NEW DIRECTIONS IN EQUITY RESEARCH€¦ · 26/05/1971 · Equity theory consists of four propositions designed to predict when individuals will perceive that they are justly treated

Journal of Personality and Social Psychology1973, Vol. 25, No. 2, 151-176

NEW DIRECTIONS IN EQUITY RESEARCH

ELAINE WALSTER2

University of Wisconsin

ELLEN BERSCHEID

University of Minnesota

G. WILLIAM WALSTER

University of Wisconsin

This article consists of four sections: The first section elucidates a general theory ofsocial behavior—equity theory. Equity theory consists of four propositions designedto predict when individuals will perceive that they are justly treated and howthey will react when they find themselves enmeshed in unjust relationships. Thesecond section summarizes the extensive research that has been conducted to lestequity theory. The third section points out the ways in which equity theory inter-locks with other major social psychological theories. The final section hints at someways in which equity theory can be applied to understanding social problems.

THE THEORETICAL FORMULATION

The proposition that individuals seek forthemselves maximum reward at minimum costis hardly startling. Theories in a wide varietyof disciplines rest on the assumption that"Man is selfish." Psychologists believe thatbehavior can be shaped by the careful appli-cation of reinforcements. Economists assumethat individuals will purchase desired productsat the lowest available price. Moral philos-ophers conclude that the ideal society canonly be one which insures the "greatest goodfor the greatest number." Politicians contendthat "Every man has his price." Equity theory,too, rests on the assumption that man isselfish.

Thus, our first proposition states:Proposition I: Individuals will try to maxi-

mize their outcomes (where outcomes equalrewards minus costs).

If everyone were unrestrained in his at-tempts to get what he wanted, everyone wouldsuffer. Every man would attempt to monopo-lize community resources; every man would becontinually confronted by rivals bent on re-claiming these resources. Only by workingout a compromise can the group avoid con-tinual warfare and maximize collective reward.

This fact is acknowledged in PropositionII A:

1 Preparation of this article was financed in part byNational Institute of Mental Health Grants MH16729 and MH 16661.

2 Requests for reprints should be sent to ElaineWalster, Department of Sociology, University of Wis-consin, Madison, Wisconsin 53706.

Proposition IIA: Groups can maximizecollective reward by evolving accepted sys-tems for "equitably" apportioning rewardsand costs among members. Thus, memberswill evolve such systems of equity and willattempt to induce members to accept and ad-here to these systems.

If group members are to be effective in in-ducing their fellow members to behaveequitably, they must make it more profitablefor the individual to behave equitably thaninequitably (see Proposition I). Thus, Pro-position IIB proposes that groups will insurethe profitability of equitable behavior in thefollowing way:

Proposition IIB: Groups will generally re-ward members who treat others equitably andgenerally punish (increase the costs for) mem-bers who treat others inequitably.

What Constitutes an Equitable Relationship?

In Proposition IIA we argued that everyculture must institutionalize systems forequitably apportioning resources among itsmembers (see De Jong, 1952). We must admit,however, that the perception of what is equi-table varies enormously between cultures. Inspite of the fact that different societies haveestablished strikingly different procedures forpartitioning their resources, we have found asingle general principle to be useful in charac-terizing these widely diverse conceptions aboutwhat is equitable (see Adams, 1965; Blau,1967; Romans, 1961; Walster, Berscheid, &Walster, 1970).

151

152 E. WALSTEE, E. BERSCHEID, AND G. W. WALSTER

We define an "equitable relationship" toexist when the person scrutinizing the relation-ship (i.e., the scrutineer—who could beParticipant A, Participant B, or an outsideobserver) perceives that all participants arereceiving equal relative outcomes from the

/ Outcomes^ Outcomes/Arelationship ( i.e.,

\ ' Inputs,i

Definition of Terms

Outcomes (0) are defined as the positiveand negative consequences that a scrutineerperceives a participant has incurred as a con-sequence of his relationship with another.Following Homans (1961), we shall refer topositive outcomes as "rewards" and negativeoutcomes as "costs." The participant's totaloutcomes in a relationship are equal to the re-wards he obtains from the relationship minusthe costs he incurs.

We denned inputs (7) as "the participant'scontributions to the exchange, which are seen(by a scrutineer) as entitling him to rewardsor costs." The inputs that a participant con-tributes to a relationship can be either assets—entitling him to rewards—or liabilities—en-titling him to costs.

Ln different settings, different inputs areseen as entitling one to rewards or costs. Inindustrial settings, assets such as "capital"or "manual labor" are seen as relevant in-puts—inputs that legitimately entitle thecontributor to reward. In social settings, assetssuch as physical beauty or kindness are gen-erally seen as assets entitling the possessor tosocial reward. Social liabilities such as boorish-ness or cruelty are seen as liabilities entitlingone to costs.

Definitional Formula

Adams proposed the simple formula -j- = -p-IA 1 H

for designating an equitable relationship, whereOA and On are A's and B's outcomes from therelationship, and IA and /« are A's and B'sinputs to the relationship. Unfortunately, thissimple notation is adequate only so long as allparticipants have positive inputs. This formulais not suitable in social relations where inputsmay be negative as well as positive. For ex-ample, according to the preceding formula, a

relationship would be calculated as equitable ifI A = 5, 0A = - 10, IB J= - 5, and 0« = 10.

0A -10Substituting in the formula : — = = — 2

Inputs,, /' was>

and — = —r = — 2. Obviously neither1 n ^

Adams nor we would feel such a relationshipin fact, equitable,place of Adams's formula, we have chosen

Allowing formula for calculating whethertheor not a relationship is equitable:

where 0 designates a scrutineer's perceptionof A's and B's outcomes, / designates his per-ception of A's and B's inputs, and J 7 desig-nates the absolute value of A's and B's inputs(i.e., the perceived value of A's and B's in-puts, disregarding sign). A participant's rela-tive outcomes will be zero if his outcomes equalhis inputs. His relative outcomes will be posi-tive if his 0 > 7, and negative if his 0 < 7.Thus, the sign and the magnitude of this mea-sure indicates how "profitable" the relation-ship has been to each of the participants.

Recall the preceding example, where I A = 5,0A = - 10, I n = - 5 and 0,, = 10. Sub-

. . . , . 0A-JA -15stituting in our formula

OH 15= -- = 3. Unlike Adams's

formula, our formula properly reveals that Bis reaping undeservedly large outcomes fromhis relationship with A.

Who Decides Whether a Relationship IsEquitable^

In Propositions 1IA and IFB we argued thatsocieties develop norms of equity and teachthese systems to their members. Thus, in anysociety there will be a general consensus as towhat constitutes an equitable relationship.However, the .preceding formulation makes itclear that ultimately, equity is in the eye of thebeholder. An individual's perception of howequitable a relationship is will depend on hisassessment of the value and relevance of thevarious participants' inputs and outcomes.

3 The restriction of this formula is that \I\> 0.

NEW DIRECTIONS IN EQUITY RESEARCH 153

Participants themselves, even after prolongednegotiation with one another, will not alwaysagree completely as to the value and relevanceof various inputs and outcomes. One personmay feel that a distinguished family name is arelevant input, entitling him to positive out-comes. His partner might disagree. Aesopacidly observed that "The injuries we do andthose we suffer are seldom weighted on thesame scales." If participants do calculate in-puts and outcomes differently—and it is likelythat they will—it is inevitable that partici-pants will differ in their perceptions of whetheror not a given relationship is equitable. More-over, "objective" outside observers are likelyto evaluate the equitableness of a relationshipquite differently than do participants.

Do People Generally Behave Equitably?

Proposition I states that individuals will tryto maximize their outcomes. We might re-phrase this proposition as:

Corollary I.I: So long as individuals perceivethat they can maximize their outcomes by be-having equitably, they will do so. Should theypsrceive that they can maximize their out-comes by behaving inequitably, they will do so.

In Propositions IIA and IIB we observedthat society will try to insure that membersrecognize that they can maximize their out-comes by behaving equitably. There is someevidence that individuals do generally behaveequitably. A number of experiments demon-strate that individuals will spontaneouslyshare rewards with others. Individuals whoare given more than their share of rewardvoluntarily surrender some benefit to theirdeprived partners. Deprived individuals, on theother hand, are quick to demand the rewardthey deserve (see, e.g., Leventhal, Allen, &Kemelgor, 1969; Harwell, Ratcliffe, & Schmitt,1969; Schmitt & Harwell, 1970").

In spite of the consistent evidence that in-dividuals do behave equitably, we must re-member that individuals will behave inequi-tably with some regularity5 (Corollary I.I).

4 Unpublished study entitled "Cooperation andInequity: Behavioral Effects," 1971. Address re-quests to D. R. Schmitt, Department of Psychology,University of Washington, Seattle, Washington 98106.

5 There are two reasons why we would expect aparticipant in a relationship to behave in a way that

We have just argued that individuals nowand then treat others in ways they perceiveto be inequitable. How should the socializedindividual react when he finds himself in aninequitable relationship?

We have argued that individuals are oftenpunished when they treat their colleagues in-equitably. We pointed out that the personwho extorts greater outcomes than he deserveswill be punished. Conversely, the person whoacquiesces in receiving fewer outcomes thanhe deserves will be punished—by deprivation(and possibly by his peers as well). Thus, in-dividuals should quickly come to associate"participating in an inequitable relationship"with punishment.

As a consequence of these inevitable sociali-zation experiences, we propose Proposition III.

Proposition III: When individuals find them-selves participating in inequitable relation-ships, they become distressed. The more in-equitable the relationship, the more distressindividuals feel.

Experimental evidence supports both ourProposition III and Homan's contention thatindividuals participating in inequitable rela-tionships do feel distress regardless of whetherthey are the victims or the beneficiaries ofthe inequity. Experiments by Walster et al.(1970), Leventhal etal. (1969), Jacques (1961),and Thibaut (1950) indicated that those whoreceive less than they deserve feel distress(usually in the form of anger). Experiments byJacques (1961), Adams and Rosenbaum(1962), Adams (1963), and Leventhal et al.(1969) demonstrated that those who receivemore than they deserve feel distress (usuallyin the form of guilt).

Evidence also exists to support the conten-tion that the greater the inequity, the moredistress participants feel (see Leventhal et al.1969; Leventhal & Bergman, 1969).

Proposition IV: Individuals who discover

he acknowledges is inequitable. Hrst, an individualshould behave inequitably whenever he is confidentthat in a given instance he can maximize his outcomesby doing so. Second, it is to the individual's long-rangebenefit to behave inequitably now and then. Only byvarying the equitableness of his behavior can a partic-ipant ascertain whether or not sanctions againstinequity are still operating. (Only by testing limitsoccasionally can one adapt to a changing world.)Thus, an individual can maximize his total outcomes ifhe tests equity norms now and then.

154 E. WALSTER, E. BERSCHEID, AND G. W. WALSTER

they are in an inequitable relationship attemptto eliminate their distress by restoring equity.The greater the inequity that exists, the moredistress they feel, and the harder they try torestore equity.

There are two ways that a participant canrestore equity to an inequitable relationship:He can restore actual equity to the relation-ship, or he can restore psychological equity.

A participant can restore "actual equity" byappropriately altering his own outcomes orinputs or the outcomes or inputs of the otherparticipants. For example, the ghetto blackwho feels his boss underpays him can re-establish actual equity by becoming a slacker(thus lowering his inputs), by forcing hisemployer to work harder (thus raising hisemployer's inputs), by demanding a raise orstealing from the company (thus raising hisown outcomes), or by sabotaging companyequipment (thus lowering his employer's out-comes). The ingenious ways individuals con-trive to bring equity to inequitable relation-ships are documented by Adams (1963).

A participant can restore "psychologicalequity" by appropriately distorting his per-ception of his own or his partner's outcomesand inputs.0 For example, an exploitativeemployer may convince himself that his in-equitable relationship with his underpaid andoverworked secretary is in fact equitable byappropriately distorting reality. He can re-store psychological equity to their relationshipif he can sufficiently minimize her inputs("You wouldn't believe how stupid she is."),exaggerate her outcomes ("Work gives her aa chance to see her friends."), exaggerate hisown inputs ("Without my creative genius thecompany would fall apart."), or minimize hisoutcomes ("The tension on this job is givingme an ulcer.").

