The Demographic Future of Humanity:
Facts and Consequences
Jesús Fernández-Villaverde1
May 31, 2025
1
University of Pennsylvania, NBER, and CEPR
The global fertility crisis
• The global fertility crisis is worse than you think.
• 2023 was likely the first year in human history when our fertility rate fell below the replacement rate.
• The world population will start falling c. 2055.
• Never seen before, even compared to wars and pandemics.
• The consequences for the world are momentous.
• Let me describe a few of them.
1
Fertility rate
The total fertility rate (TFR) of a population
The average number of children that would be born to a woman over her lifetime if:
1. She were to experience the current age-specific fertility rates throughout her lifetime.
2. She were to live through ages 15-44.
• The TFR governs whether a population reaches the replacement rate: whether enough children are
born to sustain population levels (forgetting net immigration).
• But recall “momentum.”
2
TFR around the world
• If you think this is mainly about rich, advanced economies, you are wrong.
(2024) TFR
US 1.60
Tunisia 1.56
Mexico 1.55
Turkey 1.48
Brazil 1.47
Iran 1.44
Sri Lanka 1.37
Bosnia and Herzegovina 1.20
Colombia 1.06
Thailand 0.98
3
Replacement rates
• Definition:
1 + sex ratio at birth
Replacement rate ≈
Probability of a woman to survive to 30
• For an advanced country (e.g., the U.S.):
1 + 1.05
Replacement rate rich country ≈ ≈ 2.1
0.98
• For an emerging country (e.g., Pakistan):
1 + 1.10
Replacement rate emerging country ≈ ≈ 2.33
0.90
• For the planet:
1 + 1.08
Replacement rate humanity ≈ ≈ 2.2
0.95
4
biased sex ratios at birth from 2000 to 2020
Average sex ratio at birth, or the number of male births per 100 female births, from 2000-20
Note: Globally, the natural sex ratio at birth ranges from 103 to 107 boys per 100 girls.
Source: United Nations World Population Division, 2019.
“India’s Sex Ratio at Birth Begins To Normalize”
5
PEW RESEARCH CENTER
6
7
8
Does WPP 2024 make sense?
• To Patrick and me, not a lot!
• But we are happy to be corrected.
• First: discrepancy between vital registries and UN (even for countries that the UN assesses as having
complete vital statistics).
9
10
11
12
Why does this matter? “The rule of 85”
• Imagine you have a country where life expectancy is ≈ 85 years: the highest life expectancy in the
world right now (Japan, Spain, etc.).
• Imagine that, from now on, you have 1,000 births per year, every year.
• What would be your population in about 100 years? 85,000 = 85 * 1,000.
• Thus, you can look at the current births of any given country, multiply by 85, and get a sense of the
long-run population (without migrations).
• For example, Colombia had 445k births in 2024. Long-run population: 37.8 million (85*442k).
Current population: 50.3 million.
• An equivalent way to look at it: 1000/85 = 11.76. When a country’s CBR falls below 11.76 per
1,000, births are already insufficient to maintain a constant population (this usually occurs after the
TFR falls below replacement).
13
14
Does WPP 2024 make sense?
• To Patrick and me, not a lot!
• But we are happy to be corrected.
• First: discrepancy between vital registries and UN.
• Second: Odd forecasts.
15
16
17
18
19
20
The UN explanation
WPP 2024, Summary of Results, page 23, Box II.1
“The “rebound” in future fertility for low-fertility countries is consistent with an expectation of
continued progress toward gender equality and women’s empowerment and improving social and
economic opportunities for young people and families.”
• This is not a very serious justification.
• A more subtle point: conditional vs. unconditional forecasts.
21
Does WPP 2024 make sense?
• To Patrick and me, not a lot!
• But we are happy to be corrected.
• First: discrepancy between vital registries and UN.
• Second: Odd forecasts.
• Third: Bizarre numbers.
22
23
Does WPP 2024 make sense?
