RACE AND EDUCATION

 

Educational Status and Black-White Segregation in Neighborhoods and Schools*

 

Michael O. Emerson

Rice University

and

David Sikkink

Notre Dame

Running Head: Education and Segregation

Manuscript Word Count (including tables, notes and references): 7205

Manuscript Word Count (text only): 5645

Authorship is equal.

Direct all correspondence to Michael O. Emerson, Department of Sociology—MS28, Rice University, Houston, TX 77251-1892.


DRAFT: Do Not Cite Without the Author's Permission

 


Educational Status and Black-White Segregation in Neighborhoods and Schools

 

 

Abstract

Higher levels of education are associated with more liberal attitudes on racial issues, including greater openness to residential and school integration. Does more formal education also lead to behavior that heightens residential and school integration? We argue that because education is a form of status, it does not increase the likelihood that Americans choose to live in integrated neighborhoods, or send their children to more integrated primary and secondary schools. Rather, identities formed through higher education lead to greater concern for status-enhancing strategies in residential and schooling decisions. Both the structure of status hierarchies in the United States, in which race becomes a measure of school and neighborhood status, and the structure of educational institutions, which reinforces the view that status mobility results from differences in individual ability and effort, lead us to expect that education will be associated with greater segregation. Using data from the 1996 National Household Education Survey, higher levels of education, net of other standard demographic variable and contextual effects, are associated with living in more segregated neighborhoods and sending children to more segregated schools. Based on these findings, we offer an alternative interpretation of the positive relationship between education and liberal attitudes on racial issues.

 

Educational Status and Black-White Segregation in Neighborhoods and Schools

 

Deeply implicated in individual life chances, residential and school segregation by race have been the focus of researchers and policy makers, and core areas of conflict among Americans. An important line of research inquiry attempts to understand white Americans’ attitudes toward housing and school racial segregation. One finding has been replicated so often that it is nearly a truism: education leads to liberal racial attitudes, including liberal attitudes toward integration issues (Schaefer 1996). Highly educated whites are more open to housing integration, say that they are less likely to leave a neighborhood if the minority population increases, express less discomfort with minority neighbors, and are more likely to say they would consider buying a home in a mixed-race neighborhood (e.g., Farley et al. 1994). And, increased education among whites is positively associated with more favorable attitudes toward having one's own children attend racially integrated schools (xxxxx 1997). What is more, education appears to be the single strongest predictor of such support. Though there is debate about the precise causal link between education and attitudes, there seems little doubt that increased education among whites contributes positively to racial attitudes (see Jackman 1994 and Schaeffer 1996 for exceptions).

Despite the volume of work examining connections between education and racial attitudes, we know almost nothing about education and race-related behaviors—especially regarding housing and schooling choices for children. We know that highly educated whites and blacks are about as segregated from each other as are less educated whites and blacks (Frey et al. 1994; Massey and Denton 1993). But we know little else beyond this limited finding.

At least two questions need answers: (1) Does more education among whites result in behavior that directly leads to reduced segregation in neighborhoods and in primary and secondary schools? (2) Why does education have this effect on behavior? Due to the importance of segregation for many individual and social outcomes, we must probe the mechanisms that affect the level of segregation to develop adequate public policy, especially public policy aimed at alleviating the structural effects of race. And investigations of the relationship between segregation and education must go beyond the assumption that education, through more liberal racial attitudes, leads to less segregationist behavior.

This paper develops a theory of the relationship between education and segregation. We first argue that since education constitutes status in the U.S. system of stratification, we expect a direct relationship between educational attainment and the choice of more segregated neighborhoods and schools for children. This is because individuals seek to maintain and enhance their overall status by matching their educational, neighborhood, and offspring’s school status. Moreover, because racial composition tends to determine neighborhood and school status, the process of choosing a neighborhood and school that is consistent with one’s educational status tends to increase residential and school segregation. Not only does education have this direct relationship to segregation through the structure of status hierarchies, but also processes within educational institutions mediate the relationship between educational attainment and residential and schooling choices. The second section of this paper argues that educational attainment reinforces beliefs and fosters social networks and identities that run counter to its positive effect on liberal racial attitudes.

A theoretical focus on educational practices and outcomes that oppose or mitigate the liberalizing effect of education leads us to deduce hypotheses that predict, for whites, an education-segregation relationship opposite that of the education-racial attitudes relationship. We conclude by showing how our theory and results provide a different interpretive frame for understanding the education-racial attitudes relationship.

THE STRUCTURE OF STATUS HIERARCHIES

Education is an important form of currency in systems of social stratification; it opens up avenues for social mobility (Collins 1979). Educational credentials allow one to achieve, on average, occupations with promising career ladders and greater autonomy, and higher incomes. But education is a key means to social mobility partly because it constitutes a status in and of itself (Meyer 1977). One could argue that education is one of the most important forms of cultural capital (Bourdieu 1984, 1990, 1998), and a key means of building social capital, in the U.S. system of stratification.

