[Will Wilkinson]My friend and George Mason University economist Bryan Caplan has been working on a book on "Selfish Reasons to Have More Kids" and offers a précis of his argument in today's Wall Street Journal. I've been pretty critical of Bryan's argument in the past and I see no reason in this piece to let up.
Angus (aka Kevin Grier) at Kids Prefer Cheese I think sums up Caplan's case pretty well:
1. In happiness research, the big negative hit comes from having one child. Once you've passed that, the extra unhappiness from having more is fairly small.2. Parents have little to no influence on how their children turn out, so you can relax, not bother to read to your extra kids or make costly "investments" on their behalf.3. The more kids you have, the more grandkids you might get and everyone knows grandkids are *awesome*!
I have problems with each of these lines of arguments, but let me concentrate here on happiness research. Bryan really struggles with the fact that children tend to have a negative effect on self-reported happiness. (Most economists are dismissive of survey evidence, but, to his credit, Bryan isn't.) He tries to minimize the damage this finding does to his argument by pointing out that the negative effect is small for the first kid, and even smaller for additional kids. But it remains that if one is trying to maximize happiness, no kids appears to be the best bet and fewer is better than more.
Of course, self-reported happiness is just one dubiously reliable piece of evidence about the effect of kids on well-being. The trouble with Bryan's strategy in the WSJ essay is that he resorts to even less reliable survey evidence to support his position. He cites polls that show that people tend not to report regrets about having had kids, but that a large majority of those who have not had kids say that would choose to have them if they "had it to do over again." Now, Darwinian logic suggests that the belief that one would be better off without children will not tend to be widespread. That is, as Harvard psychologist Daniel Gilbert argues, we should expect to find conviction in the satisfactions of parenthood to be strong and all-but universal whether or not those convictions reflect the truth. So one would want check them against, say, the self-reported life satisfaction of those with and without children. Or, if one is inclined to think like an economist, one might say "talk is cheap" and check these beliefs against what people actually do.