Equity theorists concur that people try tomaximize their outcomes (Proposition 1). Agroup of individuals can maximize their total

8 We can, of course, detect whether or not participantsare "distorting" by (a) examining how participants'assessments of relevant inputs and outcomes com-pare to those of observers, who are not so motivatedto distort their perceptions of participants' relativeoutcomes, or (6) by examining whether or not theperceptions of participants suddenly shift in predictableways as soon as they discover they are participatingin an inequitable relationship.

outcomes by agreeing on some equitable sys-tem for sharing resources. (A relationship isdefined as equitable when a scrutineer perceivesthat all participants are securing equal relative

(0 -I\outcomes I ~[yr~ ) from the relationship, that

• °A ~IA °" ~ Il!

Groups try to insure that members canmaximize their outcomes by behaving equi-tably; they reward members who behaveequitably and punish members who behaveinequitably. When individuals socialized bythis system participate in inequitable relation-ships, they experience distress. Participantsreduce their distress either by restoring actualequity or by restoring psychological equityto the relationship.

Applications of Equity Theory

Researchers have applied the equity frame-work to four major areas of human interaction :(a) business relationships, (b) exploitative re-lationships, (c) helping relationships, and (</)intimate relationships.

Equity Theory and Business Relationships

The vast majority of equity research hasbeen conducted in industrial settings. Sinceexcellent and recent summaries of this materialare available elsewhere (see Adams, 1965;Lawler, 1968; Opsahl & Dunnette, 1966;Pritchard, 1969; Weick & Nesset, 1969), wewill not review this literature here.

Equity Theory and Exploitative Relationships

At first researchers were preoccupied withtesting equity theory in industrial settings.Eventually they realized that the theory couldbe applied to social as well as to business ex-changes. The first social relationships to comeunder their scrutiny were "exploitative" re-lationships — relationships in which one par-ticipant received far greater relative outcomesthan did another.

The decision to focus on exploitative rela-tions was a natural one. It is easy to analyzeexploitative relations within the equity frame-work. We can reasonably define a "harm-doer"as "one who commits an act which causes hispartner's relative outcomes to fall short of his


own." The participant whose relative out-comes are reduced is the "victim" of the in-equitable action.

Let us first consider how harm-doers havebeen found to respond after treating anotherinequitably.

Reactions of the Harm-Doer

According to our theory (Proposition III),an individual who receives higher outcomesthan he knows he deserves will feel distress.Compelling evidence exists to support thecontention that individuals do feel intensedistress after injuring another. Theorists havelabeled this distress as "guilt," "fear of retali-ation," "dissonance," "empathy," or "con-ditioned anxiety." Basically, however, aharm-doer's distress is presumed to arise fromtwo sources: fear of retaliation and threatenedself-esteem. Presumably, both retaliation dis-tress and self-concept distress have their rootsin the socialization process.

Retaliation Distress

Children are usually punished if they arecaught injuring others. Soon they come toexperience conditioned anxiety when theyharm another or even contemplate doing so.(Aronfreed, 1964, provides a description of thedevelopment of conditioned anxiety in chil-dren.) By the time the normal individual reachesadulthood, he experiences some distress when-ever he harms another. How much anxiety heexperiences in a given harm-doing situationdepends first on how similar the stimuli Asso-ciated with harm-doing are to those stimulipreviously associated with punishment, andsecond on the magnitude and timing of pre-vious punishment.

The conditioned anxiety that one experienceswhen he harms others may be labeled by theharm-doer as fear that the victim, the victim'ssympathizers, legal agencies, or even Godwill restore equity to the harm-doer/victimrelationship by punishing the exploiter.

Self-Concept Distress

Harm-doing often generates a second kindof discomfort: self-concept distress. In mostsocieties, nearly everyone accepts the ethicalprinciple that "one should be fair and equitable

in his dealings with others" (cf. Fromm, 1956,for an interesting discussion of the pervasive-ness of the "fairness" principle). Harminganother violates a normal individual's ethicalprinciples and conflicts with his self-expecta-tions. When the normal individual violates hisown standards, he experiences self-conceptdistress.

In arguing that the "normal" individualaccepts norms of "fairness," we are not arguingthat everyone internalizes exactly the samecode, or internalizes it to the same extent andfollows that code without deviation. Juveniledelinquents and confidence men, for example,often seem to behave as if the exploitation ofothers was completely consonant with theirself-expectations. Evidence suggests, however,that even deviants do internalize standards offairness, at least to some extent. It is true thatthey may repeatedly violate such standardsfor financial or social gain, but such violationsdo seem to cause at least minimal distress.Exploitation evidently causes deviants enoughdiscomfort that they spend time and efforttrying to convince others that their behavioris "fair." Anecdotal evidence on these pointscomes from interviews with confidence men(see Goffman, 1952) and delinquents (seeSykes & Matza, 1957).

The distress that arises when one performsunethical or dissonant acts has been discussedin great detail by guilt theorists and by cogni-tive dissonance theorists. See, for example,Maher (1966), Arnold (1960), and Bramel(1969) for these two points of view.

The Greater the Inequity, the Greater the Distress

In Propositions III and IV we proposed thatthe more inequitable a relationship, the moredistress the participants will feel, and theharder they will try to restore equity. There issome indirect evidence to support the notionthat the more one harms another, the moredistressed he will be. Brock and Buss (1962,1964) recruited student assistants to shockother students each time the students made alearning error. The authors discovered thatthe more painfully one required the assistantto shock his fellow student, the more the stu-dent tried to foist responsibility for administer-ing the shock onto the supervisor. Lerner andSimmons (1966) found that the more one

156 E. WALSTER, E. BERSCHEID, AND G. W. WALSTEU

allows another to suffer, the more one willderogate the hapless victim. Lerner andMatthews (1967) concluded that the moreresponsible subjects feel for another's suffering,the more they will derogate the victim. If weassume that a harm-doer will be especiallydistressed when his victim suffers a great dealor when the harm-doer is obviously responsiblefor the victim's suffering, these findings areconsistent with Propositions III and IV.

There is also much compelling evidenceavailable to support the contention (Pro-position IV) that participants involved in in-equitable relations will try to eliminate theirdistress by restoring either actual equity orpsychological equity to their relationship.

Restoration of A dual Equity

Harm-doers can restore equity in a straight-forward way: They can compensate theirvictims. When we think of "compensation," weusually envision acts designed to increase thevictim's outcomes. However, one can alsocompensate the victim by allowing him tolower his inputs. The underpaid and over-worked secretary, for example, may be en-couraged to "take Monday off" by her uneasyboss.

Cynics such as Junius have acidly observedthat even "a death bed repentance seldomreaches to restitution." This pessimism is notalways warranted. Recent studies verify thefact that harm-doers do commonly compensatetheir victims: Walster, Walster, Abrahams,and Brown (1966), Walster and Prestholdt(1966), Brock and Becker (1966), Berscheidand Walster (1967), Freedman, Wallington,and Bless (1967), Carlsmith and Gross (1969),and Berscheid, Walster, and Barclay (1969).

Theoretically, a harm-doer can restore actualequity to his relationship with the victim byusing a second strategy—self-deprivation. Theharm-doer could voluntarily reduce his ownrelative outcomes to the victim's level; onecould curtail his own outcomes from the rela-tionship or increase his inputs.

The anecdotal and clinical literature providesupport for the notion that individuals some-times do react to the commission of harmfulacts by administering punishment to them-selves or by seeking punishment from others.Indeed, Sarnoff (1962) has suggested that

"punishment is the only kind of response that issufficient to reduce the tension of guilt Q>.351]." Attempts to demonstrate self-punish-ment in the laboratory have been uniformlyunsuccessful, however. In the view of our as-sumption that individuals prefer to maximizetheir rewards whenever possible (PropositionI), we would expect individuals to restoreequity by employing self-punishment only asa last resort. The data indicate that self-deprivation is not a popular strategy forequilizing our relations with others.

Restoration of Psychological Equity

As we noted earlier, by distorting reality,one can restore psychological equity to hisrelationship with another. If the harm-doercan aggrandize the victim's relative outcomesor minimize his own, he can convince himselfthat his inequitable relationship is, in fact,equitable. Some distortions which harm-doershave been detected using include derogationof the victim, minimization of the victim'ssuffering, or denial of one's own responsibilityfor the victim's suffering.

Derogation of the Victim

An act which injures another is not inequi-table if the victim deserves to be harmed. Thus,an obvious way in which a person who hasharmed another can persuade himself that hisact was equitable is by devaluating thevictim's inputs.

That harm-doers will often derogate theirvictims has been demonstrated by Sykes andMatza (1957), Davis and Jones (1960),Berkowitz (1962), Davidson "(1964), Glass(1964), and Walster and Prestholdt (1966).In a typical experiment, Davis and Jones(1960) found that students who were hired tohumiliate other students (as part of a researchproject) generally ended up convincing them-selves that the student deserved to be ridiculed.Sykes and Matza (1957) found that juveniledelinquents often defend their victimizationof others by arguing that their victims arereally homosexuals, bums, or possess othertraits which make them deserving of punish-ment. In tormenting others, then, the de-linquents can claim to be the restorers ofjustice rather than harm-doers.


Minimization of the Victim's Suffering

If a harm-doer can deny that the victim washarmed, he can convince himself that his rela-tionship with the victim is an equitable one.Sykes and Matza (1957) and Brock and Buss(1962) demonstrated that harm-doers con-sistently underestimate how much harm theyhave done to another. Brock and Buss, forexample, found that college students who ad-minister electric shock to other students sooncome to markedly underestimate the painful-ness of the shock they are delivering.

Denial of Responsibility for the Act

Tf the harm-doer can perceive that it was nothis behavior but rather the action of someoneelse (e.g., the experimenter or fate) that causedthe victim's suffering, then his relationshipwith the victim becomes an equitable one.(The person who is unjustly assigned respon-sibility for reducing the victim's outcomes willnow be perceived as the harm-doer, and it willbe his relationship with the victim, not theharm-doer's relationship, that is perceived asinequitable.)

That harm-doers often deny their responsi-bility for harm-doing has been documented bySykes and Matza (1957) and by Brock andBuss (1962, 1964). In daily life, denial ofresponsibility seems to be a favorite strategyof those who are made to feel guilty aboutexploiting others. War criminals protestvehemently they were "only following orders."

The research results enumerated in this sec-tion document the eagerness with which harm-doers restore equity after injuring others.Equity theory provides an orderly frameworkfor cataloguing the possible reactions of harm-doers. But this is not enough. Researchersare more interested in prediction than indescription.

PREDICTION OF A HARM-DOER'S RESPONSE

To predict which of many potential tech-niques harm-doers will use, researchers adopteda simple strategy. They tried to (a) condensethe multitude of potentially equity-restoringresponses into a few meaningful categories and(b) isolate variables which determine whichclass of responses a harm-doer will choose.

Condensation of Responses into a Few Categories

Prediction was facilitated when researcherseventually realized that harm-doers' responsesgenerally fall into two distinct categories:Harm-doers tend either to compensate theirvictims (and to restore actual equity) or tojustify the victims' deprivation (and restorepsychological equity). Harm-doers rarely useboth techniques in concert; compensation andjustification seem to be alternative, rather thansupplementary, techniques for restoring equity.7

Logically, it should be difficult for a harm-doer to use compensation and justificationtechniques in concert. It should be difficultfor the harm-doer to simultaneously acknowl-edge on the one hand that he is at fault for thevictim's undeserved suffering and thus exerthimself in an attempt to assist the victim whileon the other hand convincing himself that hisvictim deserves to suffer, that he is not reallyinjured, or that the harm-doer is not respon-sible for the victim's suffering.8

There is empirical evidence that individualsgenerally do not use compensation and justifi-cation in concert. Walster and Prestholdt(1966) led social work trainees to inadvertentlyharm their clients. Subsequently trainees wereasked to evaluate the clients and asked tovolunteer their free time to help them. Com-pensation and justification responses werefound to be negatively related; the moretrainees derogated their clients, the less timethey volunteered to help them. Lerner andSimmons (1966) found that observers onlyderogated a victim when they were powerless toaid him in any way. When it was clear that

7 It is possible, of course, for individuals to use com-pensation and justification techniques in sequence. Forexample, a harm-doer may attempt to make compensa-tion, find it impossible, and then resort to justification.Or he may attempt to justify his behavior, have hisrationalizations challenged, and then accede to de-mands for compensation. However, individuals do notgenerally use compensation and justification techniquessimultaneously.

8 Although victims generally do not use compensationand justification techniques in concert, it is, of course,possible for them to do so. h'or example, a harm-doercould inadequately compensate the victim (thus par-tially restoring actual equity) and minimize the harmhe had done the victim. Distorting the victim's out-comes would allow the harm-doer to deceive himselfthat this inadequate compensation had completely re-stored equity.


compensation would occur, no derogation wasdetected. Thus, we can conclude:

Conclusion I: The harm-doer will avoid usingjustification techniques and compensationtechniques in concert.

