Population Collapse Is an Economic Story: Japan, Korea, Italy and the Future of Growth

Dek: Working-age population peaks before total population peaks. By the time the headline number falls, the macro impact is a decade old.

Population collapse is not about fewer births today but a structural shift in workforce composition that began in the 1990s. Japan’s working-age population (15–64) peaked in 1995, a full thirteen years before its total population began to decline in 2008. South Korea’s fertility rate fell to 0.72 in 2023, the lowest ever recorded for a sovereign state, yet its working-age population has been shrinking since 2017. Italy’s working-age population has been in steady decline since 2010. The macro impact is not a sudden drop in consumer demand; it is a slow, grinding compression of the potential growth rate. Aggregate GDP stalls, debt-to-GDP ratios become harder to manage, pension systems face actuarial strain, and immigration becomes the primary swing variable. The headline number—total population—is a lagging indicator. The real economic damage is already a decade old by the time it makes the news.

Key takeaways

• Japan’s working-age population peaked in 1995, thirteen years before its total population peak in 2008.
• South Korea’s total fertility rate fell to 0.72 in 2023, the lowest ever recorded for any country.
• Italy’s working-age population has been contracting since 2010, with a fertility rate of 1.20 in 2022.
• China’s working-age population peaked in 2015; total population peaked in 2021, entering the depopulation regime.
• GDP per capita can still rise in a shrinking workforce, but aggregate GDP growth slows and debt dynamics worsen.
• Policy responses—pronatalist transfers, immigration, automation, and later retirement—have limited and uneven effectiveness.

Historical setup

The demographic transition that began in the 19th century in Europe and spread globally followed a predictable pattern: high birth and death rates gave way to falling mortality, then falling fertility, producing a temporary “demographic dividend” as the working-age share of the population swelled. This dividend was a one-time windfall. In East Asia, it was exceptionally large and fast. Japan’s working-age share peaked at 69.8% in 1992, South Korea’s at 73.1% in 2012, and China’s at 72.3% in 2010. The dividend ended when the share began to fall, but the absolute size of the working-age population continued to grow for a time because the overall population was still rising.

The lag between the peak of the working-age share and the peak of the absolute working-age population is typically five to ten years. The lag between the working-age peak and the total population peak is longer—often a decade or more—because the elderly population is still growing. This means that by the time a country’s total population begins to shrink, the labor force has already been contracting for a decade, and the ratio of workers to dependents has been deteriorating for even longer.

Japan is the clearest case. The working-age population peaked at 87.3 million in 1995. By 2023, it had fallen to 73.9 million, a loss of 13.4 million workers. The total population did not peak until 2008, at 128.1 million. The economic effects of the labor contraction were already visible in the late 1990s: stagnant nominal GDP, falling real estate prices, and the first of the “lost decades.” The headline population decline was a late signal.

South Korea’s trajectory is compressed and more extreme. The working-age population peaked at 37.3 million in 2017 and has since fallen to 36.5 million in 2023. The total fertility rate, already below replacement since 1983, dropped to 0.72 in 2023. The working-age share peaked in 2012 at 73.1% and has been falling since. The total population is projected to peak around 2024–2025, then decline sharply. The economic consequences are already visible in the labor market: a shrinking pool of young entrants, rising wages for low-skilled service work, and a growing dependence on foreign workers in manufacturing and agriculture.

Italy’s case is older and slower. The working-age population peaked at 39.8 million in 2010 and has since fallen to 38.0 million in 2023. The total fertility rate has been below 1.5 since the mid-1990s. The working-age share peaked earlier, in 1991, at 68.5%. Italy’s economic stagnation since the 2008 financial crisis is partly structural—low productivity growth, high public debt—but the demographic headwind is unmistakable. The labor force is shrinking, the old-age dependency ratio is rising, and the pension system requires continuous fiscal transfers from general revenue.

What the data shows

The relationship between working-age population peaks and total population peaks is not uniform across countries, but the pattern is consistent. The table below shows the key inflection points for the four countries under discussion.

Sources: UN World Population Prospects 2024; Statistics Korea; World Bank population estimates.

The gap between the working-age peak and the total population peak is 13 years for Japan, roughly 7 years for South Korea, 4 years for Italy, and 6 years for China. The gap is shorter for countries where fertility fell later and faster, because the total population stops growing sooner after the working-age peak.

The economic impact is visible in the growth of GDP per capita versus aggregate GDP. Japan’s GDP per capita (in constant 2015 USD) rose from $34,000 in 1995 to $40,000 in 2023, a compound annual growth rate of 0.6%. But aggregate GDP grew at only 0.3% per year over the same period, because the labor force was shrinking. South Korea’s GDP per capita grew at 3.5% annually from 2017 to 2023, but aggregate GDP growth slowed from 4.5% in the 2000s to 2.5% in the 2010s. The demographic headwind is not yet severe enough to offset productivity gains, but the trend is clear: as the labor force shrinks, the same productivity growth yields lower aggregate GDP growth.