Because education constitutes status, it seems likely that there is a direct connection between higher educational attainment and more segregated neighborhoods and schools. As we argue in this section, since educational status is strongly associated with neighborhood and school status, and neighborhood and school status is inversely related to percent black, those with greater educational status are more likely to choose more segregated neighborhoods and schools to maintain and enhance their overall status.

We expect that educational status is positively related to what Connelly (1999: chapter 1) calls one of the key forms of status in the U.S.: neighborhood residence. Norms and expectations of higher educational status require a correspondingly high residential status. Massey and Denton (1993) argue that within all racial groups, a key indicator of socioeconomic advancement and status, and a key way to ensure that those advancements are maintained, is to move into higher status neighborhoods. A corresponding level of residential status solidifies one’s educational status. St John and Bates’ (1990) work on attitudes toward racially integrated neighborhoods point to the important associations between socioeconomic status and neighborhood status. They find that "the effect of racial composition on neighborhood evaluation is more negative for the higher socioeconomic status respondents than for others. This is probably due to these individuals being more conscious of status than others and, hence, more sensitive to the status implications of neighborhood racial composition (p. 57)." The links across status hierarchies create associations between educational status and neighborhood status.

Educational status is also enhanced by the status of the schools one’s children attends. Highly educated parents are expected to be concerned about high status schools for their children not only because they are more likely to be aware of its implications for their child's social and economic mobility, but also because they want their schooling choices to reflect their educational status. Compared to the less educated, the highly educated are more concerned to pass along forms of cultural capital to their children through schooling, and to encounter the expectation among their peers that they should know and care about the benefits of a "quality" education. As individuals achieve greater education, they confront societal expectations that neighborhood and school status should correspond with educational status.

The second aspect of social stratification affecting the relation of education, residence, and schools is that neighborhood and school status are strongly related to the presence of African Americans. It is well known that the presence of African Americans is believed by whites to reduce home values, as well as the overall status of a neighborhood (Farley et al. 1994; Harris 1999; Massey and Denton 1993). Race is also an important factor in school status hierarchies. According to Saporito and Lareau (1999:427), in their study of parents of eighth graders from a special school choice program: "white applicants’ choices were powerfully and negatively linked to the presence of black students . . .. Race overwhelmed other factors." The higher the percentage African American, the lower the status of that school in the eyes of whites. This is partly because whites—regardless of educational level—tend to use race as an indicator of a whole host of school problems, such as violence, drugs, and discipline problems. For whites, racial composition is intimately connected to the images of class and status.

A complete explanation of the relationship of education, neighborhood residence, and schooling choices for children should also explain how processes within educational institutions shape individual action regarding residence and schools. The next section shows how part of the relationship between higher education and segregated schools and neighborhoods is mediated through socialization processes within educational institutions.

PROCESSES WITHIN EDUCATIONAL INSTITUTIONS

According to popular notions, formal education—with its emphasis on rational thought, the democratic creed, human freedom, tolerance, pluralism and related values—should increase openness to racially integrated schools and neighborhoods. If a racially segregated society results from irrationality, narrow parochialism, and undemocratic thought, the argument goes, then formal education should provide the cure (cf. Bonilla-Silva 1997; Jackman and Mahu 1984; Katz 1991; Stephan and Stephan 1984). Higher education presumably offers greater exposure to other races, provides familiarity with the ‘American Creed’ (Myrdal 1944) that inhibits the expression of prejudice, and counteracts the fear of the unknown that may facilitate intolerance (Schaefer 1996).

But some researchers are skeptical that racially tolerant attitudes among individuals overcome structural forces that lead to racial segregation. In general, attitude and behavior are tenuously linked (Schuman and Johnson 1976). The association between attitudes and behaviors/outcomes (such as the level of segregation) is highly complex and typically weak (Clark 1992; Farley and Frey 1994; Jaynes and Williams 1989), and can even be contradictory (Jackman and Crane 1986; Massey and Gross 1991). Further, some scholars have argued that educational institutions legitimate forms of inequality in society. Educational practices reinforce meritocratic values and differential rewards (Bowles and Gintis 1976; Karabel and Halsey 1977), and educational structures increase the likelihood that individuals will frame academic failure as an individual rather than a structural problem (Apple 1978; Oakes 1985). As we argue below, structural forces and identity formation within educational institutions increase the likelihood that the more educated will also see status hierarchies as outcomes of individual achievement within the context of equal opportunity. By changing social networks and creating "educational identities," formal education limits the ability of educated individuals to act on their more liberal racial attitudes.

