A Review of Psychological and Behavioral Research on Affirmative Action
A Review of Psychological and Behavioral Research on Affirmative Action
David A. Kravitz, Rice
University
David A. Harrison, University
of Texas at Arlington
Marlene E. Turner, San Jose
State University
Edward L. Levine, University of
South Florida
Wanda Chaves, University of
South Florida
Michael T. Brannick, University
of South Florida
Donna L. Denning, City of Los
Angeles
Craig J. Russell, Louisiana
State University
Maureen A. Conard, HRStrategies
Background
In September of 1995, the Scientific Affairs
Committee of the Society for Industrial and Organizational Psychology created a
subcommittee to review psychological and behavioral research on affirmative action. 1
A summary of that report follows. Research details and references are provided in the full
report. 2, 3
Affirmative action refers to a body of policies
and procedures designed to eliminate employment discrimination against women and ethnic
minorities, and to redress the effects of past discrimination. Most relevant psychological
research deals with affirmative action plans (AAPs) as required by Executive Order 11246,
promulgated by President Johnson in 1965 and subsequently revised by other presidents.
This executive order applies to federal agencies and certain private companies that do
business with the federal government. It has three key requirements: (a) The organization
must have and abide by an equal opportunity policy. (b) The organization must analyze its
workforce to assess possible underutilization of women and ethnic minorities, where
underutilization is defined in terms of qualified applicants or potential applicants. (c)
If underutilization is revealed, the organization must develop a plan of action to
eliminate it, and must make a good faith effort to execute the plan. The regulations
controlling the development and execution of AAPs emphasize recruitment and the
elimination of bias, and Supreme Court decisions have forbidden strong preferential
treatment and strict quotas except under the most extreme conditions. Within this broad
range, however, there is considerable ambiguity about what actions are legal and
appropriate.
Affirmative action is controversial; it
stimulates strong opposition and strong support. As an organization concerned with
psychological issues in the workplace, the Society for Industrial and Organizational
Psychology has a responsibility to review and report research results and theoretical
explanations that may inform this public debate. That is the purpose of this report. Most
of the psychological research deals with attitudes toward affirmative action, though some
addresses effects of affirmative action on members of the target group (e.g., self-esteem)
and implications of affirmative action for non-target group members' attributions about
target group members (e.g., stigmatization). To provide some context, this report briefly
discusses the legislative history and economic effects on organizations and target groups.
Evaluations of Affirmative
Action and Affirmative Action Plans
Much of the psychological research on affirmative
action has dealt with the antecedents of attitudes toward affirmative action. Presumably,
by understanding, predicting, and changing attitudes toward AAPs, one can substantially
increase the likelihood of understanding, predicting, and changing AAP-related behaviors.
Thus, this information could help organizations design AAPs that are likely to be accepted
while still accomplishing their objectives. This research has focused on two categories of
antecedents, with some work emphasizing structural factors and other research emphasizing
individual differences. We shall discuss these two bodies of research in turn.
Structural Influences
Much of the structural research on affirmative
action has manipulated the AAP. The assumption is that reactions to an AAP will depend on
details of the AAP -- especially on the weighting of demographic status. This research is
relevant to the public debate about affirmative action because much of that debate
involves an explicit or implicit disagreement about what affirmative action means (e.g.,
preferential treatment versus assurance of equal opportunity).
Opinion polls and experimental research indicate
that there is greater support for the principle of equal opportunity than for the
principle of affirmative action. Evaluations (i.e., attitudes and fairness judgments) of
affirmative action are strongly influenced by actual or presumed AAP structure;
favorability of evaluations is inversely related to the weighting of demographic status in
decision making. Consistent with these results, other research reveals that people respond
more positively to selection situations if the more qualified candidate is selected,
regardless of minority/majority status. Evaluations of merit selection are almost
universally positive, but evaluations of preferential treatment are moderated by such
factors as respondent race and gender (White males respond most negatively) and
self-efficacy. Although people support compensatory actions (e.g., training) and diversity
efforts (e.g., recruitment), they prefer to limit affirmative action to the elimination of
discrimination.
It is important to realize that the effect of
perceived AAP structure on attitudes does not depend on the accuracy of perception. Some
research has indicated that the public has a poor understanding of affirmative action, and
some theorists have argued that this poor understanding partly underlies the public
opposition that exists. As this logic would suggest, research has revealed that attitudes
toward affirmative action can be changed by providing respondents with information about
the structural details of the AAP or with arguments contrary to their initial positions.