Isolating Variables Which Influence a flarm-Doer's Response

Once researchers discovered that (a) ahodge-podge of equity-restoring techniquescould be classified into two distinct categories—compensation or justification—and (6) com-pensation and justification tend to be mutuallyexclusive techniques for restoring equity, theirtask was vastly simplified. It is easier to detectthe antecedents of two alternative responses—compensation or justification—than to ferretout the antecedents of a multitude of dis-organized responses.

In addition, the recognition that harm-doers generally compensate or justify made itevident that it is practically, as well as theo-retically, important to identify the variableswhich push harm-doers toward one or anotherresponse.

Any society has a vested interest in en-couraging harm-doers to voluntarily compen-sate their victims rather than derogating them.If a harm-doer refuses to make restitution, thevictim is left in sad straits. Not only has hebeen deprived of material benefits which hedeserves, but he must face both the indignityof derogation and the added difficulty that theharm-doer, because of his derogation, maycontinue to treat him unjustly (see Berscheid,Boye, & Darley, 1968). Societies should natu-ral!}' prefer that its citizens restore actualequity after committing injustices rather thanengaging in a series of justifications which endin shared bitterness and possible further harm-doing. For theoretical and practical reasons,then, we are interested in identifying thosevariables that encourage harm-doers to makevoluntary compensation and those variableswhich encourage justification.

Two situational variables have been foundto be important: (a) the adequacy of the exist-ing techniques for restoring equity and (b) thecost of the existing techniques for restoringequity. We would expect people to prefertechniques that completely restore equity totechniques tha t only partially restore equity,

and to prefer techniques with little materialor psychological cost to techniques with greatercost. More precisely we would expect:

Corollary I V.I: Other things being equal, themore adequate a harm-doer perceives an avail-able equity-restoring technique to be, the morelikely he is to use this technique to restoreequity.

Corollary 1.2: Other things being equal, themore costly a harm-doer perceives an availableequity-restoring technique to be, the less likelyhe is to use this technique to restore equity.

Adequacy of Equity-Restoring Techniques

The "adequacy" of a technique is defined asthe extent to which that technique will exactlyrestore equity to the harm-doer/victim rela-tionship. As we pointed out in Proposition III,participants in inequitable relations feel un-comfortable and the more inequitable the re-lationship, the worse participants feel. Harm-doers thus have a vested interest in reestablish-ing as equitable a relationship as possible.

Data support the contention (Corollary1V.1) that individuals are more likely to com-pensate their victim if adequate compensationis available than if it is not.

A dequacy of Compensation

By definition, an adequate compensation isone that can exactly balance the harm done.Both insufficient compensations and excessivecompensations lack adequacy; thus, harm-doers should be reluctant to make suchcompensations.

Why should a harm-doer be reluctant tomake an insufficient compensation? (Certainlyhis victim would prefer insufficient compensa-tion to no compensation at all.) If insufficientcompensation is the only way by which a harm-doer can reduce his distress, rather than donothing, he will probably choose to restore atleast partial equity to his relationship with thevictim. But insufficient compensation is notthe only equity-restoring technique open to theharm-doer. A technique incompatible withcompensation is available to him. The harm-doer can always completely eliminate his dis-tress by completely justifying the victim'ssuffering. Thus, as available compensationsbecome increasingly insufficient, justification


techniques should become increasingly appeal-ing to the harm-doer.

Making an excessive compensation is alsoan unsatisfactory way to restore equity. Anexcessive compensation eliminates one kindof inequity by producing another. The harm-doer who compensates his victim excessivelydoes not restore equity; he simply becomes avictim instead of a harm-doer—a most un-desirable transformation.

Berscheid and Walster (1967) tested thehypothesis that a harm-doer's tendency tocompensate his victim will be an increasingfunction of the adequacy of the compensationsavailable to him. Their results provide supportfor Corollary IV. 1. In this experiment, womenfrom various church groups were led to cheatfellow parishioners out of trading stamps in avain attempt to win additional stamps forthemselves. When the women were subse-quently given an opportunity to compensatethe victim (at no cost to themselves), it wasfound that adequacy of compensation wascrucial in determining whether or not thewomen chose to compensate. Women who couldcompensate with an adequate compensation(exactly restoring the number of books thepartner had lost) were much more likely tomake restitution than were women limited toinsufficient compensation (a few stamps) or toexcessive compensation (a great many stampbooks). This finding was replicated by Ber-scheid et al. (1969).

The hypothesis that individuals are pre-disposed to make adequate compensation andto resist making inadequate or excessive com-pensations has some interesting implications.In life, exploited individuals sometimes try toimpress on those in a position to make restitu-tion how much they have suffered in the hopeof eliciting increased restitution. It is naturalto assume that the better a case one makesfor his claim, the more likely it is that he will becompensated. The preceding research, how-ever, indicates that in some instances, it mightbe a more effective strategy for a victim tominimize his suffering than to aggrandize it.

The greater the inequity a victim documents,the more restitution the harm-doer should bewilling to make—up to a point. However, atsome point, the described inequity will be-come so large that the harm-doer will despair of

ever being able to make complete restitution.Once this point is reached, it is no longerprofitable for the victim to exaggerate hissuffering. Further exaggeration will not elicitincreased restitution—the harm-doer has al-ready reached his limit. In fact, the moreadditional suffering the victim describes, themore inadequate the compensations availableto the harm-doer become, and thus the moreunwilling he should become to provide anycompensation at all.

The preceding reasoning may provide someinsight into the public reaction to demands forcompensation. Most Americans probably feelthat no matter how hard they try they cannotmake adequate restitution to blacks for theircenturies of exploitation. Black leaders haveargued that citizens should at least take somestep toward restoring equity. However, theidea of making a small step toward restoringequity, making a small compensation, is notvery attractive to many citizens. The effort tomake a partial compensation mocks theirrationalizations that no harm was done or thatblacks deserved their treatment. If one can-not compensate enough to reduce his own dis-tress, he is perhaps happier with his rationali-zations. If we generalize shamelessly from thepreceding findings, we might speculate thatan effective strategy for deprived minoritiesmay be to minimize their description of theirsuffering and to make it clear that if availablecompensations are extended, it will completelyeliminate the debt owed to them. While this isnot true, it may be a profitable strategy sinceit would insure that blacks would receive atleast minimal compensation.

Adequacy of Justifications

To restore complete psychological equity inhis relationship with a victim, a harm-doermust be able to conceive of justifications that(a) adequately justify the harm done and (6)are plausible to himself, the victim, and toothers. (Only if the harm-doer believes his owndistortions will he be able to eliminate self-con-cept distress; only if he imagines that thevictim accepts his justifications will he be ableto eliminate fear of retaliation.)

Little is known about what causes an excuseto be seen as "adequate justification" for aharmful act. Once a harm-doer conceives of an


adequate justification, however, we know agreat deal about the factors which determinehow plausible or credible a given justification is(see McGuire, 1968). For a delightful and ex-tensive account of excuses that work, see ScottandLyman (1968).

In this article we describe only two factorsthat have been found to determine how plau-sible, and thus how readily used, variousjustifications are.

1. A harm-doer will perceive a potentialjustification to be more credible when it re-quires little distortion of reality than when itrequires a great deal of reality distortion. Themore serious or extensive a distortion of realityrequired by a justification, the less crediblethese justifications should be to the victim,the harm-doer, and others. A justification thatno one believes is not very effective in restoringequity. Some tangential support for thisproposition comes from Rosenberg and Abelson(1960), who provide evidence that individualsprefer to distort reality as little as possible.

2. The more contact the harm-doer has had(or anticipates having) with the victim or thevictim's sympathizers, the less likely he willbe to justify his harm-doing.

There are two reasons why one should bemore reluctant to distort an intimate's relativeoutcomes than to distort those of a stranger:First, the more intimate we are with someone,the more likely we are to have voluminous in-formation about that person. Thus, when onetries to distort an intimate's characteristics,he will soon find himself in trouble. The harm-doer's fine rationalizations will keep bumpingup against recalcitrant facts. However, it willbe easy for one's fantasies about a stranger toto proliferate boundlessly.

Virtually all of the cognitive consistencytheorists (Abelson & Rosenberg, 1958; Cart-wright & Harary, 1956; Festinger, 1957;Rosenberg, 1960; Zajonc, 1960) acknowledgeit is easier to change beliefs that exist in isola-tion. Walster, Berscheid, and Barclay (1967)demonstrated that one is more likely to avoiddistortions when future objective evidence maycontradict these distortions than when objec-tive evidence will be unavailable. Both theseobservations are consistent with the expecta-

tion that it is harder to distort the familiarthan the unknown.

There is a second reason why one should bemore reluctant to distort an intimate's relativeoutcomes than those of a stranger. One shouldexpect more difficult}- maintaining an adequatedistortion when the distortion involves an inti-mate than when it involves a stranger. If oneengages in a massive distortion of an intimate'scharacter, he must anticipate that his friendwill have more opportunities (than a strangerwould) to confront him, challenge his rationali-zations, and perhaps retaliate against him.

For two reasons, then, familiarity with thevictim should breed accuracy and discouragejustification as a distress-reducing technique.Data are available to support this hypothesis.Davis and Jones (1960) found that subjectswho ridicule another student, derogate himmore when they do not expect to see him againthan when a meeting is expected. Davis andJones assumed that they secured this resultbecause subjects who anticipate future contactplan to "neutralize their harm-doing" by ex-plaining that their negative evaluation does notrepresent their true feelings. However, since thevictim has already suffered by the timeneutralization occurs, this explanation is nottotally satisfactory. An equally plausible in-terpretation is that distortion becomes moredifficult and less likely to occur when futureinteraction is anticipated.

Ross (1965) conducted an experiment inwhich students were led to choose to consigntheir partner to electric shock to avoid painfulshock themselves. In some conditions, thestudents believed they would work with theirpartner only once. In other cases they believedthey would work with him on many tasks. Rossdiscovered that when the student allowed theother to be injured, derogation occurred moreoften when subsequent contact was not antici-pated. This finding is, of course, satisfactorilyexplained by the hypothesis that we only dis-tort the characteristics of those individualswho can be kept "out of sight and out of mind."

This conclusion suggests that the exploitedmight do well to harrass their exploiters. Forexample, so long as exploited minorities aregeographically and socially segregated, anexploiter can conveniently reduce whatevervague feelings of guilt he might have by


justifying his exploitation. It is easy for him tomaintain that minority members deserve theirexploitation ("The poor are shiftless and lazyand don't want a steady job.") or that theyare not really suffering ("A Chicano can livebetter on a dollar than a white man can onfive dollars."). We might expect that integratedhousing and forced association will make use ofsuch rationalizations more difficult. Until suchintegration occurs, however, minority memberscould arrange "symbolic integration." Welfaremothers who feel that suburban whites rein-force one another's "preposterous rationali-zations," could arrange to expose the taxpayersto reality. They might travel to the suburbs,talk to suburbanites about their plight, con-front shoppers, speak up in suburban ParentsTeachers Associations, etc.

We proposed that two variables determinehow a harm-doer chooses to restore equity tohis relationship with the victim: (a) theadequacy of existing techniques for restoringequity (Corollary IV. 1), and (b) the cost ofthe existing techniques for restoring equity(Corollary 1.2). In the last section we provideddata showing that adequate compensationsand justifications are preferred to inadequateones. In the next section, we present evidencethat the greater the cost of an equity-restoringtechnique, the less likely a harm-doer is to usethat technique to restore equity.

Cost of Equity-Restoring Techniques

In Proposition I, we stated that individualstry to maximize their outcomes. In PropositionII, we stated that groups try to arrange thingsso that individuals can maximize their out-comes by behaving equitably, that is, by insur-ing that their partner receives relative out-comes equal to their own.

On the basis of these two propositions, wecan make the following derivation:

Derivation I: Other things being equal, theharm-doer will use that equity-restoring tech-nique that allows the pair to maintain thehighest possible relative outcomes.

This conclusion is similar to Adams's (1956)assumption that a person "will reduce inequity,insofar as possible, in a manner that will yieldhim the largest outcomes [p. 284]." Adamspresented evidence supporting the validity ofhis hypothesis (Adams, 1963; Adams &

Rosenbaum, 1962). This hypothesis is alsosupported by the results of an experiment con-cerning preferences among forms of equityresolution in fictitious work situations (seeWeick & Nesset, 1969).

