R~TH~NK~N~ LEADER-MUMMER EXCHANGE:

AN ORGANIZATIONAL
JUSTICE PERSPECTIVE

Terri A. Scandura*
University of Miami

The literature on leader-member exchange (LMX) is examined from an organizational justice

perspective. The concepts of distributive, procedural, and interactional justice expand the LMX
model to consider social comparison processes operating within work groups. A model of LMX
development over time is presented that suggests that the differentiation of work groups into
in-groups and out-groups has ~mpli~tions for the emergence of organ~ation~ justice. New
research propositions based upon the mode1 are offered to encourage further research integrating
LMX and organizational justice.

INTRODUCTION
Relationships between leaders and members have been researched for over 25
years (Graen & Uhl-Bien, 1995), beginning with studies of the socialization of
organizational newcomers indicating the importance of supervisors’ attention to
new role incumbents (Dansereau, Graen, & Haga, 1975). In these studies, some
leaders treated their subordinates in different ways. Some were treated as “trusted
assistants” (in-group members) and others as “hired hands” (out-group members)
(Dansereau, Graen, & Haga, 1975). In-group members have better reIat~onships
with leaders and receive more work-related benefits in comparison to out-group
members. This “Vertical Dyad Linkage” (VDL) concept includes characteristics of
leaders, members, and the relationships between leaders and members (Dansereau,
Yammarino, & Markham, 1995). This concept was later measured differently and
labelled Leader-Member Exchange, or LMX (Graen, Liden, & Hoel, 1982a; Graen,
Novak, & Sommerkamp, 1982b). LMX has recently been defined as the unique

* Direct all correspondence to: Dr. Terri A. Scandura, Professor of Management and Psychology, 414
Jenkins Building, School of Business Administration, University of Miami, Coral Gables, Florida 33124;
e-maf scandura@miami.edu.

Leadership Quarterly, 10(l), 25-40.

Copyright 0 1999 by Elsevier Science Inc.
All rights of reproduction in any form reserved.
ISSN: 1048-9843
26 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

relationship-based social exchange between leaders and members (Graen & Uhl-
Bien, 199.5). However, this new emphasis on the relationship may obscure important
exchange-based issues (including economic exchange such as performance ratings
and pay increases) that take place in leader-member dyads. It is perhaps necessary to
rethink the LMX concept, considering what constitutes “fair exchange in leadership”
(Hollander, 1978, p. 71).
Dyadic relationships that emerge between leaders and followers in organizations
are an important aspect of leadership theory and research (Bass, 1990; Graen &
Uhl-Bien, 1995; Yukl, 1994). Moreover, measures of LMX have been related in
some studies to a number of important outcome variables in organizational research
including job satisfaction and performance ratings (Graen, Novak, & Sommerkamp,
1982b). turnover (Graen, Liden, & Hoel, 1982a; Ferris, 1985), subordinate decision
influence (Scandura, Graen, & Novak, 1986), and career progress of managers
(Wakabayashi & Graen, 1984; Scandura & Schriesheim, 1994). Yet, other studies
do not find such conclusive evidence for the relationship between LMX and produc-
tivity (cf., Vecchio & Gobdel, 1984; Duarte, Goodson, & Klich, 1994) nor turnover
(Vecchio, 1985). Such discrepancies in empirical studies conducted suggest that
there might be mediator variables that account for some of the differences reported
across studies of LMX. One possible explanation for discrepant findings across
studies is the relationship between LMX and organizational justice, which delves
into complex issues of the exchange aspect of LMX relationships. Yet, these issues
have received scant theoretical and empirical attention in the LMX literature.
The purpose of this review is to extend LMX theory to consider issues of organiza-
tional justice. Despite numerous empirical investigations of LMX, some authors
have commented that there has been limited theoretical development of the concept
(Dansereau et al., 1995; Dienesch & Liden, 1986; Miner, 1980; Vecchio & Gobdel,
1984). It seems useful to re-examine some underlying assumptions in the literature
on LMX, reintroducing relevant exchange theory concepts and treating LMX devel-
opment from a justice perspective. This article briefly reviews the LMX literature
and introduces relevant concepts from literature on organizational justice to recon-
cile some theoretical concerns regarding the usefulness of the model. Next, a model
of the role of organizational justice in the formation of in-groups and out-groups
is offered and new research propositions are developed.

REVIEW
Work Group Differentiation Process
Dansereau et al. (1975) presented a descriptive model of how work groups
become differentiated into in-groups and out-groups based upon the quality of
leader-member relationships that emerge between immediate supervisors and mem-
bers of work groups. Despite the clear indication that work group differentiation
occurs (Graen & Cashman, 1975; Vecchio, 1997), the idea that some subordinates
are treated better than others is inconsistent with norms of equality (cf., Kabanoff,
1991; Meindl, 1989). However, empirical research studies continue to document
differences in the quality of relationships and more benefits for in-group members
LMX: An Organizational Justice Perspective 27

