Affirmative Action Research Paper

This sample Affirmative Action Research Paper is published for educational and informational purposes only. Free research papers are not written by our writers, they are contributed by users, so we are not responsible for the content of this free sample paper. If you want to buy a high quality research paper on any topic at affordable price please use custom research paper writing services.


Affirmative action  refers  to  any  measure,  beyond  a simple termination  of discriminatory  practice, adopted to  correct  for  past  or  present  discrimination   or  to prevent  discrimination   from  recurring  in  the  future. This controversial social policy has inspired research in a variety of disciplines. The present research paper summarizes research on (a) the economic effects of affirmative action on target groups and on organizations, (b) stigmatization of affirmative action  target  group  members  by others and by the individuals themselves, and (c) attitudes toward affirmative action.


  1. What Is Affirmative Action?
  2. Economic Effects of Affirmative Action
  3. Affirmative Action and Stigmatization
  4. Attitudes toward Affirmative Action
  5. Conclusions

1. What Is Affirmative Action?

1.1.  Background

Humans have a propensity  to distinguish  between the in-group and the out-group, a process that often results in prejudice and discrimination  toward individuals in the out-group. Consistent with this tendency, most countries have long histories of discrimination  toward women and racioethnic  minorities.  During recent decades, many countries  have passed antidiscrimination laws, and some have gone further to require that affirmative actions be taken to enhance  opportunities  for traditionally  underrepresented groups. Although affirmative action may apply to university admissions, this research paper is restricted  to affirmative action in the workplace. It summarizes the economic, behavioral, and psychological research  on affirmative action  that  has been published in English. Most of this work has been conducted  in the United States or Canada.

1.2.  Definitions

What is affirmative action? That question can be answered conceptually, legally, and operationally.

1.2.1. Conceptual  Definition

Many conceptual  definitions  have  been  offered. For example, the U.S. Civil Rights Commission has defined affirmative action as ‘‘a term that in a broad sense encompasses  any measure,  beyond a simple termination of discriminatory  practice, adopted to correct for past or present discrimination  or to prevent discrimination [from] recurring in the future.’’

1.2.2. Legal Definitions

The  legal definitions  of affirmative action  vary from place to place and change over time. Critical dimensions on which the law varies include which groups are targeted,  which  actions  are  required  and  forbidden, what  penalties  can  be  imposed  for  violations,  and how the law is enforced. In the United States, federal affirmative action law is jointly determined by the U. S. Constitution  (5th and 14th amendments),  presidential executive  orders  (e.g.,  Executive  Order  11246  [EO 11246]),  legislation  (e.g.,  Civil  Rights Acts of 1964 and 1991), implementing  regulations promulgated  by federal agencies (e.g., Office of Federal Contract Compliance Programs [OFCCP], Equal Employment Opportunity Commission [EEOC]), and court decisions.

Organizations in the United States may be required to establish affirmative action plans (AAPs) for four reasons. First, courts may impose AAPs on organizations that have been convicted of discrimination. Second,  to  avoid  adverse  legal decisions,  organizations may set up AAPs as required by court-approved consent decrees. Third, EO 11246 and other acts require  affirmative  action  of  federal  agencies  and  of most organizations that do business with the government  (i.e., federal contractors).  Fourth,  government ‘‘set-aside’’ programs  at  the  federal,  state,  and  local levels  require  that  some  proportion   of  contracting and   procurement    expenditures    go   to   designated firms. In addition  to these legal bases for affirmative action, some organizations may voluntarily take affirmative actions to diversify their workforces. All AAPs are constrained  by the Constitution  and by the Civil Rights Acts.

It is not possible to describe here the requirements and   constraints    of   affirmative   action    in   detail. However, because most AAPs in the United States are established  by federal contractors  as required  by EO 11246, some of their key requirements  and constraints are described here. First, the organization must not discriminate  on the basis of race, color, religion, sex, or national origin. Second, the organization must complete an annual  utilization  analysis that  compares  its workforce, in terms of gender and racioethnicity,  with the available qualified workforce. Third, if underutilization  of women  or  a racioethnic  minority  group  is revealed, the organization must generate a plan for eliminating   the   underutilization.  Fourth,   the   plan should  use targeted  recruitment,  outreach,  and other nonpreferential  actions that will facilitate target group success. Fifth,  the  plan  must  not  involve the  use of demographic status (i.e., racioethnicity, gender) in making selection, promotion, or termination decisions. This last point is important:  Preferential treatment  of women  and  racioethnic  minorities  is contrary  to EO 11246. Preferences are legal only when they are specifically required  by a court  order  or consent  decree to reverse the effects of proven discrimination.

1.2.3. Operational Definitions

For current  purposes,  we must  consider  two types of operational  definitions:  organizational  implementation of affirmative action and the descriptions of affirmative action used in behavioral research. Organizational AAPs are complex. They potentially include organizational structures, written reports, proscribed and prescribed processes, and more.  Although ‘‘best practice’’ reports exist, the author  has never seen a report  summarizing the frequencies of typical practices. Importantly,  although  much  of the public  equates  affirmative action with preferential treatment, the actual frequency of such illegal procedures is unknown. Operationalizations used in behavioral research tend to be brief, simplistic, and (often) illegally strong.

1.3.  Public Beliefs

The  American  public  construes  affirmative  action  in many different ways. Some people focus on the role of affirmative action  in decreasing  discrimination  against women  and  ethnic  minorities.  Other  people  believe it involves quotas or preferential treatment.  Very few people think  of affirmative action in terms of recruitment and outreach, and virtually nobody has a realistic understanding of the law. Beliefs about what the term ‘‘affirmative action’’ means vary with respondent  characteristics, such as racioethnicity and political orientation,  and predict support or opposition to affirmative action.

