Thread by @lymanstoneky

So I'm going through results of a survey I ran last month and I've got a neat result that speaks to WHY birth rates undershoot preferences in ~all rich countries.

Expected hedonic costs of mismatch are asymmetrical!

Basically, I asked women their personal fertility desires (using the DHS standard question wording).

*But then* later on I also asked women to rate (by clicking stars, up to 7 stars, so a 0-7 scale) how happy they'd be if they had 0, 1, 2, 3, 4, 5, or 6+ kids.

So this tells us 1) what's the number women say they'd really like to have and 2) how happy do they think they'd be with that vs. other numbers.

In this graph, each LINE is women with a given personal fertility ideal, so women who say they want to have 2, or 3, or 5 kids, or whatever.

The X axis shows "a hypothetical outcome" phrased not as a number of kids, but the MISMATCH of that number and their ideal.

The y axis is the average star-rating given to that outcome. So each line shows, "For women with a given ideal, how happy do they think each parity would make them?"

WHat you can see is that there are trivial differences when parity and ideals match. Virtually all women expect they will be happiest with their ideal parity, and the level of happiness expected is very similar on average.

But when asked about parities HIGHER than their ideals, women report a VERY steep falloff in expected happiness.

The falloff in expected happiness from under-shooting preferences is a lot more gradual.

So this suggests that women think that the hit to their happiness from having extra kids will be bigger than the hit from having fewer kids than they want.

Whether this is true or not is not an empirically settled question, but it's absolutely plausible.

There are no published studies directly measuring how happiness changes after intended/unintended pregnancy, although I have one in edits right now that I'll be presenting at the Canadian Population Society's conference in May!

(the takeaway is super obvious but worth demonstrating: intended fertility is associated with increases in happiness, unintended fertility is not)

So this is suggesting of an asymmetrical happiness payoff to under/overshooting.

HOWEVER.... may be worth not overstating this. If we focus on the large majority of women with 1, 2, or 3 child ideals, the graph looks somewhat more symmetrical.

The falloff in happiness from overshooting is highest for women with zero-child ideals, and the falloff in happiness from undershooting is smallest for women with many-child ideals.

This may suggest simply that women with high-parity ideals simply have more flexible or less numerically-specific fertility preferences, while women with zero-ideals have more fixed ideas about what they want.

But this doesn't quite *seem* to be the case. The average standard deviation of happiness scores is actually low for both low- and high-ideals women. I could set some threshold like, "at least happiness 3" but that gets to be a lot of researcher degrees of freedom.

So this question isn't resolved. But at a minimum, it does seem like women tend to believe that undershooting isn't going to impact their happiness as much as overshooting.

This belief might be true!