IV. EFFECTS OF AAP ON NON-TARGET GROUP MEMBERS' PERCEPTIONS OF TARGET GROUP MEMBERS, AND ON RELATIONS BETWEEN PARTIES

It is important to understand how female and minority employees hired under an AAP are perceived by other employees. Negative perceptions could hinder the new hire's opportunities and could damage relations between the parties. In this section we review research that speaks to these issues. Several studies have been completed in which non-target members evaluate the competence of target members. Most of these studies have used undergraduate students as respondents, and have presented them with hypothetical situations in which they have judged the qualifications of the target. In such studies, some attribute of the situation (e.g., selection procedure) is manipulated. A few studies have employed adult respondents and correlational designs.

A. Evaluations of Females

Jacobson and Koch (1977) paired 72 male undergraduates with a female confederate who was assigned to a leadership position on the basis of sex (strong preferential treatment), chance, or superior performance on a test (merit). After performing a oneway communication task with the confederate, the dyad was told they had either succeeded or failed at the task. Jacobson and Koch found that females selected on the basis of gender were blamed for poor performance of the group but were not given credit for the good performance of the group. Summers (1991) asked 112 undergraduates to read materials related to promotion of a woman in a hypothetical company. The company had either willingly adopted an AAP or refused to adopt an AAP. Judgments of the woman's qualifications were affected by the interaction of Respondent Gender X AAP. Evaluations made by female respondents in the anti-AAP condition were higher than evaluations in the other three conditions. Summers interpreted this as an augmentation effect; female respondents assumed the female manager must be particularly competent to get promoted in a hostile company.

Studies by Heilman and her colleagues show that both males and females tend to assume that females hired under affirmative action programs are relatively less competent. Heilman, Block, and Lucas (1992; Study 2) asked 184 White male employees of various companies to evaluate the competence of a specific female or minority co-worker, and to indicate the extent to which affirmative action was responsible for the co-worker's selection. Judgments of competence were inversely related to the perceived importance of affirmative action in selection. Heilman et al. (1992; Study 1) asked 129 male and female undergraduates to review application materials of someone recently hired and to make predictions about their job performance. The job was said to be either highly or moderately gender-typed to be masculine. The applicants were either male or female, and if female, either were or were not associated with an AAP. Affirmative action was manipulated by placing a statement at the bottom of the applications that said either hire or hire (affirmative action hire). The results showed that women were perceived as less competent when they were associated with affirmative action than when they were not.

Heilman, Block, and Stathatos (in press) used the same affirmative action manipulation in two studies of male and female managers (Ns = 192 & 72). In Study 1, performance information was lacking, or indicated failure (lower 50%), ambiguous success (upper 50% on 2-category scale), or clear success (upper 5% on 5-category scale). Ratings of competence were affected by performance information and the interaction of performance information by employee type. In the clear success and clear failure conditions, the three types of employees obtained similar ratings; when information was lacking or ambiguous, the woman hired in the context of the affirmative action plan was rated lower than the other two employees. In Study 2, performance information was manipulated by mentioning or not mentioning that the employee had access to ongoing coaching by a senior employee. The same pattern of effects was observed: Ratings were lower in the affirmative action-ambiguous success condition than in the other five conditions. In both experiments, all significant effects in the analyses on competence ratings disappeared when the role of qualifications in selection was used as a covariate. Heilman et al. (1996) used the communication task procedure introduced by Jacobson and Koch (1977). They included one merit selection procedure and eight preferential selection procedures that varied in the justification provided for basing selection on sex and in the provision of information about participant scores on an ability pretest. Evaluations of the female leader were higher in the merit condition and when the subjects were told she had equal or superior test scores than when they were given no score information or were told she had inferior test scores.

B. Evaluations of Blacks and Other Minorities

Northcraft and Martin (1982) reported a study in which 32 participants were asked to match five resumes to five recent hires, one of whom was Black. When the participants were told the company needed to hire a Black to satisfy its affirmative action obligations, they paired the Black employee with the weakest resume at a higher-than-chance level. This did not occur when there was no mention of the company's need to hire a Black. In two studies (Ns = 168 and 135), Garcia, Erskine, Hawn, and Casmay (1981) had White males and female undergraduates evaluate minority applicants to graduate school in psychology. The applicant was either accepted or rejected, and there either was or was not an affirmative action policy statement. The minority applicant was evaluated less favorably when commitment to affirmative action was emphasized than when it was not mentioned.