Learning What Education Means

That education is a social status contributes to the dominant way that education is granted legitimacy: education is worth pursuing—and indeed "must" be pursued—because it opens opportunities for status mobility (Apple 1986). The socializing effect of education starts early as families and students negotiate the higher education status hierarchy. Through the process of choosing a list of schools to send applications, studying for entrance examinations, and discussing the positives and negatives of different higher education options, individuals tend to acquire and reinforce the view that education is of value because it affords opportunities for social mobility. This reason for the pursuit of education leads the more highly educated student to expect that education can and should lead to status mobility in other arenas, such as neighborhoods and schools for children.

The Practice of Higher Education

Educational practices shape the frames through which students understand social success, and they shape the relationship of race to social success. Some argue that increasing education fosters attitudes that challenge the dominant ideology that opportunity is open to all, but we argue that the practice of education instills the opposite. Through participation in higher educational institutions, students acquire and reinforce a belief and interest in competitive individualism. The ethos and structure of higher educational institutions generally embody cultural guidelines that privilege competition between individuals. With few exceptions, classrooms are structured so that each person competes for academic status against all other individuals. For example, the system of grading, especially "grading on a curve," fosters individual competition. Second, because higher education tends to be stratified by social class, most of the students at a college, and almost all of the students in a graduate school, are middle class. Because of this class similarity on campus, students tend to experience higher education as a relatively level playing field. Both of these effects of educational practice are reinforced by the structure and legitimation of the system of education in the United States. The cultural framework that legitimates (and to some extent structures) the educational system is a tournament model (Turner 1964), in which the "winners" at each level are advanced to the next level to compete (supposedly) on equal footing. This cultural structure is invoked to frame the more highly educated student’s experience of educational practice as an individual competition on a level playing field. And the more time that an individual spends within institutions that operate according to a competitive individualistic logic, the greater the impact of these educational institutions on individual beliefs, values, and identities.

Social networks among students in educational institutions are highly segregated by race (Hallinan and Williams 1989). It is possible, then, that through participation in higher education one's links to more racially integrated neighborhoods and schools could be attenuated or remain weak. But even if student networks are not structured by race as one moves up the educational ladder, how education itself is structured within schools reinforces the view that individual effort is more important than race in determining educational outcomes. And this may be transposed to understand larger patterns of housing and schooling segregation.

This is because, within a context of competitive individualism, the successes and failures of any individual are more easily framed as differences in individual abilities, rather than linked to social background factors, which are highly differentiated by race and linked to economic structures, residential patterns, and so forth. In an institution of higher education, a white student is more likely to come into contact with a non-white student who is highly successful, which makes equal opportunity and outcomes based on ability make more sense. And within this institutional context, it is much easier to construct the lack of success of a minority student as resulting from a lack of ability and effort. In sum, the practice of education increases the legitimacy of status hierarchies, and thus makes it more likely that one's position in neighborhood and school status hierarchies will be seen as a natural result of individual hard work and ability.

Educational Status, Networks, and Identity Formation

A given educational status has consequences for individual action because it affects identity formation and social networks. The higher the level of education, the greater role that education plays in individual identity formation. Achieving a higher level of education involves an increasing investment—in time, effort, and money—which creates an interest in the identity and status conferred by educational institutions. This identity includes a stake in the value of educational credentials; a belief that the sacrifice in achieving education is meaningful—that education matters. Not that higher education is generally a "master" identity; identities remain multiple and conflicting (Calhoun 1995). But whoever else a person is, he or she is also a college graduate, or a Ph.D. And higher educational institutions tend to foster an educational identity through community building, intercollegiate athletic competition, alumni events, and other integrating rituals, such as initiation and graduation ceremonies (Bernstein 1975).

An educational identity, like other identities, is constructed and reinforced through social relationships (White 1992). Participation in higher educational institutions alters identities and social networks simultaneously. Because the strongest predictors of friendship networks are propinquity and similarity (e.g., similar levels of education) (Hallinan and Williams 1989), individual networks of the educated are increasingly populated with those of higher educational status. Through these connections in educational and other institutions, the highly educated are linked to friends who live in high status neighborhoods and the realtors who serve them.

As identities and networks are shaped within educational institutions, the more highly educated have access to different sources of information about neighborhoods and primary and secondary schools. Because of their position in a network of highly educated individuals, they tend to be more aware of high status neighborhoods and schools. Educational status carries with it a motive for paying attention to school and neighborhood status, creates the networks in which information about school and neighborhood status is available, and provides the knowledge of where to get information on schools and neighborhoods and how to understand and use it. The less educated tend not to have the same resources (communication sources, access to information on schools) that allow the more educated to evaluate public schools and gain access to private schools. In contrast, the social capital of the more educated creates greater awareness of status hierarchies within educational institutions and neighborhoods and knowledge of how to negotiate them.