2. Economic Effects Of Affirmative Action

There has been a lively debate regarding the economic effects of affirmative action on the targeted groups and on organizations  subject to affirmative action regulations. The implicit question  is whether the benefits of affirmative action  outweigh  the  costs.  That  question cannot be answered in the abstract because it depends both  on  the  effects of affirmative action  and  on  the values one assigns to organizational  outcomes and to the  well-being  of the  target  groups.  Although  such values  are  a personal  matter,  it  is possible  to  study the effects of affirmative action. Both conceptual  and empirical studies have been published, and the conclusions are summarized  in what follows. However, it is important  to acknowledge  some unavoidable  methodological challenges.  Causal conclusions  are possible only  when  the  experimenter   can  randomly   assign experimental units to treatment conditions while holding constant all other potentially relevant factors. That is impossible when dealing with a public policy such as affirmative action, where assignment to ‘‘conditions’’ is determined  by the law and by the experimental  units (organizations  and target groups) themselves.

2.1.  Target Groups

Many studies  assess the  employment  status  of White women   and   racioethnic   minorities,   vis-a` -vis  White men,   in   federal   contractors   versus   noncontractors while controlling for relevant variables such as the type of industry.  A positive  effect of affirmative action  is inferred  if the  relative  outcomes  of target  groups  are better  in  federal  contractors  than  in  noncontractors. This research indicates that affirmative action has had a positive  effect on  the  employment  status  of White women and racioethnic minority groups, but the results are complex and some are tentative. At a minimum, the effects have varied across target groups, locations, time, and types of positions.  The initial effect of affirmative action was substantial for African Americans and slightly positive   for  White   women.   The   limited   work   on Hispanic Americans suggests that they were also helped. The clearest effects were on employment,  especially in professional  positions,  although  wages may also have been affected. However, this conclusion must be qualified by effects of region  and  time.  The largest initial impact of affirmative action (and of antidiscrimination law in general) was observed in the South, where the extent of discrimination  was most extreme and so there was the  greatest  room  for improvement.  In  addition, changes were most substantial during the decade from 1965  to  1975.  Broad  progress  stagnated  during  the 1980s, especially for African Americans, although there is some indication of continuing  effects on the status of White women.

Other studies use micro-level employer data drawn from a variety of sources, including  the organizations themselves.  This  work  has  found  that  target  groups profit more from affirmative action than from passive equal opportunity  practices. The effects of affirmative action vary across target groups, locations, industries, implementation  procedures,  and other factors.

A final indication  of the effectiveness of affirmative action  is the  finding  that  growth  in  employment  of White women and African Americans within organizations  is predicted  by organizational  AAP goals. Although the relation is positive, the actual growth in employment  of these  groups  is just a fraction  of the goals. This demonstrates  that  affirmative action goals do not function as strict quotas.

2.2.  Organizations

The public debate surrounding  affirmative action often centers on the presumed consequences of affirmative action policies for organizations.  Critics claim that  AAPs adversely affect performance by (a) forcing organizations to expend resources on implementation, (b) constraining organizational decisions, and (c) forcing organizations to hire underqualified individuals. Supporters of affirmative action  argue  that  AAPs improve  organizational  performance  by (a)  motivating  both  White  male  and  other employees and (b) providing organizations with the resources represented by a diverse workforce.

Theoretical models show that whether the net effect of affirmative action  on  organizations  is positive  or negative should  depend  on several factors, including the extent  of preexisting  discrimination,  exactly how the  AAP is implemented,  and  differences  in  human capital resources of majority and minority groups. Preexisting  discrimination  would  hinder  firm performance by limiting access to the resources represented by   the   victimized   groups.    If   affirmative   action decreased  this discrimination,  the effect on firm performance would be positive. On the other hand, if discrimination  were already absent, the constraints  on behavior posed by affirmative action could have a negative effect. Similarly, AAPs such  as focused recruitment  that  serve to  increase  organizational  access to human resources should have a positive effect on performance, whereas preferential selection should have a negative effect. Finally, the magnitude  of the negative effect of preferential selection should increase with the size of the human capital gap between the majority and minority applicant groups.

As is  clear  from  the  preceding,  there  are  logical arguments for both positive and negative effects of affirmative action  on organizations.  Thus,  we turn  to the empirical literature. The strongest research has included a wide variety of organizations  and has controlled  for industry,  job level, and other influential variables. The paper qualifications (e.g., education)  of minorities  and White women appear to be somewhat lower than those of White males in firms with AAPs, particularly  when affirmative action is used in hiring rather than when it is used only in recruiting. (The use of affirmative action in hiring does not necessarily imply preferential selection. Instead, it can involve actions such as the use of different screening mechanisms and the insistence that each set of interviews must include one person who belongs to an underrepresented group. Many organizations reported  that affirmative action was involved in hiring processes   that   resulted   in  the   selection   of  White males.) Importantly,  the minor  decrements  in educational  qualifications  are  not  associated  with  decrements   in   performance.    Several   explanations    are offered for this apparent  inconsistency,  but  all relate to the fact that firms that use affirmative action tend to have superior human resources practices. For example, virtually all AAPs include  extensive recruitment,  and this increases the overall quality of the labor pool. In addition, these organizations tend to use selection procedures   that   are  unusually   comprehensive   and valid.  Their  use  of comprehensive  selection  systems enables them to detect high-potential  minority  applicants whose competence is greater than their education would suggest. The null effect of affirmative action on organizational  performance  that has been observed in the multi-industry  research has been replicated in the few studies that have assessed a single type of organization (e.g., police departments).

3. Affirmative Action And Stigmatization

Concerns   about  stigmatization   have  been  raised  by judges rendering  decisions in affirmative action cases, by Black and Hispanic intellectuals, and by conservative opponents  of affirmative action. Empirical research has addressed   several   questions,   of  which   three   have received the lion’s share of attention.  Are members of affirmative action target groups stigmatized by others? Do members of affirmative action target groups expect to be stigmatized  by others,  and  if so, what are their reactions? Do affirmative action beneficiaries engage in self-stigmatization? Work in this area draws on concepts from stigma theory and from attribution  theory, particularly the principle of discounting.