• To Patrick and me, not a lot!
• But we are happy to be corrected.
• First: discrepancy between vital registries and UN.
• Second: Odd forecasts.
• Third: Bizarre numbers.
• Fourth: Level effects
24
Two recent censuses
• Brazil: 203 mil., not 212 mil.
• Paraguay: 6.1 mil, not 6.9 mil.
• I suspect these level effects are present in many other countries, in particular in Africa.
• Aggregate data for Africa is really bad.
• But what I have seen in Egypt, Tunisia, Lesotho, and Kenya impresses me a lot.
• Just correcting for Brazil, the “peak population” will be in 2080, not 2084.
• Remember: forecasting the peak of a concave function is really hard! Very sensitive to even minor
assumptions.
25
Our assessment
• Correcting for excess births, the world TFR in 2024 is likely to be around 2.17 instead of 2.25 (keep
in mind that a level effect would operate in the other direction, though!).
• The world replacement rate is around 2.21.
• So, we estimate that humanity is already below the replacement rate.
• We can play at the margin with these numbers, but the changes are not significant.
• Thus, a more likely scenario, given current policy, is that the peak of world population would be
around 2055.
• By the way, this is the low-fertility scenario from the WPP 2024.
26
Some economic consequences
• Like all changes, this momentous demographic shift will have good and bad consequences.
• Which ones will predominate will depend on the policy responses we offer.
• Also, please remember: we are venturing into terra incognita. Things can change.
• If you pick a social sciences textbook from the 1980s (or even 1990s), the main concern was the
population explosion.
• When I was a kid, my parents bought a book that stated that population growth in Mauritius was so
out of control that the island nation was doomed. Mauritius has had a negative population growth
for at least two years now.
• So, there is a non-trivial risk that someone might make fun of me in 25 years.
27
28
First, the good news
• With a lower population growth rate or falling population, it is much easier to design policies that
ensure the sustainability of natural resources.
• Fewer pressures on infrastructure, housing, emissions, etc.
• For example, we can redesign many of Latin America’s cities.
• Additionally, before population aging truly takes hold, many developing economies may have some
fiscal space to capitalize on lower fertility rates and implement essential reforms.
29
30
But there are also bad news
• The unpleasant arithmetic of demographics.
• A basic accounting identity:
y
y = ∗ |{z}
l
|{z} l
output
|{z} labor
labor productivity
• Thus, output growth:
gy = g yl + gl
|{z} |{z} |{z}
output growth labor productivity growth labor growth
31
The way we were
• The U.S. c. 2000:
3%
|{z} = 2%
|{z} + 1%
|{z}
output growth labor productivity growth labor growth
• 3% is the “typical” output growth that you expect:
1. When the economy is booming (i.e., unemployment is falling and labor is growing faster than average),
you see 4%.
2. When the economy is depressed (i.e., unemployment is increasing and labor is growing slower than
average), you see 2%.
32
The way we will be
• The U.S. c. 2050:
1%
|{z} = 2%
|{z} + −1%
| {z }
output growth labor productivity growth labor growth
• 1% is the “typical” output growth that you expect:
1. When the economy is booming (i.e., unemployment is falling and labor is growing faster than average),
you see 2%.
2. When the economy is depressed (i.e., unemployment is increasing and labor is growing slower than
average), you see 0%.
• There is nothing the Fed can do with additional monetary stimulus, nor the federal government with
further fiscal packages.
33
A clear example
• Japan: the poster child of bad economic performance.
• Between 1991 and 2019, GDP in Japan grew at an annual rate of 0.83%, much lower than the 2.58%
of the U.S.
Pesek (2014) and Japanization
...few lessons are more timely or critical than those offered by Japan, a once-vibrant model for
developing economies that joined the world’s richest nations, lost its way and has been struggling to
relocate it ever since.
