Ideal (Family Size) Compared to What?

I was glad to see Andrew Gelman discuss my post about the General Social Survey “ideal number of children for a family to have” question on his blog.

One thing the discussion brought up was the issue of whether you can have “a family” with zero children — which might influence how people answer the question. A while ago I noted that “start a family” only came to mean “have children” in the 1930s, but it’s pretty well established now. That’s an interesting wrinkle.

Anyway, Gelman’s post reminded me of a comment I’ve been meaning to make about the current use of “ideal” family size. In general, as I noted, this ideal should not be compared with current fertility rates to conclude that people aren’t having as many children as they want. People think a lot of things they don’t have are ideal — and the choices they actually make reflect more than their statements about ideals. (The ideal number of yachts, etc.)

Much better is research that — like this by Karen Benjamin Guzzo and Sarah Hayford — looks at intentions and desires rather than ideals. Guzzo and Hayford report that in recent decades there has been no increase in “unfilfilled” desires — which they define as people who have 0 or 1 children in their early 40s and saying they want or intend to have more.

A different problem is what to compare the “ideal” number of children to. Gallup, which has been asking ideal number of children size since the 1930s, makes what has become a common comparison: ideal versus the total fertility rate. It shows a widening gap, which many people consider troubling:

The total fertility rate is 1.6, the ideal number of children is 2.7, and the gap is growing. This seems bad! I don’t agree. It’s not just that the ideal is the wrong thing to measure for this (as someone mention in the comments on Gelman’s post, what is the ideal price for socks, $0?). It’s also that total fertility rate is the wrong thing to compare it to.

Note that Gallup incorrectly calls this the “fertility rate,” which is not the same as the total fertility rate. (Please, read this post about rates and terms!) The total fertility rate is not something you can see looking out the window, or at social media. It’s a cohort projection — a number that would come into existence if certain unrealistic assumptions hold. It’s an important measure for people with a license to practice demography, but it’s not a good measure for public discourse on fertility. I think a better referrent for the ideal family size question is completed fertility. How many children do people actually have at the end of their childbearing years? Compare the Gallup trend with these two measures, and you see the widening gap is much less dramatic with the completed fertility trend. For the logic of the comparison, this seems like the better measure to indicate how many children respondents think people are having these days.

Presumably, even with delayed fertility, the completed fertility number will decline in the coming years, and we’ll see if the ideal number of children does as well.

Either way, these comparisons are not enough to conclude that there is a problematic shortage of births in people’s lives. To assess that, you need to ask people specifcally whether they wish they had more children than they did, and why. For that question, I recommend this Pew Research Center report, which digs into the reasons people give for not having children, as well as the pros and cons of that reality. This does not offer an easy number to rest on, telling us how bad it is that birth rates have fallen. Which seems OK to me.

Note: Lyman Stone showed up in the comments of Gelman’s post to call me a liar. He has done that before. In both cases, the veracity of what I wrote can be readily checked at the links I provided. But I guess that’s the narrative he’s settled on. Sad! It’s a good reason to post this list of my commentary on his work going back to 2019, which may be of interest to people who still interact with him.

Note: Emmanuel Pont on Bluesky shared this piece: In which Tomáš Sobotka and Wolfgang Lutz make similar points about the problems with TFR in the ideal family size comparison (and other problems).

—

Previously Published on familyinequality with Creative Commons License

***