Derivation 1 is consistent with the observa-tion that self-punishment seems to be an un-popular way for harm-doers to restore equityto a relationship. Theoretically, both compen-sation to the victim and self-punishment areequally adequate techniques for restoringactual equity to a relationship. In practice,however, one rarely finds harm-doers restoringequity by self-punishment. Derivation I re-minds us of the reason for the aversion to thisequity-restoring technique; harm-doers resistlowering their outcomes unnecessarily.

Anecdotal evidence (Jon Freedman, personalcommunication, 1970) supports the proposi-tion that subjects in laboratory experimentsenergetically resist restoring equity by self-punishment. A very few experiments havedemonstrated that individuals are more willingto perform unpleasant altruistic acts followingcommitment of harmful acts (e.g., Darlington& Macker, 1966; Freedman et al. 1967). How-ever, in these experiments the experimentershave made it difficult or impossible for theharm-doer to compensate the victim. In addi-tion, the guilt-inducing procedures used in theseexperiments are such that justifications (denialof responsibility for the harm, denial that theharm was done, or perception that the act wasjust) were difficult, if not impossible. Thus,evidence of self-punishment following harm-doing has been restricted to situations in whichcompensation and justification techniques arealmost totally unavailable. The frequency withwhich use of self-punishment is found, ifeither compensation or justification is availableand perceived as adequate, is unanswered.

Restoration of Equity by the Victim amiOutside Agencies

The preceding discussion has focused on themeans by which the harm-doer may restoreequity to his relationship with the victim.However, the harm-doer is not the only possibleagent of equity restoration. The victim, thevictim's sympathizers, social workers, thecourts, etc., may all intervene to improve thevictim's lot. What effect does such intervention

162 E. WALSTEK, E. BERSCHEID, AND G. W. WALSTER

have on the harm-doer's perception of theequitableness of his relationship with thevictim? The following conclusion seems mostreasonable:

Conclusion I I : When the victim or an ex-ternal agency restores equity to a relationship,the harm-doer's distress is reduced, and he isless likely to use additional equity-restoringtechniques.

Evidence for this conclusion comes fromdiverse sources:

Retaliation

Victims do not always sit placidly by, wail-ing for the harm-doer to decide how to react.Sometimes victims take matters into their ownhands and restore equity to the relationship.They may restore equity by seizing that portionof reward they deserve. They could providethe exploiter with ready-made excuses for hisbehavior (as some Uncle Toms have beenknown to do), or they can "get even" withthe harm-doer by retaliating against him.How does a harm-doer respond under suchconditions?

The data available support Conclusion II.When the victim restores equity, it eliminatesthe harm-doer's need to do so.

Berscheid, Boye, and Walster (1968) con-ducted an experiment designed to assess theeffect that a victim's retaliation has on theexploiter's tendency to justify the victim'ssuffering by derogating him. The results of thisstudy indicate that a victim can indeed restoreequity through retaliation against the ex-ploiter. In this experiment, individuals werehired to administer severe electric shocks toanother person. If the victim could not retaliateagainst the harm-doer, the harm-doer sub-sequently derogated the victim. However,when the exploiter expected retaliation for hisharmful act, the derogation process wasarrested; the harm-doer did not derogate thevictim.

It is also interesting that the relationshipbetween retaliation and derogation was dia-metrically opposed for control subjects whomerely observed the victim's suffering; thoseobservers who expected to be hurt by thevictim in the future liked him less than didthose who did not expect to be hurt.

The insight that a victim can restore equity

to a victim/harm-doer relationship simply byretaliating has interesting implications: Thevictim naturally prefers that equity be re-stored by receiving compensation. Frequently,however, it becomes obvious to the victim thatcompensation is unlikely to be forthcoming.In such circumstances the victim must realizethat the harm-doer is likely to justify thevictim's suffering. This is not a pleasantprospect. The exploiter's justifications arepotentially dangerous to the victim. Theharm-doer who justifies his actions will endup with a distorted and unreal assessment of hisown actions. If he distorts the extent to whichthe victim deserved to be hurt, for example, orminimizes the victim's suffering as a conse-quence of the act, he may commit further actsbased on these distortions (Berscheid, Boye,& Darley, 1968). When a harm-doer usesjustification technique, then the victim is leftin sad straits. Not only has he been hurt, butas a result of the harm-doer's justification, theprobability has increased that he will be hurtagain. Thus, speculating from the little weknow, one could argue that once it becomesobvious that the harm-doer is not about tocompensate the victim, the victim might wellretaliate against the harm-doer before hejustifies what he has done.

Civil rights leaders have sometimes madesimilar speculations. James Baldwin (1964),in a statement concerning the blacks' strugglefor minority rights in this country, argued that"Neither civilized reason or Christian lovewould cause any of these people to treat you asthey presumably wanted to be treated; onlyfear of your power to retaliate will cause themto do that, or seem to do it, which was (and is)good enough [p. 34J." Black militants havetaken an even stronger position. They haveargued that widespread black violence isnecessary to restore blacks to full citizenship.The variables they discuss sound much likethose we have considered. They talk of the"white devil," his guilt, his denial of racialinjustices, and the equity-establishing effectsof violence. However, if extrapolation fromour findings is relevant, we might suggest thatretaliation, or the anticipation of retaliation,will be beneficial only if the recipient of theviolence feels that he is in some way directlyresponsible for blacks' suffering. Retaliation


against those who feel themselves to beinnocent observers of injustices would seem tobe a disastrous strategy.

Forgiveness

The victim can restore equity to the harm-doer/victim relationship in a second way: Hecan "forgive" the apologetic harm-doer.

Harm-doers often apologize to their victimin the hope that the victim will forgive him.If the victim forgives him, this implies thattheir relationship can proceed again on anequitable basis.

An "apology" is not a single strategy forrestoring equity but comprises several quitedifferent strategies.

1. An apology is often a persuasive commun-ication designed to convince the victim that theharm-doer's justifications are plausible. (Theharm-doer may point out "You hit me first";"I didn't mean it"; or "He made me do it.")If the victim agrees that these justificationsare plausible, the relationship becomes apsychologically equitable one and can proceedas before. (Scott & Lyman, 1968, provide adevastating description of how one goes aboutdevising a compelling excuse.)

2. An apology may be designed to convincethe victim that their relationship is actuallyequitable. For example, a harm-doer may pro-fusely verbalize how much personal sufferingand guilt he has endured as a consequence ofhis unjust treatment of the victim. If hisdescription is heart-rending enough, the victimmay conclude that his score with the harm-doer is settled.

The common television scenario of the care-less driver rushing to the hospital room of hisvictim to express his anguish and remorsebecomes explicable through this reasoning. Aperson not familiar with the common "I'msorry"-"You're forgiven" sequence mightwonder what benefit is conferred upon a dyingvictim or what wrong is righted by such ademonstration of the harm-doer's personalsuffering as a consequence of his act. Theanswer is none. The act is performed not forthe victim's benefit but for the harm-doer's.To say "I'm sorry" is to imply personal suffer-ing and to beg for forgiveness to end one'sdistress.

The notion that a description of one's re-morse and suffering, if convincing, may, in fact,attenuate the wrath and retaliatory intentionsof the victim and others has been supported byan experiment conducted by Bramel, Taub,and Blum (1968).

An effusive apology may make a relationshipactually more equitable in another way. Duringhis apology the harm-doer may humble him-self and exalt the victim. This redistributionof esteem may provide a valuable reward tothe victim and thus even his score with theharm-doer.

When apologies, self-derogation, and exalta-tion of the victim do not elicit forgiveness, it isinteresting that harm-doers often switch withease to another technique—often justificationtechniques.

3. Finally, an apology may be a way inwhich a harm-doer can acknowledge that theparticipants' relationship has been inequitablebut point out that nothing can be done toremedy the preceding injustice. The harm-doermay ask that the victim "forgive and forget"the injustice so that the relationship can beginanew.

Intervention by Outside Agencies

Outside agencies often intervene when rela-tionships become disturbingly inequitable (seeBaker, 1969). Legal and religious agenciessometimes insist that clients make restitutionto their victims or punish them when they donot. Social welfare agencies and insurancecompanies compensate the disadvantaged.What are the effects of such intervention? Un-doubtedly, they depend on whether the inter-vening agency encourages harm-doers to com-pensate, punishes them, preempts their plansto compensate, or simply provides a residualsource of compensation for neglected victims.

Society's first intervention attempts areusually directed toward inducing harm-doersto voluntarily compensate their victims. Thisis a wise policy. Everyone benefits when in-dividuals are motivated to voluntarily com-pensate those they have injured. The harm-doer who voluntarily decides to compensatebecomes a stauncher advocate of the equitynorm (see Mills, 1958). The voluntary com-pensator also serves as a behavioral model forothers; observers should be likely to imitate

164 E. WALSTER, E. BKRSCHEID, AND G. W. WALSTER

his equitable behavior when they find them-selves in a similar situation (see Bandura,1965).

If it becomes evident that an agency is notgoing to be able to induce the harm-doer to be-have equitably, it probably is beneficial for itto escalate and to force him to make restitution.The person who is forced to compensate atleast is dissuaded from justifying his inequit-able behavior and is prevented from servingas a negative model for others.

A state can prod an individual into makingrestitution in a variety of ways. For example,in (he Hungarian and Norwegian legal systems,whether or not harm-doers have made restitu-tion is taken into account when determiningsentences and granting paroles. When aprisoner's freedom is contingent on whetheror not he makes restitution, restitution is ob-viously not really voluntary. However, theHungarians consider it better, from a reha-bilitative point of view, to elicit semivoluntaryrestitution than none at all.

Macaulay and Walster (1971) pointed outthat in the United States both formal and infor-mal techniques are used to induce restitution:

the common-law of torts consists of rules which say awrong-doer must compensate his victim. In addition,the legal system in operation has more avenues torestitution than merely its formal rules. There is awide variety of procedures which may encourage com-pensation . . . . Some criminal sanctions are used asleverage to induce restitution: A police officer maydecide not to arrest a shoplifter if the wrong-doer isnot a professional thief and if the stolen items are re-turned ; a district attorney may decide not to prosecuteif the amounts embezzled are returned [p. 179].

In other systems, restitution is simply ex-tracted from the harm-doer (i.e., money maybe deducted from his prison earnings).

The psychological literature (i.e., Brehm &Cohen, 1962) and the observations of penaltheorists (e.g., Del Vecchio, 1959; Schafer,1960; Spencer, 1874) provide some support forthe contention that if one induces "fair" be-havior, "fair" attitudes will follow.

Sometimes restitution cannot be elicited.(For example, those who injure others areoften unknown or are indigent.) In such cases,it may be wise for the community to reconcileitself to the fact that an injustice has occurredand simply intervene to alleviate the victim'ssuffering. Such intervention is consistent with

our notion of fairness (the innocent victim isrecompensed) and is expedient (the legitimacyof equity norms is affirmed by society).

Some legal theorists (i.e., see Fry, 1956)have proposed that, in the interests of justiceand efficiency, the state should routinelyassume responsibility for compensating victimsof criminal violence. They argue that the statecould save time and money if instead of track-ing down harm-doers and prodding them intomaking restitution, it simply provided auto-matic compensation to the disadvantaged.

We argue that a society should be wary ofintroducing a compensation procedure thaterodes individuals' responsibility for restoringequity, thus weakening their adherence toequity norms. If the harm-doer knows thatthe outside agency will reestablish equity atno cost to himself, he should have little moti-vation to initiate his own equity-restoringresponses. In addition, it is probable that anagency set up to "right all wrongs" wouldsoon find that it was incapable of fulfillingthis mandate. Agencies set up to provide socialjustice are always meagerly funded. Althoughcitizens may be unanimous in their agreementthat social justice is desirable, they seldomagree that society ought to pursue this goal atall costs. Inevitably, agencies are forced to dothe best the}' can with limited funds. Socialwelfare agencies thus soon evolve into agenciesof "social compromise" rather than agenciesof perfect "social justice" (see Macaulay &Walster, 1971, for a lengthy discussion of thisproblem).

For these reasons, public compensation isseen as a residual source of equity restoration,resorted to only when attempts to induce theexploiter to compensate have failed.

Personality

Throughout the previous section we havetaken the nomolhetic approach and haveattempted to provide a framework to guideprediction of how the average individual willrespond following his commission of a harmfulact.

In spite of the fact that equity researchershave focused on the nomothetic approach, weshould not assume that personality differencesare unimportant. It is clear that personalityvariables will affect how participants evaluate


both their own and their partner's inputs andoutcomes, how much distress harm-doers feelafter injuring another, and how the harm-doerand the victim respond to their shared distress.

