Happiness researchers usually find that kids have a negative effect on happiness. By the time the result gets blogged, it's tempting to say, as Will Wilkinson does, that "children make us miserable." But how big is the estimated effect of children on happiness, anyway?
I looked at this question using the GSS, regressing happiness on marital status, job satisfaction, real income, a personality measure ("You sometimes can't help wondering whether anything is worthwhile any more."), and number of children. Children have the standard negative effect, and it's statistically significant, too. But the size is miniscule. Each child brings you down by .015 steps on a 3-point happiness scale.
In contrast, just being married gives you a boost of .286. If you take the linear model literally, that means that a married person would need 19 kids to have the expected happiness of a childless single! Of course, the linear model is pretty silly, but it does put the standard finding in perspective. The average effect of children on happiness is very tiny.
Now consider: In the real world, a very small average effect probably means that some people hate having kids, while others love the experience. So before you decide that you're too selfish to accept even a small reduction in your personal happiness, you might want to find out the best ways to beat the average. Here's my first stab at the challenge; anyone got something to add?