Executive Summary Link to heading
Liberal democracy does not mechanically produce a war between generations. It does, however, contain a structural representation problem: future citizens cannot vote, younger citizens often participate less than middle-aged voters, and many of the largest public commitments are financed across decades. Under those conditions, a present electorate can choose policies whose benefits arrive now while part of the tax, contribution, housing, or growth cost is borne later.
That possibility is real, but intergenerational extraction should be reserved for a stronger claim than ordinary age redistribution. A pension paid to an older person is not extraction merely because a younger worker finances it. Public pensions insure longevity, disability, and survivor risk; younger contributors may later receive benefits themselves. Government borrowing is not extraction merely because repayment occurs in the future; debt can finance productive investment and debt service is partly a transfer to domestic bondholders. Older households having more wealth than younger households is also not sufficient evidence, because cross-sectional wealth differences partly reflect the ordinary life cycle of saving and dissaving.
A defensible extraction claim requires evidence that policy systematically raises the lifetime net position of politically advantaged present cohorts while shifting uncompensated costs to less represented younger or future cohorts. The relevant unit is therefore not “old versus young this year,” but cohort lifetime incidence: taxes and contributions paid, transfers and public services received, asset-price effects, inherited liabilities, and the value of public investment over time.
Several mechanisms make such an imbalance plausible.
- Across OECD countries, turnout among 18–24 year-olds is on average 12 percentage points below turnout among adults aged 25–50. Australia, Belgium, and Israel are the exceptions in the OECD comparison. OECD, Society at a Glance 2024
- Pay-as-you-go pension systems expose current workers to demographic arithmetic because current contributions finance current benefits. Aging and low fertility can therefore force some combination of higher contributions, later retirement, lower replacement rates, general-revenue subsidies, or debt-financed adjustment.
- Large and persistent public deficits can reduce future national income by crowding out private investment and raising interest costs. That is a genuine burden-shifting channel, although the size and incidence depend on what the borrowing finances. CBO, The Budget and Economic Outlook: 2026 to 2036
- Housing politics can create a more direct incumbent-versus-entrant conflict. Homeownership increases local political participation, especially when zoning is at stake, and local-government meeting participants are disproportionately older, homeowners, long-term residents, and opponents of new housing. Hall and Yoder, 2022; Einstein, Palmer, and Glick, 2019; Fang, Stewart, and Tyndall, 2023
The strongest comparative conclusion is therefore conditional. Aging democracies can generate intergenerational imbalance when demographic pressure, age-skewed participation, politically protected transfers, housing scarcity, and persistent deficits reinforce one another. But the evidence does not support a universal gerontocracy thesis. Cross-national pension research finds that aging often increases aggregate pension spending because there are more retirees while simultaneously restraining or reducing benefit generosity per retiree. Institutions can force adjustment rather than allowing older voters to preserve every promise. Tepe and Vanhuysse, 2009
Among the four cases considered here, the United States and Japan show the strongest fiscal-demographic pressure; Germany shows substantial electoral aging but also stronger fiscal and pension-adjustment constraints; Israel is the clearest demographic and turnout counterexample. None of these cases, by itself, proves deliberate extraction. They show the institutional conditions under which a lifetime cohort imbalance becomes more or less likely.
1. What Would Count as Intergenerational Extraction? Link to heading
1.1 Age redistribution is not the same as cohort redistribution Link to heading
Every society redistributes across age. Children consume resources before they produce much market income. Working-age adults pay taxes and often support both children and older relatives. Retirees draw down private savings and receive pensions or health benefits. A snapshot that shows money flowing from workers to retirees therefore says little about whether one birth cohort is being exploited by another.
The stronger question concerns lifetime incidence. Let cohort \(c\)’s net public position be approximated by:
\[ L_c = PV(\text{cash transfers} + \text{public services} + \text{public investment benefits} + \text{policy-created asset gains}) - PV(\text{taxes} + \text{social contributions} + \text{policy-created asset-entry costs} + \text{debt-service burden}) \]where \(PV\) denotes a present value measured over the cohort’s life. Intergenerational extraction is a useful label only when policy persistently raises \(L_c\) for politically advantaged present cohorts by lowering the corresponding lifetime position of younger or future cohorts without an offsetting insurance, service, investment, or inheritance benefit.
