Friday, September 7, 2012

Racism is not clever

On August 19 Alex Tabarrox over at Marginal Revolution responded to a Chris Hayes quote “It is undeniably the case that racist Americans are almost entirely in one political coalition and not the other” with a blog post entitled Racism by Political Party. He checked out data available on the General Social Survey and found that there was no difference in the extent of racism between the Democrats and Republicans. This no doubt came as a surprise to many on the left. I thought it would be interesting to extend his analysis beyond party identification. In particular I wanted to see what the Smart Vote i.e. intelligent opinion, had to say about racism. Without any controls racism varies strongly with IQ. The higher a person’s IQ the less likely they are to be racist. Every 2-3 IQ points reduces the chances of some racist view by about one percent among whites. It is negligible at IQs above 132 (Mensa level), jumping to 1in 10 at average IQs and to 1 in 3 at IQs below 70. Now IQ also varies with education and income, and maybe with political outlook, so it would be a good idea to control for these variables before we deduce that intelligent opinion per se is opposed to racism. It could be that what looks like intelligent opinion is simply the effect of more education in a system that explicitly tries to eliminate racist views. Consider the Table below. The column under each heading is the list of regression coefficients and the column to the immediate right of those are the significance levels. The first two rows of results utilized linear regression on racism variables with more than two alternatives. The third row utilizes logistic regression to predict the probability of one answer on binary racism questions. Firstly note that while racist views increase with age there has been a decline in racism across all groups over the years. Secondly Political Party identity is indeed unrelated to racism on seven of eight questions. Republicans are only more likely than Democrats to think the relative lack of achievement of blacks is due to ‘lack of will’. On the other hand, conservatism is strongly related to racism on seven of eight questions. Only on the relative lack of black achievement being due to ‘inborn disability’ do conservatives not differ from liberals. In other words conservatives tend to be racists but Republicans do not so perhaps the equating of Republican with conservative is unwarranted. Women are less racist than men on two out of eight questions (even after their greater liberalism is accounted for). There are economic theories suggesting that poor whites may be more racist because they face greater job competition from blacks however income shows no relationship to racist views – except on one question where higher earners are more likely to oppose laws against interracial marriage. Lower education and lower IQs are both independently and strongly related to racism on all but one question (and that one they are just outside of statistical significance). So education does persistently and effectively work to oppose racist attitudes but it (and all the other control variables) do not account for the strong rejection of racism by intelligent opinion. Racism is truly a stupid attitude. On a more general level it seems that intelligent opinion is opposed to restricting the freedoms and rights of members of any group.

1 comment: