Explaining too much

In the last few days, I have seen several stories saying that "a new study ... found that opposition to welfare ... has grown among white Americans."  The study (by Rachel Wetts and Robb Willer) didn't actually show any figures on opposition to welfare--it focused on changes in the gap between white and non-white opinions--but I can see how journalists would have that interpretation.  The abstract says:  "we find that whites’ racial resentment increased beginning in 2008, the year of Barack Obama’s successful presidential candidacy and a major economic downturn, the latter a factor previously shown to amplify racial threat effects. . . . These findings suggest that whites’ perceptions that minorities’ standing is rising can produce periods of 'welfare backlash' in which adoption of policies restricting or curtailing welfare programs is more likely."

Has opposition to welfare grown?  Here are the means for whites and blacks in a question from the General Social Survey about whether we are spending too much [3], too little [1], or about the right amount [2] on welfare.

White opinion moved in the "too much" direction after Obama's election, but so did black opinion.  Looking over a longer period, white opinion has moved up and down, and is more favorable to welfare spending than in the 1970s and most of the 1990s.  Black opinion has also gone up and down, but there seems to be a gradual shift towards "too much."

If you look more closely, it seems that the ups and downs are related to the party of the president:  people tend to say "too much" when a Democrat is president, "too little" when a Republican is.  Presumably this is because, rightly or wrongly, they perceive the government as doing more when a Democrat is in office.  This is a well-known pattern that has been documented in research on a variety of issues.  So what mattered was Obama's party, not his race (see this post for another example).

If you regress average opinions on party in power and a time trend, you get the following predicted values:

There is a clear trend towards less opposition among whites and more opposition among blacks, so the racial gap in opinions is gradually declining.  The effects of party control are almost the same among blacks and whites.  I didn't investigate systematically, but it doesn't appear that general economic conditions have any effect among either blacks or whites.

The paper proposed that whites would also regard the rising share of non-whites in the population as a threat.  Since this changes gradually from year to year, that would lead to gradually rising opposition to welfare among whites.  The actual trend is in the opposite direction from the predicted one.

The hypothesis that underlies the paper is that when "relative advantage in the racial status hierarchy" is threatened, whites turn against programs that are seen as helping minorities.  I think the general hypothesis is probably right--the other part of their paper provides pretty convincing experimental evidence that whites express more negative views on welfare when they are made to think about the prospect of America as a "majority minority" nation.  But although perceived threat may help to explain differences among people at a given point in time (e. g., between different places), when looking at historical change it is overwhelmed by the effect of a general decline of racial prejudice. 


Technical note:
1.  Wetts and Willer use a question on welfare spending from the American National Election Studies.  I used a very similar question from the General Social Survey, mostly because the question has been asked for a longer period of time and the GSS has a convenient cumulative file.
2.  My "white" category includes people who also report that they are Hispanic.  That probably accounts for some of the trend among whites.  If I were writing a paper, I would distinguish between Hispanic and non-Hispanic whites and maybe add some other controls, but since I am writing a blog post I will just say that I don't think it accounts for much of the trend.