Debt-to-GDP ratios are also affected. Japan’s gross government debt reached 264% of GDP in 2023, the highest in the OECD. A shrinking workforce means a smaller tax base to service that debt, even as spending on pensions and healthcare rises. The IMF’s 2024 Fiscal Monitor projects that Japan’s primary deficit will persist for the next decade, driven entirely by age-related spending. Italy’s debt-to-GDP ratio was 144% in 2023, and the European Commission projects it will rise to 150% by 2026, partly because the working-age population is declining.

Country-level comparison

Japan remains the most advanced case of demographic economic compression. The working-age population has fallen by 15% since 1995. The labor force participation rate has risen, particularly among women and the elderly, but not enough to offset the absolute decline. The labor market remains structurally tight: the unemployment rate has been below 3% since 2015, yet nominal wage growth has averaged only 0.8% per year. Firms have responded by investing in automation—Japan is the world’s leading installer of industrial robots per manufacturing worker—but productivity growth has remained below 1% per year.

South Korea is Japan on a faster track. The fertility rate of 0.72 in 2023 implies that the working-age population will shrink by roughly 30% over the next 30 years, assuming no immigration. The government has spent over $200 billion on pronatalist policies since 2006, including cash allowances, parental leave subsidies, and housing support. The total fertility rate has continued to fall. The policy response has shifted toward immigration: the number of foreign workers in South Korea rose from 800,000 in 2017 to 1.5 million in 2023, but the government remains reluctant to open the door to permanent settlement. The economic consequences are already visible in the education sector, where university enrollment has fallen by 20% since 2010, and in the housing market, where demand is concentrated in Seoul and collapsing in rural areas.

Italy is a case of slow decline with structural rigidities. The working-age population has fallen by 4.5% since 2010. The labor force participation rate is low by OECD standards, at 66% in 2023, compared to 78% in Germany. The informal economy is estimated at 12% of GDP, meaning that tax revenues are lower than the official numbers suggest. The pension system consumes 16% of GDP, the highest in the eurozone. The European Central Bank’s 2023 Financial Stability Review noted that Italy’s aging population is a structural risk to its sovereign debt sustainability. Immigration has partially offset the labor decline: net migration averaged 150,000 per year from 2010 to 2023, but the inflow has been insufficient to stabilize the working-age population.

China entered the depopulation regime later but on a larger scale. The working-age population peaked at 1.001 billion in 2015 and has since fallen by 30 million. The total population peaked at 1.412 billion in 2021 and fell by 850,000 in 2022, the first decline since the 1960s. The economic consequences are already visible: the labor force is shrinking, wages are rising in coastal manufacturing hubs, and the property market is in a severe downturn. The government’s response has been to encourage later retirement and to relax the one-child policy, but the fertility rate has continued to fall, reaching 1.19 in 2023. The World Bank’s 2023 China Economic Update projected that the potential growth rate will fall from 5% in 2020 to 3% by 2030, with demographic factors accounting for half of the decline.

The political-economic mechanism

The mechanism linking population decline to economic stagnation is not a simple matter of “fewer customers.” It operates through three distinct channels: labor supply, capital accumulation, and fiscal sustainability.

The labor supply channel is the most direct. A shrinking working-age population means fewer workers, which reduces the potential output of the economy. This effect is partially offset by higher labor force participation and longer working lives, but the offset is limited. Japan’s labor force participation rate for those aged 65 and over rose from 18% in 1995 to 26% in 2023, yet the total labor force still fell by 5 million. The elasticity of output to labor supply is roughly 0.6 in advanced economies, meaning that a 1% decline in the labor force reduces potential GDP by 0.6% in the medium term.

The capital accumulation channel is more subtle. A shrinking workforce reduces the incentive for firms to invest in capacity expansion. Japan’s gross fixed capital formation fell from 30% of GDP in 1990 to 24% in 2023. South Korea’s capital formation has also declined, from 34% of GDP in 2010 to 30% in 2023. Lower investment reduces the rate of technological adoption and productivity growth, creating a feedback loop: slower productivity growth means slower wage growth, which reduces fertility further, which accelerates the labor decline.

The fiscal sustainability channel is the most politically consequential. A shrinking working-age population means fewer taxpayers to support a growing elderly population. Japan’s old-age dependency ratio (65+ per 100 working-age) rose from 20 in 1995 to 52 in 2023. Italy’s rose from 28 to 40 over the same period. South Korea’s is rising faster than any other OECD country, from 14 in 2010 to 27 in 2023. The result is that government spending on pensions, healthcare, and long-term care rises faster than GDP, forcing either higher taxes, higher debt, or cuts to other spending. The BIS’s 2023 Annual Economic Report estimated that the fiscal cost of aging in advanced economies will amount to 5–10% of GDP per year by 2040, depending on the country.