According to the discounting  principle,  an individual’s confidence  regarding  a possible  cause  of some outcome is affected by the presence or absence of plausible alternative causes. For example, if a woman is hired by a firm that has an AAP that targets women, she  could  have  been  selected  either  because  she  is highly qualified or to help the firm meet its affirmative action  goal. In the  absence of affirmative action,  the only remaining explanation is her qualifications. Thus, a  woman  hired  in  the  context  of  an  AAP will  be evaluated  less positively than  a woman  hired  in the absence of affirmative action or less positively than a man. This reasoning applies both to observers and to the selected individual herself.

Note that the logic of discounting requires the assumption that affirmative action involves preferences. Also note that this reasoning assumes there are only two plausible reasons for an individual’s selection—qualifications or affirmative action—and  that a White male’s selection must necessarily be based on his qualifications. This ignores factors such as traditional  gender or racial discrimination, nepotism, personal friendship, similar backgrounds,  and the myriad of other  factors that are unrelated  to merit  but  nonetheless  affect employment decisions.

3.1.  Stigmatization by Others

In the prototypical stigmatization experiment, the participant  is given information  about  the background  and performance  of  an  employee  who  is  either  male  or female and, if female, either is or is not an affirmative action  hire.  The  affirmative  action  condition   nearly always involves absolute  preferences.  The  participant is asked to evaluate the employee on competence  and other dimensions. Comparisons among the three conditions—male, female, affirmative action female employee—provide the key tests of stigmatization. In a second  approach,  the  male participant  believes he  is working with a female leader on a one-way communication task, with the leadership role having been assigned on the basis of either test performance or gender. Again, the affirmative action condition involves explicit preferences. After successful or  unsuccessful  completion  of the task, the participant is asked to evaluate his partner.

Consistent with the discounting  principle,  the affirmative  action  female is stigmatized.  In  the  scenario research,  she  is evaluated  less positively  than  either the male or the non-affirmative  action female. In the communication  research,  she  is evaluated  less positively when selected on the basis of gender than when selected on the basis of test scores. Although the target is usually a White female, a few studies have included racioethnic  minority targets. The stigmatization  effect generalizes to these targets.

Having  established  the  existence  of stigmatization, subsequent  research  has concentrated  on determining its limitations.  Key conclusions  include  the following. First, stigmatization can be eliminated by providing incontrovertible information regarding the employee’s competence   or  successful  performance   on  the  job.

Second, ambiguous information about employee success does not eliminate stigmatization. Third, stigmatization occurs when affirmative action is explicitly described as involving preferences and also when it is undefined. The latter finding is consistent with the common assumption that  affirmative  action  involves  preferences.  Fourth, stigmatization  does not occur  when affirmative action is explicitly limited to the elimination of discrimination. Much of this work has assessed the impact of respondent  gender  and  racioethnicity   and  has  sometimes found   that   stigmatization   is  most   evident   among respondents   who   are   not   members   of  the   target group. With this exception,  there is little research on the potential importance of other individual attributes, such as political ideology, prejudice, and belief in a just world, regarding the extent of stigmatization.  In addition, there is little use of attribution  theory other than the principle of discounting.

3.2.  Target Group Members’ Anticipation of Stigmatization

Conceptual and empirical work on self-fulfilling prophecies and stigmas suggests that the anticipation  of stigmatization will affect the individual’s affective, cognitive, and behavioral reactions. The belief that others consider one to be inferior will usually be distressing. It can stimulate either assumption-confirming  or assumption-disconfirming behavior, depending on the situation and the individual. Indeed, concerns about stigmatization have motivated several Black and Hispanic intellectuals  to speak out against affirmative action. However, little research directly addresses this issue in the context of affirmative action. The research that exists suggests that many White women and racioethnic   minorities   expect  to  be  stigmatized  by others even in the absence of affirmative action. They anticipate that affirmative action will increase the stigmatization,  although  they  may express  less concern about this possibility than do White males. The anticipation  of stigmatization,  with  or  without  affirmative action, affects their feelings and behavior. One consequence  may be a lessening of support  for affirmative action.  Nonetheless,  they apparently  believe that  the known  disadvantages  of discrimination  outweigh  the possible disadvantages  of stigmatization  because they consistently  support  nonpreferential  versions of affirmative action.  A second  possible  effect could  be an increase in stereotype  threat.  Stereotype threat  refers to the fear that  one may behave in such a way as to confirm negative stereotypes of one’s group. It is exacerbated  by factors that  draw attention  to the stereotype   and   to   the   individual’s  membership   in   the devalued  group.  Although  the  presence  of  an  AAP may be such a factor, this possibility has not yet been subjected to empirical investigation.

3.3.  Self-Stigmatization by Target Group Members

Self-stigmatization refers to doubts regarding one’s own competence,  subsequent  selection of a less demanding task rather than a more demanding task, decreased interest in the position in question, negative affect, and the like. There are two primary reasons to anticipate  self-stigmatization  as a result of preferential selection. The first relates to the discounting principle mentioned previously. If a woman knows that her selection is based on her gender rather than her qualifications, she may be less confident about her qualifications than she would be if merit were the only plausible reason for her selection. Preferential selection may also lead to self-stigmatization because receiving help can have self-threatening implications. The fact that  one is helped  suggests that  one needed to be helped, and this implies incompetence.

Most research  in this  area has used  the  following research procedure.  Male and female participants (typically White undergraduates) believe that they are about  to  partake  in  a one-way communication  task. Assignment to the leader and follower roles is said to be based either on gender or on test performance.  In reality,  assignment  to role  is predetermined  and  the participant  is assigned  to  the  leader  role.  After role assignment, participants  are asked about their leadership ability in general, their interest  in serving as the leader in the communication  task, and the like.

