Back to Index of Selected Reports

Joint Center for Political and Economic Studies 1997 National Opinion Poll--Race Relations The 1997 Joint Center for Political and Economic Studies National Opinion Poll is a national survey of 1,702 adults, conducted between March 21 and April 20, 1997. The survey's questions cover a broad range of topics including race relations, politics, devolution, social policy, and youth. The following findings on race relations represent one component of the survey results. In the coming weeks, the Joint Center will release findings on the other survey topics identified above. The survey population has three components: a sample of 850 adults from the national population; a sample of 850 adults from the national African American population; and an oversample of 100 Hispanics. (The survey methodology is described in an accompanying appendix.) [Note: Tables are not included. The complete report may be ordered from the Joint Center (click here for ordering information).]

David A. Bositis Summary of findings From the perspective of public opinion, the state of race relations in the United States remains troubled. A large number of books and magazine and newspaper articles have appeared in the past two years concerning problems of race relations and what might be done to improve them. Yet black and white Americans continue to view the race question in fundamentally different ways. These differences in perception present a special challenge for President Clinton's 'race initiative'. Black and white perceptions of racial reality-while occasionally convergent-are for the most part quite different. Based upon the most frequent responses to our survey questions, the 'white reality' may be summarized as follows: Race relations in their own communities are generally perceived by whites to be excellent or good, although race relations nationally (for the most part as seen through the mass media) are perceived to be only fair-to-poor. In the view of whites, some level of discrimination against black Americans remains common, including discrimination by the police, but adequate attention is being paid to discrimination against Hispanics. The situation of black Americans is either the same as or better than it was five years ago, whites believe. Moreover, they feel that blacks who can't get ahead in this country are mostly responsible for their own condition, and preferential treatment should not be extended to black Americans. Finally, whites believe that if given a choice, blacks will vote for a candidate of their own race, whereas whites will vote for the most qualified candidates. The 'black reality' may be summarized as follows: From the black perspective, race relations in their own communities, and in the nation as a whole, are considered fair-to-poor. Blacks believe that somewhere between 'some' and 'a lot of' discrimination against black Americans is common, including widespread discrimination by the police; moreover, not enough attention is paid to discrimination against Hispanics. Blacks believe, generally, that their situation in the United States is the same as it was five years ago. There is disagreement among blacks as to whether blacks who cannot get ahead are mostly responsible for their own condition, with a majority believing they are not. There is also disagreement regarding extending preferential treatment to blacks, with those opposed representing a slight plurality. Finally, blacks believe that if given a choice, whites will vote for a candidate of their own race, whereas blacks will vote for the most qualified candidate. Black-white differences in perceptions and attitudes toward race issues are compounded by other significant differences-generational, regional, ideological, partisan, and gender-found within both the African-American and white populations. Generational differences were especially interesting in this area. Younger age cohorts, both black and white, were comparatively similar in their views on a number of items, while at the same time, they diverged considerably from the oldest age cohorts in their respective populations. This generation gap is even more interesting because blacks and whites of retirement age were two of the most divergent subgroups in their views on race relations. In other words, black and white young people share similar views, in contrast to both the black and white elderly populations, who are substantially divided on their views of race relations. While differences between the oldest and youngest are most stark, there are also significant differences among other age groups.

Survey responses The following more detailed analysis highlights key statistical findings from the survey, organized by survey question. The tables referred to, attached at the end of this report, provide comprehensive statistics for each question, organized by a variety of demographic and other respondent characteristics.

