On May 13, 2003, Frank Newport, editor in chief of The Gallup Poll, and Jack Ludwig, director of research for Gallup Poll Social Audits, presented results from the most recent "Black-White Relations in the U.S." survey in a Web summit. Following are two of the questions asked by summit participants, together with Jack Ludwig's responses.Q. Have you asked about whether there is a preference among African Americans for being called "black" or "African American?"
A. Although we did not ask that question in the most recent update of our social audit, we did ask something close to it in the 2001 survey. Specifically, we asked black respondents whether, when identifying themselves, they preferred to use "African American," "black," "something else," or whether it didn't matter. The most frequent answer, given by 41% of our 1,003 black respondents in that survey, was "it doesn't matter." Another 27% said "African American," 20% said "black," 9% said "something else," and another 3% either volunteered "nothing," said "don't know," or refused to answer.
Historically, in asking questions about race, Gallup has followed the lead of the U.S. Census bureau, but we have periodically asked questions such as the one described above. Over a period of decades, the preferred terminology has changed substantially. Compare the following question, asked in 1969: "Which term do you like MOST -- colored people, negroes, blacks, or Afro-Americans?"
The results? A plurality of 1969 respondents (38%) preferred "negroes," another 20% preferred "colored people," 19% preferred "blacks," and 10% preferred "Afro-Americans." Although we did not explicitly offer the option, 6% volunteered the response "don't care."
A note on the importance of question wording: Although comparing the 1969 and 2001 results may suggest a large increase over the course of three decades in the percentage of blacks for whom the choice of name doesn't matter, it would probably be incorrect to conclude that there has been a large decline in sensitivity to which label is used. It is more likely that neglecting to offer the "doesn't matter" option to respondents in 1969 led many respondents who would have chosen that option if it had been explicitly presented, to restrict their choice to one of the labels offered.
Q. What areas of the United States were the respondents to this survey from? And did you see a difference in perceptions between the North and South?
A. Interviews for this survey were conducted by telephone across the continental United States, using Random Digit Dial (RDD) sampling. This survey design provided all telephone households -- including those with unlisted numbers -- with an opportunity to be included. We over-sampled black respondents in order to provide reliable estimates of the perceptions and opinions of black respondents to compare with those of white respondents. One consequence of this over-sampling is an unusually large number of interviews with residents of the South, which contains -- according to the U.S. Census Bureau -- more black residents than the remaining Census-defined regions (Northeast, Midwest, and West) combined. This over-representation was corrected statistically, in order to allow appropriate projection of survey results.
It is nearly impossible to summarize the broad variety of perceptions and attitudes that we asked about in our recent survey, and our sampling was not designed to focus on regional differences. But speaking generally, we did not find important differences by region. According to our most recent survey, Southern respondents appear to hold opinions that are very close to nationwide opinions on basic issues such as the current state of race relations, hopefulness about the future of race relations, and perceptions of how well blacks are treated in the local community. This similarity between the opinions of Southerners and those living in the rest of the country also occurs on topical issues such as perceptions of the prevalence of "racial profiling" by police and opinions toward affirmative action.