This definition imposes three disciplines on the analysis.
First, wealth by age is not wealth by generation. The Federal Reserve and Bundesbank both show large age gradients in household wealth, but both patterns are strongly consistent with life-cycle accumulation. The Bundesbank explicitly notes that wealth tends to build through middle age and then decline as older households dissave or transfer wealth to children. Federal Reserve, 2022 Survey of Consumer Finances; Deutsche Bundesbank, 2023
Second, debt is not a one-for-one invoice sent to unborn taxpayers. Borrowing can finance transfers, wars, tax reductions, infrastructure, research, or emergency stabilization. Its intergenerational effect depends on the future tax path, who holds the debt, and whether the financed activity raises future productive capacity. What can be said more securely is that persistently rising debt can reduce future private investment, output, and national income and can constrain later fiscal choices. CBO, 2025
Third, social insurance has value even when cash moves between age groups. Retirement systems insure longevity and smooth consumption; disability and survivor components pay benefits before old age; health systems pool large age-correlated risks. An extraction test has to compare benefits with lifetime contributions and risks rather than treat every transfer as a loss to contributors.
1.2 Political weight is unequal across time Link to heading
Future citizens have no electoral representation except through the preferences of present voters and institutions designed to protect long horizons. Younger citizens can vote, but in many democracies they do so at lower rates.
The OECD’s cross-national evidence is unusually useful because it measures the electoral asymmetry directly. Using Comparative Study of Electoral Systems data around 2016–2021, the OECD reports that turnout among 18–24 year-olds averages 12 percentage points below turnout among 25–50 year-olds. Only Australia, Belgium, and Israel have higher youth turnout than the 25–50 group in that comparison. OECD, 2024
A simple political-weight expression is therefore:
\[ w_{a,t} = population_{a,t} \times participation_{a,t} \times responsiveness_{a,t} \]Age structure changes the first term; turnout changes the second; party competition, lobbying, and institutional access affect the third. The expression is not an empirical estimator by itself. It identifies the mechanism that must be demonstrated before describing an aging electorate as politically dominant.
2. Pensions: The Clearest Intergenerational Contract Link to heading
2.1 Pay-as-you-go systems create demographic exposure Link to heading
A pay-as-you-go pension system is an explicit intergenerational contract: current contributions largely finance current benefits. The system can be stable for long periods, but its arithmetic changes when the ratio of contributors to beneficiaries falls. Governments then have only a limited menu: raise contributions or taxes, lower benefits relative to wages, increase retirement ages, subsidize the system from other revenue, attract additional workers, or borrow during the adjustment.
That exposure does not imply that retirees are “taking” an actuarially unfair amount. It means that demographic change forces distributional choices that political institutions must allocate.
The OECD’s current pension models illustrate how different those choices already are. For a full-career worker earning the average wage, projected gross replacement rates from mandatory schemes in Pensions at a Glance 2025 are 39.7% in the United States, 36.5% in Japan, 42.1% in Germany, and 42.8% in Israel. These modelled replacement rates are not measures of fiscal cost or generational fairness; they are useful because they show that systems facing aging pressure do not simply preserve a uniformly generous benefit promise. OECD, 2025
2.2 Aging does not automatically produce gerontocracy Link to heading
Political-economy models have long predicted that an older median voter can support larger old-age programs. Some cross-country evidence supports that mechanism. But the empirical record is more complicated than the phrase “grey power” suggests.
Tepe and Vanhuysse’s panel study of 18 OECD democracies from 1980–2002 found that population aging increased aggregate pension spending while tending to freeze or reduce benefit generosity per retiree. Their interpretation is closer to a double fiscal-and-electoral constraint than to unlimited gerontocracy: more beneficiaries raise total spending, while fiscal pressure forces smaller slices per person. Tepe and Vanhuysse, 2009
That distinction is crucial. If aggregate pension expenditure rises only because there are more retirees, the increase is demographic before it is political. Evidence of extraction requires showing that political power protects or expands lifetime net benefits beyond what the insurance contract and demographic change explain.