Perceptions of race relations Whites were considerably more likely than blacks and Hispanics to describe race relations in their own communities as excellent or good, and less likely to describe them as poor. While 58.7 percent of the white respondents said race relations in their own communities were excellent or good, only 44.2 percent of Hispanics, and 38.9 percent of African Americans answered that way. More than one in five blacks (22.4 percent) thought race relations in their own communities poor, while only 12.6 percent of whites answered that way. Hispanics were most likely to respond that race relations were poor ( 24.9 percent). Table 1 Among African Americans, men were more likely than women to respond that race relations were excellent or good (43.8 vs. 34.9 percent). Blacks also showed large generational differences, with black respondents in the 18-25 age cohort viewing race relations in their own communities more favorably (48.5 percent excellent or good and only 14.1 percent poor) than the 65-and-older cohort. More black respondents of retirement age thought race relations poor (31.5 percent) than excellent or good (27.1 percent). Table 1 Black respondents who identified themselves as Christian conservatives had relatively favorable opinions about race relations (48.9 percent excellent or good versus 23.2 percent poor), and yet black secular conservatives were the only subgroup, other than those over age 65, who were more likely to respond that race relations were poor than that they were excellent or good (48.6 percent vs. 23.0 percent). This is not entirely surprising since many black secular conservatives advocate black self-reliance partly as a response to the indifference of white society. Table 1 In the general population sample, generational differences were the reverse of those in the black sample. Whites over age 65 described race relations in their own communities as generally excellent or good (75.0 percent excellent or good vs. only 7.5 percent poor) while whites in the 18-25 cohort were much less sanguine in their responses (37.5 percent excellent/good vs. 25.7 percent poor). Table 1 When asked to describe race relations in the United States rather than in their own communities, respondents in both samples were less optimistic, and more likely to respond that race relations were poor rather than excellent or good. Only 10.4 percent of the black sample, 17.9 percent of Hispanics, and 19.0 percent of whites described race relations in the country as excellent or good, while 43.9 percent of black respondents, 42.9 percent of Hispanics, and 28.1 percent of whites described race relations as poor. The modal (most common) response for white respondents was 'fair', while the modal response for blacks and Hispanics was 'poor.' There were no especially noteworthy subgroup differences on this question. Table 2

Racial and ethnic discrimination There is a near universal opinion among African-Americans and Hispanics that discrimination against blacks remains common today. Among blacks sampled, 92.2 percent believe that discrimination is still very common; 86.8 percent of Hispanics believe likewise. A large majority (76.0 percent) of whites also believe that discrimination against blacks remains common today. Table 3 A preponderance of respondents in all subgroups of the black population believe this. Nevertheless more black men than women believe racial discrimination is common (82.7 vs. 72.4 percent), perhaps because black men are more frequently targets of certain types of discrimination, such as harassment by the police. A larger proportion (86.9 percent) of the younger cohorts (18-35 yrs.) of the mostly white general population believe racial discrimination is common than do those over 35 years. Also, among the general population, those living in the Northeast and Midwest (86.4 and 81.1 percent) are more likely to think racial discrimination is common than those living in the South or West (73.8 and 72.0 percent). Table 3 Self-identified liberals and moderates (84.7 and 81.4 percent) are more likely than either secular or Christian conservatives (70.0 and 74.3 percent) to think racial discrimination is common, and Democrats are more likely than Republicans to think that is true (82.9 vs. 71.9 percent). Table 3 There was, therefore, widespread agreement that discrimination against blacks remains common, but less agreement between the races about the extent of that discrimination. Approximately equal numbers of blacks surveyed thought there was either 'a lot' or 'some' discrimination against black Americans (48.8 vs. 50.2 percent). In contrast, a large majority of whites thought there was only some rather than a lot (76.2 vs. 21.1 percent), and most Hispanics thought likewise (66.3, some vs. 31.4 percent, a lot). Table 4 While the blacks surveyed agreed that discrimination was common, there was considerable subgroup variation on beliefs about the extent of that discrimination. The youngest black age cohort was similar to whites in believing that there was only some rather than a lot of discrimination against blacks (73.5 vs. 26.5 percent). The older black age cohorts, especially the baby boomers (ages 36-50) believed that there was a lot rather than just some discrimination (60.4 vs. 38.4 percent). Blacks in the Midwest (58.3 percent), self-identified black liberals and moderates (54.0 and 53.1 percent), and blacks with a college education (67.0 percent) were all more likely than other subgroups to believe that there was a lot of discrimination against blacks. Table 4 Among the general population, those living in the Northeast (37.2 percent), political liberals and moderates (29.0 and 31.0 percent), and those with household incomes of less than $15,000 (33.4 percent) were the subgroups most likely to think that there is a lot of discrimination against blacks. The subgroups of the general population least likely to believe this were those living in the West (16.1 percent), Christian conservatives (16.7 percent), Republicans (18.0 percent), those with college degrees (18.5 percent), and those with upper-middle ($35-60,000 annual) household incomes (14.5 percent). Table 4

