Thanks to Betsey for her thoughtful reply. I already discuss some of her main points in Selfish Reasons to Have More Kids, but this is a great chance to consider them in greater detail. Let me focus on five distinct points.
1. Betsey’s reply is extremely careful, with one exception—the title. Parents are indeed unhappier, but I doubt that Betsey wants to claim that parents are unhappy in any absolute sense. In the General Social Survey, for example, the happiness gap between the married and unmarried is vastly greater than the happiness gap between parents and non-parents—and I doubt Betsey would write a piece titled “Singles Are Unhappy. But Why?”
2. My own research confirms Betsey’s first key point: Higher-income and older parents have a smaller happiness deficit. And she is correct to claim that these are precisely the parents most likely to invest heavily in their kids. Nevertheless, we wouldn’t be so quick to conclude that over-investment is painless. Studies of momentary happiness do confirm that childcare feels like work—and Betsey is one of the very few parents I’ve encountered who suggests otherwise. The most natural interpretation of Betsey’s finding, then, is that over-parenting is positively correlated with better finances, better preparation, better coping skills, and/or other advantages that more than compensate. My prediction is that parents who combine these advantages with a less laborious parenting style will have an especially small happiness deficit – or perhaps even a happiness surplus.
3. I completely agree with Betsey that happiness is just one aspect of a good life. But this point actually strengthens my argument. If we maximize happiness alone, then potential parents who know the science might say, “Sure, parents are unhappier than they need to be. But until you show that parents are actually happier than non-parents, I’ll remain childless.” But if we maximize a weighted average of, say, happiness and kids, then all you have to do to tilt the scales in favor of child-bearing is show that kids cost less happiness than you thought.
4. Betsey is correct to point to a few well-publicized experiments finding substantial effects of early childhood intervention programs on adult outcomes. But the adoption and twin evidence is (a) much more extensive, and (b) much more relevant for parents.
5. I agree that cutting the price of kids has both substitution and income effects. But I have trouble seeing any non-circular reason why we should expect parents’ income effect to offset their substitution effect instead of compounding it. In the General Social Survey at least, family size is increasing in income after controlling for education—suggesting that elite values, not high income per se, drives the raw negative correlation between income and number of children. If you’re willing to consider more ethnographic evidence, I’ve noticed that parents who don’t want more kids are about equally likely to invoke an implicit substitution effect (“It’s too much work to have another kid”) or an implicit income effect (“I’m too tired to have any more kids”).
 Caplan, Selfish Reasons to Have More Kids, pp.31-32 points out the fact that higher-income parents have a smaller happiness deficit, but doesn’t mention the parallel finding for older parents. I was however aware of both results.
 See e.g. Caplan, pp.16-19 for a discussion of the proper interpretation of Kahneman et al. 2004. “A Survey Method for Characterizing Daily Life Experience: The Day Reconstruction Method,” Science 306, pp.1776-1780.