3. Debt: A Burden-Shifting Channel, Not a Sufficient Statistic Link to heading
Public debt is attractive in an intergenerational theory because borrowing separates the date of a political decision from the date of some of its financing. But a debt-to-GDP ratio alone cannot identify who gained and who lost.
The April 2026 IMF World Economic Outlook reports general-government gross debt of 125.8% of GDP for the United States, 204.4% for Japan, 64.6% for Germany, and 69.8% for Israel. IMF DataMapper: United States; Japan; Germany; Israel
Those numbers describe fiscal exposure, not intergenerational incidence. The more defensible burden mechanism is macroeconomic. CBO concludes that large and growing federal debt tends to raise borrowing costs, crowd out private investment, reduce future output, increase payments to foreign debt holders, and constrain future fiscal choices. In its February 2026 baseline, U.S. debt held by the public rises from roughly 101% of GDP in 2026 to 120% in 2036 and reaches 175% in the 2056 long-term projection. CBO, 2026
For an extraction claim, the next question must be what generated the debt. Borrowing to preserve current consumption while avoiding current taxation is more plausibly a transfer to the present than borrowing for a high-return asset that future citizens inherit. The gross debt stock cannot make that distinction by itself.
4. Housing: Where Incumbent and Entrant Interests Can Diverge Directly Link to heading
Housing provides a cleaner political mechanism because incumbent owners can benefit from scarcity while prospective entrants bear higher prices.
In the United States, the Federal Reserve’s 2022 Survey of Consumer Finances found that the median home was worth more than 4.6 times median family income, above the previous 2007 peak in that series. The Federal Reserve explicitly describes rising home values as a boon to owners and declining affordability for would-be buyers. Federal Reserve, 2023
Political research identifies a mechanism connecting those asset interests to policy.
- Using administrative records for more than 18 million people in Ohio and North Carolina, Hall and Yoder find that becoming a homeowner increases participation in local elections, with an even larger participation effect when zoning issues are on the ballot. Hall and Yoder, 2022
- Einstein, Palmer, and Glick find that participants in local-government meetings are disproportionately older, homeowners, long-term residents, and voters, and that almost two-thirds of participants speaking on housing projects oppose new construction. Einstein, Palmer, and Glick, 2019
- Fang, Stewart, and Tyndall connect homeowner composition to actual councillor votes in Toronto from 2009–2020: councillors representing wards with more homeowners are more likely to oppose housing bills, especially large projects inside their own wards. Fang, Stewart, and Tyndall, 2023
These studies do not show that older people as a class deliberately impoverish younger people. They do show a more specific and defensible mechanism: incumbent asset holders participate disproportionately in institutions that can restrict new supply, and political representatives respond to that constituency. Because younger households are disproportionately prospective rather than incumbent owners, the resulting scarcity can have a generational incidence even when the underlying political cleavage is formally ownership rather than age.
5. Four Comparative Cases Link to heading
The cases below are not ranked on a single “extraction index.” They illustrate different combinations of demographic pressure, fiscal exposure, pension adjustment, and electoral age structure.
| Indicator | United States | Japan | Germany | Israel |
|---|---|---|---|---|
| Latest fertility used here | 1.5995, final 2024 | 1.14, provisional 2025 | 1.32, final 2025 | 2.87, 2024 OECD harmonized |
| IMF general-government gross debt, 2026 | 125.8% GDP | 204.4% GDP | 64.6% GDP | 69.8% GDP |
| OECD 2025 gross mandatory pension replacement rate, average earner | 39.7% | 36.5% | 42.1% | 42.8% |
| Electoral-age signal | OECD youth-turnout deficit pattern | OECD youth-turnout deficit pattern | 2025 turnout highest at ages 50–69 | OECD youth-turnout exception |
Sources: NCHS final 2024 births; Japan MHLW 2025 provisional vital statistics; Destatis 2025 fertility; OECD Family Database; IMF DataMapper; OECD Pensions at a Glance 2025.