This research reveals consistent evidence of self-stigmatization by preferentially selected White undergraduate women. This effect is clearer for judgments of competence   than  for  reports   of  motivation  or  task interest.  However,  when  the  AAP involves  selection among comparable candidates, self-stigmatization is eliminated. Research on people of color is very sparse, but the work that exists has failed to observe self-stigmatization.  This  null  effect likely stems  from  the belief by people  of color  that  they are hurt  more  by discrimination   than   they  are  helped   by  affirmative action. Similarly, virtually all of this research has studied undergraduates;  the limited work on employed adults has obtained no evidence of self-stigmatization. Several studies  failed to replicate  the  effect with  White  male respondents.  This is consistent with the general lack of concern  about self-stigmatization among White males, who have profited from traditional  race and sex discrimination for centuries. These results suggest that initial self-confidence plays a critical role in reactions to preferential selection, and there is some empirical support for  that  perspective.  Finally,  psychological  processes such  as  self-protective  and  egocentric  biases  would work against the development of self-stigmatization.

3.4.  Summary

One observation applies to all of the work that has been discussed in this section: Negative effects are limited to situations  in  which  affirmative action  is operationalized as preferential treatment or is assumed to involve preferences. Preferential selection causes observers to question the competence of the selected individual and causes selected White women to doubt their own competence.  These  effects are  eliminated  or  sharply reduced  when affirmative action involves preferences among  comparable  candidates,  and  these  effects are fully eliminated  when  affirmative action  is restricted to the elimination  of discrimination  and opportunity enhancement  efforts such as targeted recruitment.

4. Attitudes Toward Affirmative Action

By far the largest body of research on affirmative action deals with the variables that predict attitudes. Support for affirmative action  is frequently  assessed in  large surveys,  and  many  dozens  of  questionnaire   studies have been published.

Although many attitudinal predictors have been studied,  this  research paper  is  limited  to  the  variables  that have received substantial empirical attention. These predictors  can be divided into three categories: structural aspects of the AAP, characteristics of the respondents, and psychological mediators. The structural factors  include  AAP strength,  identity  of the  target group, and justification. The individual difference variables fall into two categories: demographic  characteristics and opinion  variables. Finally, the respondent’s personal and collective self-interest and the perception of affirmative action fairness are presumed  to mediate the relationship between the individual difference predictors and attitudes.

4.1.  Structural Factors

The structural  factors include details of the AAP. Real AAPs are complex.  They include  utilization  analyses that compare the employer’s workforce with the corresponding  qualified labor market  as well as organizational structures that assign responsibilities for maintaining  equity.  If target  groups  are  underrepresented,  the  AAP includes  goals for target  group employment and a set of processes for implementation of the change. In contrast  to this complexity, few research descriptions of AAPs exceed a single paragraph. In addition,  much of this research includes AAPs that would be illegal if actually implemented.

4.1.1. Affirmative  Action Plan  Strength

The structural  factor that has received the most attention is the manner in which the AAP attends to demographic status. This is typically referred to as AAP strength.  The  most  common  AAPs used  in  research, from weakest to strongest, include  (a) the elimination of discrimination  against the target group; (b) opportunity enhancement activities such as focused recruitment of the target group and the provision of training; (c) a tie-break  procedure  (often referred  to as weak or soft preferential treatment)  in which members of the target group are preferred  if, and only if, their qualifications are equivalent to those of nontarget applicants; and (d) a preferential procedure (typically referred to as strong or hard  preferential  treatment)  in which members of the target group are preferred even if their qualifications are inferior to those of nontarget applicants. Quotas would also qualify as strong preferential  treatment.  Although strong preferential treatment  is normally illegal, this is the version most frequently used in behavioral research. Note that these four procedures confound the weighting of demographic status with the relevant phase(s) of the employment process.

Most research  has found  a negative effect of AAP strength  on attitudes.  The elimination  of discrimination  receives  strong  support.  Opportunity   enhancement procedures receive somewhat weaker support. Preferences are opposed, with most studies finding greater  opposition   to  strong  preferential   treatment than to weak preferential treatment.

However, the preceding conclusions must be qualified. To a great extent, they are drawn from research on White respondents.  When the effect of AAP strength on attitudes is studied within racioethnic minority groups,    the   negative   monotonic    effect   observed among  Whites  is not  always observed.  Although  all groups express opposition to strong preferences, racioethnic  minorities  make less of a distinction  than do Whites  among the  weaker versions of affirmative action.  Indeed,  two studies  on  minority  respondents have found greater support for recruitment  than for the elimination  of discrimination  or any other  AAP. Very few studies have included multiple racioethnic groups and  AAPs of varying strength,  so this  conclusion  is tentative.

4.1.2. Identity  of Target  Group

A few studies have manipulated the identity of the target group. There is a tendency for White respondents’ attitudes to be less positive when the AAP targets minorities   (especially   African   Americans)   than   when   it targets White women or individuals with disabilities. Explanations  for this effect could include racial prejudice (which would lower evaluations when the AAP targets  racial  minorities)  and  inclusion  of the  target group   in  the  respondents’  scope  of  justice  (which would  raise evaluations  when  the  AAP targets  White women or individuals with disabilities). However, there is such little work in this area that  these conclusions must be tentative. One would expect the impact of target group status on attitudes to vary with demographic status of the respondents,  but this interaction has received too little attention to permit any conclusion.

4.1.3. Justification

A number of studies have assessed the effects of explicit justification on attitudes. This work has found a modest positive  effect of justifying  the  AAP on  the  basis  of fairness (it is needed to remedy past discrimination)  or economic value (it increases diversity, which has a positive impact  on  the  organization’s success).  Justifying the AAP by stating that the target group is underrepresented appears to have a neutral or very small positive effect.

4.2.  Individual Difference Variables

Dozens  of studies  have  correlated  affirmative action attitudes with a host of individual difference variables. The  conclusions  drawn  in  what  follows  are  largely based on a meta-analysis of this research that was presented  by  David  A. Harrison,  David  A. Kravitz, and   Dalit   Lev-Arey  at   the   2001   meeting   of  the Academy of Management.