In addition, providing a justification for the AAP leads to more positive attitudes. The
content of the justification (redressing past discrimination versus improving
organizational profitability via enhancing diversity) does not strongly affect attitudes,
possibly because reactions to different justifications are moderated by other factors.
Researchers have attempted to explain why
weighting of demographic status influences attitudes. Most research has dealt with two
possible mediators: perceptions of fairness and implications for self-interest.
Strong support has been obtained for theories
that incorporate concepts of fairness. Qualitative research reveals that people think of
affirmative action in terms of fairness. Consistent with the research discussed above,
they consider affirmative action to be fair in principle but often unfair in practice.
Fairness ratings and attitudes are highly correlated, and judgments of fairness have been
shown to fully mediate the effects of an AAP manipulation on attitudes. Finally, some
findings regarding fairness are qualified by interactions with the respondent's
demographic status, perhaps due to the presence of egocentric bias (effects of
self-interest).
Self-interest has been studied both at the
personal level and at the collective level, where the latter refers to implications of
affirmative action for the respondent's demographic group. Personal and collective
self-interest are highly correlated. Attitudes toward affirmative action are correlated
with both personal and collective self-interest, and there is some evidence that the
former is more important. Comparative research has revealed that attitudes are more
closely related to fairness judgments than to perceptions of personal or collective
self-interest. Furthermore, one study found that effects of an AAP manipulation on
attitudes were only partly mediated by implications of the AAP for self-interest.
In addition to weighting of demographic status,
several other structural variables have been studied. This work has dealt with the
identity of the target group, the setting, and the apparent need for affirmative action.
Attitudes toward AAPs are influenced by the
identity of the target group. Whites tend to evaluate programs directed at Blacks or
ethnic minorities less positively than programs directed at women or people with
disabilities. This effect is relatively small, however, and additional research is needed
to confirm it. The effect is moderated by details of the AAP and by respondent education,
political perspective, and gender. Results of several studies suggested effects of
ingroup-outgroup differences or self-interest, and research discussed above has shown that
self-interest considerations influence attitudes.
Research on the effects of setting has most often
compared college admissions, employment in private organizations, and employment in public
organizations. This research has revealed little or no effect of setting on attitudes.
Interestingly, although research on AAP structure revealed positive attitudes toward
training programs, only two of the five relevant studies found more positive attitudes
toward affirmative action in college admissions than in employment, and in both cases the
setting was confounded with the description of affirmative action.
There is little research on the question of how
attitudes toward affirmative action are affected by the target group's need for
assistance. One study found that the belief that Blacks suffer discrimination is
positively associated with attitudes toward affirmative action. Several studies have
manipulated the organization's history of discrimination, and this sometimes affects
attitudes. Survey research suggests that many White Americans believe discrimination is a
problem of the past, and that Blacks have themselves to blame for their economic
disadvantages. This perception might tend to decrease Whites' support for affirmative
action targeted at Blacks. The emphasis on need is consistent with court decisions, which
have emphasized that AAPs must be remedial.
Individual Differences Bases
The research described above has focused on
structural aspects of the situation and of the AAP. Other research has addressed the role
of individual differences. The assumption in much of this work is that individual
differences will affect the person's attitude toward affirmative action in general, and
this in turn will affect the person's attitudes toward specific AAPs. This research has
included the respondent's role, demographic variables, dispositional variables, opinion
variables (prejudice, relative deprivation, and political ideology), and personal
experiences. When considering this research it is important to realize that it, like all
passive observation research, does not logically imply specific causal conclusions.
Two characterizations of a respondent's role have
received some investigation. The first approach deals with differences between those who
make decisions about or administer AAPs versus those who do not. Two general conclusions
can be drawn from this research: (a) The most positive attitudes toward affirmative action
will be observed among those whose jobs involve the maintenance of AAPs. (b) Opinions of
decision makers who have thought about the issue are less flexible than opinions of
typical citizens. The second approach deals with differences between those who are or are
not members of the AAP target group. Results of this research are clear: People respond
more positively toward AAPs that are targeted at their own group. This trend is clear for
ethnicity, but less clear for gender. It is, of course, consistent with self-interest.