[cf. (Graen & Uhl-Bien, 1995) for review]. For example, Lagace, Castleberry and
Ridnour (1993) found that in-group members (with higher quality LMX relation-
ships) were higher on motivational factors and evaluations of their bosses and
experienced less role-related stress (role overload, role insufficiency, role ambiguity
and conflict). The literature has indicated, in some studies, that in-group members
(i.e., those with higher quality LMX) receive more attention and support from the
leader than out-group members (those with with lower quality LMX) (cf., Graen
et al., 1982b). Also, out-group members (with lower LMX) are more likely to file
grievances (Cleyman, Jex, & Love, 1993). These results seem relevant to the concept
of organizational justice since out-group members might see their leader as treating
them unfairly. A careful review of the literature indicates that the LMX literature
has referenced issues related to the fair treatment of members.
Dansereau, Alutto, and Yammarino (1984) presented a general model of ex-
change theory discussing the role of equity perceptions in the development of
leader-member relationships. They defined investments as “. . . what one party gives
to another party” (p. 98) and returns as “. . . what one party gets back from another”
(p. 98). Their formulation suggests that investments trigger returns and vice versa
and that, over time, stable patterns of exchange emerge between leaders and mem-
bers, based on the ratios of investments to returns by both parties. Equity is thus
maintained by changes in either what is invested or returned to attain an overall
optimal level for both parties. Relationship development, over time, was proposed
to be a function of these investment-return cycles.
As an illustrative example of the use of multiple-levels of analysis in theory-
building, Dansereau et al. (1984) then elaborated their model to include multiple-
relationships which invokes the concept of social comparisons between work group
members. Investment-return cycles were compared for two different hypothetical
work group members, noting that equity can be maintained at different levels
of investments and returns. Social comparison processes emerge as one member
compares his/her investments and returns to a comparison other in the work group.
However, as long as the leader attends to the appropriate level of returns for
investments, feelings of inequity should not emerge for the member receiving lower
returns. This theoretical example presented by Dansereau and his colleagues cap-
tures the complexity (and also the necessity) of examining issues of organizational
justice in studies of LMX.
As noted by Graen and Scandura (1987) one of the requirements for the develop-
ment of high quality leader-member exchanges in organizations is that *‘. . . each
party must see the exchange as reasonably equitable or fair” (p. 182). Yet, current
theoretical approaches may limit the potential of LMX theory, because they place
too much emphasis on social exchange and do not develop aspects of economic
exchange (Graen & Uhl-Bien, 1995). Both social and economic exchange should
perhaps be given more weight in future studies. It seems that exchange can involve
both social aspects (such as availability and support) and economic aspects (such
as pay raises).
Most studies of work group differentiation into in-groups and out-groups are
descriptive, and not intended to instruct managers on how to manage their work
groups. This differs from prescriptive or normative theory where guidelines for
18 LEADERSHIP QUARTERLY Vol. IO No. 1 1999

managerial practice are developed. Without concerns for organizational justice,

LMX may have limited contributions in terms of normative theory, because percep-
tions of organizational justice are necessary for the leadership process. From an
organizational justice perspective, the LMX model might be criticized as reinforcing
the special treatment of some work group members over others (Vecchio, 1997).
Hence, supervisors may be reluctant to discuss the work group differentiation
process; concern for organizational justice may explain lower variance in some
supervisor reports of LMX (Scandura et al., 1986).
Lack of attention to theoretical concerns related to organizational justice have
perhaps limited the theoretical development of LMX. Yet, the empirical literature
on LMX is expanding (cf., Graen & Uhl-Bien, 1995). Several researchers have
employed longitudinal research designs (cf., Liden, Wayne, & Stilwell, 1993), which
have illuminated the development of LMX relationships over time. These studies
support the premise that in-group members receive more benefits compared to out-
group members, yet the question of whether or not this results in deprecation of
team-level outcomes remains. This study reviews the literature on organizational
justice to further explain the “fair exchange” aspect of LMX development (Danser-
eau et al., 1984; Graen & Scandura, 1987; Hollander, 1978) by revisiting important
exchange-based issues in LMX (including economic exchange) that may be lost in
current treatments of LMX as relationship-based and predominantly grounded in
relationship-based social exchange (Graen & Uhl-Bien, 1995).

Organizational Justice: A Brief Review of Relevant Concepts

Cropanzano and Folger (1991) present a two-component model of justice, which
includes distributive and procedural forms of justice. Distributive justice is defined
as the individual’s perception that the outcomes that they receive are fair (Adams,
1965; Greenberg, 1990). Examples of distributive outcomes are pay increases, pro-
motions, and challenging work assignments. Procedural justice is defined as an
employee’s perception that the procedures followed by the organization in determin-
ing who receives benefits are fair (Folger & Greenberg, 1985; Greenberg, 1987;
Lind & Tyler, 1988). Examples of procedural justice are the degree of voice the
person has in decision making and whether or not consistent rules are followed in
making decisions.
Research on justice has indicated that a decision will be accepted by subordinates
if procedural justice is followed, even if the distributive outcome is less that what
an individual desires (Tyler, 1986). For example, a low pay raise would still be
accepted if the organization’s procedures of performance appraisal and rewards
are seen as being followed in the determination of the raise. Also, communications
about what is fair to organizational members, labelled interactional justice, ‘has
been proposed as a third aspect of justice at the workplace (Bies & Moag, 1986;
Moorman, 1991). Interactional justice involves the manner in which organizational
justice is communicated by supervisors to followers. Moorman (1991) demonstrated
that distributive, procedural, and interactional justice are correlated, but distinct
aspects of organizational justice. Following this conceptualization, organizational
justice is defined as distributive, procedural, and interactional justice.
LMX: An Organizational Justice Perspective 29

