Isolating the racial component of symbolic racism

I blogged here about inferential problems in the typical symbolic racism research study, which measures symbolic racism with these items:

1. Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.

2. Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.

3. Over the past few years, blacks have gotten less than they deserve.

4. It's really a matter of some people not trying hard enough; if blacks would only try harder they could be just as well off as whites.

One inferential problem is that the above items cannot differentiate racial bias from non-racial beliefs and attitudes; for example, a respondent might oppose special favors for blacks because that respondent is racist or because that respondent opposes special favors in general.

Symbolic racism research has typically addressed this inferential problem through statistical control, tossing into a regression such variables as partisanship, self-identified ideology, or specific conservative beliefs. But the working manuscript here provided evidence that statistical control does not always isolate the racial component of symbolic racism.

The TESS proposal that I recently submitted tried to isolate the racial component of symbolic racism through a survey experiment: a randomly-selected half of white respondents would receive the traditional symbolic racism items, and the other half would receive an adjusted set of items, such as "Irish, Italians, Jewish and many other groups overcame prejudice and worked their way up. Poor whites should do the same without any special favors." The idea would be to compare support for the traditional item to support for the adjusted item to assess how much responses differ due to the target mentioned in the item (blacks or poor whites).

I did not originate this idea: Paul Sniderman and Edward Carmines reported a survey that did the same exact thing, but which used "new immigrants from Europe" in the special favors item (see p. 199 here, but the survey is reported in other publications, too). However, as far as I can tell, that survey experiment concerned only the special favors symbolic racism item; the purpose of the proposed experiment is to assess the racial component of the entire battery of symbolic racism items.

Benefits of the survey experiment measurement of symbolic racism are resources saved (no need to include control items) and stronger inference. Ideally, adjusted items could reflect the race or ethnicity of each respondent, such as poor Hispanics or poor Asians instead of poor whites.

Here is the proposal for TESS, which was rejected. Comments are welcome.