In addition to this interaction of Target Group
Status x Respondent Status, research on demographic differences has revealed some main
effects. Blacks clearly feel more positively about affirmative action in general than do
Whites. Hispanics appear to be intermediate in attitude, but this conclusion is based on
very few studies. Demographic research has not distinguished among subgroups (e.g.,
Cuban-Americans versus Mexican-Americans), and has ignored Asians and Native Americans.
Conclusions about gender differences are more tentative. When respondents (especially
White respondents) are asked about affirmative action in general, women appear to have
more positive attitudes than do men. When males and females are involved in or mentally
place themselves in a situation in which they benefit from an AAP, however, females tend
to report more negative affect and attitudes. Such gender effects often are moderated by
the structural properties of the AAP; women tend to be more responsive to these factors
than are men.
Researchers occasionally assess the relations
between affirmative action attitudes and demographic variables other than gender and
ethnicity. Only age, education, and income have been included with any regularity.
Correlations between these variables and attitudes toward affirmative action are
invariably small and usually statistically nonsignificant. Three studies incorporated
measures of self-efficacy; they found that AAP structure, perceived fairness, and
self-efficacy are complexly interrelated.
Most research on the effect of prejudice has
studied racism rather than sexism. There has been considerable disagreement over the
conceptual and operational definitions of racism. Despite this disagreement, empirical
results consistently demonstrate an inverse relation between racism and attitudes toward
affirmative action. Most comparative research has revealed that racism has somewhat less
explanatory power than does self-interest, and much less explanatory power than does
fairness.
Relative deprivation refers to the belief that a
group suffers unfairly low opportunities or outcomes. Relevant research on affirmative
action has dealt with collective relative deprivation (which refers to the respondent's
own demographic group) and relative deprivation on behalf of others (which refers to some
other demographic group). Almost all this research suffers from severe measurement
problems, so conclusions must be very tentative. It is probably safe to conclude that
people who believe their own demographic group suffers from collective relative
deprivation will tend to have positive attitudes toward affirmative actions designed to
decrease that deprivation. This, of course, is confounded with self-interest. There is
also some indication that a sense of relative deprivation on behalf of others is
associated with positive attitudes toward affirmative action targeted at the other group.
Research on the perceived need for affirmative action, discussed above, is consistent with
this conclusion.
Most research on the relationship between
political perspectives or ideology and affirmative action attitudes has operationalized
the concept of political ideology in terms of stratification beliefs, self-identification
by political party, or self-identification on a conservative-liberal response scale. The
stratification approach considers several dimensions of the dominant ideology -- the
belief that everyone has an equal opportunity to succeed, that success or failure are due
to individual factors rather than structural influences, and that inequality is
appropriate because it reflects unequal contributions and serves as a motivator. Survey
research has sometimes found that people who adhere to this dominant structural ideology
tend to oppose affirmative action. Egalitarians, in contrast, are more likely to support
affirmative action. Other research reveals that opposition to affirmative action is
stronger among Republicans than Democrats, and among conservatives than liberals.
There is little research on how attitudes toward
affirmative action are affected by prior experiences. Attitudes toward affirmative action
seem to be positively associated with having experienced discrimination and having worked
at an organization with an AAP, at least if the AAP resulted in positive experiences.
Negative experiences with affirmative action appear to decrease attitudes.
The research summarized above dealt with people's
attitudes toward affirmative action and AAPs. We now turn to research on the psychological
effects of affirmative action on involved parties -- members and non-members of the target
group.
Effects of Affirmative Action
Plan on Non-target Group Members' Perceptions of Target Group Members, and on Relations
Between Parties
Some research has dealt with effects of
affirmative action on non-target group members' perceptions of target group members. This
research shows that non-target group members view women and minorities selected in the
context of an AAP as less competent than those selected without benefit of affirmative
action. Such findings occur when affirmative action is operationalized as strong
preferential treatment and when affirmative action is simply mentioned and not defined
procedurally. Some limited research suggests that the attribution of incompetence can be
moderated by providing clear evidence of the individual's competence. It is likely that
evaluations of target group members by others will be influenced by the type of selection
process (i.e., weighting of merit and demographic status).
