Another Year, Another Record-Low American Fertility Rate – Plus, A Dependency Ratio Surprise

Readers, it is but mere coincidence that the day after once again revisiting old-age dependency ratios in light of the coming insolvency of the Social Security Trust Fund, the CDC released its 2019 fertility statistics, in which, once again, the total fertility rate dropped even further: from 1.7295 in 2018 to 1.705 in 2019. In terms of the total number of births in the country, this is the lowest it’s been since 1985. In terms of the actual fertility rate, this is lower than at any point in the past; even at the prior trough in the 1970s, the lowest the rate had been was 1.738 in 1976. What’s more, although the rates and numbers of births continue to decline most dramatically for younger women, the uptick for older ages is by no means sufficient to compensate, either with respect to younger women today, or their younger selves at the start of this decline 12 years ago — 38,000 fewer births among women under 35 and only 8,000 more among women 35 and older.

It remains just as true, as it has been each of the past times I’ve written about the issue, that no one knows how far American fertility rates will sink. (Honestly, I know there are many people who not only claim that they do know the trajectory, but have the solutions — I don’t believe this at all.) Yes, in a static sort of way, there are all manner of explanations for why women might want to have relatively few children, late in life, or not at all, but no satisfactory explanations as to why 2019’s rate is worse than 2018’s, and that worse than 2017’s, and so on.

At the same time, readers will recall that I raised the question of whether the usual marker of the “old age dependency ratio” is really the right perspective to take on questions of fertility/childbearing and our country’s ability to support citizens in their old age, and, as it turns out, the United Nations, in its population forecasting, develops multiple versions of dependency ratios, including ratios with both the too-old-to-work and the too-young-to-work population combined.

Here, for example, are three different ratios side-by-side, projected out to 2040: the ratio of those under 25 or 65 and older, relative to those 25 – 64; the ratio of those under 25 or 70 and over to the rest, and the ratio of those under 20 or 70 and older to the rest of the population.

As one might expect, this is a much more level development than Tuesday’s chart of solely the old-age ratio:

especially if one considers an increasing expectation that greater numbers of those aged 65 – 70 will shift from dependents to workers over this time.

Is it reasonable to lump in children and the elderly in terms of the “cost” to society? There’s no simple answer here — obviously the medical costs for the elderly are greater, as well as the living expenses of separate households, but schooling costs and subsidized childcare are not trivial, either.

What’s more, looking at the ratio in the middle of these three, the US is suddenly no longer the outlier as it had been with its relatively higher birth rate than such historically low fertility countries as Germany and Japan:

(Small nit-pick of a note: in calculating these ratios, the UN does in fact use the current low US fertility rate rather than assuming an up-tick as with the Social Security Trustee’s report, but does assume that countries with even lower rates will have a small rebound; graphs illustrating this can be viewed at their website.)

For comparison, here is the same three-country comparison for the “traditional” old-age dependency ratio, looking at ages 15 – 64 vs. age 65 and older, again using UN projection data:

What does this all add up to? No simple answers, and no easy way to forecast, or plan for, the future. But it is worth asking how to make sense of these two types of dependents, and their future costs for the workforce.

As always, you’re invited to comment at JaneTheActuary.com!