Chapter 21 Racial Identity and Experimental Methodology
Interest in how social group attachments translate into political attitudes and behavior motivated much of the early attention
to social identity. According to Tajfel (1981), the underlying foundation in the development of political and social beliefs was “the shared perceptions of social reality
by large numbers of people and of the conditions leading to these shared perceptions” (15), as opposed to personality and
environmental characteristics. Although Tajfel was not referencing the forces that make one's racial identity relevant to
an individual, but instead focused on the development of social identity around the prejudice toward Jews, racial identity
(at least in regard to political and social behavior research) has been among the most powerful explanations of behavior.
Despite the multitude of identities a person may possess and the events that make such identities more or less salient, social identity theory has had special insight and significance into the connection between racial identity and political behavior. In the field
of political behavior, no form of identity has received nearly the amount of attention and scrutiny as racial identity (e.g.,
group consciousness, racial consciousness, linked fate, race identification).1
At the same time, however, research on racial identity and political behavior could benefit from adherence to the conceptual
foundations of social identity theory, as well as from greater reliance on experimental research. Several problems exist in
the conceptual development and measurement of racial identity research that beg out for a reexamination of racial identity.
These problems are closely tied to a heavy reliance on survey-based research that essentially ignores the two-stage development of identity and thus overstates the importance of
racial identity in influencing political behavior. Controlling the accessibility of racial identity among a multitude of identities
a person might possess, as well as the level of identity salience, is critical to the study of racial identity. Otherwise,
racial identity may be viewed as somewhat artificial because researchers impose an identity and assert a certain level of psychological importance. Such artificiality
of identity is compounded by the fact that racial identity is almost always measured contemporaneously with other political
and social attitudes, which makes causal statements tenuous.
In the sections that follow, I review the essence of social identity theory, explore the survey-based approaches to studying
racial identity and political behavior, and then propose an experimentally based research agenda for overcoming such limitations.
1. Social Identity Theory
Social identity theory was originally developed to explain the psychological basis of intergroup discrimination. The core
idea is that people tend to simplify the world around them and that a particularly important simplifying device is the categorization
of individuals, including themselves, into groups according to their similarities and differences. Describing one's self and
others as African American, conservative, a woman, and so forth is a way in which categories are created and maintained. People
recognize at an early age their differences and similarities to others.2 Such a simple and automatic classification process is often assumed to be sufficient to produce distinctive group behavior
and prejudice. Early on, scholars recognized that the mere categorization or designation of group boundaries could provoke
discrimination; however, Turner (1978, 138–39) would later show that social categorization per se (the “minimal group” paradigm) was, by itself, not sufficient for in-group
favoritism.
Once people have categorized themselves and others into distinct groups, self-esteem is enhanced by creating favorable comparisons
of their own groups vis-à-vis other groups, thus making their own groups appear superior. Motivating individuals are the need
for self-esteem and the desire for a positive self-evaluation.
This process linking categorization and self-esteem to group attachments is simple enough, but it should be clear that not
all identities are equally accessible and important at the same time. Identities contribute to our self-concept, but, for
the most part, they need to be activated and made salient in order to be useful in political and social decision making. For
the positive distinctiveness of group identities to become politically and socially important to the individual, a mechanism
must exist for activating or making salient the psychological attachment to social categories. Information and political and
social events that increase the salience of different identities at different times abound.
However, a different set of assumptions and processes seems to characterize the role of racial identity among African Americans.
Without separating the components of social identity theory within racial identity, the primacy of a racial identity among
African Americans is assumed to be a dominant identity and, as a result, an African American racial identity is considered
more easily activated and sustained than other identities. This might or might not be true, but it is a testable proposition,
just the same. The safest assumption, and one that should guide methodological approach in this area, is that racial identity
is highly variable and highly contextual among African Americans. A person might think of him- or herself as African American
and receive positive self-esteem from such a racial identity, but it is important to recognize that the African American identity
competes with other identities. It might come as a surprise to some, but African Americans might also think of themselves
as Americans, parents, teachers, middle class, and so on.
2. Racial Identity in Political Behavior Research
Most treatments of racial identity in political behavior research, such as with racial consciousness, group consciousness, linked fate, and racial identification, seem to focus on a contrived or artificial identity and fail
to sufficiently capture the esteem that comes from preferring one group over another or to account for how one goes from identification
to an embodiment of the group.3 The various elements of social identity theory exist independently of each other in the literature. Although racial or group
consciousness could be considered to capture the salience of racial identity, linked fate, common fate, or group identity
could also capture the categorization of racial identity.4
Consider the public opinion literature that seeks to connect racial and group consciousness to political behavior. Verba and Nie (1972) recognize that racial consciousness or the “self-conscious awareness of one's group membership” among African Americans could be a potent force in political participation. The authors offer few details about the origins and
activation of group consciousness among African Americans, and their measurement of racial consciousness is far removed from
the essence of social identity theory. Using responses to public opinion survey questions, they measure black consciousness
by whether blacks voluntarily raised the issue of race in response to a series of open-ended questions asking about the presence
of any conflict within their communities or any problems they perceived in their personal lives, the community, or the nation.