4.2.1. Demographic Variables

Two demographic variables, racioethnicity and gender, have received the bulk of the attention,  although there is also enough research on education to permit conclusions. Other variables such as age, income, and religion have occasionally been studied, but they have typically been included  as controls  rather  than  as variables of central interest. It is important to emphasize that demographic  variables  cannot  serve  an  explanatory role.  If one  finds  an  effect of gender,  for  example, then one must ask why the effect occurred. The answer necessarily  lies in  the  opinion  variables  that  covary with gender and with the psychological mediators,  as detailed in what follows. Racioethnicity

 One of the most consistent and powerful predictors  of affirmative action attitudes is the  respondent’s  racioethnicity.  African Americans are the most supportive,  Whites are the least supportive, and Hispanic Americans usually fall between the other  two  groups.  The  very limited  work  on  Asian Americans  suggests  that  their  attitudes   are  usually more  favorable toward  affirmative action  than  those of Whites  but  are less favorable than  those  of other minorities. Despite these mean differences, it is important to note that disagreement exists within each of the groups. Many White Americans support affirmative action,  and many African Americans oppose it. Also, as mentioned  previously, attitudes vary with the interaction  of racioethnicity  by AAP strength.  There  is a large effect of racioethnicity  in evaluations  of strong preferential treatment, a weaker effect in evaluations of weak preferential treatment  and opportunity  enhancement plans, and no effect in evaluations of the elimination of discrimination.

Why is the effect of racioethnicity  so consistent and substantial,  at least in reactions to the stronger AAPs? One simple explanation is that attitudes are a function of self-interest, and affirmative action has different implications  for the  interests  of the  various  groups. There is no doubt some truth  to this explanation,  but it  is  much  too  simple.  For  historical  and  cultural reasons, individuals in the various racioethnic  groups see the world in different ways, and these differences affect their affirmative action attitudes.  If we consider Blacks and Whites in the United States, for example, we find that  African Americans are more likely than Whites to be politically liberal and to identify with the Democratic party. They are more likely than Whites to have experienced  racial discrimination  and to believe that it is an ongoing problem. They are less likely than Whites to be racially prejudiced  against Blacks and to believe  that  affirmative  action  involves  preferences. Each of these  differences is associated  with attitudes toward affirmative action. Gender

Numerous  studies  have assessed gender   differences   in   attitudes   toward   affirmative action.  In general,  women  are more  supportive  than men. However, this conclusion must be qualified. First, it is not clear whether  it applies equally to all racial groups.  It  is only  among  Whites  that  gender  determines target group  membership,  so self-interest  considerations would suggest that the gender effect should be larger among Whites than among other racioethnic groups. At least one study has reported a reverse effect of gender among African Americans. Second, the author’s reading of the literature suggests that the gender  effect has  shrunk  over  the  past  20  years or  so. However, the correlation  between effect size and year of publication  has not yet been reported,  so this supposition is tentative. Education

The third demographic variable that has received the most attention is education. Some of the most significant work on education has used it as a moderator  of other predictors  in the context  of tests between competing  theories.  When the main effect of education  has been assessed, the effect has varied with respondent  race. The relation  between  education  and support  for affirmative action is usually positive among African Americans and is usually nonsignificant or negative  among  Whites.  This  latter  result  is anomalous because  education  is  also  associated  with  decreased racial prejudice and increased political liberalism, both of which  predict  support  for affirmative action.  This relationship between education and support for affirmative action among Whites merits closer investigation.

4.2.2. Opinion Variables

The two theoretical perspectives that have received the most research  attention  emphasize  the importance  of political ideology and prejudice. Distinguishing between these two approaches is complicated by the positive  correlation   between   political   conservatism and most measures of prejudice. A third theory posits that prejudice and political orientation serve as mediators between a desire for group dominance  and affirmative  action  attitudes.  Each  of  these  three perspectives is described in what follows.  Prejudice

Research indicates  that  prejudice is among the most powerful predictors  of White Americans’ attitudes  toward  affirmative action.  However, the effect of racial prejudice varies with the measure  used.  The  correlation  is larger  when  racial prejudice is assessed with a measure of contemporary racism than when it is assessed with a measure of old-fashioned  racism,  and  it  is  smaller  yet  when  it  is assessed as pure anti-Black affect. Because of a dearth of research,  comparable  distinctions  among measures of sexism cannot be made.

The distinction  between  contemporary  racism and old-fashioned  racism requires  some explanation.  The core of old-fashioned racial prejudice, which prevailed among Whites  until  the  past  few decades,  is the  assumption  of inherent  inferiority  of non-White  racial groups. Blacks, in particular, were assumed to be inferior  in  intelligence  and  other   characteristics.   This belief was accompanied  with support  for racial segregation   and   for   discrimination.    Such   beliefs  have become relatively uncommon.  Instead, many scholars argue,  such  old-fashioned  racism  has  been  replaced with contemporary  forms of racism. The author  uses the  term  ‘‘contemporary racism’’ to represent  several different but related concepts, including ambivalent racism,  aversive racism,  laissez-faire racism,  modern racism, racial resentment,  subtle racism, and symbolic racism. These various conceptualizations  differ in important  ways, but  most  incorporate  the  following components.   First,  they  assume  that  contemporary racists  have  underlying  anti-Black  feelings.  Second, they assume that  contemporary  racists do not accept old-fashioned racist beliefs in inherent  inferiority, segregation, and discrimination. Instead, contemporary racists  believe  that  African  Americans  violate American   values   such   as  individualism   and   self-reliance. Third, contemporary  racists acknowledge that racial discrimination  existed in the past but believe that it is no longer a problem. Thus, they conclude that any economic disadvantages experienced by Blacks are due to   their   cultural   inferiority.   Finally,   contemporary racists reject racism and any suggestion that their own beliefs are racist. One implication of this is that they feel most free to express their  prejudice  when some nonprejudicial reason for their action can be offered.