Immigration is the most obvious offset, but it is politically constrained and economically partial. The OECD’s 2023 International Migration Outlook found that net migration to OECD countries reached a record 6.1 million in 2022, but the distribution is uneven. Japan’s net migration has been less than 200,000 per year, and the government has resisted large-scale permanent immigration. South Korea’s net migration has risen to 200,000 per year, but the inflow is mostly temporary workers. Italy’s net migration has been positive but insufficient to offset the natural decline. The economic effect of immigration on GDP per capita is ambiguous: immigrants increase aggregate GDP, but their impact on GDP per capita depends on their skill levels and labor market integration.

Automation offers a partial productivity offset. Japan has the highest robot density in the world, at 392 industrial robots per 10,000 manufacturing workers in 2022, according to the International Federation of Robotics. South Korea is second, at 1,000 per 10,000 workers. Yet productivity growth in both countries has been below the OECD average. Automation can replace workers in routine tasks, but it does not automatically generate the kind of innovation-driven productivity growth that raises living standards. The ECB’s 2022 Occasional Paper on demographic change and productivity found that automation offsets about half of the labor supply decline in advanced economies, but the effect is concentrated in manufacturing, not services.

Longer working lives are the least controversial policy response, but the effect is limited by health and labor market conditions. Japan raised the pension eligibility age from 60 to 65 between 2001 and 2013, and the labor force participation of older workers rose sharply. Italy raised the retirement age to 67 in 2019, but the participation rate for those aged 55–64 is still only 64%, compared to 77% in Japan. The OECD estimates that raising the effective retirement age by one year increases the working-age population by roughly 1%, a small offset relative to the scale of the decline.

Why it matters today

The demographic regime shift is not a future risk; it is a present reality for Japan, South Korea, Italy, and China. The economic consequences are already embedded in the data: slower growth, higher debt, tighter labor markets, and rising fiscal pressure. The policy response space is narrow, and the tools available—pronatalist transfers, immigration, automation, and later retirement—have limited and uneven effectiveness.

The most important implication is for fiscal sustainability. As the working-age population shrinks, the tax base contracts, while age-related spending rises. The IMF’s 2024 Fiscal Monitor projects that the median advanced economy will see its primary deficit widen by 2% of GDP by 2030 due to aging alone. For countries with high initial debt levels—Japan, Italy, and increasingly South Korea—this creates a vulnerability to interest rate shocks. A rise in borrowing costs of one percentage point would increase Japan’s debt-to-GDP ratio by 25 percentage points over five years, assuming no fiscal adjustment.

The second implication is for the global growth outlook. The UN World Population Prospects 2024 projects that the working-age population will decline in 60 countries by 2050, including all of East Asia, most of Europe, and several Latin American countries. The aggregate effect on global GDP growth is uncertain but likely negative. The World Bank’s 2023 Global Economic Prospects report estimated that demographic factors will reduce the potential growth rate of the global economy by 0.5 percentage points per year over the next decade, relative to the 2000–2010 average.

The third implication is for political stability. The combination of slower growth, higher taxes, and rising dependency ratios creates distributional conflicts between generations. Japan’s elderly population votes at higher rates than the young, and the government has prioritized pension and healthcare spending over education and child care. South Korea’s government has reversed course, increasing spending on children and families, but the fertility rate has not responded. Italy’s political system has been paralyzed by the trade-off between fiscal consolidation and social spending. The demographic pressure is not deterministic, but it narrows the range of feasible policy choices.

The data is clear: the working-age population peak is the real inflection point. The total population peak is a lagging indicator. By the time the headline number falls, the economic damage is a decade old. The question is not whether growth will be lower, but how much lower, and which policies can soften the landing.

Data notes

• Working-age population (15–64): UN World Population Prospects 2024, accessed June 2024.
• Total population: UN World Population Prospects 2024; Statistics Korea, Population Census (2023).
• Total fertility rate: World Bank, World Development Indicators (2023); Statistics Korea, Vital Statistics (2023).
• GDP per capita (constant 2015 USD): World Bank, World Development Indicators (2023).
• Gross government debt (% of GDP): IMF, Fiscal Monitor (April 2024).
• Old-age dependency ratio (65+ per 100 working-age): UN World Population Prospects 2024.
• Labor force participation rate (65+): OECD, Society at a Glance (2023).
• Robot density per 10,000 manufacturing workers: International Federation of Robotics, World Robotics Report (2023).
• Net migration: OECD, International Migration Outlook (2023).
• Potential growth estimates: World Bank, China Economic Update (June 2023); IMF, World Economic Outlook (April 2024).