Members of work groups often interpret the behavior of their immediate supervi-
sor in terms of organizational justice. Studies of procedural and distributive justice
indicate that leaders who are perceived as procedurally fair are rated favorably by
subordinates even when resource allocation is unequal (Folger & Konovsky, 1989;
Tyler, 1986; Tyler, 1989; Tyler & Caine, 1981; Tyler & Lind, 1992). Organizational
justice has implications for LMX theory since the focus of LMX is on the devefop-
ment of differentiated (~-group/out-groups leader-member relationships,
Despite suggestions by Hollander (1978), Dansereau et al. (1984) and Graen
and Scandura (1987) that equity matters for LMX development, the issue has
received little empirical attention. Notable exceptions are studies by Vecchio.
Griffeth, and Horn (1986) and Manogran, Stauffer, and Conlon (1994). These studies
suggest LMX is s~gni~can~y related to perceptions of org~izatio~al justice by
subordinates. Vecchio et al. (1986) showed a relationship between LMX and distrib-
utive justice. Those who had high quality relationships with their immediate supervi-
sor viewed the workplace as being more fair than those with low quality relationships
with their boss. Manogran et al. (1994) showed positive and significant correlations
between LMX and procedural and interact~onal justice, in addition to distributive
justice, using measures developed by Moorman (1991). These empirical findings
are intriguing, yet it is unclear whether the correlations reported between LMX
and organizational justice variables reflect that organiza~o~al justice is a tangential
outcome to the LMX process or a more central element in the development of
LMX relationships. A theoretical framework is needed to further elucidate the role
of organizational justice in the LMX development process.

LMX and Orga~i~ationai Justice

Hollander (1978) noted that a “psychological contract” (p. 73) emerges between
leaders and followers that depends upon expectations and actions of both parties
to the dyad. He noted that equitable treatment of subordinates is often one of
the most valued behaviors of a leader, since social comparison processes are so
fundament~ to human nature (Festinger, 1957). Hollander’s (1978) treatment of
leadership and exchange focused mostly on issues of distributive justice, ensuring
that reward distribution is fair. Meindl (1989) contrasted equity with parity, noting
that equity refers to “. . . entitlement based on relative contributions” (p. 254). He
also noted that the most frequently used alternative to equity is parity (also referred
to as equality). Yet, equal d~st~bntion of rewards would not totally avoid inequity
perceptions, since those whose investments are high might feel that they are not
receiving appropriate (i.e., higher) levels of returns in comparison to others (Danser-
eau et al., 1984). Equal reward distribution may harm those who are the hardest
workers in the group.
Now we come to the crux of the argument: can we have work group di~erentiation
and organizational justice as well? LMX and work group differentiation may be
perceived very differently, based upon whether norms of equity or parity (equality)
are operating in the leader’s decisions regarding allocation of work group resources.
Viewing LMX from an organizational justice perspective expands the model, in
that LMX must be viewed as a system of interdependent relationships rather than as
30 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

a set of independent dyads (Graen & Scandura, 1987). Social comparison processes
operate at the unit, team, or network level. Although it makes the model more
complex, justice in LMX is a theoretically rich framework through which the distri-
bution of benefits (both economic and social) within the LMX process may be
studied. Understanding role development within complex systems of interlocked
roles necessitates incorporation of multiple levels of analysis (Dansereau et al..
1984). For example, social comparison processes between work group members
(i.e., between dyads) must be addressed. At the group level of analysis, we must also
consider the interdependencies between subordinate and subordinate in addition to
leader-member interdependencies.
The differentiation process of in-groups and out-groups is not discrepant with
the concept of organizational justice. Procedural justice suggests that as long as a
leader is perceived as fair by all work unit members (fair procedures for allocating
rewards are followed), then a fair exchange of inputs to rewards might be maintained
for all members of the work unit. Also, interactional justice seems to play a role
in member’s perceptions of the reasons for reward distribution in the work group
as these are communicated to them by the leader.
In the development of work group perceptions of organizational justice, it is
necessary to consider whether fair procedures are followed (procedural justice) and
how this is communicated to members (interactional justice). Although it makes
the model more complex, viewing LMX through a broad justice perspective (distrib-
utive, procedural, and interactional) provides a rich theoretical framework from
which some interesting and non-obvious hypotheses can be generated. First, distrib-
utive justice enables us to understand how leaders distribute both economic and
social benefits. Second, procedural justice and interactional justice provide an under-
standing as to how employees in the in-group and the out-group react to the
distribution of benefits.
Figure 1 suggests how organizational justice issues may affect the LMX develop-
ment process over time. This time-based model suggests the points at which organi-
zational justice concepts become relevant to the development of LMX and perfor-
mance (other possible outcomes will be discussed later). Time in this model
specifically refers to the tenure of the leader-member relationship. Once leader
and member begin to interact, a process unfolds which results in the differentiation
of the member into an in-group or an out-group member.