Discrimination against Hispanics When asked about discrimination against Hispanics, similar proportions of both the black sample (68.1 percent) and the Hispanic sample (68.3 percent) indicated that there should be more attention paid to this form of discrimination. More whites disagreed (43.8 percent) than agreed (41.7 percent) with this statement. Table 5 While there were no substantial subgroup differences among the black population, there were several differences among subgroups of the general population. Those blacks least likely to agree that there was not enough attention paid to discrimination against Hispanics were those in the 65-and-older bracket (36.9 percent agree vs. 41.7 percent disagree), those living in the West (39.0 vs. 46.9 percent), self-identified conservatives both secular (31.8 vs. 60.1 percent) and Christian (38.5 vs. 47.0 percent), and Republicans (34.3 vs. 52.1 percent). Those blacks most likely to agree that this issue needs more attention were the youngest age cohort (56.4 agree vs. 40.8 disagree), self-identified liberals (62.9 vs. 21.3 percent), Democrats (57.0 vs. 29.7 percent), those with less than a high school education (59.2 vs. 33.0 percent), and the lowest income group (58.5 vs. 34.8 percent). Table 5

Police harassment When asked about one particular form of discrimination, namely police harassment and discrimination, there was widespread agreement among both blacks and Hispanics. Both groups agreed that the police were much more likely to harass and discriminate against blacks than against whites (81.4 percent of blacks and 82.8 percent of Hispanics agreed with this). A majority of whites-albeit a much smaller majority-also believed that police were much more likely to discriminate against blacks than whites (55.6 percent). Table 6 African Americans living in the Midwest (93.3 percent), political moderates (88.9 percent), those with college degrees (92.8 percent), and those with upper-middle ($35-60,000) incomes (88.4 percent) were most likely to agree that police discriminated against blacks; black Republicans were the subgroup least likely to agree (60.1 percent). Table 6 Among the general population, those subgroups who were most likely to believe that police discrimination existed include the youngest age cohorts (71.5 percent of those 18-25 and 66.4 percent of those 26-35), those living in the Northeast (69.4 percent), and self identified liberals (71.1 percent). The people least likely to believe that it existed were those of retirement age. In fact, this last subgroup was the only one where almost as many people disagreed as agreed (43.7 vs. 40.1 percent). Table 6

Have race relations improved? On the question of whether the situation of African-Americans in the United States over the past five years had improved, worsened, or stayed the same, major differences in perception appeared across races, and within subgroups of each race. Most African Americans (55.7 percent) thought the situation stayed the same (55.7 percent), with slightly more thinking things got better (24.5 percent) than got worse (17.1 percent). Approximately as many whites, on the other hand, thought things were better (42.9 percent) as thought they stayed the same (44.2 percent), with about one in eight (12.2 percent) believing things were worse. Hispanics' views were closer to those of whites, with 38.2 percent thinking the situation was better, 43.6 percent that it was the same, and 12.3 percent it was worse. Table 7 The youngest age cohort in the black population was considerably more optimistic than the oldest age cohort. Among those 18-25 years, those thinking the situation of blacks in the U.S. was better outnumbered those thinking it was worse by about three-to-one (38.0 vs. 13.7 percent). In contrast, the ratio of better to worse responses was small among blacks of retirement age (21.7 vs. 15.5 percent), and among blacks ages 50 to 64 more thought things actually got worse than better (22.9 vs. 18.2 percent). Table 7 Blacks living in the South (30.4 percent better vs. 17.4 percent worse), secular conservatives (31.4 vs. 12.0 percent), those with some college or a technical education (30.7 vs. 14.6 percent), and the lowest income group (33.4 vs. 10.2 percent) voiced the most optimistic assessments of the changing situation of black Americans. The most pessimistic were blacks living in the Northeast (19.1 percent better vs. 22.5 percent worse), those with college degrees (12.8 percent vs. 20.8 percent), and those in the highest income group (14.3 vs. 20.4 percent). Table 7 Among the mostly white general population, those viewing the situation of black Americans as most improved were people of retirement age (59.1 percent better vs. 13.6 percent worse) and secular conservatives (56.6 vs. 10.7 percent. Table 7