HYPOTHESES

We have argued that the construction of education as a status-enhancing strategy counters its liberalizing effect on racial attitudes. Both the direct connection between educational and neighborhood and school status, and the effect of processes within educational institutions on identities and networks, lead us to expect that education will be related to more segregated schools and residence. We hypothesize the following: (1) Increases in education among whites will be associated with living in more segregated neighborhoods. (2) Increases in education among whites will be associated with placing children in more segregated schools. (3) In the choice of assigned public school versus other school, highly educated whites will be more sensitive to the percentage black in the surrounding area than will less educated whites.

 

 

DATA AND METHODS

The data to test the hypotheses come in part from the 1990 U.S. Census. Using the census tracts within SMSAs, we constructed a dissimilarity index for whites at each educational level. This allows us to compare whites average level of segregation from blacks by white educational level. To move to the individual level, and to test education’s effect on schooling choices, we use the 1996 National Household Education Survey (NHES). Conducted by the National Center for Education Statistics, this national telephone survey of the noninstitutionalized civilian population has the most direct and detailed data for our purposes. For our analysis of education and residence, we use the Household and Library data file, which contains 55,708 completed interviews. For our analysis of education and schooling choice, we use the Parent and Family Involvement in Education and Civic Involvement data file, which contains 20,792 completed interviews with parents having children aged 3 through 20 and in grade 12 or below. For both files, zip-code-level census data were merged with the household-level data to allow for complete hypotheses testing.

Because the NHES relies on stratification of telephone numbers and persons within the households, resulting in unequal probabilities of selection, sampling weights are needed. Further, because the data collection methodology results in the underestimation of actual standard errors by common statistical packages such as SAS or SPSS, we used the WESVAR statistical package. This package creates estimates through the use of the replicate method. This method splits the sample into groups, or replicates, based on the actual sample design of the survey, and then calculates estimates for each group by estimating replicate weights that parallel the actual sample design and estimation procedures employed in the full sample. The variation in the estimates derived from the replicate weights is then used to estimate the sampling errors of the estimates from the full sample.

 

FINDINGS

Residential Patterns

To appropriately test residential patterns, we need to know where whites live, by educational level, relative to African Americans. Based on our hypothesis, our question is this: Does the average level of segregation between blacks and whites increase as the educational category of whites increases? To answer this question, we obtained tract data for all tracts in all metropolitan areas of the United States. Using the most common measure of segregation, we calculated the average index of dissimilarity between blacks and each of five educational categories of whites.

Table 1 reports our results. We provide both the mean and median index of dissimilarity, which ranges from 0 (complete integration) to 100 (complete segregation), by the five white educational categories. We do this for all metros, and following Farley and Frey (1994), for just those metros that are at least three percent African American. In each case, the results are nearly identical. Highly educated whites are more segregated from African Americans than are less-educated whites. Further, the average segregation score rises monotonically by white educational category, producing a 10-point difference between the highest and lowest category. Given that a 4-point decline in the average index of dissimilarity between 1980 and 1990 was viewed as a positive step toward an integrated society (Farley and Frey 1994), the 10-point increase across educational categories is obviously substantial.

--Table 1 about here--

The effect of white education on black-white segregation levels may be entirely indirect. That is, though more educated whites are more segregated from African Americans, this may be due not to higher education per se, but to higher income or other factors associated with education. To assess whether there is a direct education effect, net of variables such as income, we predicted the percentage non-Hispanic black in the zip codes of white households, which we defined as any household where the oldest two people are non-Hispanic white. We controlled for other household-level factors, region, and community size. The results of this regression are in Table 2, and education is measured using the same categories as in Table 1, for the highest educated person in the household.

Among whites and on average, greater income leads to living in zip codes with a smaller percentage black, an unsurprising result. Also, among whites and on average, the greater the number of people under 18 in the household, the smaller the percentage black in the zip code. Turning to the variable of primary theoretical interest, Table 2 suggests that education does indeed have a direct effect on segregation. But it is not linear. That is, when we control for other important variables, education has a direct negative effect on the percentage black in the zip code of white households, but perhaps due to ceiling effects, the increase is smaller at higher educational levels. On average, and net of the other model variables, more education among whites leads to a declining percentage black in the zip codes of their residences.

--Table 2 about here--

We should also note that all available evidence indicates that high-status whites (whether the measure is education, income, or occupation) are no less segregated from high-status African Americans than low-status whites are segregated from low-status African Americans (Denton and Massey 1988; Farley 1977; Massey 1979; Massey and Denton 1987; 1993). And also that high status whites are far more segregated from black Americans then they are from lower status white Americans (Massey and Denton 1993). Highly educated whites, then, appear to contribute to the high levels of black-white segregation evidenced in the nation’s metropolitan areas.