Role Specification
Early in the LMX development process, leaders send roles to members and
members respond to these sets of expectations. Specifically, the leader specifies the
tasks to be performed by the member. Issues of organizational justice may emerge
in this early phase, since the leader and member are essentially strangers and levels
of trust are probably low (Graen & Uhl-Bien, 1995). Distributive justice may become
a concern. for example, if a member is asked to perform a task that he/she feels is
beyond the formal job description. The member may feel that his/her level of
compensation is too low for the task and that they are being asked to perform
work that is more appropriate to a higher job classification. This involves issues of
procedural justice, because if the member’s perceptions are correct, then the formal
LMX: An Organizational Justice Perspective 31

procedures of the organization for assigning work have been violated. These early
perceptions of roles develop into patterns of intitial exchange as the member pro-
vides feedback to the leader regarding whether or not roles are accepted.

Initial Exchange and Feedback

As the leader makes requests and the member responds, the leader begins to form
a perception of the member based upon his/her responses to requests. Dansereau et
al. (1984) referred to these exchange patterns as investment-return cycles. In the
above example in which the leader asks the member to perform an extra-role
behavior (i.e., a task that is outside of the formal job description), the member can
provide different forms of feedback to the leader. In response to the request the
member can (a) complete the task without questions, but expect a reward (a distribu-
tive justice response), (b) not do the task or do it poorly because it is not in the
job description and he/she is not compensated for it (another variation on the
distributive justice theme), (c) file a grievance because the leader has asked for an
inappropriate task for his/her level (a procedural justice response), or (d) ask the
leader for an explanation regarding why he/she has been asked to perform the task
(an interactional justice response). These responses have clear implications for the
next step, which is the emergence of a leader-member relationship of either an in-
group or out-group type. It is important to note that this decision is not necessarily
the sole judgement of the leader; it is, in part, based upon the feedback provided
to the leader by the member. As Vecchio (1997) notes, some group members may
want to be in the out-group. Some members may not want to invest extra effort
in their work, and out-group status is perceived as equitable.

LMX and
/I Initial Exchange
and Feedback Interactional

Time (Relationship Tenure)

Figure 1. The Role of Organizational Justice in
In-Group/Out-Group Differentiation
32 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

LMX and lnteractional Justice

Aspects of interactional justice such as honesty (Bies & Moag, 1986) are impor-
tant in the development of LMX. In Figure 1, LMX and interactional justice are
included in the same box because they represent aspects of the leader-member
relationship. The leader must be consistent, and not hide things from members-
even those with low LMX. This is critical to the process of LMX development,
since the first test of the leader by the member will often be his/her honesty in
dealing with the member. For example, if the leader promises a reward, the leader
must come through or else the member may perceive the leader as dishonest.
Interactional justice pervades the LMX-organizational justice process because com-
munication is such an integral part of LMX relationships (Fairhurst & Chandler,
1989; Schiemann, 1977), and the construction of meaning regarding what constitutes
fair exchange (Sias & Jablin, 1995).
While interactional justice is a distinct concept from LMX, it is expected that
they will be positively and significantly correlated (Manogran et al., 1994). LMX
and interactional justice are variables that measure aspects of the quality of the
leader-member relationship. Based upon the quality of the relationship that
emerges, a decision is made regarding each member’s in-group/out-group status in
the work group, as shown in Figure 1. This process is described in detail by Vecchio
(1997):

The boss tries out each new employee by offering small but challenging
assignments. The subordinate’s reaction to these additional responsibilities is
then closely watched. If the employee reacts negatively (by saying, “It’s not
my job”) or positively (by replying, “I’m happy to help”), then a cycle of trust
or distrust is begun. In short, supervisors learn quickly who is reliable and
who is not (p. 275).

In this manner, members in a work group are sorted into in-group or out-group
members.
It is important to note that LMX and interactional justice are measured along
a continuum (Dienesch & Liden, 1986), and some work group members’ status may
therefore be ambiguous (maybe in-group or out-group). As depicted in Figure 1,
the LMX and interactional justice process have not yet resulted in an in-group or
out-group decision, and the exchange and feedback processes continue (this is
shown as recursive arrows between the initial exchange and feedback box and the
LMX/interactional justice box).

In-Group/Out-Group Decisions
Dansereau et al. (1975) presented a descriptive model of how units become
differentiated into in-groups and out-groups based upon the degree of negotiating
latitude offered by the immediate supervisor to members of their work groups.
From a distributive justice perspective, LMX is equity-based. From procedural and
interactional justice perspectives, in-groups and out-groups may peacefully coexist,
if the leader maintains fairness in procedures and interactions with all work group
members (Tyler, 1986). Research has indicated that if a leader is procedurally fair,
LMX: An Organizational Justice Perspective 33

his/her resource allocation decisions will be accepted by all work group members.
even the out-group (Tyler & Caine, 1981).
In-group members are more likely to understand procedural justice issues due
to the higher quality LMX and communication with supervisors. In-group members
perform at higher levels, based upon perceptions that their leader is being procedur-
ally just and explains decisions (interactional justice). Out-group members may be
more likely to focus on dis~ibutive justice and perform at the level that is appropriate
to the rewards they receive based upon the formal employment agreement (Graen &
Scandura, 1987). This does not mean that procedural justice is irrelevant for out-
group members, however, because their performance may be more related to reward
distribution (distributive justice).

Performance and Other Outcomes

Since LMX has been linked to a number of outcome variables in organizational
research [cf. (Graen & Uhl-Bien)], it can be expected that job satisfaction, organiza-
tional commitment, and extra-role behavior (such as organizations citizenship)
would have similar relationships to organizational justice. Also, decision influence
and delegation (Scandura et al., 1986; Schriesheim, Neider, & Scandura, 1998)
would be outcomes of the LMX-organizational justice model. Also, absenteeism
and turnover might be negatively related to LMX and organizational justice vari-
ables, since those who perceive their leader as being fair may be less likely to
psychologically and/or physically withdraw from work. Role conflict and role ambi-
guity might be lower, since the process of why the work group has become differenti-
ated might be better unde~tood by work group members. For simplification, the
model and research propositions refer to performance as an outcome. However,
propositions with the additional outcome variables noted above as dependent vari-
ables can be tested as well.

DIRECTIONS FOR FUTURE RESEARCH

Propositions
Some research questions on LMX and justice must be examined from multiple
levels of analysis (Dansereau, Alutto, & Yammarino, 1984, Dansereau et al., 1995;
Klein, Dansereau, & Hall, 1994). Between-dyad justice issues emerge at the unit
level, when social compa~son processes are operative (Adams, 196.5). For example,
a person may perceive that their raise is fair until they learn that another person’s
raise is higher than theirs. Leaders must try to achieve perceptions of procedural
justice through communi~tion (interactional justice) with all unit members to
accept distributive outcomes. The examination of organizational justice and LMX
will require multiple levels of analysis (Dansereau et al., 1984; Klein et al., 1994).
In the propositions that follow, these issues are marked with a “D” to designate
that the issue incorporates a between-dyad. level of analysis perspective. Issues at
the work group level are designated with a “G”. Research issues that cut across
organizational ders are designated with a iiT”. The use of these superscript distinc-
34 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

tions is new in multiple-level research, yet may improve theorizing by more directly
specifying the nature of the multiple-level effect.
In testing the following propositions, it is also important to bear in mind that
data from both supervisors and subordinates should be collected. While these
reports are sometimes discrepant in the literature on LMX (Scandura & Schries-
heim, 1994) this carries important information for researchers. Hence, Propositions
1 through 9 (Pl-9) can be examined from the subordinates’ perspective, the supervi-
sors’ perspective, or both, to determine whether perceptions are shared. More
research is needed that examines multiple perspectives to more fully understand
the implications of disagreement on the relationship. No published research could
be located in this review on LMX and organizational justice has employed multiple
levels and this appears to be a key area for future research. Proposition 10 requires
the collection of data from both leaders and members in order to examine the
cross-level effect posited.
The first six propositions involve between-dyad issues relating to the role of
the organizational justice variables in the emergence of in-group and out-groups.
Investments (Dansereau et al., 1984) or contributions (Dienesch & Liden, 1986)
are an element of distributive justice during work group differentiation process.
The member contributes to the work unit and norms of equitable allocation of
rewards for these contributions are established over time.
One implication of the model presented in Figure 1, is that out-group membership
is based upon distributive justice concerns (what is invested and received as per-
ceived by members):

P1: Distributive justice contributes to the out-group decision.”

For in-group members, however, procedural justice is more important than what
is received (distributive justice). One key aspect of procedural justice is “voice” or
the ability to participate in decisions (Greenberg, 1990). In-group members, due
to their higher quality relationships with the leader (LMX), are more likely to
perceive the leader as following fair procedures, and allowing decision influence
(Scandura et al., 1986). Hence,

P2 Procedural justice cont~butes to the in-group decision.D

The communication process is central to the development of LMX, and thus

interactional justice is proposed to determine in-group membership. It can be ex-
pected that leaders and members who communicate about organizational justice
issues are more likely to evolve into in-group relationships:

P3: Interactional justice contributes to the in-group decision.D

It can be expected that procedural justice mediates the link between in-group
membership and performance. A mediating variable is an inde~ndent variable
that explains all the variance in a dependent variable previously explained by
another independent variable in a model. Once a mediator is introduced, no unique
LMX: An Organizational Justice Perspective 35

variance is explained by the first independent variable (Baron & Kenny, 1986). If
procedural justice is experienced by members, the effect of in/group or out-group
memberships on performance becomes nonsignificant.

P4: Procedural justice mediates the relationship between in-group/out-group

membership and performance.D

Similarly, distributive justice perceptions are expected to mediate the relationship

between in-group/out-group membership and performance. Those who view their
rewards as fair will accept their status and thus the effect on pe~o~~ce becomes
nonsignificant:

P5: Distributive justice mediates the relationship between in-group/out-

group membership and performance.*

LMX and interactional justice should have unique effects on performance. Aug-
menting effects in leadership research are additive effects of one variable over
another (Bass, 1990). Having a high quality relationship plus being able to discuss
organizational justice issues when they arise (interactional justice) should result in
higher performance:

P6: Interactional justice augments the relationship of in-group/out-group

membership and performance.D

The next three propositions address the group level of analysis because social
comparisons between members of the work group are involved. Based upon social
comparisons with all other members of the work group, an individual makes judg-
ments regarding distributive, procedural, and interactional justice. As suggested by
Dansereau et al. (1984), the dete~ination of what is fair in leadership is often a
function of social comparisons. Out-group members are more likely to base social
comparisons on what is received (inputs). They focus on what is received more
than the procedures employed to determine reward distribution. They may even
believe that outcomes should be equally distributed. Since out-group members are
less likely to communicate with the boss regarding these issues, they may be unaware
of procedural justice:

P7: Out-group members make social comparisons based upon distributive

justice.G

For in-group members, the relationship between the member’s performance and
rewards is clear, although immediate rewards are not necessary, due to the in-group
member’s commitment to the leader (Graen & Scandura, 1987). For example, a
member may even forgo a reward to preserve what’s perceived as fair by other
work unit members. There is a high degree of interdependence between leader and
in-group member, so the member might do what’s best for the group. Hence, in-
36 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

group members are more likely to base their social comparisons on the leader’s
procedural justice:

P8: In-group members make social comparisons based upon procedural jus-
tice.G

In-group members communicate more frequently with their leaders on issues of

organizational justice (Manogran et al., 1994). Since there is a higher level of
interdependence between leader and member, issues in the work group are more
openly discussed with in-group members. Thus, these members will rely on informa-
tion from the leader (interactional justice) in making social comparisons involving
others in the work group:

P9: In-group members make social comparisons based upon interactional

justice.G

The final proposition involves cross-level effects because the leader’s relationship
with his/her boss must be studied in addition to relations within and between work
group members. As noted earlier, data from both leaders and members must be
collected to test these relationships (data from the leader regarding his/her relation-
ship with the boss must also be obtained). Members receive rewards in the work
group based upon their investments, however this assumes that the leader has
sufficient rewards to distribute among members. This is a multi-tier issue for re-
search, suggested by a study conducted by Graen, Ginsburgh, and Schiemann (1977).
In this study, the quality of the relationship with the leader’s supervisor was related
to the quality of the relationships the leader developed with his/her work group
members. It is expected that a higher quality relationship with the leader’s boss
will result in (a) more rewards for members (distributive justice), (b) better under-
standing of procedures to follow in reward administration (procedural justice), and
(c) better communications regarding organizational justice concerns (interactional
justice):

PlO: The leader’s relationship with his/her supervisor determines the level
of (a) distributive, (b) procedural, and (c) interactional justice perceived
by the work group.T

Implications for Practice

Perceptions of organizational justice within work groups must be maintained
throughout the LMX development process. For example, a critical incident in which
the leader or the member perceives the other’s action as violating the norms that
have emerged, justice may be questioned. The idea that attributions of leaders (or
members) may bound the process of leader-member exchanges has been proposed
by Dienesch and Liden (1986). For example, a leader may fail to come through on
a promised reward (a distributive justice concern), without appropriately communi-
cating the reasons to the member (interactional justice). Without communications
LMX: An Organizational Justice Perspective 37

about organizational justice, the member begins to question the leader’s actions,
which over time destroys the established norms of procedural justice. These percep-
tions may send the relationship back to the role-specification phase (see Figure 1).
A second implication of the model outlined in this review is that leaders should
offer in-group relationships to all work group members initially. Also, out-group
members should be re-tested periodically by the leader making offers of in-group
roles. Work group differentiation should not be based on factors other than perfor-
mance (such as race, sex, or handicap status). The assumption should be made that
all members can become in-group members if given the opportunity to contribute
to the work group and the research base on LMX supports this (Scandura & Graen,
1984). Thus, the leaders’ offering of in-group tasks and benefits to all members has
clear ethical (and perhaps legal) implications. Thus, the integration of organizational
justice and LMX moves LMX theory in the direction of normative theory which
provides clearer guidelines for leaders in the management of work groups. Access
must be provided to the leadership process for all members and out-group status
should be based upon member’s decisions not to participate and/or performance
and not other factors.
Failure to recognize the important role that organizational justice plays in LMX
can help explain why some high quality LMXs disintegrate over time. Recognition
of distributive, procedural, and interactional justice is necessary to maintain long-
term LMX relationships. Key issues for future empirical study are attribution pro-
cesses and interactional justice variables (such as perceptions of honesty) in the
process of communicating justice issues to members (Lind & Lissak, 1985). Percep-
tions of justice may operate at multiple levels of analysis-both between unit
members as they compare their inputs to rewards, but also within the dyad as
individuals compare their current level of outcomes with previous outcomes from
the LMX relationship (Dansereau, et al., 1984; Klein, et al., 1994).

SUMMARY
Issues of organizational justice appear central to further refinement of the LMX
model. A conceptual framework was offered in this review that integrates LMX
and organizational justice theories. This framework highlighted (D,G,T)some multiple
level research propositions that might be pursued in future research on LMX
and organizational justice. The empirical examination of some research questions
regarding LMX and justice requires multiple levels of analysis perspective because
a leader may have an overall approach to justice (between unit) and may also
develop unique justice norms with members one-on-one through the LMX develop-
ment process (within unit). Also, the collection of data from multiple perspectives
to determine the degree of agreement between leaders and members is encouraged.
The LMX model is perhaps one of the more promising developments in leadership
research. Grounded in initial descriptive studies of work group differentiation
(Dansereau et al., 1975), the model has been integrated with a number of other
theories and variables in organizational research over the past 25 years (Graen &
Uhl-Bien, 1995). Although the importance of equity was noted by some leadership
theorists [cf. (Dansereau et al., 1984)], research on this aspect of LMX has been
38 LEADERSHIP QUARTERLY Vol. 10 No. 1 1999

sparse. This review hopefully addresses this deficiency in the literature by further
developing a conceptualization of LMX and organizational justice.

