The overlapping generations model originates with Allais (1947) and Samuelson (1958). Samuelson's pure-exchange consumption-loan economy, in which people live three periods, showed that a competitive equilibrium can be improved upon and that fiat money - a 'social contrivance' - can support saving across generations. Diamond (1965) embedded capital accumulation into the framework, in its now-standard two-period form. Diamond's contribution was decisive: by letting young agents save through physical capital rather than paper money, he showed that a decentralized economy can accumulate more or less capital than the Golden Rule level, and that over-accumulation (r<n) is Pareto-inefficient. The model became the canonical tool for intergenerational economics.
The core mechanism is life-cycle saving. Each cohort lives for two periods: young workers earn wages, consume some, and save the rest; old retirees consume their savings plus interest. Because generations overlap - the old coexist with the young - the capital stock in any period is exactly the saving of the previous young cohort, and capital per worker is that saving divided among the current, (1+n) times larger, cohort of young workers. This link between cohort saving and aggregate capital is the engine of the model. Each period is a whole stage of life, young or old, so every rate on this page (population growth n, the return r∗ and the efficiency gap) is per period, not per year: if a period is read as 30 years, the default return of about 0.025 per period is about 0.08% a year. The same reading explains full depreciation (δ=1): capital built in one stage of life is worn out by the next.
OLG is the workhorse for Social Security analysis, pension reform, demographic transition modeling, and public debt sustainability at institutions like the Congressional Budget Office, the IMF Fiscal Affairs department, and central bank long-run projection units. Auerbach and Kotlikoff (1987) scaled the two-period structure to a 55-period realistic life-cycle model that remains the template for applied fiscal policy analysis.
Key extensions include Blanchard's (1985) perpetual youth model with a constant probability of death (tractable for continuous-time analysis), Gertler (1999), who adds random transitions from work to retirement so that life-cycle saving survives in a tractable aggregate model, and Krueger and Ludwig (2007) on demographic transition effects. The stochastic OLG literature (Rios-Rull 1996, Conesa and Krueger 1999) adds idiosyncratic risk and incomplete markets, bridging toward the heterogeneous-agent tradition.