Shingles (1981) subsequently repeated this measure to conclude that black consciousness is grounded in low political efficacy and political
mistrust. A problem with this approach to racial identity is that, although African American identity may be related to attributing
racial explanations, identity is not required to make such assessments.
Miller et al. (1981) define group consciousness as a “politicized awareness, or ideology, regarding the group's relative positions in society,
and a commitment to collective action aimed at realizing the group's interests” (495). This measure supposedly differs from
group identification, which “connotes a perceived self-location within a particular social stratum, along with a psychological
feeling of belonging to that particular stratum.” Group consciousness is considered a multidimensional concept integrating
group identification, polar affect (i.e., a preference for members of one's in-group and a dislike for the out-group), polar
power (i.e., dissatisfaction with the status of the in-group), and system blame (i.e., a belief that inequities in the system
are responsible for the status of the in-group). Miller et al.'s conceptualization of racial consciousness encompasses many
of the consequences of identity, but it leaves unresolved how an individual decides for him- or herself which identities are
relevant and how group identity evolves from simple attachment to consciousness (or salience). Miller et al. suggest that,
through behavior and interactions, individuals learn of the discontent of one's group position, which makes the group salient
or personally meaningful.
Other prominent attempts to assess racial identity among African Americans have been equally assertive in giving individuals
an identity. African Americans are assumed to possess a racial identity, and it is assumed to take precedence over all other
possible identities. No other identities or competing attachments are considered. For instance, Gurin, Hatchett, and Jackson (1989) initially conceive of identity as a multidimensional construct with different behavioral consequences. Based on a common
fate and an exclusivist identity, racial identity reflected an implicit affiliation with the in-group. Building off this measurement
approach, Dawson (1994) intended his “linked fate” to be a simpler construct of racial identity: as African Americans observe an attachment to other African Americans, they
also come to believe that their interests, mostly economic, are linked to the economic interests of their racial group (77).
Unfortunately, this measure seems to be driven more by available survey-based items than by an understanding of social identity, as the mechanism through which group affiliation or even shared
fate becomes salient is not explicit. As a result, it may be premature to suggest that affiliation automatically leads to
linked fate (especially along an economic dimension) and that it is always a salient evaluative consideration. Within this
same tradition, Tate (1994) equates common fate to racial identification.
Relating objective group membership to psychological attachment, Conover's (1984) concept of group identification closely mirrors Tajfel's (1981) treatment of identity: a self-awareness of one's objective membership in the group and a sense of attachment to the group.
Beginning with an awareness of their group affiliations, individuals’ salience or attachment to a group (perhaps from past
experiences or in response to political events) becomes a component of their self-concept. Group identity, then, becomes a
point of reference in organizing and interpreting information and guides how individuals process information concerning others
(Conover 1984, 763). In following Tajfel's initial conceptualization, Conover is able to show that objective group membership acting in concert
with a sense of psychological attachment produces distinctive perceptual viewpoints. This survey-based measure first determines
respondents’ objective group membership from available survey-based measures, such as class, gender, age, and race. Although
somewhat artificial, this objective measure does consider a range of identities. Then, Conover determines whether respondents
feel especially psychologically attached to the objective groups to which they presumably belong. She accomplishes this by
asking respondents which groups they feel particularly close to – people who are “most like you in their ideas, interests,
and feelings about things.” Once the respondents finish rating how close they feel to all groups, they are asked to pick the
one group to which they feel closest.
More recent research by Transue (2007) examining identity salience and superordinate identity is also instructive. Using an experiment embedded in a public opinion
survey, Transue examines the salience of multiple identities, both subgroup and superordinate, on policy preferences. Identities
are primed through the random assignment of respondents to two different question treatments: one group is asked about their
closeness to their ethnic or racial group (subgroup salience), and the other group is asked about their closeness to other
Americans (superordinate group salience). Respondents are also assigned to two different dependent variables, willingness
to improve education (superordinate treatment) and willingness to improve educational opportunities for minorities (subgroup
treatment).
It is clear from these studies that racial identity research reflects more of an afterthought than an intentionally designed
research agenda. Racial identity has not been the focus of specialized attention, but rather it has been an idea superimposed
on existing data. As is often the case in these circumstances, such an approach creates many problems. Because attitudinal
measures occur roughly at the same time in survey research, it is problematic to make causal statements about racial identity.
Measures assumed to be influenced by racial identity could actually prime racial identity. And, because survey-based approaches
require questions ahead of time, racial identity or a set of identities are usually imposed on respondents, which might or
might not be how they view themselves. In short, this imposed racial identity might well be viewed as contrived or artificial.