In this book I explore what the world can learn from a Japanese economic funk that began more than 20
years ago and has never really ended. That means exploring where Japan went wrong, how it sank under
the weight of hubris and political atrophy, and missed opportunity after opportunity to scrap an insular
model based on overinvestment, export-led growth, and excessive debt.
34
Japan is doing as well as they can
Table 1: G7 plus Spain: Output and Population Growth Rates
1991-2019 Canada France Germany Italy Japan Spain UK U.S.
GDP 2.47 1.61 1.38 0.70 0.83 2.05 2.08 2.58
GDP per capita 1.40 1.10 1.25 0.52 0.76 1.35 1.53 1.63
GDP per w.-age adult (15-64) 1.48 1.33 1.47 0.79 1.39 1.41 1.62 1.65
GDP per w.-age adult (15-69) 1.39 1.25 1.40 0.74 1.21 1.38 1.60 1.58
GDP per hour worked 1.23 1.28 1.31 0.71 1.26 0.67 1.37 1.53
Population 1.05 0.50 0.14 0.18 0.08 0.68 0.54 0.94
Working-age adults 0.98 0.27 -0.09 -0.08 -0.54 0.63 0.46 0.91
Total hours worked 1.23 0.33 0.08 0.00 -0.43 1.40 0.71 1.04
35
What is the problem?
• Don’t we care about output per capita?
• Yes and no.
• Yes, output per capita is the primary measure of individual welfare but...
• our ability to service debt and social security obligations depends on total output.
• Also, labor productivity is unlikely to grow at 2% any longer.
• Why?
1. Firm dynamism.
2. New technologies.
36
even for the largest firms in BDS data.
50% 20%
18%
16%
45%
14%
12%
40%
10%
8%
35%
6%
4%
30% 2%
1985 1990 1995 2000 2005 2010 2015 1940 1950 1960 1970 1980 1990 2000 2010 2020
(a) Share of Firms Age 11+ (b) Entry Rate
Figure 3
Sources. (a) BDS. (b) Entry rate 1940-1962: Survey of Current Business. Entry rate 1978-2014: BDS.
Notes. The entry rate from 1963 to 1977 is linearly interpolated. 37
1 32
Effective number of
1/2 researchers (right scale) 16
1/4
Index (1930 = 1)
Index (1930 = 1)
8
1/8
Research productivity 4
1/16 (left scale)
2
1/32
1/64 1
1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s
38
What about immigrants?
• Not so easy as it seems.
• First, TFR is falling everywhere in the world: net migration into the Earth is zero.
• Second, the number of immigrants that the advanced would need to bring is phenomenal, causing
multiple strains in society.
• We live in a welfare state, and this is unlikely to change anytime soon.
• Roughly speaking, the bottom 70% of the population receives net payments (and services) from the
government over their lifetime, while the 70%-90% percentile receives approximately zero net
payments, the top 10% pays it all.
• Most immigrants worsen the fiscal position of the government.
• Each immigrant into a rich country makes the position of poor countries harder.
39
(Economic) strategies
• Most of the current drop in TFR below 2.1 is driven by the extensive margin, not the intensive margin.
• Affordable housing:
1. Cheaper housing, yes...
2. ... but also “more” convenient housing (size, transportation, etc.).
3. Modern city planning is deeply anti-family and anti-kids.
• Changes in education.
40
A few concluding thoughts
• I have skipped, in the interest of time, tons of other aspects.
• We are entering a whole new world: fertility and births are falling much faster than anyone
anticipated.
• There is a good chance that the TFR of middle-income countries (e.g., Colombia, Chile, Thailand,
Turkey) will fall below the TFR of rich countries (at least outside East Asia). Most likely, it is already
the case that TFRUS > TFRMexico .
• Migration will exacerbate the situation: Colombia’s population is already declining due to high
migration to the U.S. and Spain.
• This creates a whole range of challenges.
• Middle-income countries may not have the margins of maneuver that Japan has.
• Once you start thinking about these issues, it is hard to think about anything else: demography is
destiny.
41
