue to a point where the threshold is reached over a period of time. Inaddition, this time-lagged effect of procedural justice on turnover intentionsmay reflect a process similar to what some of the turnover literature (e.g.Sheridan, 1985; Wright & Bonett, 1993) refers to as catastrophe or discon-tinuous effects. These abrupt changes in employee reactions or responses maybe a result of finally reaching a threshold.

Unlike the effects of procedural justice, the effects of distributive justiceon turnover intentions were more immediate. This result may be related toprevious suggestions (Greenberg, 1987) that people are more likely to takeaction against injustice if they believe this type of injustice is likely to con-tinue in the future. Employees may believe that outcome injustice, along withthe more structural aspects of procedural justice, are likely to remain rela-tively constant in the future. As a consequence, intentions to leave maydevelop immediately in response to perceptions of pay inequity as opposedto eventually in the case of interpersonal treatment inequity. This pattern is

Robbins et al. Intra- and inter-justice relationships 1 3 4 7

03robbins(ds) 2/8/00 3:01 pm Page 1347

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 21: Human Relations 2000 Robbins 1329 55

consistent with our findings of a cross-sectional effect of distributive justice,but a lagged-effect of procedural justice, on intentions to leave the organiz-ation.

The results of this study suggest that the effects of attitudes on justiceperceptions may also strengthen over time. For example, the turnover inten-tions-to-procedural justice effect, which was also contingent on a time lag,may reflect employees’ reluctance to immediately attribute these intentionsto unfair treatment. Over time, however, as they get closer to actually leaving,employees may feel the need to find more reasons to justify their impendingactions. Similarly, the commitment-to-distributive justice relationship, whichwas also contingent on a time lag, may take a history of commitment, orcumulative levels of commitment, before it begins to influence perceptions ofoutcome justice. The more immediate commitment-to-procedural justiceeffect, and the reciprocal nature of this relationship, may explain why pro-cedural justice has been found in prior research to be more significantlyrelated to commitment than is distributive justice, especially in studies thatwere not designed to capture effects that develop over time (e.g. Folger &Konovsky, 1989; McFarlin & Sweeney, 1992).

Practically, the overall reciprocal causation results of this study suggestthat not only must employers be aware that employees’ perceptions of justiceinfluence attitudes but also that attitudes influence perceptions of justice.Employees’ perceptions of changes (e.g. in pay system, and other procedures)that are implemented to address equity concerns will be biased by their pre-viously held attitudes. As a result, managers may find it difficult to overcomealready established perceptions of injustice. Not only would organizationshave to change employee justice perceptions but they would also have to over-come the biasing effect of attitudes that were previously influenced by thosesame perceptions. The reciprocal relationship between justice and turnoverintentions suggests that it may be very difficult to sway employees once theydecide to leave since an attributional process may distort subsequent percep-tions. This reciprocal relationship creates a self-fulfilling prophecy, because theemployees use these attributions of unfairness to justify their intentions.

Furthermore, prior perceptions of inequity may not be forgotten andwill have some influence on later reactions, regardless of whether the injus-tice was resolved (Cosier & Dalton, 1983). This existence of cumulativeeffects of inequity further explains why, in some cases, it is extremely diffi-cult to change employee attitudes. Furthermore, our cross-lagged findingssuggest that managers should not expect attitudes measured shortly after anew employee has joined the company, or soon after employees have experi-enced some change, to be enduring. Early positive reactions may not endure.On the positive side, early negative or neutral reactions may actually

Human Relations 53(10)1 3 4 8

03robbins(ds) 2/8/00 3:01 pm Page 1348

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 22: Human Relations 2000 Robbins 1329 55

improve. In sum, employers cannot expect to have a true picture of reactionsto change until after some period of time.