Individual Responsibility and Governmental Action Two questions solicited views about who is responsible for the current status of black Americans, and each revealed significant cleavages, racial and otherwise, in public perceptions and responses. Are blacks mostly responsible for their own condition? When asked if blacks who cannot get ahead in this country are mostly responsible for their own condition, among those in the black sample slightly more disagreed (50.2 percent negative) than agreed (47.6 percent positive), while majorities of whites and Hispanics agreed. Among whites, 52.6 agreed (40.7 percent disagreeing), while among Hispanics 51.4 percent agreed (45.7 percent disagreeing). Table 8 Black men disagreed with the statement that blacks are mostly responsible for their condition (52.0 percent disagree to 32.7 percent agree) considerably more often than black women (48.6 vs. 46.4 percent). Blacks age 65 or older substantially disagreed with the statement (59.5 percent disagree vs. 33.9 percent agree) while those in 'generation X' (26 35 years) were more likely to agree than disagree (57.1 percent agree vs. 42.9 percent disagree). Black Christian conservatives tended to disagree with the statement (56.6 percent disagreeing vs. 39.4 percent agreeing), in opposition to black Republicans (58.6 vs. 36.9 percent) who tended to endorse it. Table 8 Blacks with the highest levels of education also tended to disagree with this statement about responsibility. Specifically, among blacks with some college or technical school, 62.3 percent disagreed and 35.4 percent agreed, while among black college graduates, 55.0 percent disagreed and 36.0 percent disagreed. In contrast, blacks with the highest incomes tended to endorse its sentiments (57.4 percent agree vs. 42.6 percent disagree). Table 8 In the general population, the subgroups most likely to agree with the statement were those of retirement age (68.0 agree vs. 27.2 percent disagree), conservatives both secular (70.9 vs. 24.2 percent) and Christian (60.7 vs. 34.9 percent), Republicans (63.3 vs. 32.3 percent), and high school graduates (68.4 vs. 28.3 percent). Self-identified liberals (40.0 percent agree vs. 51.0 percent disagree) and moderates (42.4 vs. 49.1 percent), and those with higher levels of education (some college: 44.5 vs. 47.2 percent; college graduates: 45.9 vs. 48.4 percent) were the subgroups most in disagreement with the statement. Table 8 Should preferential treatment be accorded to blacks? The second question on responsibility is related to society's responsibility to black Americans (even without reference to the past) given the generally acknowledged prejudice and discrimination against black Americans. Respondents were asked if every possible effort should be made to improve the position of blacks and other minorities even if it means giving them preferential treatment. A small plurality of blacks responded to this question in the negative (48.8 percent said no vs. 45.3 percent yes), while an overwhelming majority of whites disagreed with it (83.0 percent no vs. 15.3 percent yes). Hispanics also responded negatively, but their views were much closer to those of the black population than those of the white population (53.8 percent disagree vs. 41.9 percent agree). Table 9 Within the black population, the youngest age cohort was the most supportive of preferential treatment (54.9 percent for it vs. 42.9 percent opposed to it) while the baby boomers were most opposed (32.3 percent for vs. 64.2 percent against). Those most opposed were blacks in the Northeast (36.9 vs. 56.3 percent) and those in the West (41.1 vs. 56.7 percent). Liberals (40.5 percent for vs. 56.5 percent against), college graduates (33.0 vs. 64.8 percent), and those in the highest income group (35.6 vs. 60.5 percent) were the most opposed to preferential treatment. Surprisingly, secular conservatives (58.4 vs. 30.6), Republicans (63.9 vs. 38.1), those without a high school degree (54.2 vs. 32.4 percent), and the lowest income group (57.3 vs. 37.7 percent) were the black subgroups most in favor of preferential treatment. Table 9 In the general population, there was much subgroup variation since such a large majority opposed preferential treatment. However, of the subgroups in the general population, those without a high school degree (46.3 percent for vs. 48.7 percent against) and those in the lowest income group (35.8 vs. 61.7 percent) were (comparatively) the most supportive of preferential treatment for black Americans and other minorities. Table 9

