PREDICTING RETALIATION IN THE WORKPLACE THE THEORY OF PLANNED BEHAVIOR AND ORGANIZATIONAL JUSTICE

Dr. NW

Calgary, AB, Canada, T2N 1N4

INTRODUCTION

Counterproductive behavior in the workplace can be costly for organizations; up to 75% of

employees engage in theft, sabotage, unexcused absenteeism, or vandalism (Harper, 1990). Of

the many predictors of deviant workplace behavior, perceived unfairness is among the strongest

(Bennett & Robinson, 2000). Counterproductive behavior that occurs in response to unfair

treatment often represents an attempt to restore justice by “getting even” with a person or

organization, and is referred to as retaliation (Skarlicki & Folger, 1997).

In this study, I investigated how employees’ perceptions of unfair treatment predict their

decisions to retaliate against their supervisors and organization. Specifically, I tested whether

individuals who perceive different types of injustice direct their retaliation at different targets. To

account for other contextual factors, I also tested whether the theory of planned behavior (Ajzen,

1991) offers a useful theoretical framework for understanding retaliation in the workplace.

The Theory of Planned Behavior

The theory of planned behavior (TPB, Ajzen, 1991) states that behavioral intentions are the

most immediate precursor of behavior and, in turn, intentions are predicted by behavioral

control, expected utility, and perceived norms. As applied to retaliation, behavioral control refers

to beliefs about the existence of safe opportunities to “get back”. Support for behavioral control

as a predictor of retaliation comes from studies showing that employees who perceived a low,

rather than high, chance of punishment were more likely to model coworkers’ anti-social

behavior (Robinson & O’Leary-Kelly, 1998), and employees who perceived a low, rather than

high, possibility of detection were over three times as likely to steal (Hollinger & Clark, 1983).

Expected utility refers to the extent to which an individual believes that the outcomes of

retaliating are worth the potential costs. Retaliation offers a means of restoring self-esteem and

satisfying a desire for revenge (Bies & Tripp, 1996). Retaliation also allows one to maintain a

belief in a just world (Lerner, 1980) by causing something “bad” to happen to a “bad” person

who perpetrated an injustice. Individuals calculate the costs and benefits of retaliation (Bies &

Tripp, 2001) and may be more likely to retaliate if such an action is deemed worthwhile.

Perceived norms refer to beliefs about whether “significant others” might approve of

retaliation. Counterproductive behaviors are predicted by antisocial climates (Robinson &

O’Leary-Kelly, 1998) and employees are absent more often when it is common among their

coworkers (Johns, 1997). Moreover, theft is influenced by employees’ beliefs about how a

thieving coworker would be judged (Shaprio, Trevino, & Victor, 1995).

The above review suggests that the TPB predictors may influence decisions to retaliate. I

hypothesized that behavioral control, expected utility, and perceived norms predict intent and,

ultimately, retaliation against supervisors (H1a, H1b, H1c) and organizations (H2a, H2b, H2c).

Academy of Management Best Conference Paper 2003 OB: L1

Organizational Justice and Retaliation

Researchers have examined the perceived fairness of outcomes (distributive justice),

procedures (procedural justice), and treatment from authorities (interactional justice) (Folger &

Cropanzano, 1998). Interactional justice comprises perceptions of interpersonal justice, which

refers to the extent that authorities treat employees with dignity and respect, and informational

justice, which is the extent to which authorities provide adequate explanations for decisions

(Colquitt, 2001; Greenberg, 1993). Regardless of its form, injustice represents a loss of

something to which individuals feel entitled (Folger & Cropanzano, 1998); hence, people seek

retribution to restore justice (Bies & Tripp, 1996).

Bennett and Robinson (2002) proposed that researchers should examine the relative impact

of different types of justice on various forms of deviance. The agent-system model (Masterson,

Lewis, Goldman, & Taylor, 2000) suggests that the source of procedural justice is typically the

organization, whereas the source of interactional justice is often a supervisor. Thus, given social

exchange principles, procedural justice should predict retaliation against organizations (RAO),

and interactional justice should be related to retaliation against supervisors (RAS).