There are, of course, some limitations to the present study. As in moststudies relying primarily on questionnaires, this one is subject to error fromthe use of a common method for most measures. We mitigated this by varyingthe response formats (three distinct sets of scales and anchors to measure fourperceptual constructs), by mixing of items that used the same response format(e.g. the items measuring commitment and turnover intentions), by reversescoring of several items, and by including two measures from another source– the company’s personnel records on reprimand actions and performanceevaluations. The inclusion of longitudinal data almost certainly mitigated thepotential of common method bias, as common methods, used several monthsapart, are not likely to influence responses or correlations between predictorand criterion variables. Interestingly, in some cases, we found even strongerlagged effects than cross-sectional effects, which not only limits the ability ofcommon methods to explain variance but also provides evidence that thecross-lagged findings were not simply attributable to stability of measuresacross time. Further, method variance is unlikely to account for all of thestudy’s relationships, because we sometimes found high, sometimes foundlow, and sometimes found no relationships among our predictor and criterionvariables (i.e. as indicated by path coefficients in the models), a complexpattern difficult to explain by common method variance (Konovsky &Cropanzano, 1991). For all these reasons, we believe that our results are notattributable to common method variance.

The failure to find more significant relationships between justice andbehaviors may have been attributable to the low variability of our measuresof performance and compliance behaviors in this study. The overwhelmingmajority of respondents had not been reprimanded, and the majority whohad been reprimanded had only one pending. Similarly, although perform-ance was rated on a five-point scale, very few employees were ‘ones’ or ‘twos’.As a result, the variables representing performance and compliance did nothave the same amount of variance as the other variables, and this is likely tobe one reason why relationships with these variables did not yield findingsas strong as the others. However, as noted in the results section, the factorloadings for the indicator variables were all statistically significant and loadwell on the underlying constructs.

Although causation cannot be established in an uncontrolled field studysuch as this one, our method did permit inferences beyond the making ofassumptions about which variables are ‘independent’ and which are ‘depen-dent’. In a review of studies using longitudinal causal modeling, Williams &Podsakoff (1989) concluded that these designs were valuable ‘because they

Robbins et al. Intra- and inter-justice relationships 1 3 4 9

03robbins(ds) 2/8/00 3:01 pm Page 1349

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 23: Human Relations 2000 Robbins 1329 55

use time to empirically untangle the issue of direction of relationship betweenorganizational variables’ (Williams & James, 1994). Our analyses and longi-tudinal data permitted tests of ‘backward’ paths, and we found many movingin this backward, or attributional, direction. Our pattern of results suggeststhat the relationships between justice perceptions and attitudes and inten-tions are reciprocal, and in turn are consistent with the interactional modelregarding behaviors and attitudes (e.g. Terborg, 1981). At minimum, theseresults should serve as a reminder of the inadvisability of assuming causaldirections regarding justice relationships.

Not only were these longitudinal data useful for assessing direction ofrelationships, the comparison of cross-sectional with time-lagged results per-mitted us to identify relationships that are more likely to develop over timeand therefore go undetected in a static study. Although criticized in earlywork, most of the justice research has continued to rely on a static view ofthe process by which justice perceptions influence attitudes and behaviors.Our findings of significant lagged effects suggest that much of this prior cross-sectional work may not be adequately capturing potential justice relation-ships. Further, ‘. . . since employment relations are not characterized by singleinstances of inequity existing independently of past incidents or conditions’(Cosier & Dalton, 1983), tests for lagged, accumulated effects, such as weperformed, are critical for understanding justice effects.

Notes

1 For distributive justice, the two items addressing each type of referent com-parison (internal, external, and system) were paired. With proceduraljustice, the two items addressing supervisory treatment during performanceevaluations were paired, the two items tapping supervisory feedback andgoal/plan development were combined, as were the two items assessinggeneral (global) supervisory support and treatment.

2 We analyzed potential interactions between distributive and proceduraljustice in some post-hoc analyses. We found significant interactions in pre-dicting both performance and reprimands. Consistent with prior research(e.g. Brockner & Wiesenfeld, 1996) the relationships between proceduraljustice and both performance and reprimands were strongest when dis-tributive justice was low.