Schooling Patterns and Choices for Children

As our first task, we examined whether highly educated whites are more likely to have their children in schools of 75 percent or more white (a continuous variable was not available). From the previous section, we know that highly educated whites live in more homogenous neighborhoods. To appropriately test our hypothesis, then, we must control for the percentage black in the zip code.

We report the results of our logistic regression in Table 3. As the table shows, we control for non-white representation in the zip code—the percentage black, Latino, and Asian. We also include controls for the percentage poor and the percentage owning their own homes in the zip code, plus community size and region. Even when controlling for these contextual effects, plus important household-level variables, the education effect is statistically significant and in the hypothesized direction. On average, the children of highly educated whites are more likely to be attending schools that are at least 75 percent white than are the children of less educated whites, net of other important factors.

--Table 3 about here--

To understand part of the process that produces this result, and to test our third hypothesis, we explored the influences on white parents’ choice of schools. As the percentage black in the surrounding area increases, the percentage black in public schools also increases. Those who are uncomfortable with the percentage black in schools, for whatever reason, will likely select alternative schooling, at least insofar as they are able. Thus, we would expect that as the percentage black in an area (in our case, the zip code) increases, so too would whites’ likelihood of choosing alternative schooling. Alternative schooling may mean selecting a public school other than the local, assigned public school, if that choice is available, such as under open enrollment policies or magnet schools. Or it may mean taking children out of public schools altogether, opting for private or home schooling. We expect that whites in general will be more likely to select alternative schooling for their children in areas with a higher percentage black. But according to our hypothesis, this choice pattern should be even stronger for highly educated whites.

With our dependent variable being those who selected alternative schooling of any kind, we ran logistic regressions first to explore whether white parents are more likely to select alternative schooling as the percentage black in the zip code increases (Model 1, Table 4), and then to examine the effect of education in interaction with the percentage black (Model 2, Table 4). According to Model 1 of Table 4, even when attempting to control for area and school quality—with measures such as percentage of poor families, median household income, and percentage of homes owner-occupied—the higher the percentage black in the zip code, the greater the likelihood that whites select schooling other than the assigned public school.

Model 1 of Table 4 also reveals that it is the percentage African American that matters. Percentage Asian and percentage Latino have no net effect on alternative school choice, whether tested in separate equations or, as shown here, in the same equation as percentage black. Be it race per se, or a proxy for expected school quality or other reason, it is only the proportion of African Americans that is influential in whites’ schooling choices.

--Table 4 about here--

Model 2 tests our third hypothesis. It includes an interaction term (percentage black in the zip code multiplied by the highest educational level in the household). The interaction term is positive and statistically significant. That is, net of other variables, the likelihood of choosing alternative schooling as the percentage black in the zip code rises increases as education increases. Thus, on average, the greater the education of the white parents, the greater the likelihood that they respond to a high percentage black in the surrounding area by removing their child(ren) from the assigned public school. Making the reasonable assumption that they then place their child(ren) in a school with a greater proportion of white students, we see part of the active process by which the children of highly-educated whites end up in more homogenous schools, even net of the typically smaller proportion African American in their area.

DISCUSSION

In an advanced industrial society such as the United States, education is a key means to increasing status. The practice of education heightens concern for social status-enhancing strategies, and generates legitimacy among the more educated for the status hierarchies in the surrounding society. Participation in higher education shapes individual identities, social networks, and expectations toward higher status neighborhoods and schools for children. The presence of racial minorities, especially African Americans, is associated with lower neighborhood and school status. Thus, education leads to greater segregation in schools and neighborhoods because of the interaction of a system of status hierarchies and identity formation within educational institutions, including changes in social networks and expectations.

From these theoretical considerations, we hypothesized (1) that highly education whites, compared to less educated whites, would be more residentially segregated from African Americans; (2) that highly educated whites, compared to less educated whites, would be more likely to have their children in racially homogenous schools; and (3) that in choosing schools for their children, highly educated whites would be more affected by the percentage black in the surrounding area than would less educated whites. Using data from the 1990 U.S. Census, and from the 1996 National Household Education Survey, we found support for each of these hypotheses. At the bivariate level, we found that the greater the educational level of whites, the greater their segregation from African Americans. When we turned to a multivariate analysis, we found that highly educated whites have a smaller percentage black living in their zip code than do less educated whites. In schooling for their children, highly educated whites are more likely to have their children in homogenous schools than are less educated whites, even when controlling for the smaller percentage non-white living in their zip code. Part of the reason for this outcome appears to be that highly educated whites, compared to less educated whites, are more likely to choose alternative schooling for their children as the percentage African American in the surrounding area increases.