5.1 United States: strong fiscal pressure, but incidence must be separated from rhetoric Link to heading
The United States has a clear long-horizon financing problem. Final NCHS data put the 2024 total fertility rate at 1.5995 births per woman, a record low. The 2026 Social Security Trustees project that the Old-Age and Survivors Insurance Trust Fund can pay full scheduled benefits until the fourth quarter of 2032; after reserve depletion, continuing income would cover 78% of scheduled OASI benefits absent legislation. The combined OASDI reserves are projected to last until 2034. NCHS, 2026; Social Security Trustees, 2026
CBO’s 2026 baseline adds the broader fiscal context. Social Security and Medicare spending rises as the population ages, while large deficits and interest costs push federal debt higher. CBO projects Social Security and Medicare outlays together rising from 8.7% of GDP in 2027 to 10.1% in 2036. CBO, 2026
This is strong evidence of intergenerational fiscal pressure. It is not sufficient evidence that current retirees receive excessive lifetime returns. A publication-grade extraction claim would require cohort-specific tax-benefit incidence under the reform path eventually chosen.
Housing provides the stronger distributional mechanism. High entry prices and homeowner-dominated local participation create a channel through which incumbent asset protection can raise costs for later entrants. In that domain, the political mechanism is directly observed rather than inferred from a debt ratio.
5.2 Japan: extreme demographic and debt exposure Link to heading
Japan has the most severe demographic-fiscal combination in the four-case set. The Ministry of Health, Labour and Welfare’s June 2026 provisional annual summary reports 671,236 births in 2025 and a total fertility rate of 1.14, both record lows in the reported series. The IMF’s April 2026 WEO places general-government gross debt at 204.4% of GDP. MHLW, 2026; IMF, 2026
Those conditions make pension and health financing unusually exposed to the ratio of workers to older beneficiaries. But Japan also demonstrates why “high debt + old population = extraction” is too crude. The OECD’s 2025 model gives an average earner a projected mandatory gross pension replacement rate of 36.5%, below the OECD average. Fiscal pressure can therefore appear as constrained future generosity as well as as protected retiree benefits. OECD, 2025
Japan fits the thesis strongly as a case of intergenerational adjustment pressure. Demonstrating extraction would require showing which cohorts ultimately absorb pension, tax, debt-service, and growth adjustments relative to the benefits they receive.
5.3 Germany: electoral aging with visible adjustment constraints Link to heading
Germany is a useful counterweight because its electorate is old but its debt burden is much smaller than in the United States or Japan. Destatis reports a final 2025 fertility rate of 1.32, down from 1.35 in 2024. The IMF reports 2026 general-government gross debt of 64.6% of GDP. Destatis, 2026; IMF, 2026
The 2025 Bundestag representative electoral statistics show the political-age structure directly. Turnout was lowest among 21–24 year-olds at 78.3%, while voters aged 50–69 had the highest turnout at 85.5%. People aged 60 and over made up 42.6% of eligible voters, up sharply from 36.3% in 2017. Yet turnout among those 70 and older fell to 79.3%, so the simple proposition “the older the voter, the greater the participation” is false. Federal Returning Officer, 2025
Germany’s pension architecture also contains an explicit sustainability factor, and OECD modelling places the future mandatory gross replacement rate for an average earner at 42.1%. This is evidence that aging pressure can be converted into rule-based adjustment rather than unlimited benefit preservation. Germany therefore fits the report better as a case of contested burden allocation under electoral aging than as straightforward gerontocracy.
5.4 Israel: a direct challenge to demographic inevitability Link to heading
Israel is the clearest counterexample to the demographic chain. The OECD Family Database update of April 2026 reports Israel at 2.87 births per woman in 2024, the only OECD member above the approximate replacement level in that dataset. The OECD voting comparison also places Israel among the three countries where 18–24 year-olds are more likely to vote than adults aged 25–50. OECD Family Database; OECD, 2024
Israel still has old-age transfers and public debt, and the IMF places 2026 gross debt at 69.8% of GDP. But the two variables central to the simplest gerontocracy story—sub-replacement fertility and a youth turnout penalty—are absent in the OECD comparison. Israel therefore demonstrates that democratic institutions do not inherently produce the demographic and electoral structure required by the extraction thesis.
6. What the Evidence Does and Does Not Establish Link to heading
6.1 Established with reasonable confidence Link to heading
Several empirical statements survive the audit.
Aging creates real fiscal pressure in pay-as-you-go systems. Fewer workers per beneficiary require some combination of higher contributions, lower relative benefits, later retirement, subsidies, migration, productivity growth, or borrowing.
