Do People vote their self-interest?

George Mason's Bryan Caplan has a provocative piece in this last Sunday's Washington Post. His first point is:

1. People vote their self-interest.

In fact, there is only the tiniest correlation between income and party. The country is not divided into two camps: the poor, who vote Democrat, and the rich, who vote Republican. If you consider your own experiences, this is hardly surprising: Are your rich friends really Republicans and your poor friends Democrats?

Self-interest is also a bad predictor of views about specific issues. Yes, the elderly heavily support Social Security and Medicare, but so does almost everyone else. The old bumper sticker says, "If men could get pregnant, abortion would be a sacrament," but men are actually slightly more pro-choice than women. And so on. Pollsters have found a few exceptions where self-interest really matters, such as smoking restrictions, which smokers obviously tend to oppose. But overall, where voters stand has little to do with where they sit.

I guess that I have a simple explanation for this claim regarding abortion based upon my own research. Single men support abortion because it makes it more likely that women are willing to engage in pre-marital sex. This op-ed is based upon Bryan's book "The Myth of the Rational Voter: Why Democracies Choose Bad Policies." His point on "Voters' errors balance out" is very similar even to some research that I have done in the past.