There is little empirical work on effects of
affirmative action on relations between target and non-target group members. Research on
reactions of involved parties suggests that affirmative action would have deleterious
effects on relations unless the AAP was positively evaluated by both target group members
and non-members. That is, to prevent impaired relations, the AAP must be seen as fair,
negative implications for non-target group members' self-interest must be minimized, and
qualifications of the selected target group members must be emphasized. It is apparent
that these effects will be strongly affected by details of the AAP.
Psychological and Behavioral
Effects of Affirmative Action on Target Group Members
Research on target group members has included
three categories of outcomes: (a) measures of motivation and task interest, (b)
selfevaluations of ability and performance, and (c) performance. Almost all this
research has focused on effects of gender-based selection on women; there is very little
work on effects of race-based selection on racial minorities.
This research suggests that affirmative action
programs have mixed consequences for recipients. Women's task motivation, interest, job
satisfaction, and job commitment do not appear to be strongly affected by the selection
procedure. Women's task and job choice and selfevaluations of ability and performance
can be impaired by gender-based selection procedures that provide no evidence of recipient
qualifications, but these effects can be eliminated by provision of explicit, unambiguous,
and focused evidence of recipient qualifications. Research further suggests that the
relation between selection procedure and individual task performance is complex, and may
be moderated by the individual's level of self-efficacy and by task characteristics. There
are, however, few studies on performance. In sum, it appears that implementation strategy
assumes primary importance in determining the nature of women's reaction to gender-based
affirmative action.
Economic Effects of Affirmative
Action on Target Groups
Research on target group attainment has examined
three categories of outcomes: (a) employment rates, (b) income attainment, and (c)
promotion rates and occupational attainment. It appears that affirmative action has led to
improvements in all three categories for both women and ethnic minorities. This
improvement has been neither consistent nor substantial, however, and methodological
problems inherent in this research limit the ability to attribute changes solely to
affirmative action. Finally, none of the Committee members are economists, and our review
did not include a full scale investigation of the economics or financial literature.
Economic Effects of Affirmative
Action on Organizations
Research on organizational performance has dealt
with organizational effectiveness and financial equity. It is sometimes argued that
affirmative action will hurt organizational effectiveness by forcing organizations to hire
unqualified women and ethnic minorities. Others argue that increasing diversity will
improve organizational performance. In several studies there was no support for either of
these predictions; this research revealed no difference in effectiveness between
organizations that appeared to emphasize affirmative action and those that did not. Other
research has found that formal charges of discrimination led to a decrease in stock
prices, and public recognition of affirmative action excellence led to a temporary
increase in stock prices. Again, however, we must point out that we did not do a
comprehensive review of the economics literature.
Conclusions, Limitations of
Current Knowledge, and Needed Research
There are, of course, limitations to the research
on affirmative action. Affirmative action is a complex policy, and reactions are equally
complex. No single study can simultaneously consider all the relevant variables, so the
results of every study must be qualified by what was not included. Other limitations also
exist. For example, almost all research on reactions of target group members has dealt
with women's reactions to gender-based selection procedures, ignoring reactions of
minority group members to race-based procedures. In addition, research on some topics has
focused almost entirely on the reactions of students to situations described in scenarios.
Obviously, additional research is needed. Nonetheless, we feel it is safe to draw certain
conclusions.
The clearest finding is that attitudes toward
affirmative action are strongly influenced by details of the AAP, and this influence is
probably mediated by perceived fairness and implications for the respondent's
self-interest. Attitudes toward affirmative action are inversely related to the weighting
of demographic status by the AAP. Furthermore, women selected under the auspices of an AAP
are likely to be stigmatized by themselves and others unless it is made clear to everyone
that they are qualified. Although individual difference variables such as racism and
political ideology are also related to attitudes, they are less important than details of
the AAP. These results have implications for the manner in which AAPs should be
structured, and demonstrate the need for clear communication about organizational AAPs.
Footnotes
1. Subcommittee members were: Rabi S. Bhagat,
Michael T. Brannick, Maureen A. Conard, Donna L. Denning, David A. Harrison, David A.
Kravitz (chairperson), Edward L. Levine, Craig J. Russell, and Marlene E. Turner.
2. Neither this summary nor the full report are
to be taken as the official or unofficial position of the Society for Industrial and
Organizational Psychology.
3. The full report is available either through
the SIOP administrative office or on the SIOP Web Page (http://www.siop.org/siopinfo.aspx).