Having kids move blacks to the left: In an earlier post, I presented numbers from the General Social Survey (GSS) that squared with Steve Sailer's thesis that low housing prices facilitate family formation among whites which then turns them into political conservatives. So my next question is: what does a large family do to black parents? According to the GSS, lots of kids moves blacks to the left. Data on the 2000 presidential election (the most recent available to the public) reveals that blacks who voted for Gore had a mean of .75 children living at home with them, while Bush supporters averaged only .38. (The numbers are low compared to those in my analysis of whites because I included older parents here to maximize sample size).

Conservative values might explain why some blacks voted for Bush, but not why they have fewer kids. Having even one child seems to cause a shift: 14% of those with no kids voted Republican, but that drops to 7% for those with at least one. Whites are a middle-income group: perhaps concern for your children leads people to see big government as an obstacle, while black parents, as a low-income group that has turned to government for answers for decades, look even more to Democrats in order to help their children.

So Steve's thesis is a race-specific one that does not seem to apply to blacks. He also argues that whites, as a group, avoid having kids until they own a nice house--the kind that is safe and enjoyable for kids, in a good neighborhood. This attitude does not seem widespread among blacks, so Democrat strategists should not draw the lesson from this analysis that cheap housing in good neighborhoods will increase the number of blacks with children, thereby attracting more black supporters. (As if that were possible--the only thing harder to find than a black Republican is me, a conservative social scientist).

This analysis reminds me of something I have seen many, many times in research studies: findings are almost never the same for blacks and whites. Social scientists--the most politically correct people in the world--routinely analyze the two groups separately. They force black and white data points to live in different neighborhoods because they know that they are so distinct from each other that they must do it.

By the way, I also looked to see what factors other than number of kids might distinguish black Republicans from Democrats. I was surprised that the two groups do not differ much in terms of income, education, IQ, occupational prestige, or religiosity. The only thing I found was gender: men were almost two times as likely as women to vote for Bush.