Acknowledgments: I sincerely appreciate the constructive comments of three anony-

mous Leadership Quarterly reviewers. I would also like to thank Gail Fairhurst,
George Graen, and Robert Moorman for helpful comments on earlier versions of
the manuscript.

REFERENCES
Adams, J. S. (1965). Inequity in socialexchange.In L. Berkowitz (Ed.), Advances in experi-
mental social psychology, 2, 267-300.
Baron, R. B., & Kenny, D. A. (1986).The moderator-mediatorvariable distinction in social
psychologicalresearch:Conceptual,strategicand statisticalconsiderations.Journal of
Personality and Social Psychology, 51, 1173-1182.
Bass,B. M. (1990). Bass and Stogdill’s handbook of leadership, 3rd. ed. New York, NY:
Free Press.
Bies, R. J., & Moag, J. S. (1986). Interactional Justice: Communicationcriteria of effective-
ness.In R. J. Lewicki, B. H. Sheppard, & B. H. Bazerman (Eds.), Negotiation in
Organizations, 1, 43-55, Greenwich, CT: JAI Press.
Cleyman, K. L., Jex, S. M., & Love, K. G. (1993). Employee grievances: An application of
the leader-member exchange model. Paper presentedat the 9th Annual Meeting of the
Society for Industrial and Organizational Psychology, Nashville, TN.
Cropanzano, R., & Folger, R. (1991). Procedural justice and worker motivation. In R. M.
Steers and L. W. Porter (Eds.), Motivation and work behavior. New York, NY:
McGraw-Hill.
Dansereau,F., Alutto, J. A., & Yammarino, F. J. (1984). Theory testing in organizational
behavior: The varient approach. Englewood Cliffs, NJ: Prentice-Hall.
Dansereau, F., Graen, G., & Haga, W. J. (1975). A vertical dyad linkage approach to
leadershipwithin formal organizations:A longitudinal investigation of the role making
process.Organizational Behavior and Human Performance, 13. 46-78.
Dansereau,F., Yammarino, F. J., & Markham, S. E. (1995).Leadership:The multiple-level
approaches.Leadership Quarterly, 6, 97-109.
Dienesch,R. M., & Liden, R. C. (1986).Leader-memberexchangemodel of leadership:A
critique and further development.Academy of Management Review, II, 618-634.
Duarte, N. T.. Goodson,J. R., & Klich, N. R. (1994).Effects of dyadic quality and duration
on performanceappraisal.Academy of Management Journal, 37,499-521.
Fairhurst, G. T., & Chandler, T. (1989). Social structure in leader-memberinteraction.
Communication Monographs, 56, 215-239.
Ferris, G. R. (1985).Role of leadershipin the employeewithdrawal process:A constructive
replication. Journal of Applied Psychology, 70, 777-781.
Festinger, L. (1957).A theory of cognitive dissonance. Evanston, IL: Row, Peterson.
Folger, R., & Greenberg, J. (1985).Proceduraljustice: An interpretive analysisof personnel
systems.In K. Rowland & G. Ferris (Eds.), Research in personnel and human resources
management, Vol. 3 (pp. 141-183).Greenwich, CT: JAI Press.
Folger, R., & Konovsky. M. A. (1989). Effects of procedural and distributive justice on
reactionsto pay raise decisions.Academy of Management Journal, 32, 115-130.
Graen, G., & Cashman,J. (1975).A role-makingmodel of leadershipin formal organizations:
LMX: An Organizational Justice Perspective 39

A developmental approach. In J. G. Hunt & L. L. Larson (Eds.), Leadership frontiers.

