Does Wealth Make Blacks More Conservative?
Zoltan Hajnal says no:
What implications does the growing economic divide between poor and middle class blacks have for the political arena? Traditional accounts suggest that increased economic diversity should lead to increased political division as the middle class becomes more conservative. Others maintain that race will continue to trump class because of ongoing racial inequality and widespread racial discrimination. I argue for a third alternative. I suggest that for blacks and possibly for other racial minorities increasing class status reinforces race. Class gains may increase the salience of race because economic success often means working in a predominantly white world and experiencing discrimination more regularly. I test these theories using the vote in direct democracy. I find that middle class blacks are more rather than less likely to support a liberal or black agenda. Class works differently for African Americans than for whites.
The published paper is here (gated). A final draft version is here (in MS Word). The data are from a pooled set of Los Angeles Times exit polls in California, which produces an unusually large sample of African-Americans.
If this article’s conclusions are correct, the future economic advancement of blacks will not necessarily lead them to a more conservative politics.
Comments
“The first conclusion to emerge from Table 2 is that among African Americans increasing class status does not lead to greater conservatism “
Is this a fairly typical statistical analysis in political science?
There isn’t an causal effect estimated in this paper. All this says is that currently richer blacks are on average just as conservative as poorer blacks, at least for the relevant subsample that votes.
That’s not at all the object of interest. What is instead the object of interest is the marginal effect on being conservative from increasing black wealth. This survey doesn’t come close to doing that.
Posted by: OneEyedMan | December 12, 2007 11:01 AM
OneEyedMan, this is one common form of statistical analysis — to be precise, a multivariate model of cross-sectional data (survey data, in this case). Can it definitively show causation? No. It shows patterns of association. But can we learn something from patterns of association? Yes, I think so. Hajnal shows in Table 2 that high-income blacks are more likely to support liberal positions on ballot initiatives than low-income blacks — which does shed some light on the questions in which he is interested. Of course, it would be nice to buttress this conclusion with panel data that could show the impact of changes in wealth on changes in opinion.
Posted by: John Sides | December 12, 2007 12:40 PM