Falling global fertility, how far and how fast?
The charts for this post were generated with my AS Demographics Chart Creator v 1.2 GPT which has the July 2024 UN global demographic data uploaded as its database. Go check it out.
It has been a while since I last delved into one of my favorite topics: global demographics. In this piece, I revisit the subject by examining high-level data on key global demographic indicators from the UN’s July 2024 Population Prospects database. I will begin with birth rates.
Scarcely a day goes by without an article, podcast, or both highlighting the accelerating decline in global fertility. As I explain in my essay on the fertility wars, this discussion tends to divide interlocutors into two increasingly polarized factions. On one side are those who believe falling fertility is a grave problem; on the other are those who remain more sanguine, viewing declining birth rates as a natural consequence of modernity—or perhaps postmodernism—and less of a threat to economic growth, government budgets, or humanity’s survival. If women choose to have fewer children and prioritize careers and personal freedom—long the exclusive domain of men—shouldn’t we support that choice?
Regardless of where one stands on the economic and social implications of declining global birth rates, one fact is becoming increasingly clear: global period fertility—the total fertility rate (TFR)—is falling more rapidly than official forecasts suggest, a point I have previously argued here. And if my analysis isn't convincing, The Lancet has made a similar observation, which I discuss here.
The chart below illustrates the global total fertility rate (TFR), with UN forecasts extending beyond 2023 and a trendline extrapolating the trajectory of the past decade. The UN assumes a slowing rate of decline, projecting that period fertility will remain above the replacement level until 2050—an assumption that contradicts the most recent empirical evidence. According to the UN, the global TFR stood at 2.25 in 2023, down from 2.57 a decade earlier. If this trend continues, the global TFR will fall below the replacement level by 2027, as the chart suggests. The pressing question now is when the UN will adjust its near-term forecasts to reflect this reality. Will it happen in the July 2025 World Population Prospects (WPP) update? I don’t know—but when it does, it will undoubtedly add further intensity to the ongoing debate on global birth rates.
Source: UN Population Prospects, July 2024. Chart created by ChatGPT
The debate over whether declining global birth rates should be a cause for concern is often preceded by a discussion of why they are falling in the first place. This is a topic I have analyzed extensively in my research. While the issue is complex, the key theme is simple: we must differentiate between period fertility (TFR) and cohort fertility. TFR measures the number of live births per woman in a given year—or more precisely, the average number of children a woman would have over her lifetime if she experienced the current age-specific fertility rates throughout her reproductive years (typically ages 15 to 45). Cohort fertility, by contrast, tracks the actual number of children born to a specific group (cohort) of women over their entire reproductive lifetime. The distinction between the two is summarized below:
Total Fertility Rate (TFR) vs. Cohort Fertility Rate (CFR)
| Metric | Total Fertility Rate (TFR) | Cohort Fertility Rate (CFR) |
|---|---|---|
| Type | Period measure (hypothetical) | Cohort measure (actual) |
| Timeframe | Based on fertility rates in a given year | Based on lifetime fertility of a birth cohort |
| Volatility | Can fluctuate due to short-term factors | More stable, long-term measure |
| Purpose | Estimates future trends | Analyzes past fertility trends |
The key distinction between period fertility (TFR) and cohort fertility (CFR) lies in the fact that TFR can be influenced by temporary fluctuations, which may not necessarily be reflected in CFR. In the present context, this suggests that the sharp decline in global period fertility might overstate the actual decline in total cohort fertility for women currently in their reproductive years. Broadly, the academic literature—spanning fields from evolutionary anthropology and economics to the social sciences concerned with demographics—maps the distinction between period and cohort fertility into two major conceptual frameworks for analyzing fertility patterns:
The quantum effect of fertility (Hillard Kaplan and Gary S. Becker) - This framework links shifts in fertility to the trade-off between the quantity and quality of offspring, emphasizing the changing relative returns on investment in these two areas. In essence, it explains declining fertility as part of the broader demographic transition, arguing that as societies develop, parents increasingly prioritize investing in the quality of their children (e.g., education and human capital) over having more children. Put simply, the quantum effect of fertility seeks to explain the paradox of why fertility declines as income rises.
The tempo effect of fertility or birth postponement (The Second Demographic, SDT, theory) - This framework focuses on declining birth rates in the post-World War II period, particularly from the 1970s and 1980s onward. It attributes the rapid decline in fertility across various countries—first in Europe, then in Asia and North America in the 1990s and 2000s—to birth postponement, or an increase in the mean age of first childbirth. The SDT framework highlights how cultural, economic, and policy shifts have contributed to later family formation, further accelerating declines in period fertility.
As I demonstrate in my paper on the tempo effect, this trend is evident in the data—not only in individual countries but increasingly in global aggregate data as well. The chart below illustrates the mean age of childbearing (MAC), defined as the average age of women at childbirth, weighted by age-specific fertility rates. More simply, it represents the average age of mothers at the birth of their children.