We recognize the need to continue studying the relationship between education and segregation behavior. The definition of "surrounding area" may matter, but due to data limitations, we were forced to use zip codes. Census tracts or block groups may be more appropriate levels at which to test our hypotheses. More complete controls would also aid in understanding the process that leads to more homogenous neighborhoods and schools for educated whites. For example, measures of school quality—especially if they were measured separately from variables highly associated with racial composition—would allow a more nuanced testing of what exactly educated whites are reacting to when they remove their children from high percentage black schools. However, previous work on schooling decisions suggests that race overpowers concern for school quality (Saporito and Lareau 1999). Controls for occupation (a variable not available to us in our household-level data) would allow a more complete testing of the paths by which education leads to more segregated neighborhoods. Time series data would allow for more complete and accurate examination of the process of segregation by educational status. And tests of education’s influence on the behavior of African Americans and other racial and ethnic groups will all expand our theoretical understanding of education’s relationship with segregation.

Despite these limitations, we think the results suggest that education among whites does not have a positive effect overall on reducing residential and school segregation. Highly educated whites, like others, make decisions about schooling and residence within a particular structure of status hierarchies. These hierarchies link educational and neighborhood and school status, and negatively relate minority presence with status. Education does not free people from this structure. In fact, education increases people’s investment in status hierarchies, and their ability to successfully negotiate them. As such, education may actually heighten segregation levels.

We also think this paper suggests an alternative frame for interpreting the positive education-racial attitudes relationship so commonly found in the literature. At least three theoretical interpretations have been offered. According to the profound view (see Jackman and Mahu 1984), formal education, with its emphasis on rational thought, the democratic creed, human freedom, tolerance, and related values, bestows a more enlightened perspective to students. As such, it makes them less vulnerable to intergroup negativism.

The superficial view argues that education heightens commitment to democratic norms, but only at the abstract, generalized level. When concrete, specific race questions or behaviors are studied, education has no effect. As one result, education refines support for general principles, but does not raise support for specific policies aimed at the realization of those principles, or for specific questions that call for sacrifice of other principles.

The intergroup ideology view (Jackman 1994; Jackman and Mahu 1984) claims that groups at the top of the social structure create an interpretation of reality and a set of normative prescriptions that serve their interests. Because the educated want to see themselves as enlightened benefactors of a fair and just social system, when their views are challenged, the educated lead the way in avoiding racist language and stated prejudice. At the same time, they stress the validity and centrality of freewill individualism. Thus, they both support the principle of equality between racial groups, but oppose concrete measures to alter racial inequality.

We think our theory and findings in this paper suggest another factor that should also be considered. If prejudice is at least in part rational, then educated whites should exhibit less prejudice not only (or even primarily) out of "enlightenment," or as a creative cover-up to protect their advantage, but simply because they are more structurally free to exhibit less prejudice than less educated whites.

Well-educated whites are in less direct economic and cultural competition with African Americans. This is in part due to the socioeconomic structure of the United States. It is also due to the relationship between education and segregation that we find above. Highly educated whites are able to and do isolate themselves and their children from high concentrations of African Americans. Given that the smaller the out-group presence at the macro level, the more positive the racial attitudes (Quillen 1996; Taylor 1998), this isolation allows the white educated to be more "open" in their attitudes.

In sum, two major implications stem from this work. A rethinking of the relationships between education and prejudice is needed. A major weakness of the work to date is the sole focus on attitudes rather than also on behavior. Doing so misses whatever the education-behavior relationships may teach us about the education-attitudes relationships. In part this focus on attitudes reflects the different "definitions of racism" held by U.S. racial groups (Schaffer 1996). Whites tend to define racism as bad thoughts, an overt doctrine of racial superiority. Blacks and other racial groups tend to define racism more in terms of the results and outcomes. Although the debate over the relationship between education and racial attitudes is important, behaviors may be even more important for understanding the mechanisms of segregation. It is for this reason that theory and research must also be redirected to examine the relationship between education and behaviors. This paper is but one step in that direction.

ENDNOTES

REFERENCES

Apple, Michael W. 1978. "Ideology, Reproduction, and Educational Reform." Comparative

Educational Review 22:367-87.

______. 1986. Teachers and Texts. New York: Routledge.

Bernstein, Basil. 1975 Class, Codes, and Control. Vol. 3. London: Routledge and Kegan Paul.

Bonilla-Silva, Eduardo. 1997. "Rethinking Racism: Toward a Structural Interpretation."

American Sociological Review 62:465-80.

Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgment of Taste. Cambridge, MA: Harvard.

______. 1990. Reproduction in Education, Society, and Culture. London: Sage.

______. 1998. "The New Capital." Pp. 19-34 in Practical Reason: On the Theory of Action, by Pierre Bourdieu. Cambridge: Polity Press.

Bowles, Samuel, and Herbert Gintis. 1976. Schooling in Capitalist America: Educational Reform

and the Contradictions of Economic Life. New York: Basic Books, Inc., Publishers.

Calhoun, Craig. 1995. Critical Social Theory: Culture, History, and the Challenge of Difference.