For example, we have proposed that afterinjuring another, a normal harm-doer experi-ences distress from two sources: (a) He experi-ences self-concept distress, since his harmfulact is inconsistent both with his own goodopinion of himself and with his moral prin-ciples; and (6) he experiences fear of retalia-tion distress. Interest' in individual differencesleads to the obvious conclusion that harm-doerswho have high self-esteem (e.g., harm-doerswho think they are foresightful, kind, non-exploitative people), and who have stronglyinternalized ethical standards will feel moreself-concept distress after harming anotherthan will individuals who have low self-esteemand poorly internalized ethical standards. Ahigh-self-esteem individual, then, will be ex-pected to make a greater effort to restoreequity—either actual or psychological—afterharming another than a low-self-esteemindividual.

There is some experimental evidence to sup-port this derivation. In an experiment by Glass(1964), an individual's self-esteem was experi-mentally raised or lowered by providing himwith "authoritative" information about hisown personality. After harming another, high-self-esteem subjects justified their harm-doing(by derogating their victims) more than didlow-self-esteem subjects.

Profitable and interesting attempts toidentify some of the individual difference vari-ables that are related to or affect a harm-doer'sreaction have been made by Aronfreed (1961),Weinstein, De Vaughan, and Wiley (1969),Lawler and O'Gara (1967), Blumstein andWeinstein (1969), Schwartz (1968), Glass andWood (1969), and Tornow (1970). Such studiesmay be expected to considerably sharpen pre-diction in individual situations.

REACTIONS OF THE EXPLOITED

Distress

If an inequitable relationship is distressingto the exploiter, it is doubly distressing for theexploited. Although an exploiter must endurethe discomfort of knowing he is participating

in an inequitable relationship, he at least hasthe consolation that he is benefiting materiallyfrom his discomfort. The victim has no suchcomfort—he is losing in every way from theinequity. He is deprived of deserved outcomes,he must endure the discomfort of participatingin an inequitable relationship, and he is facedwith the unsettling realization that unless hecan force the harm-doer to provide compensa-tion, he is likely to justify the inequity byderogating him.

Propositions I, I f , and ITT lead to the follow-ing derivation:

Derivation JI: A participant will be more dis-tressed by inequity when he is a victim thanwhen he is a harm-doer.

Several theorists have noticed that thosewho materially benefit from inequity are moretolerant of inequity than are those whomaterially suffer from it (see Adams, 1965;Blumstein & Weinstein, 1969; Homans, 1961;Lawler, 1968).

Researchers have also documented thatthose who materially suffer from inequity arequicker to demand a fair distribution of re-sources than arc those who do not (see, e.g.,Andrews, 1967; Leventhal & Anderson, 1970;Leventhal & Lane, 1970; Leventhal, Weiss, &Long, 1969).

Demands for Compensation

Undoubtedly the victim's first response toexploitation is to seek restitution (see Leven-thal & Bergman, 1969; G. Harwell, D. P.Schmitt, & R. Shotola; 1970"). If the victimsecures compensation, he has restored therelationship to equity, and he has benefitedmaterially. It is easy to see why this is apopular response.

Retaliation

A second way a victim can restore equity isby retaliating against the harm-doer (andthereby reducing the harm-doer's outcomesto the level he deserves). Ross, Thibaut, andEvenbeck (1971) demonstrate that when giventhe opportunity, victims will retaliate against

9 Unpublished study entitled "Cooperation andInterpersonal Risk," 1970. Requests should be sent toG. Marwell, Department of Sociology, University ofWisconsin, Madison, Wisconsin.

166 E. WALSTEE, E. BERSCHEID, AND G. W. WALSTER

those who have treated them inequitably.The more inequitably they perceive they were

treated, the more they will retaliate. Evidencefrom Berscheid, Boye, and Walster (1968) sug-gests that appropriate retaliation will causethe harm-doer (as well as the victim) to per-ceive that the relationship is again an equitableone.

Justification of the Inequity

Sometimes a victim finds that it is impossibleeither to elicit restitution or to retaliate againstthe harm-doer. The impotent victim is thenleft with only two options: He can acknowledgethat he is exploited and that he is too weak todo anything about it, or he can justify hisexploitation. Often, victimized individualsfind it less upsetting to distort reality andjustify their victimization than to acknowl-edge that the world is unjust and that theyare too impotent to elicit fair treatment(Lerner & Matthews, 1967).

Victimized individuals have been found torestore psychological equity in several ways:Victims sometimes console themselves byimagining that their exploitation has broughtcompensating benefits ("suffering brings wis-dom and purity"), or they console themselvesby thinking that in the long run the exploiterwill be punished as he deserves ("The mill ofthe Lord grinds slowly, but it grinds exceed-ingly fine."). Victims may also convince them-selves that their exploiter actually deserves theenormous benefits he receives because hepossesses previously unrecognized inputs. Re-cent data demonstrate that the exploited willjustify the excessive benefits of others. Jeckerand Landy (1969), Walster and Prestholdt(1966), and A. Hastorf and D. Regan (per-sonal communication, February 1962) pres-sured individuals into performing a difficultfavor for an unworthy recipient. They foundthat the abashed favor-doer tries to justifythe inequity by convincing himself that therecipient is especially needy or worthy.

Reformers who have worked to alleviatesocial injustices, at great personal sacrifice, areoften enraged to discover that the exploitedthemselves sometimes vehemently defend thestatus quo. Black militants encounter "UncleToms," who defend white supremacy. Women'sliberation groups must face angry housewives

who threaten to defend to the death the cur-rent status of women. Reformers might havemore sympathy for such Uncle Toms and"Doris Days" if they understood the psycho-logical underpinnings of such reactions. Whenone is treated inequitably but has no hope ofaltering his situation, it is often less degradingto deny reality than to face up to one's humil-iating position.

Equity Theory and Helping Relationships

In spite of the fact that social observers re-proach individuals for not helping others asmuch as they "ought" to help, the fact remainsthat individuals do help one another to a re-markable extent. Public assistance agencieshelp welfare recipients, parents care forchildren and elderly parents, Boy Scouts helplittle old ladies across the street, Congress aidsunderdeveloped nations, and eager suitorsurge gifts on overdeveloped maidens.

On the surface, relationships between bene-factors and recipients seem quite different fromrelationships between victims and exploiters.Conceptually, however, both relationshipscan be analyzed in the same terms. Let usassume that initially participants are in anequitable relationship. Then the benefactorhelps the recipient. Now the previously equi-table relationship becomes an inequitable one.The benefactor, like the victim, is now a partic-ipant in an inequitable—and unprofitable—relationship. The recipient, like the exploiter,is now a participant in an inequitable—andprofitable—relationship. Equity theory leadsus to expect that benefactor and recipient, likeharm-doer and victim, should experience vaguediscomfort when they discover they are par-ticipating in an inequitable relationship. AsProposition IV indicates, they should alleviatetheir distress by restoring either actualequity or psychological equity to theirrelationship.

Research indicates that those in philan-thropic relations do respond much as do thosein harm-doing relations (see, e.g., Leventhalet al., 1969).

Much of the research investigating the im-pact of helpful acts on the gift giver and hisrecipient has been conducted by "actionpsychologists." The research of these psychol-ogists was generally not designed to test


theoretical propositions but to enable practi-tioners to devise better welfare programs, con-struct better foreign aid programs, etc. As aconsequence of their practical orientation, help-ing researchers focused on very different vari-ables than have harm-doing researchers. Thisdifference in focus has produced unusual theo-retical benefits. Helping researchers enable usto document the importance of variables whichharm-doing theorists have contended weretheoretically important but which they hadtotally neglected to investigate empirically.Let us consider two variables which helpingtheorists have found to be important deter-minants of participants' responses: inten-tionality and ability to repay.

1' ntentionality

Proposition III stated that when individualsfind themselves participating in inequitablerelationships, they will become distressed. Itseems plausible to argue that an individualwho feels responsible for creating an inequityshould feel more distressed than an individualwho inadvertently finds himself in an in-equitable relationship. The intentional harm-doer should experience both self-concept dis-tress and fear of retaliation distress. At most,the inadvertent harm-doer should experienceonly retaliation distress.

A few harm-doing theorists observed that theperson who deliberately provokes an inequityshould become more distressed than the personwho accidentally behaves inequitably (see,e.g., Davis & Jones, 1960; Glass, 1964). Un-fortunately, harm-doing researchers can call uplittle evidence to document this contention.Few studies have investigated the reactions ofvoluntary harm-doers. Instead of focusing onthe reactions of voluntary harm-doers, re-searchers have focused on the reactions ofindividuals who were prodded into injuringothers. The reason for this is simple. Labora-tory researchers found it almost impossible toinduce subjects to voluntarily harm others.Even intensely provoked subjects would re-fuse to treat another inequitably. To inducesubjects to harm another, the experimentershad to practically force them to behave in-equitably. By necessity, then, rather than bydesign, harm-doing researchers were confined

to studying the reactions of inadvertent andreluctant harm-doers.

Helping researchers, on the other hand, havenot encountered such problems. Their researchfinally enables us to document the importanceof intentionality in determining how a bene-factor and a recipient respond to inequity,The data they provide lead to the followingconclusion:

Conclusion III: When the inequity is inten-tionally produced, participants in an inequi-table relationship will experience more distressand will have stronger desires to restore equityto the relationship than if the inequity occursinadvertently.

Some support for Conclusion III comes fromThibaut and Riecken (1955), Goldner (1960),Goranson and Berkowitz (1966), Greenberg(1968), and Greenberg and Frisch (in press).Their research demonstrates that when one isintentionally helped, he has a much strongerdesire to reciprocate (and thus to restoreequity to his relationship with the benefactor)than when he is helped accidentally.

A typical experiment was conducted byGreenberg and Frisch (in press) in which theyrecruited subjects to participate in an experi-ment ostensibly designed "to identify per-sonality characteristics associated with successin the business world." Subjects were promisedextra credit if they were successful on a task.During the course of the experiment, the sub-jects discovered that they had little chance ofsucceeding unless they received help from theirpartners. (Their partners possessed graphcards which the subjects needed to success-fully complete their tasks.) In all cases thepartner helped the subject: In some cases thishelp was intentional; in other cases it was not.In the high-intentionality condition, thepartner sent the subject the graph cards heneeded along with a note: "I have someduplicates that probably belong to you. I'msending them over since you can probably usethem." In the low-intentionality condition, thepartner sent the needed graph cards, but theywere accompanied by a note which made itclear that the partner did not realize that hewas helping the subject. The note said: "Someof my cards don't have the month on them.Can you help me and identify them for me?"Before the subjects could reply, however, the


inadvertent benefactor sent him another notesaying: "Forget it. I found the cards T wasmissing."

The subject's eagerness to restore equity tohis relationship with the partner (by repayingthe help he had received) was then assessed.As predicted, subjects were more eager to payback the intentional helper than the inadver-tent helper.

Ability to Repay

A second variable has an important impacton how an altruistic act affects the benefactorand his recipient, that is, the beneficiary'sability to make restitution. Earlier (see Corol-laries IV.l and 1.2) we argued that the ade-quacy and the cost of available equity-restor-ing techniques would determine how indivi-duals would restore equity. The same rationaleis applicable here: If a recipient has no abilityto repay his benefactor or has the ability tomake only a most inadequate compensation, heis unlikely to try to restore equity via com-pensation techniques. He is more likely tojustify his windfall instead. On these grounds,we would expect ability to make restitution tobe a theoretically important variable.

A variety of researchers have testified thatability to repay is a potent determinant ofhow helping affects the benefactor-recipientrelationship. Those researchers who have in-vestigated the reactions to help on the part ofwelfare clients, underdeveloped nations, andthe physically handicapped have dealt withrecipients who know they will never be ableto repay their benefactors. Researchers whohave investigated the reactions to help ofreceivers of holiday gifts, members of the kularing (a complex institution of internationalceremonial exchange), and the kindness ofneighbors have dealt with donors and recipientswho know that each helpful act will be recipro-cated in kind. The differing reactions of partici-pants in reciprocal versus nonreciprocal rela-tions underscore the importance of the re-cipient's "ability to repay" in determining howhelp affects a relationship. Ability to repayseems to determine whether favor-doinggenerates pleasant social interactions or dis-comfort and reationalization.

Research supports Conclusion IV: Unde-served gifts produce inequity in a relationship.

If the participants know the recipient can andwill reciprocate, the inequity is viewed astemporary, and thus it produces little distressand lit t le need to justify the inequity. If theparticipants know the recipient cannot or willnot reciprocate, however, a real inequity isproduced; the participants will experiencedistress and will therefore need to restore actualor psychological equity to the relationship.