References

Abelson, R. Are attitudes necessary? In B.T. King and E. McGinnies (Eds), Atti-tudes, conflicts and social change. New York: Academic Press, 1972, p. 25.

Human Relations 53(10)1 3 5 0

03robbins(ds) 2/8/00 3:01 pm Page 1350

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 24: Human Relations 2000 Robbins 1329 55

Adams, J.S. Toward an understanding of inequity. Journal of Abnormal andSocial Psychology, 1963, 67, 422–36.

Adams, J.S. Inequity in social exchange. In L. Berkowitz (Ed.), Advances inexperimental social psychology, Vol. 2. New York: Academic Press, 1965,pp. 267–99.

Alexander, S. & Ruderman, M. The role of procedural and distributive justice inorganizational behavior. Social Justice Research, 1987, 1, 177–98.

Anderson, J.C. & Gerbing, D.W. Structural equation modeling in practice: Areview and recommended two-step approach. Psychological Bulletin,1988, 103, 411–23.

Ball, G.A., Trevino, L.K. & Sims, H.P. Just and unjust punishment: Influences onsubordinate performance and citizenship. Academy of ManagementJournal, 1994, 37, 299–322.

Bem, D.J. Self-perception theory. In L. Berkowitz (Ed.), Advances in experimentalsocial psychology, Vol. 6. New York: Academic Press, 1972, pp. 267–99.

Bentler, P.M. & Bonett, D.G. Significance tests and goodness of fit in the analy-sis of covariance structures. Psychological Bulletin, 1980, 88, 588–606.

Bies, R.J. Identifying principles of interactional justice: The case of corporaterecruiting. In R.J. Bies (Chair), Moving beyond equity theory: New direc-tions in research on justice in organizations. Symposium conducted at themeeting of the Academy of Management, Chicago, IL, August 1986.

Bies, R.J. & Moag, J.S. Interactional justice: Communication criteria of fairness.In R.J. Lewicki, B.H. Sheppard & B.H. Bazerman (Eds), Research on negoti-ation in organizations, Vol. 1. Greenwich, CT: JAI Press, 1986, pp. 43–55.

Brockner, J. & Wiesenfeld, B.M. An integrative framework for explaining reac-tions to decisions: Interactive effects of outcomes and procedures. Psycho-logical Bulletin, 1996, 120, 189–208.

Buchanan, B. Building organizational commitment; The socialization of man-agers in work organizations. Administrative Science Quarterly, 1974, 19,533–46.

Carrell, M.R. & Dittrich, J.E. Equity theory: The recent literature, methodologi-cal considerations and new directions. Academy of Management Review,1978, 3, 202–10.

Cohen, A. Organizational commitment and turnover: A meta-analysis. Academyof Management Journal, 1993, 38, 1140–57.

Conlon, D.E., Lind, E.A. & Lissak, R.I. Nonlinear and nonmonotonic effects ofoutcomes on procedural and distributive fairness judgments. Journal ofApplied Social Psychology, 1989, 19, 1085–99.

Cosier, R.A. & Dalton, D.R. Equity theory and time: A reformulation. Academyof Management Journal, 1983, 8, 311–19.

Crino, M.D. Employee sabotage: A random or preventable phenomenon? Journalof Managerial Issues, 1994, 6, 311–21.

Dailey, R.C. & Kirk, D.J. Distributive and procedural justice as antecedents ofjob dissatisfaction and intent to turnover. Human Relations, 1992, 45,305–17.

DeCotiis, T.A. & Summers, T.P. A path analysis of a model of the antecedentsand consequences of organizational commitment. Human Relations, 1987,7, 445–70.