Nacoste and Fender (1993, cited in Nacoste, 1994) replicated and extended Garcia et al. (1981). Their respondents read a scenario about a Black student who had applied for admission to a graduate program in psychology, and there was or was not an affirmative action statement. In addition to evaluating the applicant, respondents reported their beliefs about the nature of typical affirmative action programs, the evaluations of such programs relative to available options, and the relative advantage of groups with and without affirmative action programs. Nacoste and Fender replicated the findings of Garcia et al. (1981). In addition, they showed that respondents who believed Blacks were at a disadvantage in competing for slots in a graduate program were responsible for the elevated evaluation of the Black student in the no-affirmative action condition.

The previous studies dealt with reactions to individual members of the target group. Maio and Esses (1996) asked whether the stigmatization effect would generalize to the target group as a whole. Their 51 Canadian undergraduate respondents were given information about a little-known group that might be forced to emigrate to Canada to escape a natural disaster. In both conditions the instructions stated that the group would contribute to the Canadian economy, but in the experimental condition they also stated that the group members would profit from affirmative action. Analyses revealed significant effects of condition on five of the seven dependent variables, with ratings being less positive when affirmative action was mentioned. There was some weak evidence that this stigmatization effect was larger among the respondents who initially had negative attitudes toward affirmative action; a stronger test of this effect would be provided by a larger sample. Due to its emphasis on stigmatization of groups rather than individuals, this study represents a novel and important extension of previous work.

C. Summary

Majority members typically view women and minorities selected through AAPs to be less competent than those selected without affirmative action, and this effect may generalize to evaluations of the target group as a whole. Such findings occur when affirmative action is operationalized as strong preferential treatment and when affirmative action is not defined procedurally, that is, when affirmative action is simply mentioned. This stigmatization may be eliminated by providing clear and compelling evidence of the woman or minority member's competence. A frequent criticism of affirmative action is that non-target group members will stigmatize target groups members, as found in this research. Thus, it is interesting to ask whether people believe this stigmatization occurs. Witt (1990) asked her university faculty respondents whether they believed affirmative action: (a) perpetuates the myth of minority and female inferiority, and (b) robs successful women and minorities of a sense of accomplishment. Among her White males, 21% agreed with the first point, and 17% with the second. Agreement rates were lower among minorities; only 8% of the Black females agreed with each of the statements. Research on self-evaluations, discussed below, finds that self-evaluations are strongly affected by information about the specific AAP and qualifications of the selected individual. The effects of AAP detail and qualifications on judgments made by non-target group members merit additional attention.

D. Effects on Relations Among Parties

Little research has addressed the effect of affirmative action programs on relations among target and non-target groups. Heilman (1994) summarized anecdotal evidence of angry and hostile reactions by non-target groups. For example, she noted an article reporting violence during a rally protesting the layoff of White male police instead of Black and female officers with less seniority. Heilman, McCullough and Gilbert (1996) investigated males' reactions to strong preferential treatment of females in a laboratory task in which a female was appointed to be the leader and the male was appointed to be a follower. They found males' reactions to be negative unless (a) males believed themselves inferior to the females in task relevant ability or (b) the males thought themselves equal to the females in ability and were also given a historical rationale for the female being appointed as the leader.

Barnes Nacoste (1992) presents a model of the relations among the enacting agency, target group, and non-target group. He argues that harmonious relations between target and non-target group members can only be expected when both have positive reactions to the AAP; that is, when target group members have low evaluation apprehension and non-target group members believe that target group members are qualified. Interactions will be strained if either group has negative reactions to the AAP. Consistent with the research summarized above, Barnes Nacoste (1992) argues that details of the AAP will determine these reactions. Barnes Nacoste (1994) extends this model with his policy schema approach, and argues that relations will be determined by the parties' policy schemas. These schemas will be based, in part but not completely, on details of the AAP.

A related question concerns the effect of affirmative action on relations between individuals and the organizations by which they are employed. There is very little research of this type. Witt (1990) reports the results of three discriminant analyses on the job satisfaction of White male university faculty, and concludes that the effects of attitudes toward affirmative action are minimal when compared to the effects of time demand and other types of stress. Konrad and Linnehan (1995b) found that employee ratings of career opportunities and organizational commitment were not related to the number of affirmative action procedures employed by the company. In a laboratory experiment, Heilman et al. (1996) found that willingness to engage in citizenship behaviors (help the experimenter code data) was affected by selection procedure, being higher when selection was based on merit than when it was based on sex. Information about relative qualifications and justification for the use of sex did not moderate this effect. Leck, Saunders, and Charbonneau (1995) asked White male and female employees of a Canadian publishing company how they would respond if female or minority employees were added to their work groups. Their intentions were associated with resistance to integration, support for equal opportunity, belief that some positive action is needed to hire more women and minority employees, and support for employer rights. Finally, in a study combining laboratory and field samples Bell (1996) found that intentions to perform a broad set of AAP-related behaviors were strongly related to attitudes toward the AAPs.