Evidence in support of Conclusion IV comesfrom three diverse sources:

1. On the basis of ethnographic data, Mauss(1954) concluded that three types of obliga-tions are widely distributed in human societiesin both time and space: (a) the obligation togive, (b) the obligation to receive, and (c) theobligation to repay. Mauss (1954) and Dillon(1968) agreed that when individuals are pre-vented from discharging their obligations,mutual distress is the result. They noted thatwhile reciprocal exchanges breed cooperationand good feelings, gifts that cannot be recipro-cated breed discomfort, distress, and dislike.The authors observed that some societies haveworked out exchange systems in which every-one can be both a donor and a receiver. (Thekula ring is an example.) Harmonious stablerelations are said to be the result. They con-trasted these societies with those in which nomechanisms for getting rid of obligations byreturning gifts is provided. For example,Dillon (1968) noted:

Instead of the kula principle operating in the Mar-shall Han, the aid effort unwittingly took on some ofthe characteristics of the poilalch ceremony of the 19thCentury among North Pacific Coast Indians in whichproperty was destroyed in rivalry, and the poorhumiliated [p. 15].

Volatile and unpleasant relations are said to bethe result of such continuing inequities. Theseauthors, along with Blau (1955) and Smith(1892), agreed that the ability to reciprocate isan important determinant of how nations willrespond to help from their neighbors.

2. There is evidence that individuals aremore likely to accept gifts that can be recipro-cated than gifts that cannot.

Greenberg (1968), Berkowitz (1968), andBerkowitz and Friedman (1967) demonstratedthat people are reluctant to ask for help ifthey cannot repay it.


Greenberg (1968) told subjects that theywould be participating in a study of the effectsof physical disability on work performance.Subjects were told that on the first task theywould have restricted ability to use theirarms. It was obvious to them that this restric-tion would make it difficult for them to per-form the task they had been assigned. If theincapacitated subject wished, however, heknew he could solicit help from another sub-ject on this task. Half of the subjects believedthat the fellow subject would need their helpon a second task and that they would be ableto provide assistance. The remaining subjectsbelieved that the fellow subject would not needtheir help and that, in any case, they would beunable to provide much help. The subjects'expectations about whether or not they couldreciprocate any help provided to them stronglyeffected their willingness to request help. Sub-jects in the no-reciprocily condition waitedsignificantly longer before requesting helpthan did those in the reciprocity condition.Greenberg and Shapiro (1971) replicated thesefindings.

There are three reasons why individuals maybe reluctant to accept help when they are un-able to reciprocate in kind: (a) Individualsprobably avoid accepting undeserved benefitsbecause such benefits place them in an in-equitable relationship with the benefactor. Aswe indicated in Proposition III, inequitablerelationships are unpleasant relationships, andindividuals avoid unpleasantness, (ft) In-dividuals ma)' avoid accepting help which theycannot repay in kind, because to accept suchhelp means one is obligated for an indefiniteperiod to repay the benefactor in unspecifiedways. The recipient might reasonably beworried that his benefactor may attempt toextract greater repayment than the recipientwould have been willing to give, had the con-ditions of the exchange been known ahead oftime. Democritis (in the Fourth Century B.C.)said: "Accept favors in the foreknowledge thatyou will have to give a greater return forthem." The recipient may be unwilling toextend such unlimited blanket credit to hisbenefactor. (See Blau, 1967, for a discussionof this point. ) (c) Or, the recipient may havemore specific fears. He may worry that thebenefactor wil l demand excessive gratitude or

constant acknowledgement of his social ormoral superiority from the recipient.

Homans (1961) observed that "anyone whoaccepts from another a service he cannot repayin kind incurs inferiority as a cost of receivingthe service. The esteem he gives the other heforegoes himself [p. 320]." The recipient maybe unwilling to risk being assigned so meniala status as a consequence of accepting help.(See Blau, 1967, for further discussion of thispoint.)

3. Research also demonstrates that gifts thatcan be reciprocated are preferred to gifts thatcannot be repaid. Gergen (1968) questionedcitizens in countries that had received UnitedStates aid as to how they felt about the assis-tance their country received. Gergen notedthat international gifts, when they are accom-panied by clearly stated obligations, are pre-ferred either to gifts that are not accompaniedby obligations or gifts that are accompaniedby excessive "strings." Presumably, gifts thatcan be exactly reciprocated (by fulfilling clearlystated obligations) are preferred to gifts thatcannot be reciprocated or to gifts which re-quire excessive reciprocation.

In laboratory research, Gergen (1968) foundadditional support for the conclusion thatindividuals like a benefactor more when theyknow they can reciprocate his help than whenthey know they cannot return his generosity.

Gergen investigated the reactions of malecollege-age students in the United States,Japan, and Sweden to inequitable situations.Suudents were recruited to participate in anexperiment on group competition. Thingswere arranged so that during the course of thegame, the subject discovered that he waslosing badly. At a critical stage (when thestudent was just about to be eliminated fromthe game), one of the "luckier" players in thegame sent him an envelope. The envelope con-tained a supply of chips and a note. For one-third of the subjects (low-obligation-conditionsubjects), the note explained that the chipswere theirs to keep, that the giver did not needthem, and that they need not be returned.One-third of the subjects (equal-obligation-condition subjects) received a similar note,except that the giver of the chips asked thesubject to return an equal number of chipslater in the proceedings. The remaining sub-


jects (high-obligation-condition subjects) re-ceived a note from the giver in which he askedfor the chips to be returned with interest andfor the subject to help him out later in thegame.

At the end of the game, subjects werequeried about their attraction toward variouspartners. The results support Conclusion IV:Those partners who provided benefits withoutostensible obligation or who asked for excessivebenefits were both judged to be less attractivethan were partners who proposed that thestudent make exact restitution later in thegame.

J. Gergen, P. Diebald, and M. Seipel10 con-ducted a variation of the preceding study.Just as subjects were about to be eliminatedfrom a game because of their consistent losses,another "player" in the game loaned the sub-ject some resources. The donor loaned thechips with the expectation that they would bepaid back. However, in subsequent play, onlyhalf of the subjects managed to retain theirchips. Thus, half of the subjects were unableto return the gift; half were able to do so. Insubsequent evaluations of the donor, recipientsthat were unable to repay the donor evaluatedhim less positively than did recipients thatwere able to repay. These results were repli-cated in both Sweden and the United States.These results are consistent with Tacitus'sobservation that "Benefits are only acceptableso far as they seem capable of being requited:Beyond that point, they excite hatred insteadof gratitude."

Equity Theory and Intimate Relations

When equity theorists argue that businessor neighborly relationships will endure only solong as they are profitable to both participants,few demur. Yet, when one argues that intimaterelations—relations between husband and wife,parent and child, or best friends—might besimilarly dependent on the exchange of re-wards, objections are quickly voiced. Peopleinsist their intimate relations are "special"relations—relations untainted by crass con-siderations of social exchange.

For example, Liebow (1967) reported thesentiments of Tally, a black "streetcornerman":

The pursuit of security and self-esteem push him toromanticize his perception of his friends and friend-ships. .. . He prefers to see the movement of money,goods, services and emotional support between friendsas flowing freely out of loyalty and generosity andaccording to need rather than a mutual exchange rest-ing securely on a quid pro quo basis. . .[p. 34].

Yet, in spite of Tally's insistence that hisrelationship with Wee Tom transcended selfishconsiderations, an outsider would be skepticalof this contention. Liebow, for example,pointed out that Tally and Wee Tom's rela-tionship quickly disintegrated when mutualreinforcement faltered. When Tally won $135on numbers and refused to lend his "walkingbuddy" any more than $5, their friendshipbegan to disintegrate (Liebow, 1967, pp. 176-177).

Anticipating inevitable opposition, then, westill contend that even in the most intimaterelations, considerations of equity will influencestrongly the viability and pleasantness of arelationship.

A variety of equity theorists have voicedsimilar conclusions: Blau (1967) argued it isinevitable that people generally end up pairedwith those partners they "deserve." He pointedout that if one wants to reap the benefits ofassociating with another, he must offer hispartner enough to make it worthwhile for himto stay in the relationship. The more thepartner has to offer, the more demand therewill be for the partner's company, and themore one will have to offer before he can hopeto win the other's friendship. Thus, marketprinciples insure each person will get as de-sirable a friend as he "deserves."

On the basis of such reasoning, Backmanand Secord (1966), Homans (1961), and Blau(1968) proposed a "matching hypothesis"—they predicted that the more equitable a rela-tionship is, the more viable it will be.11

Backman and Secord (1966) argued that(in groups) "The final structure that emerges isalways a compromise. The group structuremoves toward an equilibrium in which eachperson's position in the affect structure is thebest he can obtain in terms of his reward-cost

10 Study in preparation entitled "Intentionality andAbility to Reciprocate as Determinants of Reactionsto Aid."

11 In previous sections, we discussed established re-lationships which are disturbed by an inequitable act.We then examined various techniques by which partici-pants can restore equity to their relationships. Now weare suggesting that equity considerations may have astrong impact on whether or not a group will even formand whether or not it will disintegrate once formed.


outcomes [p. 190]." Thus, they proposed thatpartners of similar value tend to pair up.

Homans (1961) argued that people chooseintimates who are about equal to them instatus.

Theorists have even suggested that equityconsiderations affect one's most intimatechoices—one's choice of a romantic andmarriage partner.

A number of experimenters have investi-gated this proposition. Their data lead to thefollowing conclusion:

Conclusion V: (a) Individuals' romanticchoices are influenced by equity considerations.They tend to choose and prefer partners ofapproximately their own "social worth." (6)There is a constant upward bias in one'schoices. Individuals persist in trying to formrelations with partners who are somewhatmore desirable than themselves.

One's romantic choices thus seem to be adelicate compromise between the realizationthat one must accept what he deserves andthe insistent demand for an ideal partner.

Evidence for Conclusion V comes fromseveral sources: Walster, Aronson, Abrahams,and Rottman (1966) found evidence thatdating preferences are sometimes influencedby equity considerations and are sometimesdetermined by the unlimited aspirations ofparticipants.

The authors had predicted that equity con-siderations would affect all dating choices.They initially proposed two hypotheses: (a)The more "socially desirable" an individualis (i.e., the more physically attractive, person-able, famous, or rich, etc., he is), the moresocially desirable he or she will expect a"suitable" romantic partner to be; (6) coupleswho are similar in social desirability will moreoften continue to date one another and willbetter like one another than will couples whoare markedly mismatched.

Figure 1 depicts graphically the predictionthat participants will prefer dates of approxi-mately their own attractiveness.

The authors' hypotheses were tested in afield study. College freshman were invited toattend a dance. They were told that theirpartner would be assigned by computer.

Physical attractiveness was chosen as theindicant of participants' social desirability.(Data indicate that physical attractiveness isstrongly correlated with popularity, self-

i

EXTREMELY EXTREMELYLOW HIGH

DATES PHYSICAL ATTRACTIVENESS

I''IG. 1. Amount of liking predicted for dates ofvarious attractiveness by ugly, average, and attractivesubjects.

esteem, and other indexes which comprise"social desirability.") The freshman's physicalattractiveness was then evaluated by fourstudents while he or she was purchasing aticket. Whether or not students expected andpreferred partners of approximately their ownsocial desirability was assessed in the followingways.

First, when freshmen signed up for thedance, they were asked how socially desirablethey expected their date to be. (They wereasked how physically attractive, how person-ally attractive, and how considerate they ex-pected their date to be.) Equity theory pre-dicts that the more attractive the freshman,the more desirable his date should be expectedto be. This prediction was confirmed.

Second, freshmen were randomly assigned todates, whom they met for the first time at thedance. Equity theory predicts that the moresimilar the dates are in attractiveness, the moreviable their relationship will be. The validityof the relationship was assessed in three ways:First, during intermission, students were askedhow much they liked their partner and,second, how eager they were to continue thedating relationship. Third, whether or notcouples actually continued to date was deter-mined by interviewing all participants 6months after the dance.

Once partners had met one another, equitytheory predictions were not supported. Every-one, regardless of his or her own social desir-ability, best liked and most often attemptedto continue to date the most desirable datesavailable. Equity considerations seemed not


to limit the participants' aspirations in anyway.

To make things even worse for the theory,these findings were replicated by Brislin andLewis (1968). Walster (1970) secured addi-tional support which further weakens the con-clusion that equity considerations influenceromantic preferences. In accord with equitytheory, she predicted that when an individual'sself-esteem is lowered, he lowers his romanticaspirations. When his self-esteem is raised, heraises his romantic aspirations. Two studiesfailed to find any support for this contention.