Robbins et al. Intra- and inter-justice relationships 1 3 5 1

03robbins(ds) 2/8/00 3:01 pm Page 1351

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 25: Human Relations 2000 Robbins 1329 55

Deutsch, M. Equity, equality and need: What determines which value will be usedas the basis of distributive justice? Journal of Social Issues, 1975, 31, 137–49.

Dipboye, R.L. & dePontbriand, R. Correlates of employee reactions to perform-ance appraisals and appraisal systems. Journal of Applied Psychology,1981, 66, 248–51.

Eisenberger, R., Fasolo, P. & Davis-LaMastro, V. Perceived organizationalsupport and employee diligence, commitment and innovation. Journal ofApplied Psychology, 1990, 75, 51–9.

Eisenberger, R., Huntington, R., Hutchison, S. & Sowa, D. Perceived organiz-ational support. Journal of Applied Psychology, 1986, 71, 500–7.

Fasolo, P.M., Eisenberger, R. & Michaelis, E.D. The effects of distributive andprocedural justice on organizational performance. Unpublished manu-script, 1990.

Folger, R. & Greenberg, J. Procedural justice: An interpretive analysis of per-sonnel systems. Research in Personnel and Human Resources Manage-ment, 1985, 3, 141–83.

Folger, R. & Konovsky, M.A. Effects of procedural and distributive justice onreactions to pay raise decisions. Academy of Management Journal, 1989,32, 115–30.

Gilliland, S.W. The perceived fairness of selection systems: An organizational per-spective. Academy of Management Review, 1993, 18, 694–734.

Gooding, R.Z. & Kinicki, A.J. Interpreting event causes: The complementary roleof categorization and attribution processes. Journal of ManagementStudies, 1995, 32, 1–22.

Goodman, P.S. An examination of referents used in the evaluation of pay.Organizational Behavior and Human Performance, 1974, 12, 170–95.

Goodman, P.S. & Friedman, A. An examination of Adams’ theory of inequity.Administrative Science Quarterly, 1971, 16, 271–88.

Greenberg, J. Determinants of perceived fairness of performance evaluations.Journal of Applied Psychology, 1986, 71, 340–2.

Greenberg, J. Reactions to procedural injustice in payment distributions: Do themeans justify the ends? Journal of Applied Psychology, 1987, 72, 55–61.

Greenberg, J. Organizational justice: Yesterday, today and tomorrow. Journal ofManagement, 1990a, 16, 399–432.

Greenberg, J. Employee theft as a reaction to underpayment inequity: The hiddencost of pay cuts. Journal of Applied Psychology, 1990b, 75, 561–8.

Greenberg, J. Stealing in the name of justice: Informational and interpersonalmoderators of theft reactions to underpayment inequity. OrganizationalBehavior and Human Decision Processes, 1993, 54, 81–103.

Greenberg, J. Using socially fair treatment to promote acceptance of a work sitesmoking ban. Journal of Applied Psychology, 1994, 79, 288–97.

Greenberg, J. & Tyler, T.R. Why procedural justice in organizations? SocialJustice Research, 1987, 1, 127–42.

Hair, J.F., Anderson, R.E., Tatham, R.L. & Black, W.C. Multivariate Data Analy-sis with Readings. New York: Macmillan, 1992.

Homans, G.C. Social behavior: Its elementary forms. New York: Harcourt, Brace& World, 1961.

Karambayya, R. & Brett, J.M. Managers handling disputes: Third-party roles

Human Relations 53(10)1 3 5 2

03robbins(ds) 2/8/00 3:01 pm Page 1352

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 26: Human Relations 2000 Robbins 1329 55

and perceptions of fairness. Academy of Management Journal, 1989, 32,687–704.

Kiesler, C.A., Nisbett, R.E. and Zanna, M.P. On inferring one’s belief from one’sbehavior. Journal of Personality and Social Psychology, April 1969,687–704.

Konovsky, M.A. & Cropanzano, R. Perceived fairness of employee drug testingas a predictor of employee attitudes and job performance. Journal ofApplied Psychology, 1991, 76, 698–707.

