How Do We Know When Fertility Is Too Low?

A briefing paper by Leslie Root, Institute of Behavioral Science, University of Colorado Boulder, together with Shelley Clark, Department of Sociology, McGill University, Jennifer Beam Dowd, Department of Population Health and Leverhulme Centre for Demographic Science, University of Oxford, Alison Gemmill, Department of Epidemiology, University of California, Los Angeles, Karen Benjamin Guzzo, Department of Sociology and Carolina Population Center, University of North Carolina at Chapel Hill, Sarah R. Hayford, Department of Sociology and Institute for Population Research, Ohio State University, Laura D. Lindberg, School of Public Health, Rutgers University, and Amanda Jean Stevenson, Department of Sociology and Institute of Behavioral Science, University of Colorado Boulder.

As birth rates have fallen both in the U.S. and globally, it’s common for the media to refer to a country’s total fertility rate (TFR) as “below replacement” if it falls below 2.1 children per woman.

Intuitively, this makes sense—to maintain the same population size, each woman in a population must have about 2 children, to replace herself and her partner. The number is a little bit above 2 to account for children who die before maturity and the fact that male births slightly outnumber female births.

The news media, policymakers, and even some demographers often take this simple number and use it to interpret contemporary birth rates. Five years ago, J.D. Vance took the stage at a conservative conference to say, “Our people aren’t having enough children to replace themselves. That should bother us.”  More recent comments from him about “childless cat ladies” as a threat to our nation highlight his—and others’—ongoing concern with declining fertility.  This type of inflammatory speech, though, doesn’t reflect demographic realities.

In reality,the relationship between the current fertility rate and population decline is not straightforward.

The TFR is a “snapshot” birth rate, summarizing patterns of birth by age in a given year. But the idea that the TFR directly leads to population decline when it falls below replacement level relies on lots of assumptions about the future. After all, women don’t literally need to replace themselves to keep the population the same size, since they can expect to live for decades after giving birth.

Going from short-term birth rates to long-term population decline requires three big assumptions, all of which are mostly wrong. First, we have to assume the TFR reflects how many babies today’s women will have over their lifetimes. Second, we have to assume that birth rates have been stable over long periods of time and will persist for generations to come. Third, we have to assume no migration into or out of a country.

The first point is a big one: the TFR is rarely a good indicator of how many babies women will eventually have, especially as more and more women postpone childbearing until their 30s or even 40s. With a 20-30 year window in which to have children, lots of things can affect the timing of births (economic conditions, finding a partner, feeling ready), without affecting the ultimate number of children women have. 

As people delay childbearing compared to previous generations, annual “snapshot” measures of fertility like the TFR are artificially lower than the total number of children women eventually have. We can see this by comparing recent trends in annual measures of fertility (the TFR) in the U.S. with the number of children women actually have by the time they hit their early 40s (see Figure 1). Because of later first births, since 2010 the TFR has been consistently lower than completed family size, which tends to be much more stable over time.

Figure 1. U.S. Total Fertility Rate and Completed Fertility

Most demographers think that as people wait longer to have children, the average number of children per woman will be slightly lower in the U.S. by the end of their childbearing years than in past generations—but not by as much as the TFR would lead us to believe.

The second common mistake is ignoring the impact of past population trends. Even as the TFR drops in the U.S., the population can’t shrink right away, because of something called population momentum. The number of Americans now in their childbearing years is unusually large, because many are the children of the Baby Boomers, another very large generation. They, in turn, will likely produce another large cohort.

Furthermore, sustained immigration, which tends to bring young adults into the country, has made the current population in their childbearing years even bigger. Because the number of young people is larger than the previous generation, it doesn’t actually matter if they are “replacing themselves”—they are having plenty of babies to replace the older adults who are dying. And indeed, the U.S. hashad a positive rate of natural increase (or growth from births outnumbering deaths) every year since vital statistics began to be reliably collected, though this is expected to become negative in 2034. Even so, the rate of U.S. population growth is projected to remain positive for the next three decades once immigrants are factored in. To the extent that migration is declining in response to the current administration’s anti-immigration stance, that’s a self-inflicted wound that only serves to hurt U.S. population size and stability.

Below replacement fertility isn’t a population death knell

That said, we should be prepared for some population decline this century. We don’t have a crystal ball, but it’s not unreasonable to plan for a future where population growth tapers off, as projections show it may. Slower or negative growth will undoubtedly have consequences, just as high growth did. But not all of them are bad. Slower growth can make economic and infrastructure planning easier. And birth rates that reflect people’s ability to choose whether, when, and how many children to have are a win for everyone. The bottom line is that we should understand that current fertility rates that fall below “replacement fertility” are not strong indicators of actual future population decline. While demographic models offer a useful framework for short term planning, the real world rarely unfolds over the long run according to tidy assumptions—and it’s the messy, changing realities we need to plan for.

FOR MORE INFORMATION, PLEASE CONTACT:

Leslie Root, Institute of Behavioral Science, University of Colorado Boulder.

leslie.j.root@colorado.edu