Kiesler and Baral (1970) did find supportfor the equity theory predictions. The authorsrecruited male college students for a study onintelligence tests. The experimenter told themen that he was perfecting a new intelligencetest that had already been successfully used onhundreds of students. Men were then given adifficult test. Men in the high-self-esteem con-dition were led to believe that they were doingextremely well on the test. (The experimenternodded and smiled at their answers and men-tioned that other men had much more troublewith the questions.) Men in the low-self-esteem condition were led to believe that theywere doing badly on the test. (The experi-menter made it apparent that he was displeasedwith their performance. He frowned, lookedaway, and mentioned that other subjects hadperformed better.)

During a break in testing, the experimenterand the subject visited a nearby canteen. Whenthey entered the canteen, the experimenterrecognized a girl (actually an experimentalconfederate). In one condition (the attractivecondition), the confederate was made-up to bevery physically attractive. She wore becomingmake-up and fashionable clothing. ]n the un-attractive condition, she was far less attractive.She wore no make-up, heavy glasses, and hadher hair pulled back with a rubber band. Herskirt and blouse clashed and were arrangedsloppily.

The girl sat down and chatted with the ex-perimenter and the subject. After a minute,the experimenter excused himself to make aphone call. While he was gone, the confederatecontinued to engage the subject in conversationfor | hour. She acted in a friendly and interest-ing way toward all subjects.

The dependent variable was the extent towhich the male indicated to the female con-

federate that he was romantically interested inher and the extent to which he attempted toprolong their relationship. The confederatekept track of whether the man asked her for adate, asked for her phone number, offered tobuy her a snack or coffee, offered her a ciga-rette, complimented her, or, finally, ignoredher when at the end of the prescribed time shesaid that she should get back to work.

Kiesler and Baral found strong support forthe matching hypothesis. When the man'sself-esteem had been lowered, he behaved mostromantically with the moderately attractiveconfederate. When the man's self-esteem hadbeen raised, he behaved in a far more romanticway with the attractive confederate than withthe unattractive one.

Other support for the matching hypothesiscomes from Berscheid, Dion, Walster, andWalster (1971).

At the present time, then, data do not con-sistently support either equity theory or thenotion that individuals' intimate social choicesare unchecked by reality. Thus, Conclusion Vcan merely state that individuals' socialchoices are a compromise between two con-flicting pressures.

The conclusion that one's social choices area compromise between fantasy and realityseems to be consistent with our own observa-tions in the daily world.

Sometimes individuals talk and act as if theyhave unlimited social inputs and thus are de-serving of perfection. They talk as if the factthat they had to compromise in selecting amarriage partner is an inequity. For example,we can all think of prestigious but aging pro-fessors who leave their wives and marrybeautiful, young, graduate students. Often,within a short time, our professor may beginto lament his protege's shortcomings. "If onlyshe were more intelligent and more consid-erate," he complains. Observers sometimessmile, because they are more atuned to theoperation of exchange processes in determiningsocial pairings than is the participant in therelationship. They are smugly aware that ifhis lady were smarter, she would not have tosettle for the company of the aging professor.

Sometimes, however, individuals are awareof equity considerations. The man with theundeservedly beautiful wife often manifestsvague uneasiness—whether the uneasiness isgenerated by his own recognition that he has


married a better woman than he deserves, orwhether his uneasiness is generated by thefact that she constantly reminds him that hehas married too well, we do not know.

Intimate relations, then, seem to be in-fluenced in part by equity considerations andin part by fantasy.

EQUITY THEORY AND SOCIAL PSYCHOLOGY

Social psychological theory is generallyacknowledged to be in a chaotic state. A num-ber of mini-theories exist. Rigorous researchsupports all of them, at least some of the time.Yet, the relationships between these theoriesare vague. Thus, when we wish to predicthow people in a given situation will respond,we often discover that several mini-theoriescan make predictions—and often they makedifferent predictions. We are often left in theembarrassing state of being able to predictanything, anytime.

Obviously, what is needed is a general com-prehensive theory and rules of transformationwhich tell us how the various mini-theories fitinto the general framework. In equity theorywe have made an effort to synthesize varioustheoretical approaches rather than follow themore usual (and more entertaining) procedureof setting up crucial confrontations betweenthe mini-theories.

For example, equity theory proposes thatindividuals enmeshed in inequitable relation-ships feel distress. When discussing the genesisof this distress, we tried to synthesize the in-sights of learning theory, cognitive consistencytheory, and Freudian theory. In the future, wemust try to continue this process of synthesis.We must try to formally relate existing socialpsychological theory to equity theory. (Twoareas of research which could be profitablyrelated to equity theory may have alreadyoccurred to the reader.)

On Establishing the Relationship

Equity theory describes how individualsenmeshed in inequitable relationships respond.To predict how an individual will respond, onemerely has to ascertain whether the scrutinizerperceives participants to be in a relationshipand how he calculates the participants' rela-tive outcomes.

The theory has no need to know why thescrutineer perceives individuals to be in a rela-

tionship; it is enough to know that he does.Yet, a mini-theory, telling us when individualswill perceive participants to be in a relationshipand when they will not, could be a useful addi-tion to the theory.

A body of literature exists which tells uswhen individuals perceive themselves in a rela-tionship with, and compare their outcomeswith, others. In the future, it would be profit-able to attempt to formally relate this mini-theory to equity theory.

Theorists such as Stouffer, Suckman, De-vinney, Star, and Williams (1949), Festinger(1954), Merton (1957), Romans (1961), Gurr(1970), and Latane (1966) provide us withsome insights as to when people compare in-dividual's (or group's) outcomes and when theydo not.

Homans (1961) said that when "Person" istrying to decide whether or not distributivejustice prevails, he goes through the followingmental procedures:

Am I getting as much as other men in some respectlike me would get in circumstances in some respect likemine? And is Other giving me as much as other men, insome respect like Other, would give? When it comes,moreover, to comparisons with other men, the mostimportant other man is the particular one with whomexchange is now taking place [p. 76],

Festinger (1954) in discussing the questionof whom a person chooses to compare his ownopinions with said:

Given a range of possible persons for comparison,someone close to one's own ability or opinion will tiechosen for comparison [p. 121].

The more attractive a group is to a member, the moreimportant that group will be as a comparison group forhim [p. 131].

Merton (1957) argued that soldiers comparethemselves with [a] "Others with whom theywere in actual association, in sustained socialrelations. . . ." [6] "Men who are in somepertinent respect of the same status or in thesame social category." [_c] "Men who are insome pertinent respect of different status or ina different social category. . . ."

Merton argued that

some similarity in status attributes between the in-dividual and the reference group must be perceived orimagined, in order for the comparison to occur at all.Once this minimal similarity is obtained, other similari-ties and differences pertinent to the situation will pro-vide the context for shaping evaluations [p. 242].

He also suggested that people compare them-selves with those they admire.


Finally, Merton argued that structural vari-ables determine with whom one compareshimself:

If the structure of a rigid system of stratification, forexample, is generally defined as legitimate, if the rights,prerequisites and obligations of each stratum are gen-erally held to be morally right, then the individualswithin each stratum will be the less likely to take thesituation of the other strata as a context for appraisalof their own lot. They will, presumably, tend to con-fine their comparisons to other members of their ownor neighboring social stratum. If, however, the systemof stratification is under wide dispute, then members ofsome strata are more likely to contrast their ownsituation with that of others, and shape their self-appraisals accordingly. .. the range of groups taken aseffective bases of comparison in different social systemsmay well turn out to be closely connected with thedegree to which legitimacy is ascribed to the prevailingsocial structure [p. 267].

In brief, these writers suggest that compari-sons take place with those who are (a) in anactual physical relationship with Person at thetime, (b) perceived as similar to Person alongsalient dimension, or (c) attractive, looked upto, or admired in some way. There is the furthersuggestion that the less structured the situa-tion, the broader the spectrum of individualswith whom it is possible to make a comparison.

In the future, it might be valuable to restatethese insights into a prepositional form whichis consistent with equity theory. This set ofpropositions could then be used to tell us whenthe scrutineer is likely to see participants asbeing "in a relationship."

On Evaluating Rewards and Costs

A second important element in equity theoryis the scrutineer's perceptions of the rewardsand costs that participants are securing as aconsequence of interacting with one another.In experiments one can manipulate thescrutinizer's perception of the rewards andcosts participants are incurring, and thendetermine the effect these manipulations haveon his behavior. In field studies, one can mea-sure participants' perceptions of the outcomesthey are receiving as a consequence of theirrelationship with another. Such procedureswork well when one is testing a theory. Butwhat about applying the theory? How doesone calculate inputs and outcomes then?

Once again, an impressive body of researchalready exists to guide us in calculating howcostly or rewarding various stimuli will beperceived to be (see Homans, 1961; Thibaut &

Kelley, 1959). This formulation could bevaluably incorporated into equity theory.

Given the relevance of many social psycho-logical findings to equity theory, the next stepin equity research should be to integrate thesesocial psychological findings into equity theoryin a formal way, in the hope of further increas-ing the breadth of the equity theorypredictions.

REFERENCES

ABELSON, R. P., & ROSENBERG, M. J. Symbolic psycho-logic: A model of attitudinal cognition. BehavioralScience, 1958, 3, 1-13.

ADAMS, J. S. Toward an understanding of inequity.Journal of Abnormal and Social Psychology, 1963,67, 422-436.

ADAMS, J. S. Inequity in social exchange. In L. Berko-witz (Ed.), Advances in experimental social psychology.Vol. 2. New York: Academic Press, 1965.

ADAMS, J. S., & ROSENISADM, W. E. The relationshipof worker productivity to cognitive dissonance aboutwage inequity. Journal of Applied Psychology, 1962,46, 161-164."

ANDREWS, I. R. Wage inequity and job performance:An experimental study. Journal of Applied Psychol-ogy, 1967, 51, 39-45.

ARNOLD, M. B. Kmolion and personality. Vol. 2. Psycho-logical aspects. New York: Columbia UniversityPress, 1960.

ARONFREED, J. The nature, variety and social pattern-ing of moral responses to transgression. Journal ofAbnormal and Social Psychology, 1961, 63, 223-240.

ARONFREED, J. The origin of self-criticism. Psycho-logical Renew, 1964, 71, 193-218.

BACKMAN, C. W., & SECORD, R. F. The compromiseprocess and the affect structure of groups. In C. W.Backman & P. I1'. Secord (Eds.), Problems in socialpsychology. New York: McGraw-Hill, 1966.

BAKER, K. A. An experimental analysis of third-partyjustice behavior. Unpublished doctoral dissertation,University of Wisconsin, 1969.

BALDWIN, J. The fire next time. New York: Dial, 1964.BANDURA, A. Vicarious processes: A case of no-trial

learning. In L. Berkowitz (Ed.), Advances in experi-mental social psychology, Vol. 2. New York: AcademicPress, 1965.

BERKOWITZ, L. Aggression: A social psychologicalanalysis. New York: McGraw-Hill, 1962.

BERKOWITZ, I,. The frustration-agression hypothesisrevisited. Paper presented at the meeting of theWestern Psychological Association, San Diego,March 1968."

BERKOWJTZ, L., & FRIEDMAN, P. Some social classdifferences in helping behavior. Journal of Personalityand Social Psychology, 1967, 5, 217-225.

BERSCHEID, E., BOYK, D., & DARLEY, J. M. Effects offorced association upon voluntary choice to associate.Journal of Personality and Social Psychology, 1968,8, 13-19.

BERSCHEFD, E., BOYE, D., & WALSTER, E. Retaliationas a means of restoring equity. Journal of Personalityand Social Psychology, 1968, 10, 370-376.


BERSCHEID, E., DION, K., WALSTER, E., 8: WALSTER,G. W. Physical attractiveness and dating choice:A test of the matching hypothesis. Journal of Experi-mental Social Psychology, 1971, 7, 173-189.

BERSCHEID, E., & WALSTER, E. When does a harm-doercompensate a victim ? Journal of Personality and SocialPsychology, 1967, 6, 435-441.

BERSCHEID, E., WALSTER, E., & BARCLAY, A. Effectof time on tendency to compensate a victim. Psycho-logical Reports, 1969, 25, 431-436.

BLAU, P. M, The dynamics of bureaucracy: A study ofinterpersonal relations in two government agencies.(Rev. ed.) Chicago: University of Chicago Press,1955.

BLAU, P. M. Exchange and power in social life. NewYork: Wiley, 1967.