Konovsky, M.A. & Pugh, S.D. Citizenship behavior and social exchange.Academy of Management Journal, 1994, 37, 656–69.

Korsgaard, M.A. & Roberson, L. Procedural justice in performance evaluation:The role of instrumental and non-instrumental voice in performanceappraisal decisions. Journal of Management, 1995, 21, 657–69.

Landy, F.J., Barnes, J.L. & Murphy, K.R. Correlates of perceived fairness andaccuracy of performance evaluation. Journal of Applied Psychology, 1978,63, 751–4.

Leventhal, G.S. Fairness in social relationships. In J.W. Thibaut, J.T. Spence andR.C. Carson (Eds), Contemporary topics in social psychology. Morris-town, NJ: General Learning Press, 1976, pp. 211–39.

Lind, E. & Lissak, R. Apparent impropriety and procedural fairness judgements.Journal of Experimental Social Psychology, 1985, 21, 19–29.

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

McFarlin, D.B. & Sweeney, P.D. Distributive and procedural justice as predictorsof satisfaction with personal and organizational outcomes. Academy ofManagement Journal, 1992, 35, 626–37.

Martin, C.L. & Bennett, N. The role of justice judgments in explaining therelationship between job satisfaction and organizational commitment.Group and Organization Management, 1996, 21, 84–104.

Mayer, R.C. & Schoorman, F.D. Predicting participation and production out-comes through a two-dimensional model of organizational commitment.Academy of Management Journal, 1992, 35, 671–84.

Miller, G.A. & Ross, M. Self-serving bias in the attribution of causality: Fact orfiction? Psychological Bulletin, 1975, 82, 213–25.

Mobley, W.H. Intermediate linkages in the relationships between job satisfactionand employee turnover. Journal of Applied Psychology, 1977, 62, 237–40.

Mobley, W.H. Employee turnover: Causes, consequences and control. Reading,MA: Addison-Wesley, 1982.

Moorman, R.H. The relationship between organizational justice and organiz-ational citizenship behaviors: Do fairness perceptions influence employeecitizenship? Journal of Applied Psychology, 1991, 76, 845–55.

Mowday, R.T., Porter, L.W. & Steers, R.M. Employee–organization linkages.New York: Academic Press, 1982.

Pearce, J.L. & Porter, L.W. Employee responses to formal appraisal feedback.Journal of Applied Psychology, 1986, 71, 211–18.

Robinson, S.L., Kraatz, M.S. & Rousseau, D.M. Changing obligations and thepsychological contract: A longitudinal study. Academy of ManagementJournal, 1994, 37, 137–52.

Robbins et al. Intra- and inter-justice relationships 1 3 5 3

03robbins(ds) 2/8/00 3:01 pm Page 1353

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 27: Human Relations 2000 Robbins 1329 55

Ronen, S. Equity perceptions in multiple comparisons: A field study. HumanRelations, 1986, 39, 333–46.

Sapienza, H.J. & Korsgaard, M.A. Procedural justice in entrepreneur–investorrelations. Academy of Management Journal, 1996, 39, 544–74.

Segars, A.S. & Grover, V. Re-examining perceived ease of use and usefulness: Aconfirmatory factor analysis. MIS Quarterly, 1993, 17, 517–25.

Sheridan, J.E. A catastrophe model of employee withdrawal leading to low jobperformance, high absenteeism and job turnover during the first year ofemployment. Academy of Management Journal, 1985, 28, 88–109.

Staw, B.M., McKechnie, P.I. & Puffer, S.M. The justification of organizationalperformance. Administrative Science Quarterly, 1983, 28, 582–600.

Steers, R.M. Antecedents and outcomes of organizational commitment. Admin-istrative Science Quarterly, 1977, 22, 46–56.

Summers, T.P. & DeNisi, A.S. In search of Adams’ other: Reexamination of ref-erents used in the evaluation of pay. Human Relations, 1990, 43, 497–511.