In addition to RAS, interactional justice might also predict RAO. Individuals who are treated

unfairly by a supervisor might retaliate against the organization that their supervisor represents

(Ambrose, Seabright, & Schminke, 2002), which might also prevent a cycle of “tit for tat” with

an unfair supervisor (Aquino, Lewis, & Bradfield, 1999). Indeed, studies have shown that

interactional justice predicts retaliation toward both the organization and its members (i.e.,

coworkers and supervisors), but procedural justice tends to predict only retaliation against the

organization (Ambrose et al., 2002; Aquino et al., 1999; Henle, 2002). In general, interactional

justice might be a stronger predictor of retaliation because it is easier to attribute blame for

mistreatment by a supervisor than for an unfair procedure (Folger & Cropanzano, 1998), and

blame is a critical component of revenge (e.g., Bies & Tripp, 1996).

In this study, interactional justice was assessed through informational and interpersonal

justice (Colquitt, 2001). I hypothesized that informational and interpersonal justice would predict

RAS (H3a, H3b) and RAO (H4a, H4b). Procedural justice was expected to predict RAO (H5),

and informational and interpersonal justice were hypothesized to be stronger predictors of RAS

(H6a, H6b) and RAO (H7a, H7b) than procedural justice. Because justice-retaliation

relationships are well-supported, all hypothesized effects were tested stringently by examining

them incremental to the other justice constructs and the TPB variables.

METHODOLOGY

Participants, Procedure, and Measures

Participants (N = 96) were employed individuals taking courses at a Canadian university

(67% female, age: M = 22.4, SD = 3.99). On average, participants had held over four jobs,

resulting in 6.5 years of work experience. Participants had been employed in their present

organization for over two years in various industries including service (37.5%), sales (23%),

professional (13.5%), and others (26%). I sampled from multiple organizations because it might

result in: (a) greater variance in, for example, the perceived risk of being detected retaliating, (b)

increased generalizability (Bennett & Robinson, 2002), and (c) reduced response distortion that

might have resulted from concerns about admitting to retaliation in an organization-wide survey.

Academy of Management Best Conference Paper 2003 OB: L2

Participants completed questionnaires in a quiet room and were not asked to name

themselves or their employer. The order of whether the justice measures were completed before

(n = 49) or after (n = 47) the retaliation and TPB measures was counterbalanced.

Organizational justice variables were assessed using Colquitt’s (2001) items. Response

options ranged from 1 (to a small extent) to 7 (to a large extent) for interpersonal (four items,

Cronbach’s alpha = .92), informational (five items, alpha = .90), procedural (seven items, alpha

= .76), and distributive justice (four items, alpha = .96).

Each TPB item was assessed on a scale from 1 (strongly disagree) to 7 (strongly agree). TPB

items for RAS and RAO included “my immediate supervisor” or “the organization for which I

work”, respectively, and consisted of intentions (I intend to get back at or settle the score

with…), behavioral control (If I was mistreated by…, I could easily do things to settle the score

with little chance that I would be caught), expected utility (If I was mistreated by…, the

satisfaction of getting even would outweigh the risks of getting caught), and perceived norms (If

my coworkers were mistreated by…, many of them would try to get back in some way).

Items used to assess retaliation were very similar to items used in previous research (Aquino

et al., 1999; Robinson & O’Leary-Kelly, 1998; Skarlicki & Folger, 1997), and were assessed

through self-report because retaliation is often enacted in private (Bennett & Robinson, 2000).

Seven items assessed RAO (alpha = .72), five of which were taken from Bennett and Robinson

(2000), and the response format ranged from 1 (never) to 7 (daily): “Taken property from work

without permission”, “Taken an additional or longer break than is acceptable at your workplace”,

“Put little effort into your work”, “Intentionally worked slower than you could have worked”,

and “Spent too much time fantasizing or daydreaming instead of working”. Two items

constructed for this study were assessed

...