Global fertility will soon fall below the replacement level
I am currently writing the third chapter on fertility for my book on demographics—see here— which focuses on on the onset of sub-replacement fertility in one country after the other since the 1970s, what's driving this shift, and whether the decline—to a large extent driven by birth postponement or so-called tempo effects—is reversing or accelerating. It is a treatise on the notion of a second demographic transition, including empirical case studies, and a discussion about whether sub-replacement fertility is something to worry about, cherish, or perhaps to approach with indifference. This is an enquiry that is defined just as much by what it excludes as what it includes. Once we dip into the multidisciplinary academic work on demographics which has emerged in the last two-to-three decades, we go from a large literature to an almost unmanageable one. I will hopefully be able to present a first draft on the chapter soon.
Global fertility is set to drop below 2.1 by 2026
In the mean time, however, one thing that as struck me during my recent work is that global fertility will soon fall below the replacement level, defined as just over two children per women, despite what the latest UN projections would have you believe. To be clear, I am talking about period fertility here—the total fertility rate in any given period—as oppose to cohort fertility. The latter measures the completed fertility rate for women throughout their fertile period, usually defined as the late teens to the mid-40s. The distinction between period and cohort fertility is important in the context of tempo effects/birth postponement driving down period fertility, but not necessarily cohort fertility. This is in contrast to the so-called quantum effect of fertility decline, which is assume to operate through a reduction in the overall number of children women have.
The charts below shows the UN's 2019 medium variant projection for five-year global period fertility, TFR, through 2100, and a similar projection in the 2022 UN dataset, with one-year periods. In the first dataset, global fertility is estimated at 2.47 in the five years ending 2020, with a projected decline to 2.42 by 2025. These numbers are well out of date with the latest UN population update in 2022. In the 2022 data, and forecasts, global fertility is set at 2.31 in 2021, with an estimated decline to 2.30 by 2025. These shifts, even if verified by subsequent estimates, are almost surely not statistically significant. But what is interesting in both datasets is that the UN continues to assume that global fertility will remain above the replacement level for several decades. In the 2019 dataset, the UN's medium variant forecast assumes that global fertility will be above replacement levels until 2065, similar to the assumption in the 2022 dataset.
Chart 01 - Global fertility rate and forecasts, UN 2019
Chart 02 - Global fertility rate and forecasts, UN 2022
In other words, UN continues to assume that the decline in global fertility rates is about to stall. This assumption flies in the face of even the simplest empirical analysis. For instance, we see clear evidence of a quicker fall in global period fertility in the two years ending 2021 than expected by the UN in 2019. In addition, from 2016 to 2021, the UN data shows that the global TFR fell by an average of 0.04 per year, quicker than the five-year averages in the preceding five years. Put differently, the fall in the global fertility is accelerating. If we extrapolate this trend , the TFR will hit the replacement level, 2.1, by 2026, a cool 49 years before schedule. I am open to suggestions why this won't happen, but relying on the UN assumption of an imminent slowdown in the pace of the decline in global birth rates seems fanciful.
The reality that global birth rates are falling quicker than official forecasts assume is also clear by taking a cursory look at fertility trends in key regions. Recent data suggest that the decline in African birth rates is accelerating, fertility in China is collapsing, and it remains rock bottom in other Asian countries such as Japan and South Korea. We see similar signs in Latin America. In the developed world, birth rates in the Anglo-Saxon countries—which have tended to have relatively stable fertility rates at or close to replacement levels—is now falling well below 2.1 children per women. And finally in Europe, fertility remains below the replacement level, with few signs of tempo-reversal pushing fertility back up towards two. In short, the number of countries in the world with fertility rates below the 2.1 replacement level is rising, fast. According to the UN, two thirds of people in the world now live in countries or regions with sub-replacement fertility.
Chart 03 - The low fertility club is growing in membership, author’s calculations
Are falling fertility rates a problem?
Pointing fingers at official long-term forecasts for fertility is a time-honoured practice in demographics research. In the seminal work on the idea of a low fertility trap, Lutz et al. (2006) ponder why UN and Eurostat forecasts, from 1999 and 2005 respectively, assume stabilisation in fertility rates, despite evidence to the contrary. Focusing on why official forecasts imply a sudden and sustained reversal in long-term trends, Lutz et al. (2006) says:
“This deviation from the conventional rules of trend analysis must have to do with strong beliefs that somehow there is a powerful force that will stop and even reverse the trend, i.e., that at the individual level, people will always want children, and that at the aggregate level, human populations would not voluntarily shrink and age to an extent that would be socially disruptive or in the very long run might even mean shrinking to insignificance. From an evolutionary perspective, these are seemingly reasonable assumptions because a species without a drive to reproduce would not have survived to this day.”
