“And Then A Miracle Occurs:" Birth Rates and Social Security
With fertility statistics in the news, thanks to the New York Times Upshot column claiming (seemingly misleading) that birth rates aren’t in fact falling, it seemed a good time to weigh in on the Social Trustees’ assumptions for fertility – which seem, um, optimistic.
One of my favorite cartoons features a scientist at a blackboard scribbling equations, with a crucial middle step preceding the conclusion: “Then a miracle occurs.” I thought of this cartoon in considering the Social Security Trustees’ assumptions for birth rates.
Fertility is really important for Social Security’s financing: more babies equals more workers equals more taxes collected by Social Security equals a smaller long-run funding gap. Moving from the current birth rate of 1.6 children per woman to a fertility rate of 1.9 would cut Social Security’s 75-year funding gap by about 15%.
There’s just one problem: the Social Security Trustees already assume that fertility will rise from 1.6 to 1.9. You can see it in the chart below, where the break-point between experience and assumptions is very clear.
The only age group for which the fertility projections appear to follow a pre-existing upward trend is among 40- to 44-year-olds, where the increased fertility this age group has experienced since 1980 is assumed to continue at a roughly similar rate.
But for 14- to 19-year-olds, 20- to 24-year-olds and 25- to 29-year-olds, downward trends in fertility are assumed to cease. And for 30- to 34-year-olds and 35- to 39-year-olds, fertility is assumed to increase—not along some pre-existing upward trend, but in sharp discontinuities from the past several decades.
Other government agencies don’t share this optimism. The Congressional Budget Office projects that fertility will continue to decline, reaching 1.56 children per woman in 2036 and remaining roughly stable after that. The U.S. Census Bureau projects that fertility will decline to 1.62 in 2036 and 1.60 in 2056. Independent researchers Melissa Kearney and Phillip Levine, writing for the Brookings Institution, are skeptical that birth rates will rebound.
In explaining the assumption of rising fertility, SSA’s actuaries state, “the cultural and economic climate in the U.S. makes it highly unlikely that our TFR will remain below a level of 1.60 for any sustained period.”
The obvious cultural difference between the US and other low-fertility developed countries is America’s greater religiousness. And within the U.S., the link between religion and fertility is clear: among those who call religion “very important” in their lives, average fertility was 2.3 children per woman. Among those calling religion “somewhat important,” fertility averaged 2.1. And among those calling religion “not important,” fertility averaged 1.8 children per woman,
But here’s the problem, according to Gallup: “U.S. Drop in Religiosity Among the Largest Ever Measured in OECD.” In 2007, 65% of Americans agreed that “religion is an important part of your daily life,” In 2025, only 49% agreed with that statement. Among Americans aged 18 to 29, prime childbearing ages, only one-third say religion is very important.
And with affordability concerns over housing and education costs rising, I wouldn’t count on America’s economic climate to bring birth rates back up.
Is it possible that fertility will increase? Sure, and it would be great to see.
But the danger of an unsupported assumption of rising birth rates, if that assumption indeed turns out to be unsupported, is Social Security’s funding problems will be underestimated, leading to delayed and incomplete solutions. It is so, so difficult for Congress to take on an issue like Social Security. When that opportunity comes, Congress needs to fix the Social Security problem we’ve got, not the one we wish we had.
First off, one of my FAVORITE cartoons - I've used it often. Honestly, until this whole immigration thing settles out (and I don't think we'll know until after the next presidential election), it's going to be hard to be accurate on those trends. My gut says that the birth rates are slower and later - not good for Social Security funding. on the other hand, didn't COVID "help" on the benefits side?
I can't help thinking that a chunk of the problem is the insistence on a Social Insurance approach (younger generations fund the older ones) rather than an actuarial model (each generation funds its own benefits during its working lifetimes). An actuarial approach doesn't preclude progressivity. I think the triggering cliche of Social Insurance being Ponzi-like has some merit, and, as you frequently point out, it's no longer true that older population segments are more destitute. So how about DB pension-like benefits and funding into a trust invested solely in Treasuries/TIPS with benefits skewed more (as a % of pay) towards the less well-off. This seems kind of like a middle ground between individual accounts and current generational cross-subsidization? And costs would be clearer than the current confusing rolling 75-year solvency basis. And fertility is a non-issue. (Just don't ask me about transition.)
Easy for me to say...