Cambridge, MA: Blackwell.

Clark, William A.V. 1992. "Residential Preferences and Residential Choice in a Multiethnic Context." Demography 29:451-66.

Collins, Randall. 1979. The Credential Society: An Historical Sociology of Education and Stratification. New York: Academic Press.

Conley, Dalton. 1999. Being Black, Living in the Red: Race, Wealth, and Social Policy in America. Berkeley: University of California Press.

Denton, Nancy and Douglas S. Massey. 1988. "Residential Segregation of Blacks, Hispanics,

and Asians by Socioeconomic Status and Generation." Social Science Quarterly 69:797-818.

Farley, Reynolds. 1977. "Residential Segregation in Urbanized Areas of the United States in 1970: An Analysis of Social Class and Racial Differences." Demography 14:497-518.

Farley, Reynolds, and James Frey. 1994. "Changes in the Segregation of Whites From Blacks

During the 1980s: Small Steps Toward a More Integrated Society." American Sociological

Review 59:23-45.

Farley, Reynolds, Charlotte Steeh, Maria Krysan, Tara Jackson, and Keith Reeves. 1994. "Stereotypes and Segregation: Neighborhoods in the Detroit Area." American Journal of Sociology 100:750-80.

Hallinan, Maureen T., and Richard A. Williams. 1989. "Interracial Friendship Choices in Secondary Schools." American Sociological Review 54:67-78.

Harris, David. R. 1999. "’Property Values Drop When Blacks Move In, Because . . .’: Racial and

Socioeconomic Determinants of Neighborhood Desirability." American Sociological Review

64:461-79.

Jackman, Mary R. 1994. The Velvet Glove: Paternalism and Conflict in Gender, Class, and Race Relations. Berkeley: University of California Press.

Jackman, Mary R., and Marie Crane. 1986. "Some of My Best Friends are Black . . .: Interracial

Friendship and Whites’ Racial Attitudes." Public Opinion Quarterly 50:459-86.

Jackman, Mary R. and Michael J. Mahu. 1984. "Education and Intergroup Attitudes: Moral

Enlightenment, Superficial Democratic Commitment, or Ideological Refinement?" American Sociological Review 49:751-69.

Jaynes, Gerald D., and Robin M. Williams (eds.) 1989. A Common Destiny: Blacks and American Society. Washington, D.C.: National Academy Press.

Karabel, Jerome, and H. Halsey. 1977. "Educational Research: A Review and Interpretation." Pp.

1-85 in Power and Ideology in Education, edited by Jerome Karabel and H. Halsey. New York:

Oxford.

Katz, Daniel. 1991 "Gordon Allport’s The Nature of Prejudice." Political Psychology 12:25-57.

Massey, Douglas S. 1979. "Effect of Socioeconomic Factors on the Residential Segregation of

Blacks and Spanish Americans in the United States Urbanized Areas," American Sociological Review 44:1015-22.

Massey, Douglas S. and Nancy Denton. 1987. "Trends in the Residential Segregation of Blacks,

Hispanics, and Asians: 1970-1980." American Sociological Review 52:802-25.

______. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge,

MA: Harvard University Press.

Massey, Douglas S., and Andrew B. Gross. 1991. "Explaining Trends in Racial Segregation, 1970-1980." Urban Affairs Quarterly 27:13-35.

Meyer, John W. 1977. "The Effects of Education as an Institution." American Journal of

Sociology 83:55-77.

Myrdal, Gunnar. 1944. An American Dilemma. New York: Harper & Row, Publishers, Inc.

Oakes, Jeannie. 1985. Keeping Track. New Haven: Yale.

Pettigrew, Thomas. 1971. Racially Separate or Together? New York: McGraw Hill.

Quillian, Lincoln. 1996. "Group Threat and Regional Change in Attitudes toward African-Americans." American Journal of Sociology 102:816-60.

Saporito, Salvatore, and Annette Lareau. 1999. "School Selection as a Process: The Multiple Dimensions of Race in Framing Educational Choice." Social Problems 46:418-438.

Schaefer, Richard T. 1996. "Education and Prejudice: Unraveling the Relationship." The Sociological Quarterly 37:1-16.

Schuman, Howard, and Michael Johnson. 1976. "Attitudes and Behavior." Annual Review of Sociology 2:161-207.

St. John, Craig, and Nancy A. Bates. 1990. "Racial Composition and Neighborhood Evaluation." Social Science Research 1:47-61.

Stephan, Walter G. and Cookie White Stephan. 1984. "The Role of Ignorance in Intergroup

Relations." Pp. 229-55 in Groups in Contact: The Psychology of Desegregation, edited by M. Brewer and N. Miller. New York: Academic Press.

Taylor, Marylee C. 1998. "How White Attitudes Vary with the Racial Composition of the Local Population: Numbers Count." American Sociological Review 63:512-35.