Survey-based approaches are often not conducive to studying racial identity. Experimentally based methodology, in contrast,
can provide the control necessary to measure racial identity properly and to make convincing causal statements.
3. Value of Experiments in the Study of Racial Identity
My argument, so far, has been that the reliance on, or dominance of, the survey research enterprise in political behavior
research has had a profound impact on the study of racial identity. Survey research is invaluable, but the approach to studying
racial identity requires more attention. I now turn to how an experimental approach can produce more valid measures of racial
identity, which would permit stronger assessment of the direction of causality.
Individuals belong to multiple groups and they possess multiple identities. In addition to racial groups, individuals may
also identify with their gender, country, schools and universities, organizations and clubs, and occupations. The possible
identities are too numerous to list, and doing so would be futile because the most important groups are those that individuals
select for themselves. It is almost impossible to determine which of the identities are important for an individual's self-concept,
but this has not prevented those who study the connection between identity and political behavior from doing so. For African
Americans, a racial identity and racial consciousness are assumed to be the most prominent identity and the identity from
which they receive the most esteem. African Americans are seen as fixating on racial identity as a consequence of their history,
culture, and perceptions of racism and discrimination. I am not suggesting that a racial identity is irrelevant to a person's
self-concept, but I am questioning the common assumption that racial group identity is always the most important. The reality
is that racial group identity is one of many identities.
Experimental methodology seems more flexible than survey research in allowing a multiple identity approach. Similar to the
salience approach, subjects can be presented with multiple identities that might conflict or be incompatible. Subjects would
then be expected to identify with their most salient and relevant identities. Because there would be a choice among social
groups, individuals would not be forced to respond to a priori social groups with whom they might not have a strong attachment.
Such an experimental feature would make it possible for subjects themselves to identify their most salient social group.
4. Experimental Opportunity
The argument that individuals should be allowed to choose the identities that they consider relevant and salient is grounded
in the political tolerance literature. Beginning with the work by Stouffer (1955), political tolerance was conceived as the willingness to extend democratic rights (i.e., being allowed to speak publicly,
teach in public schools, or publish books) to groups on the political left (i.e., suspected communists, atheists, and socialists).
As it turned out, Stouffer's measure of tolerance assumed that certain groups in American society, particularly those on the
political right, would not be extended democratic rights. By not realizing that many individuals may not find such groups
as threatening, the measure of tolerance would be contaminated by ideology. Individuals on the political right would be mistaken
as political. To correct this conceptual and measurement issue, Sullivan, Piereson, and Marcus (1979) suggested that political tolerance implies willingness to permit the expression of those ideas or interests that one opposes
or finds objectionable. Following this line of reasoning, Sullivan et al. proposed a content-controlled measure of tolerance,
whereby individuals were allowed to identify functionally equivalent unpopular groups they opposed. Operationally, individuals
in a public opinion interview were provided with a list of groups on both sides of the political spectrum (i.e., atheists,
proabortionists, Ku Klux Klan members) from which they were to select the groups that they liked the least. After the selection
of their functionally equivalent groups, individuals were then presented with a series of statements about a range of democratic
activities in which members of that group might participate.
The take-away from Sullivan et al.'s (1979) content-controlled measure is that functionally equivalent groups are important considerations in comparing how individuals
perceive groups. Instead of assigning or assuming an identity based on some predetermined characteristic, such as race or
gender, individuals must be allowed to choose their own identities.
An interesting question is how such an approach would work for racial identity. For starters, it would be important for individuals
to select groups with whom they closely identify (of course, without using the ambiguous term “identify”). Borrowing from
the racial identity literature (Conover 1984), individuals could be asked about the groups “they feel particularly close to and people who are most like them in their ideas, interests, and feelings about things.” Similar to the tolerance measure, which was asked about four groups that individuals like least, this identity measure
could also ask about the top four identity groups.5
Next, for each group it would be necessary to determine the identity salience or psychological attachment. Assuming that racial
identity is among the selected identities, it would be important to distinguish the salience of racial identity from the salience
of other identities. Thus, priming identities by assigning the same treatment to everyone (asking the same follow-up questions across the board)
would be problematic because each respondent would have each identity primed or made salient in the same survey context and
over a matter of seconds. Because such an approach is taxing on the individual and each identity would be primed temporarily,
this is not an ideal approach to assessing the role of identity. Actually, this approach would be worse than imposing a single
identity.6
An interesting approach would entail randomly assigning high salience and low salience primes for each identity that an individual
selects. In this way, each individual receives only one primed identity (either high or low), which can then be compared to
similar identities or compared to a similarly primed alternate identity. Such an approach would be a direct test of the salience
of racial identity over an alternate identity. Equally important, such an approach would be a direct test of racial identity
against itself and at different levels of salience.
