For birth cohorts 1935–44, 1945–62, and 1964–74, we estimate the contribution of edu-cation; permanent heterogeneity in wage rates, employment, and hours; labor market shocks; spouse characteristics ...