While this global perspective overlooks significant regional and national variations in the tempo effect, the broader trend is unmistakable: birth postponement is now a global phenomenon—and it is accelerating. Given this, it is puzzling that the UN continues to predict a relatively slow decline in period fertility over the next five to ten years, despite simultaneously forecasting a sharp rise in the MAC—a key proxy for accelerated birth postponement and a more rapid decline in the total fertility rate (TFR).
Granted, MAC is a somewhat crude measure. A more nuanced perspective comes from examining age-specific fertility rates (ASFR), a demographic indicator that tracks the number of births per 1,000 women within specific age groups over a given period. Typically analyzed in five-year intervals, ASFR provides a detailed view of fertility patterns across different reproductive age groups. The paper linked above uses ASFR data to map the tempo effect of fertility, tracking shifts in the distribution over time across different countries and regions.
Source: UN Population Prospects, July 2024. Chart created by ChatGPT
The concept of adjusting period fertility for birth postponement is a well-established theme in demographic literature. This adjustment is typically approached using either a general measure—such as the Bongaarts and Feeney (1998) tempo-adjusted TFR—or parity progression ratios (PPRs), which estimate the probability of a woman progressing from one child to the next. The charts below illustrate one method for calculating a global tempo-adjusted TFR using the Bongaarts and Feeney approach, applied to data from the UN population database. This calculation suggests that the world is currently experiencing a relatively pronounced tempo effect. However, it also indicates that the gap between recorded fertility and the tempo-adjusted index remains relatively small.
Source: UN Population Prospects, July 2024 and Bongaarts and Feeney (1998). Chart created by ChatGPT
Source: UN Population Prospects, July 2024. Chart created by ChatGPT
While more sophisticated methods for tempo-adjusting fertility exist beyond the calculations above, the challenge of accurately accounting for birth postponement in fertility analysis is more complex than simply devising an empirical methodology to adjust period fertility upward. In theory, the quantum and tempo effects should be independent: the former is governed by the desired number of children—shaped by income levels and the quantity/quality trade-off—while the latter dictates only the timing of these births. In practice, however, the two are intricately linked. Tempo effects exert a lagged impact on quantum, meaning that significant birth postponement can ultimately affect cohort fertility over time.
This occurs for three key reasons:
Biology - Even the best-laid plans are subject to biological constraints. Female fecundity peaks in the 20s and early 30s but declines sharply in the late 30s and early 40s, before menopause eliminates the possibility of natural conception. Estimates suggest that the probability of conception per cycle drops from 25–30% in the 20s to 15% by the mid-30s, 5–10% by age 40, and just 1–3% by age 45. Modern birth postponement is particularly poorly timed in this regard—women often delay childbirth during their peak reproductive years, only to attempt conception at precisely the time when their fertility declines dramatically. On a large scale, this dynamic suggests a significant lagged impact of tempo effects on quantum fertility.
Resource constraints - Having children requires financial, emotional, and time-related resources, all of which entail opportunity costs. Indeed, it is often these very costs—career ambitions, financial stability, and housing concerns—that lead to birth postponement in the first place. However, waiting for the “perfect” conditions does not always work out. If couples continuously delay childbirth in pursuit of financial and career security, they may eventually run into the biological constraints outlined above, or never have children, simply because the perfect resource conditions never arise.
Preference (in)stability - A woman’s stated fertility preferences at a given point in time do not necessarily remain stable over her lifetime. Put differently, what a woman desires in her 20s is not always what she will want in her 30s or 40s. Shifts in fertility intentions can be influenced by both biological and resource constraints, as discussed above, or by broader socioeconomic changes that operate independently of these factors. This raises an important question: Is early-life desired fertility a reliable predictor of final cohort fertility? This question warrants its own detailed analysis, but suffice it to say that stated fertility preferences are often a poor predictor of actual fertility outcomes, both for individuals and across cohorts. This discrepancy arises in part due to survey limitations in accurately capturing fertility intentions, but also because preferences evolve over time.
As global period fertility continues its descent toward sub-replacement levels, the distinction between tempo and quantum effects will become increasingly critical for policymakers, researchers, and commentators engaged in the global birthrate debate. The future of fertility will be shaped by the intersection of these two forces, and those who grasp this dynamic will have the best framework for analyzing and understanding global fertility trends.
If the first step is recognizing that tempo effects can have powerful lagged impacts on cohort fertility, then the next logical step is to determine how these impacts can be mitigated—whether through public policy or technological advancements. One particularly intriguing possibility is that technology could enable a significant "catch-up" effect for women currently in their reproductive prime. Advances such as egg freezing and assisted reproductive technologies (ART) could, in both theory and practice, facilitate a late-reproductive baby boom among the very women who are currently lowering period fertility by postponing childbirth. Policymakers might take note here. The evidence suggests that direct financial incentives to encourage childbirth yield poor returns on investment. But what if governments redirected resources toward helping couples who already want children? This could involve subsidizing ART treatments or funding egg freezing for younger women to provide them with greater reproductive flexibility in the future.
What happens next?
Two things, I think. First, global fertility will soon fall below the replacement level, and secondly, when it does, those with a stake in the debate on global birth rates—policymakers, researchers and commentators alike—must look for the truth in the intersection between quantum and tempo effects.