Kent, OH: Kent State University Press.
Graen, G., Ginsburgh, S., & Schiemann,W. (1977). Effects of linking-pin quality on the
quality of working life of lower participants. Administrative Science Quarterly, 22,
491-504.
Graen, G., Liden, R., & Hoel, W. (1982a).Role of leadershipin the employeewithdrawal
process.Journal of Applied Psychology, 67, 868-872.
Graen, G., Novak, M., & Sommerkamp,P. (1982b).The effects of leader-memberexchange
and job designon productivity and satisfaction:Testing a dual attachment model.
Organizational Behavior and Human Performance, 30, 109-131.
Graen, G. B.. & Scandura,T. A. (1986). A theory of dyadic career reality. In G. Ferris
and K. Rowland (Eds.), Research in Personnel and Human Resource Management, 4:
147-181.
Graen, G. B., & Scandura,T. A. (1987).Toward a psychologyof dyadic organizing.In L. L.
Cummings,& B. Staw (Eds). Research in Organizational Behavior, 9: 175-208.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-basedapproachto leadership:Develop-
ment of Leader-memberexchange(LMX) theory of leadershipover 25years:Applying
a multi-level-multi-domain perspective.Leadership Quarterly, 6, 219-247.
Greenberg,J. (1987).A taxonomy of organizationaljusticetheories.Academy ofManagement
Review, 12, 9-22.
Greenberg, J. (1990). Organizational justice: Yesterday, today and tomorrow. Journal of
Management, 16, 399-432.
Greenberg, J.. & Folger, R. (1983). Procedural justice, participation, and the fair process
effect in groupsand organizations. In P. B. Paulus(Ed.), Basic group processes (pp.
235-256).New York, NY: Springer-Verlag.
Hollander. E. P. (1978). Leadership dynamics: A practical guide to effective relationships.
New York, NY: Free Press.
Kabanoff. B. (1991).Equity, equality, power, and conflict. Academy of Management Review,
16.416-441.
Klein, K. J., Dansereau,F., & Hall, R. J. (1994).Levels issuesin theory development,data
collection, and analysis.Academy of Management Review, 19, 195-229.
Lagace,R. R.. Castleberry, S. B., & Ridnour, R. E. (1993).An exploratory salesforcestudy
of the relationship betweenleader-memberexchangeand motivation, role stress,and
managerevaluation. Journal of Applied Business Research, 9, 110-119.
Liden, R. C., Wayne, S.J.. & Stilwell, D. (1993).A longitudinalstudy on the early development
of leader-memberexchanges.Journal of Applied Psychology, 78, 662-674.
Lind, E. A., & Lissak,R. I. (1985).Apparent impropriety andproceduralfairnessjudgements.
Journal of Experimental Social Psychology, 21, 19-29.
Lind. E. A.. & Tyler, T. R. (1988). The social psychology of procedural justice. New York,
NY: PlenumPress.
Manogran, P.,Stauffer, J., & Conlon, E. J. (1994). Leader-member exchange as a key mediating
variable between employees’ perceptions of fairness and organizational citizenship be-
havior. National Academy of Managementmeeting proceedings,Dallas, TX.
Meindl, J. R. (1989).Managing to be fair: An exploration of values,motives, and leadership.
Administrative Science Quarterly, 34, 252-276.
Miner, J. B. (1980). Theories of organizational behavior. Hinsdale,IL: Dryen Press.
Moorman, R. H. (1991). Relationship between organizational justice and organizational
citizenship behavior: Do fairnessperceptionsinfluence employeecitizenship?Journal
of Applied Psychology, 76, 845-855.
Scandura,T. A., & Graen. G. B. (1984).Moderating effectsof initial leader-memberexchange
40 LEADERSHIP QUARTERLY Vol. 10 No. 1 19W

status on the effects of a leadership intervention. Jourrzul of Applied P.sychology. hY

4288436.
Scandura,T. A., Graen, G. B., & Novak, M. A. (1986).When managersdecidenot to decide
autocratically: An investigation of leader-memberexchangeand decisioninfluence.
Journal of Applied Psychology, 71, 579-584.
Scandura,T. A., & Schriesheim,C. A. (1994). Leader-memberexchange and supervisor
career mentoring as complementary concepts in leadershipresearch. Academy of
Management Journal, 37, 15881602.
Schiemann,W. A. (1977). Structural and interpersonal effects on patterns of managerial
communications: A longitudinal investigation (Ph.D. dissertation)University of Illinois.
Urbana-Champaign,IL.
Schriesheim,C. A., Neider, L. L.. & Scandura,T. A. (1998). A within- and between-groups
analysisof leader-memberexchangeas a correlate of delegationand a moderator of
delegationrelationshipswith performance and satisfaction.Academy of Management
Journal, 41, 298-318.
Sias,P. M., & Jablin, F. M. (1995).Differential superior-subordinaterelations, perceptions
of fairness,and coworker communication.Human Communication Research, 22, S-38.
Tyler, T. R. (1986). The psychology of leadership evaluation. In H. W. Bierhoff, R. L.
Cohen, & J. Greenberg (Eds.), Justice in social relations. New York, NY: Plenum.
Tyler, T. R. (1989). The psychology of procedural justice: A test of the group-value model.
Journal of Personality and Social Psychology, 57, 830-838.
Tyler, T. R., & Caine, A. (1981). The role of distributive and procedural fairnessin the
endorsementof formal leaders.Journal of Personality and Social Psychology, 41,
643-655.
Tyler, T. R., & Lind, E. A. (1992).A relational model of authority in groups.In M. P. Zanna
(Ed.), Advances in experimental social psychology, 25, 115-191.
Vecchio. R. P. (1985).Predictingemployeeturnover from leader-memberexchange:A failure
to replicate. Academy of Management Journal, 28, 478-485.
Vecchio, R. P. (1997).Are you in or out with your boss?In R. P. Vecchio (Ed.), Leadership:
Understanding the dynamics of power and influence in organizations. Notre Dame, IN:
University of Notre Dame Press.
Vecchio, R. P., & Gobdel, B. C. (1984). The vertical dyad linkage model of leadership:
Problemsand prospects.Organizational Behavior and Human Performance, 34,5-20.
Vecchio, R. P., Griffeth, R. W., & Horn, P. W. (1986). The predictive utility of the vertical
dyad linkage approach.Journal of Applied Psychology, 126, 617-625.
Wakabayashi,M., & Graen, G. (1984).The Japanesecareerprogressstudy: A 7-year follow
up. Journal of Applied Psychology, 69, 603-614.
Yukl, G. (1994). Leadership in organizations, 3rd ed. Englewood Cliffs, NJ: Prentice-Hall.