Households live for two periods: young and old.
That keeps the intergenerational transfer logic visible.
Cookies on Macro by Mark.We use essential cookies for sign-in, locale, and security. With your permission, we also use optional cookies for preferences, for analytics, and for advertising measurement with the Meta (Facebook) Pixel. Read our cookie policy.
A two-period overlapping-generations framework linking saving, population growth, and capital accumulation across cohorts.
This theoretical page is a short sketch of two-period overlapping-generations saving with its own illustrative calibration, including partial depreciation. The DSGE OLG page solves the Diamond (1965) model with each period read as one generation and full depreciation, so its capital and return figures differ from this page's.
How do saving choices and demographics change capital accumulation across generations?
Paul Samuelson (1958) introduced the overlapping-generations structure to study money and intergenerational exchange. Peter Diamond (1965) generalized it into a full capital-accumulation model where finitely lived agents save for retirement, and the aggregate capital stock emerges from their choices.
OLG answers questions that representative-agent models cannot pose: why do the young save and the old dissave? How do unfunded pension systems alter the capital stock? What happens to growth when the population ages and the ratio of workers to retirees shrinks?
A distinctive result: when the interest rate falls below the population growth rate (), the economy is dynamically inefficient: it has accumulated more capital than is optimal, and reducing saving would make every generation better off. This possibility does not arise in infinitely-lived-agent models.
Capital installed for use by the next cohort.
It is not drawn as a separate element: it works through the equations behind the chart.
One horizontal-axis unit is one cohort transition; the plotted points are generations 0 through 20.
Depreciation changes the old-age net return but not this Diamond asset-market-clearing map; young saving purchases the full next-period stock.
Slower demographic growth lets the capital stock thicken more easily.
Two generations coexist: young workers who earn wage income and save for retirement, and old retirees who consume their savings plus the return. In this version the young save a fixed share of the wage; with log utility and a discount rate , that share would be .
Next period's capital stock is the saving of today's young. Young cohorts build it up; old cohorts draw it down. Steady-state capital per young worker depends on the saving rate, population growth, the capital share and productivity, ; the interest rate is then pinned down by .
A pay-as-you-go pension system transfers income from the young to the old, so the young save less and the steady-state capital stock falls. Comparing with tells you whether the system improves welfare (it can when , dynamic inefficiency), not whether it reduces capital.
That keeps the intergenerational transfer logic visible.
Wages and returns both move with capital deepening. Young-cohort saving purchases the full next-period capital stock: . Undepreciated capital belongs in the old cohort's gross payoff, so delta changes the return rather than adding a separate term to the capital-market law.
Pension reform runs through OLG logic. A pay-as-you-go system stays solvent when population growth supports a favorable worker-to-retiree ratio. When fertility drops, the ratio inverts, and the system faces pressure to raise contribution rates, cut benefits, or increase the retirement age.
In this two-period Diamond model, slower population growth raises steady-state capital per worker and the wage and lowers the return on capital, because each cohort's saving is spread over fewer young workers. How a rising old-age dependency ratio changes aggregate saving needs a model with many overlapping cohorts, where the net effect on capital per worker is debated.
OLG models evaluate policies whose costs and benefits fall on different generations, such as child tax credits, education subsidies and public debt. A policy that benefits today's old at the expense of tomorrow's young can be a Pareto improvement only if the economy is dynamically inefficient (). Under dynamic efficiency it helps today's old at the expense of later generations: a transfer, not a free gain.
Population growth falls because of declining fertility. How does the steady-state capital per worker, interest rate, and wage rate adjust? Is the economy more or less likely to be dynamically inefficient?
Compare pay-as-you-go versus fully funded pension systems. Which one reduces the steady-state capital stock, and why?
Should a dynamically inefficient economy (r < n) run a permanent public debt? Why does the relationship between the interest rate and the growth rate determine the answer?
The retirement age rises unexpectedly for workers in their 30s. How do they adjust saving, and what is the aggregate effect on capital accumulation?