Turner, Ralph. 1964. The Social Context of Ambition. San Francisco: Chandler.

U.S. Department of Education. National Center for Education Statistics. 1997. A Guide to Using Data from the National Household Education Survey, NCES 97-561, by Mary A. Collins and Kathryn Chandler. Washington, DC.

White, Harrison C. 1992. Identity and Control: A Structural Theory of Social Action. Princeton: Princeton University Press.

 

Table 1: Average Index of Dissimilarity Between Blacks and Whites, by Educational Level of Whites, 1990a

 

All Metropolitan Areas

Metros 3% or Greater Black

Educ. Level

Mean

Median

Mean

Median

< High Schl.

56

55

58

58

HS Graduate

58

58

60

62

Some College

59

60

62

63

College Grad

63

63

66

67

Post-College

65

65

68

68

N

335

335

236

236

a Index of Dissimilarity ranges from 0 (complete integration) to 100 (complete segregation).

A value over 60 is generally considered high segregation (Massey and Denton 1993).

Source: Calculated from 1990 Summary Tape Files (U.S. Bureau of the Census).

Table 2: OLS Regression of Percentage Black in the Zip Code,

White Americans, 1996

 

Beta

Standard Error

Household Level

   

Education (logged)

-.29*

(.13)

Income

-.13**

(.02)

Age, oldest adult (x10)

-.05

(.03)

Number of people < 18

-.19**

(.06)

In Post-Secondary

Schooling, Full Time

1.67**

(.45)

Region

   

Central

2.09**

(.12)

Northeast

3.19**

(.12)

South

10.52**

(.14)

Community Pop Size

-.43**

 

(.03)

Zip Code Pop Size

(x 100)

.01**

(.00)

Intercept

3.76**

 

(.38)

     

R2

.15

 

N

38,502

 

Source: National Education Household Survey, Household and Library Date File, 1996.

*p < .05 **p < .01 (2-tailed tests)

 

Table 3: Logistic Regression of 75 Percent or More in Child’s School,

White Parents, 1996

 

Beta

Std. Error

Household Level

   

Education

.16**

.03

Income

.05**

.01

Age of Oldest Adult

(x10)

.01

.04

Number of Siblings

.01

.03

Religious Attendance

-.02*

.01

     

Zip Code Level

   

% African American

-.05**

.00

% Latino

-.04**

.01

% Asian

-.05**

.02

% Poor

.03**

.01

% Owning Own Home

.01**

.00

% in Same Home,

1985

-.00

.00

Median Household

Income

.00

.00

     

Region

   

Central

.72**

.10

Northeast

.57**

.12

South

-.01

.11

     

Community Size

-.08**

.01

Intercept

1.99**

.33

     

Pseudo R2

.12

 

N

11,514

 

Source: National Education Household Survey, Parent and Family Involvement in Education and Civic Involvement Data File, 1996.

*p < .05 **p < .01 (2-tailed tests)

Table 4: Logistic Regressions of Alternative School Choicea,

White Americans, 1996

 

Model 1

Model 2

Household Level

   

Education

.23** (.03)

.20** (.04)

Income

-.01 (.01)

-.01 (.01)

Age Oldest Adult (10)

.04 (.05)

.04 (.05)

Number of Siblings

.05 (.04)

.05 (.04)

Religious Attendance

-.02 (.01)

-.02 (.01)

     

Zip Code Level

   

% African Am. (x10)

.09** (.02)

-.02 (.06)

% Latino (x10)

-.05 (.03)

-.05 (.03)

% Asian (x10)

.17 (.10)

.16 (.10)

     

% Poor (x10)

.09 (.06)

.10 (.06)

% Owning Own Home

-.01** (.00)

-.01** (.00)

% Same Home, 1985

(x10)

.06 (.03)

.06 (.03)

     

Region

   

Central

-.01 (.09)

-.01 (.09)

North

-.13 (.09)

-.13 (.09)

South

.01 (.08)

.00 (.08)

     

Community Size

.10** (.01)

.10** (.01)

     

Educ. x % Black (x10)

 

.04* (.02)

     

Intercept

-.98** (.26)

-.91** (.26)

     

Pseudo R2

.04

.04

Number

12,371

12,371

a Alternative school includes non-assigned public, private, and home schools.

Note: Numbers in parentheses are standard errors.

Source: National Education Household Survey, Parent and Family Involvement in Education and Civic Involvement Data File, 1996.

*p < .05 **p < .01 (2-tailed tests)

 

 

 

© David Sikkink 2000
The Diversity of Christian Schools | God, Politics, and Protest | Education and Black/White Segregation in Neighborhoods and Schools | Ethics in the Workplace
Vita | Current Research | Courses and Syllabi | Personal | Links to Other Resources | David Sikkink Homepage