Cross-racial perceptions of voting How black and white Americans perceive the behavior of members of their own as well as other racial groups is an important underlying dimension of race relations. To assess these cross-racial views, two questions were asked about voting and race. One asked how black voters would vote in an election where one candidate was black and the other white; the second asked how white voters would vote in a similar election. The choices were for the candidate of their own race or the most qualified candidate. How do black voters behave? When asked how black voters would vote, a clear majority of the black sample responded that they would vote for the most qualified candidate (56.5 percent) rather than the black candidate (35.0 percent). Nevertheless, whites, by a large majority, believe that black voters would vote for the black candidate over the most qualified candidate (70.0 vs. 23.6 percent). Table 10 In the black sample, the most noteworthy subgroup differences on this question were generational. The youngest age cohort believed, by a large margin, that black voters would vote for the black candidate over the most qualified one (64.9 vs. 30.6 percent). The oldest age cohorts, by an even larger margin, believed that black voters would vote for the most qualified candidate (51-64 years: 68.1 percent most qualified vs. 16.8 percent black candidate; 65 years and older: 68.5 vs. 19.0 percent). Blacks living in the Northeast (43.6 percent black candidate vs. 52.3 percent most qualified) and black Republicans (45.4 vs. 49.6 percent) were more likely than other subgroups to believe that black voters would vote for the black candidate over the most qualified one. Table 10 In the general population, the youngest age cohort was the subgroup most likely to believe that black voters would vote as described along racial lines (80.1 vs 15.0 percent). People living in the Midwest (75.8 percent black candidate vs. 18.3 percent most qualified) were much more likely to believe this was the case than were people living in the West (51.5 vs. 38.9 percent). This view was also held more by secular conservatives (79.8 vs. 17.5 percent) and those in the highest income group (74.7 vs. 19.3 percent) than by others in the general population. Table 10 How do white voters behave? When asked how white voters would vote, a majority of the black sample believed that whites would vote for the white candidate over the most qualified one (55.3 vs. 34.5 percent). Whites, on the other hand, believed that white voters would pick the most qualified candidate over the white candidate (55.3 percent most qualified vs. 39.5 percent white candidate). Table 11 It should be noted here that Hispanics believe that both white and black voters prefer candidates of their own race-whites even more so than blacks. Hispanics said that black voters would vote along racial lines, 57.1 to 37.0 percent, and that whites would do so as well, 68.8 to 26.8 percent. Table 11 In the black sample, the subgroups most likely to believe that whites would vote along racial lines were the youngest age cohort (66.3 vs. 24.0 percent), secular conservatives (65.0 vs. 35.0 percent), self-identified independents (76.0 vs. 19.4 percent), and those with some college (66.1 vs. 25.4 percent). Black Republicans (39.9 percent white candidate vs. 50.1 percent most qualified) and high school graduates (46.2 vs. 46.2 percent) were the least likely to think this would be the case. Table 11 The single most noteworthy subgroup of the mostly white general population believing whites would vote along racial lines was people ages 18 to 25 (66.3 percent agreeing vs. 28.1 disagreeing). Those in the general population least convinced that whites would vote this way were baby boomers (32.1 percent for white candidates vs. 61.7 percent for most qualified), westerners (34.0 vs. 60.2 percent), and secular conservatives (37.1 vs. 60.8 percent). Table 11

Methodology The survey was designed and the questionnaire developed at the Joint Center for Political and Economic Studies. The overall study design consists of three groups: a national general population sample of 850, a national sample of African Americans of 850, and a national sample of 100 Hispanics. There are 80 African-American and 18 Hispanic respondents in the general population sample who are also part of the national samples of African-Americans and Hispanics. Thus, in total, there are 1,702 adults, 18 years of age or older, included in this study. The survey results are based upon telephone interviews with 1,702 adults conducted between March 21, 1996, and April 20, 1997. The field work was done by Research America) with a sample provided by Survey Sampling. Randomized procedures were used to select respondents within each household reached by telephone. After the initial call, there were at least three "call-backs" if no interview was completed. The results of this survey for both the general population sample and the black population sample should be interpreted with a statistical margin of error of ±4 percentage points. That is, one can say with 95 percent confidence that the statements made based upon the procedures employed have a random error (sampling error, random measurement error, etc.) component of ±4 percentage points. Actually, this survey like all others does not have a margin of error. The individual items in the survey have margins of error based upon their sample variance, the level of confidence desired (e.g., 95 percent), and sample size.[ The ±4 percentage points is a conservative estimate of margin of error; i.e., many items-especially those in which majorities of either sample hold similar positions-have a margin of error much smaller than ±4 percentage points. In addition to the random error component in surveys, non-random errors may be present. While this survey is based upon random digit dialing techniques that effectively deal with potential problems in telephone surveys such as unlisted numbers, new numbers, etc., non-response in telephone surveys produces a variety of known biases, and probably some unknown ones. Further, a telephone survey by definition has a population of individuals with some reasonable expectation of being reached by telephone. Such a definition, of course, eliminates certain populations; for example, most homeless people and others living in poverty who are not reachable by phone are not part of the sample population. Thus, the statements made based upon this survey are most likely not generalizable to homeless people, black or white. During the field work phase of the survey, an effort was made to maximize the use of same-race interviewers, and a majority of the interviews were conducted with same-race interviewers, i.e., black interviewers for black respondents and white interviewers for white respondents. The sample data from the overall survey are weighted in the analyses to population parameters for a variety of demographic factors. The parameters used in this weighting are from the U.S. Bureau of the Census, 1992 Current Population Survey, and prior Joint Center surveys. [ A random sample is a random subset of a population. One makes observations on suitable units of a random sample in order to make statements about the population and to estimate the error associated with such statements. A common misconception regarding surveys and associated statistical theory, is that 'population' (or population size) is a factor in margin of error. This is not true. Population does not appear in the statistical formulation for margin of error.

Joint Center for Political and Economic Studies