Consider the example depicted in Figure 21.1, in which individuals are allowed to select a number of identities that they consider important (Race, Identity-1, Identity-2,
Identity-3) and a randomized assignment of salience for each identity group (High, Low).7 Individuals would be randomly selected, and only one identity per person randomly primed (although it would facilitate matters
if, across a certain number of individuals, a racial identity could be selected). It is often the case that we are interested
in examining racial identity at different levels of salience. The expectation is for high racial identity salience to be more
powerful than low racial identity salience in predicting some form of political behavior. If there were no differences between
them, then we could conclude that racial identity was unimportant. Another important test involves the extent to which racial
identity is more influential than other identities. Thus, instead of assuming that racial identity is more salient than other
identities, it could be tested empirically.
Figure 21.1. Example of Experimental Design for Racial Identity
5. Conclusion
This chapter is about how survey-based approaches can contribute to a flawed conceptualization of racial identity in political
behavior research and how experimental methodology might involve a better approach. Perhaps the most serious problem takes
the form of imposing an artificial or contrived identity. Individuals possess a multitude of identities that become more or
less salient with information and in interactions with others. Instead of seeking to capture a range of these identities among
African Americans, there has been a tendency for researchers to impose a racial identity, regardless of whether such an identity
is relevant to the individual.
Causal statements are made concerning racial identity when all attitudinal measures are measured contemporaneously. The political
and social attitudes that racial identity has been expected to influence are just as likely to determine racial identity.
The greatest value of an experimental methodology is its capacity to make stronger claims about the causal relationships of
racial identity.
References
Conover, Pamela Johnston. 1984. “The Influence of Group Identifications on Political Perception and Evaluation.” Journal of Politics 46: 760–85.
Dawson, Michael C. 1994. Behind the Mule: Race and Class in African-American Politics. Princeton, NJ: Princeton University Press.
Gurin, Patricia, Shirley Hatchett, and James S. Jackson. 1989. Hope and Independence: Blacks’ Response to Electoral and Party Politics. New York: Russell Sage Foundation.
Miller, Arthur H., Patricia Gurin, Gerald Gurin, and Oksana Malanchuk. 1981. “Group Consciousness and Political Participation.” American Journal of Political Science 25: 494–511.
Shingles, Richard D. 1981. “Black Consciousness and Political Participation: The Missing Link.” American Political Science Review 75: 76–91.
Stouffer, Samuel. 1955. Communism, Conformity, and Civil Liberties. New York: Doubleday.
Sullivan, John L., James Piereson, and George E. Marcus. 1979. “An Alternative Conceptualization of Political Tolerance: Illusory Increases 1950s–1970s.” American Political Science Review 73: 781–84.
Tajfel, Henri. 1981. Human Groups and Social Categories. Cambridge: Cambridge University Press.
Tate, Katherine. 1994. From Protest to Politics: The New Black Voters in American Elections. New York: Russell Sage Foundation.
Transue, John E. 2007. “Identity Salience, Identity Acceptance, and Racial Policy Attitudes: American National Identity as a Uniting Force.” American Journal of Political Science 51: 78–91.
Turner, John C. 1978. “Social Categorization and Social Discrimination in the Minimal Group Paradigm.” In Differentiation between Social Groups: Studies in the Social Psychology of Intergroup Relations, ed. Henri Tajfel. London: Academic Press, 235–50.
Verba, Sidney, and Norman H. Nie. 1972. Participation in America: Political Democracy and Social Equality. New York: Harper & Row.
1 Research has indeed focused on other identities, such as patriotism, nationalism, gender, and social class. But research on
these identities appears to lag behind research on racial identity. Equally important, social identity theory has not been
as readily applied to those identities.
2 Categorization leads to the formation of stereotypes to aid in the processing of information, but the positive and negative
attributions underlying discrimination occur when individuals interact with others.
3 By contrived, I mean that it is almost impossible to determine a priori the multitude of identities one may possess. But,
in the construction of survey research, the researcher has to decide which identities to measure. Thus, these two processes
seem somewhat incompatible.
4 Although one can argue that racial identity is different from racial consciousness, group identity, or linked fate, I see
those concepts as tapping different aspects of the same multicomponent of racial identity. They simply tap different dimensions
of racial identity. Whereas group identity and racial identity may be viewed as assessing the identity component of racial
identity, group consciousness and racial consciousness may be viewed as assessing the salience component.
5 Another way of measuring this first part of identity could also involve linked fate or common fate measures.
6 The likelihood of individuals selecting the same identities is very low, but this approach requires only that individuals
select a racial identity. Because the alternate identities would only be used for comparison, the actual content of those
identities is not important.
7 These identities can be any identities, as long as the individual selects them. With the exception of a racial identity, the
alternative identities do not have to be identical across individuals. The interest is only in a racial identity.