Young adults often have less electoral weight. The OECD turnout gap is large enough to matter for political representation, although it varies substantially by country and is not universal.
Persistent debt can shift economic costs forward. CBO’s mechanism is not a moral claim: more government borrowing can reduce national saving and private capital formation and raise future interest costs.
Housing institutions can privilege incumbent owners over entrants. The participation and councillor-voting evidence is unusually direct and gives the report its strongest observed political mechanism.
6.2 Not established by the current evidence Link to heading
Several stronger propositions should be rejected or left unclaimed.
Age-wealth gaps do not prove extraction. A 30-year-old and a 70-year-old are at different points in the saving life cycle. Cross-sectional net-worth tables are not cohort accounts.
High pension spending does not prove elder political domination. An older population mechanically raises aggregate pension expenditure even if benefits per retiree are being restrained.
High debt does not identify the beneficiary cohort. The incidence depends on the policy that produced the deficit and the assets, services, or economic stabilization it financed.
Liberal democracy is not a sufficient cause. Demography, housing institutions, pension design, turnout rules, fiscal institutions, growth, migration, and welfare architecture vary enormously among liberal democracies.
7. A Better Empirical Test Link to heading
A serious test of intergenerational extraction should proceed in four layers.
| Layer | Required evidence | What would count against extraction? |
|---|---|---|
| Lifetime fiscal incidence | Present value of taxes/contributions and age-specific transfers/services by birth cohort | Similar or improving lifetime net positions for later cohorts after risk adjustment |
| Political mechanism | Cohort/age turnout, policy responsiveness, lobbying, and reform votes | No systematic relationship between age-weighted political power and burden allocation |
| Asset-policy incidence | Housing supply, tax treatment, entry prices, and incumbent asset gains | Supply and tax institutions that do not systematically protect incumbents at entrant expense |
| Debt/investment decomposition | What deficits financed and their effect on future productive capacity | Borrowing that creates assets or growth benefits at least commensurate with future servicing costs |
The strongest design would compare successive birth cohorts, not age groups observed in one year. It would track lifetime taxes and benefits, housing-entry conditions, inherited public liabilities, public capital, and policy changes as each cohort passes through the same ages. Quasi-experimental reforms—pension-index changes, retirement-age increases, zoning reforms, tax changes, or voting reforms—can then test whether politically stronger age groups systematically avoid adjustment.
This approach also makes the thesis falsifiable. If later cohorts receive comparable lifetime net benefits, if aging triggers automatic pension restraint, if debt finances productive assets, and if housing supply remains responsive, then the extraction thesis weakens even when the population is old.
Conclusion Link to heading
The evidence supports a conditional theory of intergenerational imbalance, not a universal theory of liberal-democratic generational warfare.
Democracy creates a genuine long-horizon representation problem because future citizens cannot vote and young adults frequently vote less than older groups. Pay-as-you-go pensions create explicit cross-cohort financing. Persistent deficits can reduce future fiscal space and economic output. Housing politics can empower incumbent owners in ways that make entry more expensive for later households. These mechanisms are real and, in some cases, directly observed.
But the stronger language of extraction requires more. It requires showing that politically advantaged present cohorts obtain higher lifetime net benefits while externalizing costs to less represented cohorts after accounting for social insurance, public services, public investment, inheritance, and normal life-cycle wealth accumulation.
That standard changes the comparative judgment. The United States and Japan face severe intergenerational adjustment pressures, but those pressures do not themselves reveal the final cohort incidence. Germany shows that an aging electorate can coexist with substantial pension adjustment and fiscal constraint. Israel shows that neither low fertility nor a youth turnout penalty is inherent to liberal democracy.
The most defensible claim is therefore narrower and more useful: liberal democracies are vulnerable to intergenerational extraction when institutions allow politically powerful present cohorts to separate current benefits from their full lifetime costs, and the vulnerability grows when demographic aging, turnout asymmetry, housing scarcity, and persistent deficits reinforce one another. Whether that vulnerability becomes actual extraction is an empirical question about cohort lifetime incidence, not a conclusion that can be read directly from age, debt, or wealth statistics.