Framing the Future of Race Relations

By David C. Wilson

In the measurement of racial attitudes, opinions, and beliefs, survey researchers face a complex trade-off in which they must choose either to track historical questions or seek more reflective measures of contemporary opinions. The general rule of thumb is to select items that work. In the United States, however, where race is an especially sensitive topic, responses may be influenced by social desirability and survey context.

Context effects can take many forms, including those due to question order, content, and wording. Many consider such effects to be forms of bias or sources of error, yet these “artifacts” of surveys can serve as useful indicators of social relations of a broader nature. For instance different question wordings may elicit different considerations that affect how respondents judge racial topics. Researchers such as Howard Schuman and Stanley Presser have shown that Americans are more willing to “not allow” speeches than they are to “forbid” them. These response differences across variations in the tone of wording for similar items likely highlight the extent to which Americans are sensitive to the importance of civil liberties. Other studies show that the public expresses greater support for affirmative action for women than for racial minorities, indicating that racial and gendered contexts likely bring about different substantive considerations about who deserves what.

The point is that survey questions require constant attention, and must periodically be reviewed and evaluated to determine their utility. If different wordings elicit different considerations, then researchers must understand what they are getting when they ask a question one way versus another.

Yet it’s no secret that minor changes in question wording can have large effects on survey results. As early as 1941, results from a Roper survey showed that support for free speech was higher when question tone was more restrictive (“forbid public speeches against democracy”) versus less restrictive (“allow public speeches against democracy”). Subsequent studies have shown similar results. Reported support for helping economically disadvantaged Americans, for example, can be influenced by how the groups are framed. The public expresses a greater willingness to help “poor” persons than “people on welfare.” Thus, one can easily argue that the public is attentive to the specific wording of questions, rather than the broader concepts that social scientists are measuring.

Research suggests that these effects are most likely to occur when the issue under consideration is ambiguous, and when attitudes toward the issue are less extreme. The reasoning is that because ambiguous items are less specific and more general, and extreme attitudes are more crystallized than less extreme attitudes, responses are less likely to be influenced by subtle differences in question wording. While the idea of race relations is an ambiguous concept rather than a concrete phenomenon, individual attitudes toward race are arguably crystallized in the minds of both blacks and whites. Therefore, whether or not more ambiguous racial items are influenced by question wording is still inconclusive.

To what extent, then, are beliefs about the future of black-white race relations influenced by the tone of the questions being presented? I would begin this inquiry expecting variations in the question wordings related to race relations to have a significant influence on response. I would predict that the more questions are framed in terms of race being a problem, the more considerations of race relations as a problem will arise. Alternatively, if race relations are framed positively, then negative considerations are less likely to be brought to mind. Thus, the more negative the presentation, the more negative the outlook, and vice versa.

For more than a decade, the Gallup Organization has been asking the public a question that probes its expectations about the future of black-white relations in the United States. Phrased very straightforwardly, it asks the respondent to choose between hope (optimism) and despair (pessimism) on this defining issue in the historical, political, and social life of our nation.

Like many items that address race relations, this “future” question reveals quite a bit of pessimism. Although differences were generally less than ten percentage points, for the most part, up to 2001, the majority of American adults believed race would always be a problem for the United States. After 2001, the majority believed that problems would eventually be worked out. Based on these results, one might conclude that things are headed in the right direction but remain close. This trend can be seen in Figure 1.

But to what extent is the public’s expressed optimism or pessimism due to the wording of the question? Perhaps while appropriate in the past, framing race relations as a problem that will either always exist or eventually fade away does not fit the way in which many people view the issue today. By and large, race problems will likely never go away; there is little doubt, however, that for most black Americans, relations have improved and are likely to continue to get better. Thus, an alternative approach would be to frame race relations positively, presenting optimistic and pessimistic alternatives in a tone of gradual positive improvement, rather than conclusive resolution.