Summers, T.P. & Hendrix, W.H. Modelling the role of pay equity perceptions: Afield study. Journal of Occupational Psychology, 1991, 64, 145–57.

Sweeney, P.D. & McFarlin, D.B. Workers’ evaluations of the ‘ends’ and the‘means’: An examination of four models of distributive and proceduraljustice. Organizational Behavior and Human Decision Processes, 1993, 55,23–40.

Terborg, J. Interactional psychology and research on human behavior in organiz-ations. Academy of Management Review, 1981, 6, 569–76.

Thibaut, J., Friedland, N. & Walker, L. Compliance with rules: Some socialdeterminants. Journal of Personality and Social Psychology, 1974, 30,792–801.

Tyler, T.R. Why people follow the law: Procedural justice, legitimacy and com-pliance. New York: Plenum, 1988.

Tyler, T.R. The psychology of procedural justice: A test of the group-value model.Journal of Personality and Social Psychology, 1989, 57, 830–8.

Tyler, T.R. & Bies, R.J. Beyond formal procedures. The interpersonal context ofprocedural justice. In J. Carroll (Ed.), Advances in applied social psychol-ogy: Business settings. Hillsdale, NJ: Erlbaum, 1989, pp. 77–98.

Vecchio, R.P. Predicting worker performance in inequitable settings. Academy ofManagement Journal, 1982, 7, 103–10.

Weick, K.E. & Nesset, B. Preferences among forms of equity. OrganizationalBehavior and Human Performance, 1968, 3, 400–16.

Williams, L.J. & James, L.R. Causal models in organizational behavior research:From path analysis to Lisrel and beyond. In J. Greenberg (Ed.), Organiz-ational behavior: The state of the science. Hillsdale, NJ: Erlbaum, 1994,pp. 181–205.

Williams, C.R. & Livingstone, L.P. Another look at the relationship between per-formance and voluntary turnover. Academy of Management Journal, 1994,37, 269–89.

Williams, L. & Podsakoff, P. Longitudinal field methods for studying reciprocalrelationships in organizational behavior research: Toward improved causalanalysis. In L. Cummings & B. Staw (Eds), Research in OrganizationalBehavior, Vol. 11. Greenwich, CT: JAI Press, 1989, pp. 247–92.

Human Relations 53(10)1 3 5 4

03robbins(ds) 2/8/00 3:01 pm Page 1354

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from

Page 28: Human Relations 2000 Robbins 1329 55

Wright, T.A. & Bonett, D.G. Role of employee coping and performance in volun-tary employee withdrawal: A research refinement and elaboration. Journalof Management, 1993, 19, 147–61.

Zuckerman, M. Attribution of success and failure revisited, or: The motivationalbias is alive and well in attribution theory. Journal of Personality, 1979,47, 45–87.

Robbins et al. Intra- and inter-justice relationships 1 3 5 5

Tina L. Robbins is an Associate Professor in the Department ofManagement at Clemson University where she teaches organizationalbehavior and general management courses. She received her PhD in Busi-ness Administration from the University of South Carolina. Her currentresearch interests are in the areas of organizational justice and employeeempowerment.[E-mail: [email protected]]

Timothy P. Summers is an Associate Professor of Management,College of Business and Public Affairs, Clemson University. ProfessorSummers received his PhD in Business Administration from the Uni-versity of South Carolina. His area of teaching is human resourcesmanagement. His current research interests are strategic humanresources management, compensation and organizational justice.[E-mail: [email protected]]

Janis L.Miller received her PhD degree in Business Administration fromthe University of Missouri. She is currently an Associate Professor ofManagement at Clemson University. Her research interests are in theareas of quantitative methods and service recovery.[E-mail: [email protected]]

03robbins(ds) 2/8/00 3:01 pm Page 1355

at Jazan University on March 3, 2014hum.sagepub.comDownloaded from