This idea of some baseline rate of reproduction below which the rate of fertility becomes evolutionarily suboptimal has long been dogging demographic research. Evolutionary literature, see for example Mulder (1998), raises the question of whether modern fertility trends are maladaptive, a question I discuss in more detail here. By contrast, some researchers have long since come to the conclusion that replacement level fertility is , in the famous words* of Demeney (1997), an "implausible end-point of the demographic transition". Lutz et al. (2006) counter the idea of some unbreakable evolutionary lower bound of fertility by noting that the advent of modern contraception has broken the link between sex and procreation, rendering the latter a simple function of changeable, and fickle, “individual preferences and culturally determined norms”. To me, Kaplan (1996) and (2002) are the best attempts to reconcile this debate.
The debate will continue**, however, mainly because the question of an inherently optimal level of fertility invariably leads to the discussion about whether the accelerating decline in global birth rates is a good thing or not. To the extent that the accelerating decline in fertility is driven by countries moving from high-to-low fertility regimes—sub-saharan Africa, and many countries in LatAm and Asia in the past two decades—almost everyone would argue that it is a good thing. This is because falling fertility in most of those cases are strongly correlated with rising economic wealth and GDP per capita. The relatively simple hypothesis here is that falling fertility is a necessary, but not sufficient, conditions for moving up the economic value chain. More recently, I guess we should add to this argument that falling birth rates in high fertility countries, and perhaps even in general, is good for the environment. This, according to some, is especially true in the context of climate change, because it reduces the human ecological foot print, or something. For the record, I am deeply sceptical of this argument, mainly because it, writ large, easily devolves into regressive Neo-Malthusianism. Which people is it exactly that are surplus to requirements, and what are we supposed to do with them?!
On the flip side, falling fertility in countries with already low fertility—which implies a sustained decline in both period and cohort fertility below the replacement level—raises a number of complex questions. If this is happening as a result of women's entry into the labour market, inferring a conscious choice by women to allocate resources to their own development rather than reproduction—it is difficult to argue that it is a negative trend. This is to say, you can, if you're willing to argue that women ought to spend more time reproducing than working; good luck with that! More traditionally, another argument is that modern society ought to make it attractive for women to work and procreate, and if it doesn't, we end up with a sub-optimal outcome. This is the idea that a key driver of sustained sub-replacement fertility—especially birth postponement—is because women end up having less children than they really want. This stands in contrast to the idea that the fall in fertility to below replacement levels is a structural result of shifts in family-formation and cohabitation trends, marriage rates and women's control over their reproduction. In truth, these two lines of thought are complementary in the sense that they meet in the discussion about how much public policy can effect birth rates in developed economies, and whether they should. Finally, population ageing is the most obvious widely agreed negative (economic) impact from a sustained decline in fertility rates, mainly via a strain on the financing of public services, low economic growth and the negative externality from too much savings and too little investment, in an open economy.
does it matter if global fertility falls below the replacement level?
It will easy to brush away a premature, from the point of view of official forecasts, decline in global TFR to below 2.1 if it is assumed driven by tempo effects. But empirical evidence from countries who have experienced the largest and most sustained declines in period fertility suggest that ongoing birth postponement exert a powerful force on long-term trends in birth rates and cohort fertility. In other words, sustained tempo effects have a lagged impact on quantum effects. More generally, given the increasingly clear evidence of an accelerated demographic transition in so-called developing economies—moving from high fertility to low fertility regimes—and similarly clear evidence that birth rates in countries with sub-replacement fertility are not rebounding, it is easy to see how once global fertility falls below the magic 2.1 level, it will continue to decline for some time.
To me, the question of whether a global fertility rate below the replacement level is appropriate, or maladaptive if you will, is as much philosophical as it is a question of defining optimality in the face of the unbreakable laws of evolutionary biology. After all, evolution works over timeframes so long that it seems to me perfectly plausible to assume that it could in fact be evolutionarily optimal for the human population to exist with total fertility rates below 2.1 for a few generations. This, however, won’t prevent researchers from asking the question of whether the observed and perceived socioeconomic and biological drivers of falling fertility are in fact, socially, economically and evolutionarily optimal. If I am right about the trend in global fertility, this discussion will continue with undiminished strength in the next few years. I am looking forward to it
—
* Famous in the context of demographics research.
** A new paper fresh off the press, slated to become a chapter for The Oxford Handbook of Cultural Evolution seems to offer a nicely condensed sweep of many of the major ideas on the fertility transition.
^ Links to the academic sources cited above are available in my continuously updated list of references for my book; see here.