Given this distinction between old-fashioned racism and contemporary racism, one can understand  why the latter  predicts  opposition  to  affirmative action  more strongly  than  does the  former.  The  anti-Black affect assumed  by contemporary  racism predisposes  individuals to oppose affirmative action. The assistance that affirmative action offers to target group members decreases the importance  of self-reliance. In addition, affirmative action is designed to eliminate the effects of discrimination  that  contemporary  racists  do  not  believe exists. Finally, because they can justify their opposition  to affirmative action by arguing about the importance  of self-reliance  and  the  lack  of ongoing discrimination,    contemporary    racists   feel   free   to oppose  it.  On  the  other  hand,  old-fashioned  racists who genuinely believe that Blacks are genetically inferior  may conclude  that  African Americans  need  the assistance offered by affirmative action and so lessen their opposition.  In addition, the ideological argument for individualism  that is at the center  of the contemporary approaches, and that is a core reason for opposition    to   affirmative   action,    is   not    central    to old-fashioned racism. When both old-fashioned racism and symbolic racism are included  as predictors  of affirmative  action  attitudes,   the  latter  is  appreciably more important and typically explains incremental variance beyond the former.

One  problem  with  the  contemporary   racism  approach  is that  it overlaps conceptually  and (in  some cases) empirically  with  political  conservatism.  Thus, an  individual  who  espouses  a conservative  ideology but   genuinely   lacks  any  prejudice   toward   Blacks would score relatively high on most measures of contemporary racism.  Political  Orientation

Beliefs associated with respondents’ political orientation consistently predict  attitudes  toward  affirmative  action. Specifically, numerous studies have found that support for affirmative action is greater among Democrats than among Republicans and is greater among liberals than among conservatives.

From  the  beginning  of the  modern  civil rights  era during  the  1950s,  the  major  political  parties  in  the United States have taken opposing  positions  on many racial issues. This divergence was solidified in the 1964 presidential race, when the Republican party nominated Barry Goldwater to run against President Lyndon Johnson.  The  civil rights  movement  played  a central role in that  election,  and this established  a pattern  of Republican opposition  to, and Democratic support  for, government policies designed to assist racioethnic minorities. Indeed, in recent national elections, opposition to affirmative action, in particular,  has been a formal or informal plank in the Republican party platform. Republican   politicians   typically  refer  to  affirmative action as ‘‘preferential treatment’’ or ‘‘quotas’’ and argue for its elimination.  With this history, it is no surprise that party identification predicts public attitudes toward affirmative action.

Political conservatism is an even stronger predictor of affirmative action attitudes  than is political party identification. Political conservatism includes several beliefs that  engender  opposition  to  affirmative action.  First, conservatism posits that government  should not intervene in the marketplace. Affirmative action is a federally mandated   policy   that   constrains   business   actions. Second,  conservatism  is  associated  with  support  for self-reliance  and  individualism  in  both  its  normative and descriptive meanings. Normatively, individualism argues  that  individuals  should  take  responsibility  for their  own  success  and  should  not  rely on  assistance from others (e.g., the government).  Descriptively, individualism argues that success in the United States actually is  determined  by  individual  actions  rather  than structural forces. Affirmative action, with its underlying assumption that success is affected by structural  factors and its targeted assistance to women and minorities, is contrary  to individualism.  A related point  is that conservatives would support actions designed to ensure equality of opportunity but not those designed to ensure equality of outcomes. For these and associated reasons, conservatism is a strong predictor of opposition to affirmative action, particularly when it is construed as involving preferences.

The past decade has seen the development of a principled  conservatism  theory  of affirmative  action attitudes.  The argument  is that opposition  to affirmative action is largely a function of race-neutral conservative principles rather than of racial prejudice. These scholars do not deny the importance of prejudice, but they argue that  it is more  important  among the  less educated  than  among  the  more  educated,  who  are better  able  to  behave  in  a  manner  consistent  with their  principles.  One important  point  made by some principled  conservatism  scholars  is that  members  of the public may be less than candid in their responses to  survey questions  regarding  sensitive racial issues. When these researchers use both overt and covert measures  of anger about  affirmative action, they find the usual effect of political ideology on the overt measures, but this effect is greatly reduced or entirely eliminated  on the covert measures. The effect of political  orientation   decreases  because  liberals  express more anger on the  covert measures  than  they do on the overt measures. A related point is that the effect of prejudice  on affirmative action attitudes  was stronger among liberals than among conservatives.  Social Dominance Theory

A number  of scholars  have  developed  models  that  emphasize  the role of group  conflict in motivating  attitudes  toward affirmative action. These approaches assume that societies are organized as hierarchies, with some social groups being dominant  and others being subordinate. To maintain  its position,  the dominant  group creates an  ideology  that  legitimizes  the  status  quo.  In  the current  case, this  ideology incorporates  meritocracy, political conservatism, classical racism, and a belief in the cultural inferiority of racial minority groups. These beliefs serve as legitimizing myths; they legitimize the differences in status between the dominant  and subordinate groups. They also legitimize opposition to affirmative  action,  which  threatens  the  status  hierarchy. This theory explains the correlation between prejudice and political conservatism, both of which serve as legitimizing myths, as well as the correlations  of each with opposition  to affirmative action.

Social dominance  theory  differs from  some  other conflict-based  theories  in its assumption  that  people vary in the extent to which they have internalized this perspective. This personality trait is called social dominance orientation.  Supporting social dominance theory, there is a consistent correlation between social dominance orientation and affirmative action attitudes.  Comparison  among Theories

There  has been  a rigorous  debate  within  the  affirmative action area among scholars who espouse the prejudice, principled conservatism, and social dominance perspectives.  The  resulting  literature  is  quite  complex  and cannot be summarized  briefly. In addition,  much of it suffers from a significant limitation, that is, the equating of affirmative action with preferential treatment. As mentioned   previously,   attitudes   toward   affirmative action  are  strongly  affected by  AAP strength.  More importantly,  there  is evidence that  the importance  of certain predictors varies with AAP strength. For example,  research  suggests  that  principled   conservatism may be especially important  in predicting reactions to strong forms of affirmative action that run counter  to the  basic  principles  of  conservatism.  On  the  other hand,  racial prejudice  seems to be most important  in predicting  opposition  to weaker forms of affirmative action that are less inconsistent with conservative principles. To the author’s knowledge, no research has assessed the relative importance  of social dominance orientation  across AAP strength.