BLAU, P. M. Social exchange. In D. L. Sills (Ed.),International encyclopedia of the social sciences. Vol. 7.New York: Macmillan, 1968.

BLUMSTEIN, P. W., & WEINSTEIN, E. The redress ofdistributive injustice. American Journal of Sociology,1969, 74, 408-418.

BRAMEL, D. Interpersonal attraction, hostility, andperception. In J. Mills (Ed.), Experimental socialpsychology. Toronto: Macmillan, 1969.

BRAMEL, D., TAUH, B., & BLUM, B. An observer's re-action to the suffering of his enemy. Journal ofPersonality and Social Psychology, 1968, 8, 384-392.

BREHM, J. W., & COHEN, A. R. Explorations in cognitivedissonance. New York: Wiley, 1962.

BRISLIN, R. W., & LEWIS, S. A. Dating and physicalattractiveness: A replication. Psychological Reports,1968, 22, 976.

BROCK, T. C., & BECKER, L. A. Debriefing and sus-ceptability to subsequent experimental manipula-tions. Journal of Experimental Social Psychology,1966, 2, 314-323.

BROCK, T. C., & Buss, A. H. Dissonance, aggression,and evaluation of pain. Journal of Abnormal andSocial Psychology, 1962, 65, 192-202.

BROCK, T. C., & Buss, A. H. Effects of justification foraggression in communication with the victim on post-aggression dissonance. Journal of Abnormal and SocialPsychology, 1964, 68, 403-412.

CARLSMITH, J. M., & GROSS, A. E. Some effects ofguilt on compliance. Journal of Personality and SocialPsychology, 1969, 11, 232-239.

CARTVVRIGHT, D., & HARARY, F. Structural balance.A generalization of Heider's theory. PsychologicalReview, 1956, 63, 277-293.

DARLINGTON, R. B., & MACKER, C. E. Displacement ofguilt-produced altruistic behavior. Journal of Per-sonality and Social Psychology, 1966, 4, 442-443.

DAVIDSON, J. Cognitive familiarity and dissonance re-duction. In L. Festinger (Ed.), Conflict, decision, anddissonance. Stanford: Stanford University Press,1964.

DAVIS, K. E., & JONES, E. E. Changes in interpersonalperception as a-means of reducing cognitive disson-ance. Journal of Abnormal and Social Psychology,1960, 61, 402-410.

DE JONG, J. P. B. DE JOSSELIN. Levi-Strauss's theoryon kinship and marriage. Leiden, Holland: Brill, 1952.

DEL VECCHIO, G. The problem of penal justice (im-prisonment or reparation of damage). Remsla

Juridica de la Universidad de Puerto Rico, 1959, 27(Trans, by Silving).

DILLON, W. S. Gifts and nations. The Hague: Mouton,1968.

FESTINGER, L. A theory of social comparison processes.Human Relations, 1954, 7, 117-140.

FESTINGER, L. A theory of cognitive dissonance. Evan-ston, 111.: Row, Peterson, 1957.

FREEDMAN, J. L., WALLINGTON, S. A., & BLESS, E.Compliance without pressure: The effect of guilt.Journal of Personality and Social Psychology, 1967, 7,117-124.

FROMM, E. The art of loving. New York: Harper & Row,1956.

FRY, M. Justice for victims. Journal of Public Law,1956, 8, 155-253.

GERGEN, J. Altruism from the recipient's viewpoint:Individual and international perspectives. Swarth-more College: Author, 1968. (Mimeo)

GLASS, D. C. Changes in liking as a means of reducingcognitive discrepancies between self-esteem andaggression. Journal of Personality, 1964, 32, 531-549.

GLASS, D. C., & WOOD, J. D. The control of aggressionby self-esteem and dissonance. In P. G. Zimbardo(Ed.), The cognitive control of motivation. Glenview,111.: Scott Foresman, 1969.

GOPFMAN, E. On cooling the mark out: Some aspectsof adaptation to failure. Psychiatry, 1952,15,451-463.

GOLDNER, A. W. The norm of reciprocity. AmericanSociological Review, 1960, 25, 165-167.

GORANSON, R. E., & BERKOWITZ, L. Reciprocity andresponsibility reactions to prior help. Journal ofPersonality and Social Psychology, 1966, 3, 227-232.

GREENBERG, M. S. A preliminary statement on atheory of indebtedness. Paper presented at themeeting of the Western Psychological Association,San Diego, March 1968.

GREENBERG, M. S., & FRISCH, D. M. Effect of inten-tionality on willingness to reciprocate a favor.Journal of Experimental Social Psychology, in press.

GREENHERG, M. S., & SHAPIRO, S. P. Indebtedness:An aversive aspect of asking for and receiving help.Sociometry, 1971, 34, 290-301.

GURR, T. R. Why men rebel. Princeton: PrincetonUniversity Press, 1970.

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

JACQUES, E. Equitable payment. New York: Wiley, 1961.JECKER, J., & LANDY, D. Liking a person as a function

of doing him a favor. Human Relations, 1969, 22,271-378.

KIESLER, S. B., & BARAL, R. L. The search for aromantic partner: The effects of self-esteem andphysical attractiveness on romantic behavior. InK. Gergen & D. Marlowe (Eds.), Personality andsocial behavior. Reading, Mass.: Addison-Wesley,1970.

LATANE, B. (Ed.) Studies in social comparison. Journalof Experimental Social Psychology, 1966, 2(1, Pt. 2).

LAWLER, E. E., III. Equity theory as a predictor ofproductivity and work quality. Psychological Bul-letin, 1968, 70, 596-610.

LAWLER, E. E., & O'GARA, P. W. Effects of inequityproduced by underpayment on work output, workquality, and attitudes toward work. Journal ofApplied Psychology, 1967, 51, 39-45.


LERNER, M. J., & MATTHEWS, G. Reactions to thesuffering of others under conditions of indirect re-sponsibility. Journal of Personality and SocialPsychology, 1967, 5, 319-325.

LERNER, M. ]., & SIMMONS, C. H. Observers reactionto the "innocent victim": Compassion or rejection?Journal of Personality and Social Psychology, 1966,4, 203-210.

LEVENTHAL, G. S., ALLEN, J., & KEMELOOR, li. Re-ducing inequity by reallocating rewards. PsyclwnomicSciences, 1969, 14, 295-296.

LEVKNTHAL, G. S., & ANDERSON, D. Self-interest andthe maintenance of equity. Journal of Personalityand Social Psychology, 1970, 15, 57-62.

LEVENTHAL, G. S., & BERGMAN, J. T. Self-deprivingbehavior as a response to unprofitable inequity.Journal of !']xperimenlal Social Psychology, 1969, 5,153-171.

LEVENTHAL, G. S., & LANE, D. W. Sex, age, and equitybehavior. Journal of Personality and Social Psychol-ogy, 1970, 15, 312-316.

LEVENTHAL, G. S., WEISS, T., & LONG, G. Equity,reciprocity, and reallocating the rewards in the dyad.Journal of Personality and Social Psychology, 1969,13, 300-305.

LIEMOW, E. Tally's corner. Boston: Little, Brown, 1967.MACAULAY, S., & WALSTER, K. Legal structures and

restoring equity. Journal of Social Issues, 1971, 27,173-187.

MAHER, B. A. Principles of psycliopalliology: An experi-mental approach. New York: McGraw-Hill, 1966.

MARWELL, G., RATCLIEF, K., & SCHMITT, D. K.Minimizing differences in a maximizing differencegame. Journal of Personality and Social Psychology,1969, 12, 158-163.

MAUSS, M. The gift: Forms and functions of exchange inarchaic societies. Glencoe, 111.: Free Press, 1954.

McGuiKE, W. J. The nature of attitudes and attitudechange. In G. Lindzey & E. Aronson (Eds.), Thehandbook of social psychology, (2nd ed.). Reading,Mass.: Addison-Wesley, 1968.

MERTON, R. K. Social theory and social structure, (Rev.ed.) Glencoe, 111.: 1'ree Press, 1957.

MILLS, J. Changes in moral attitudes following tempta-tion. Journal of Personality, 1958, 26, 517-531.

OPSAHL, R. L., & DUNNETTE, M. D. The role of financialcompensation in industrial motivation. PsychologicalBulletin, 1966, 66, 108-118.

PRITCIIARD, K. D. Equity theory: A review and critique.Organizational behavior and human performance,1969,4, 176-211.

ROSENBERG, M. J. An analysis of affective-cognitiveconsistency. In M. J. Rosenberg & C. I. Hovland(Eds.), Altitude organization and change. \ew Haven,Vale University Press, 1960.

ROSENBERG, M. J., & AHELSON, L. P. An analysis ofcognitive balancing. In M. J. Rosenberg & C. I.Hovland (Eds.), Attitude organization and change.New Haven: Vale University Press, 1960.

Ross, A. Modes of guilt reduction. Unpublisheddoctoral dissertation, University of Minnesota, 1965.

Ross, M., THIBAUT, J., & EVENBECK, S. Some deter-minants of the intensity of social protest. Journalof Experimental Social Psychology, 1971, 7, 401-418.

SARNOFE, I. Personality dynamics and development.New York: Wiley, 1962.

SCHAFER, S. Restitution to victims of crime. London:Stevens & Son, 1960.

SCHWARTZ, S. H. Words, deeds, and the perception ofconsequences and responsibility in action situations.Journal of Personality and Social Psychology, 1968,10, 235.

SCOTT, M. B., & LYMAN, S. M. Accounts. AmericanSociological Review, 1968, 33, 46-62.

SMITH, A. The theory of moral sentiments. London: Bell,1892.

SPENCER, H. Essays: Moral, political and aesthetic.New York: Appleton, 1874.

STOUFFER, S. A., SUCKMAN, 1C. A., DEVINNEV, L. C.,STAR, S. A., & WILLIAMS, R. M. The Americansoldier: A djustmenl during army life. Vol. 1. Princeton:Princeton University Press, 1949.

SYKES, G. M., & MAT/A, D. Techniques of neutrali-zation: A theory of delinquency. American Socio-logical Review, 1957, 22, 664-670.

THIHAUT, J. W. An experimental study of the cohesive-ness of underprivileged groups. Unman Relations,1950, 3, 251-278.

THIHAUT, J. W., & KELLEY, H. H. The social psychologyof groups. New York: Wiley, 1959.

THIHAUT, J. W., & RIECKEN, H. W. Some determinantsand consequences of the perception of social causality.Journal of Personality, 1955, 24, 113-133.

TORNOW, W. W. Differential perception of ambiguousjob characteristics as inputs or outcomes moderatinginequity reduction. Unpublished doctoral dissertationUniversity of Minnesota, 1970.

WALSTER, E. Effect of self-esteem on liking for dates ofvarious social desirabilities. Journal of ExperimentalSocial Psychology, 1970, 6, 248-253.

WALSTER, E., ARONSON, V., AHRAHAMS, D., & ROTT-MAN, L. Importance of physical attractiveness indating behavior. Journal of Personality and SocialPsychology, 1966, 4, 508-516.

WALSTER, E., BERSCHEID, E., & BARCLAY, A. A deter-minant of preference for modes of dissonance reduc-tion. Journal of Personality and Social Psycl/ologv,1967, 7, 211-215.

WALSTER, E., BERSCHEID, E., & WALSTER, G. W. Re-actions of an exploiter to the exploited: Compensa-tion, justification or self-punishment? In J. R.Macaulay Si L. Berkowitz (Eds.), Altruism and help-ing behavior. New York: Academic Press, 1970.

WALSTER, E., & PRESTHOLDT, P. The effect of mis-judging another: Overcompensation or dissonancereduction? Journal of Experimental Social Psycliology,1966, 2, 85-97.

WALSTER, E., WALSTER, G. W., ABRAHAMS, D., &BROWN, Z. The effect on liking of tmderaling oroverrating another. Journal of Experimental SocialPsychology, 1966,2, 70-84.

WEICK, K. E., & NESSET, B. Preferences among formsof equity. Organizational Behavior and Human Per-formance, 1968, 3, 400-416.

WI-UNSTEIN, K. A., DEVAUGIIAN, W. L., & WILEY, M.G. Obligation and the flow of deference in exchange.Sociometry, 1969, 32, 1-12.

ZAJONC, R. B. The concepts of balance, congruity, anddissonance. Public Opinion Quarterly, 1960, 24,280-286.

(Received May 26, 1971)

Documents

NEW DIRECTIONS IN EQUITY RESEARCH€¦ · 26/05/1971 · Equity theory consists of four propositions designed to predict when individuals will perceive that they are justly treated