The consistent correlation between measures of prejudice and opposition to affirmative action supports both the prejudice  and social dominance  approaches. However, it is important  to emphasize that the racism literature is complex and that the strength of the racism–attitude  correlation varies with the conceptualization and measure of racism. The correlation is most substantial in studies that employ measures of contemporary racism that often overlap with measures of political  conservatism.  A limitation  of the  prejudice approach  is that it does not explain attitudes of target group members.

Much of the research designed to competitively test these theories assesses the impact of potential moderators. For example, research has found that correlations between affirmative action attitudes  and all three predictors increase with respondent  education and political sophistication. This is fully consistent with the logic of social dominance theory, but the principled conservatism  approach  would  predict  a  negative  effect of education   on   the   correlation   between   affirmative action attitudes and racism. Also contrary to the principled conservatism approach, the correlation between racism and political conservatism increases with education. However, these results are based on overt measures  of racism  and  affirmative  action  attitudes.  As mentioned previously, the principled conservatism scholars have demonstrated  that the use of covert measures can profoundly affect results.

Some results supportive of social dominance theory were mentioned  previously. In addition,  research  has shown that the significant correlation  between racism and  conservatism  disappears  when  social dominance orientation  is controlled.  This result is consistent with the   assumption   that   racism  and   conservatism   are merely legitimizing myths that render  the dominance hierarchy acceptable. Other supportive research has involved  the   use  of  structural   equation   modeling analyses of the social dominance  and principled  conservatism theories.

Finally, at least one study found that all three concepts—racism, conservatism, and social dominance orientation—contributed unique variance to the prediction of affirmative action attitudes. It is not now possible to say that  one  of these  approaches  is right  and  the others are wrong; it is likely that each is at least partly correct. It is also likely that the validity of these approaches  varies with the operationalization  of affirmative action. Unfortunately, little can be said about the moderating  effect of AAP strength  because,  as noted previously, the vast majority of studies in the principled conservatism and social dominance areas have equated affirmative action with preferential treatment. Beliefs About the Extent of Discrimination

More than 20 studies have correlated beliefs about the extent of target group discrimination  with affirmative action attitudes.  The result is clear: Support for affirmative action  increases  with  the  perception  that  the target group  has suffered from discrimination.  There are at least two reasons for this relationship. First, most people  support  equality  of opportunity  for all racial groups and actions designed to ensure such opportunity. Consistent with this reasoning, the association of perceived discrimination  with support  for affirmative action appears to be most substantial for the strongest AAPs. The second  reason  for this  correlation  is that perceptions  of target  group  discrimination  are  associated with other predictors  of affirmative action attitudes. For example, political liberals generally believe that  discrimination  is still a major  problem,  whereas conservatives  tend  to  believe  that  discrimination   is now uncommon.  Along the  same lines, social dominance theory would argue that disbelief in discrimination serves as a legitimizing myth.

4.2.3. Psychological Mediators

Finally, research has focused on the psychological mediators   of  affirmative  action   attitudes.   Precisely how do AAP strength, racioethnicity,  political orientation, and other variables affect attitudes? The two answers  that  have  received  the  most  attention   are that they affect the respondent’s perceptions regarding (a) the impact of the AAP on his or her interests and (b) the fairness of the AAP.  Self-Interest  (Personal  and  Collective)

Assumptions regarding  self-interest underlie  the conflict theories,  such as social dominance  theory,  mentioned previously. Self-interest has been assessed both indirectly  and  directly.  An  example  of  an  indirect approach  would be to use respondent  age as an indicator of economic self-interest, assuming that younger workers are more strongly affected by affirmative action  policies  than  are older  workers.  The  direct  approach  asks the respondent  to rate the likely impact of the  AAP on  his  or  her  outcomes.  Each  of these approaches   is  flawed.  The  indirect   measures  suffer greatly  from  construct  contamination  and  deficiency, whereas the direct measures suffer from common method variance and experimental demands. Research using indirect  measures  of self-interest  finds that  it is only weakly related to affirmative action attitudes,  whereas work employing direct measures finds a strong relation.

Furthermore, research using direct measures finds that self-interest partly mediates the impact of other predictors such as gender and racioethnicity.

Self-interest has been studied  at both  the  personal and collective levels. Personal self-interest refers to the impact of the AAP on the respondent’s own outcomes. Collective self-interest refers to the impact of the AAP on the respondent’s demographic group, which is typically defined  in  terms  of racioethnicity  and  perhaps gender. The few studies that have included  both predictors  have found  that  they correlate  similarly with affirmative action attitudes. Perceived Fairness

Attitudes toward affirmative action are more closely related to perceived fairness  of  affirmative  action  than  they  are  to  any other variable. Indeed, research on evaluative reactions to affirmative action has often used measures  of fairness rather  than  attitudes  as the  dependent  variable. People report positive attitudes toward AAPs they consider  fair and  report  negative attitudes  toward  AAPs they consider unfair.

Fairness perceptions mediate many of the effects described previously. Strong AAPs that provide rewards on the basis of demographic characteristics rather than merit are disliked because they are seen as unfair. That is a central  argument  of the  principled  conservatism perspective. Justifications affect attitudes in part because they affect the respondent’s beliefs about the fairness of the  AAP. In  short,  support  for  affirmative  action  is strongly predicted by beliefs about its fairness.

4.3.  Complications and Limitations

As detailed previously, attitudes toward affirmative action are affected by actual or perceived details of the AAP, by the target group, and by justification  for the plan. Attitudes are also associated with various demographic factors and  opinion  variables. Many of these effects are mediated by judgments of AAP fairness and self-interest implications. Most scholars would acknowledge the  importance  of all these predictors.  There  is disagreement, however, regarding their interrelationships and their relative importance. Unfortunately, most studies include relatively few predictors. In these cases, one cannot know how important  the predictors would be in the context of additional variables. A related issue is that many of the most important  predictors  are individual  difference  variables  that  are  intercorrelated and over which the experimenter  has no control. This exacerbates the ‘‘missing variables’’ problem  and also means that causal conclusions  cannot  be drawn from the results. Of perhaps greater significance, not enough is known about how the predictors  interact. We know that  the  relative  importance  of racioethnicity  varies with AAP strength, but because of a dearth of research, it is impossible to draw strong conclusions about other interactions. This limitation is due in part to an unfortunate  tendency  for research  to include  only a single AAP.

The  operationalizations of affirmative action  have tended  to  be  brief,  simplistic,  and  (often)  illegally strong. In many studies, affirmative action is not even defined. Instead, the stimulus is simply the term ‘‘affirmative action.’’ In such situations, responses are determined  by the  participant’s  beliefs about  what affirmative action entails, and these beliefs are rarely assessed. In addition,  this work relies on respondent self-reports. Self-reports on controversial issues may be less than candid, especially if the research procedures do  not  guarantee  anonymity.  Furthermore, much  of this research  has used only White respondents.  Even more  disturbing,  a  few studies  pool  data  across  all racioethnic   groups,  and  many  studies  pool  results across minority groups. There is evidence that the minority  groups  differ in  their  attitudes.  Of greater importance,  it appears that minority  groups may also differ in the factors that predict those attitudes.

Affirmative action is a complex social policy, and the determination    of  attitudes   toward   it   is  complex. Theories and research designs must be similarly complex  if  we  are  to  obtain  a  fuller  understanding   of attitudes  toward  affirmative  action.  There  is  ample room for progress in this area.

5. Conclusions

One common thread weaves throughout the tapestry of research  on affirmative action:  Many negative effects are associated with the use of strong preferences but not  with  weak forms of affirmative action.  Negative effects of affirmative action  on organizational  performance  should  be limited  to  situations  in  which  the organization  gives preference to underqualified  target individuals. Stigmatization of target group individuals by others  and  by the  individuals  themselves  occurs when affirmative action involves (or is believed to involve)  preferences  but  not  when  preferences  are known  to be absent. Opposition  to affirmative action and the resulting conflict exist primarily in response to preferences  and  quotas.  The good news is that  such strong  preferences  are  illegal, at  least in  the  United States. The bad news is that  a substantial  segment of the population  equates affirmative action with preferences.  A major  challenge  for  the  future  is  to  craft affirmative action procedures that are widely perceived to be fair and legitimate and that improve employment opportunities  for traditionally underrepresented groups. Furthermore, the nonpreferential  nature of these plans must be publicized, and the tendency to equate affirmative action  with  preferences  should  be eliminated  in public  discourse  and academic research.  Such actions should decrease social conflict while leading to a more equitable society.


  1. Bobocel, D. R., Son Hing, L. S., Davey, L. M., Stanley, D. J., & Zanna, P. (1998).  Justice-based  opposition  to  social policies: Is it genuine?  Journal of Personality and Social Psychology, 75, 653–669.
  2. Crosby,   J. (2004).  Affirmative action is  dead; long live affirmative action.  New Haven, CT: Yale University Press.
  3. Crosby, F. J., & VanDeVeer, C. (2000). Sex, race, and merit: Debating affirmative action in education and employment.Ann Arbor: University of Michigan Press.
  4. Donohue, J. J., III, & Heckman, J. (1991). Continuous versus episodic change: The impact of civil rights policy on the economic status of Blacks. Journal of Economic Literature, 29, 1603–1643.
  5. Doverspike, , Taylor,  M. A., &  Arthur,  W.,  Jr. (2000). Affirmative action: A psychological perspective. Huntington,  NY: Nova Science Publishers.
  6. Federico, M., &  Sidanius, J. (2002).  Sophistication  and the antecedents of Whites’ racial policy attitudes: Racism, ideology, and affirmative action in America. Public Opinion Quarterly, 66, 145–176.
  7. Heilman, E., Battle, W. S., Keller, C. E., &  Lee, R. A. (1998). Type of affirmative action policy: A determinant  of reactions  to  sex-based  preferential  selection?  Journal of Applied Psychology, 83, 190–205.
  8. Holzer, H. J., & Neumark, (2000).  Assessing affirmative action. Journal of Economic Literature, 38, 483–568.
  9. Holzer, H. J., & Neumark, D. (2000). What does affirmative action do?  Industrial  and  Labor  Relations Review, 53, 240–271.
  10. Kravitz, D. A., Harrison, D. A., Turner, E., Levine, E. L., Chaves, W., Brannick, M. T., Denning, D. L., Russell, C. J.,
  11. & Conard,  A. (1997).  Affirmative action: A review of psychological and behavioral research. Bowling Green, OH: Society for Industrial and Organizational Psychology.
  12. Kravitz, D. A., & Klineberg, S. L. (2000). Reactions to two versions of affirmative action among Whites, Blacks, and Hispanics. Journal of Applied Psychology, 85, 597–611.
  13. Major, B., Feinstein, , & Crocker,  J. (1994).  Attributional ambiguity  of affirmative action.  Basic and Applied Social Psychology, 15, 113–141.
  14. Rubio, P. F. (2001). A history of affirmative action, 1619–2000.  Jackson: University Press of Mississippi.
  15. Sears, D. O., Van Laar, C., Carrillo, , &  Kosterman,  R. (1997).   Is   it   really   racism?   The   origins   of   White Americans’ opposition   to  race-targeted   policies.  Public Opinion Quarterly, 61, 16–53.
  16. Selmi, M. (1995). Testing for equality: Merit, efficiency, and the affirmative  action    UCLA Law  Review, 42, 1251–1314.
  17. Sniderman, M., & Carmines,  E. G. (1997).  Reaching beyond race.  Cambridge, MA: Harvard University Press.
  18. Spann, G. A. (2000). The law of affirmative action: Twenty-five years of Supreme Court  decisions on race  and  New York: New York University Press.

See also:

Free research papers are not written to satisfy your specific instructions. You can use our professional writing services to order a custom research paper on any topic and get your high quality paper at affordable price.

Need a Custom Research Paper?