Individuals develop adverse health conditions as they get older, which limit their earnings, resources, and well-being. Workers also face uncertainty about their life spans, which affect their savings and consumption decisions. Many well-known economic characteristics of these risks prevent private markets from pooling these risks adequately. Pay-as-you-go social insurance programs – by levying payroll taxes on workers and paying benefits to disabled workers, retirees, and survivors – can substitute for these missing markets. In addition to sharing these two types of risks, such programs will affect the aggregate capital stock, consumption, and the elderly poverty rate. This paper models the risk sharing and the aggregate effects using a calibrated stochastic overlapping generations general equilibrium model. The paper uses as a benchmark the stationary equilibrium of an economy in which individuals self-insure themselves against these risks by adjusting their own saving, and compares this benchmark to the stationary equilibriums of economies with simplified old-age insurance (OASI) and disability insurance (DI) programs, alone and in combination (OASDI). The representative worker’s lifetime welfare is computed in the benchmark and the alternatives. Optimal replacement rates for the model’s OASI and DI programs are also calculated. The main findings of the paper are that the OASI and DI programs in the model economy together improve the lifetime welfare of workers, reduce the variation of consumption and asset holdings over the life cycle, and virtually eliminate elderly poverty that may arise from these two sources of income shocks. The welfare-equivalent increase in consumption at all ages in the benchmark would be about 14 percent for the DI program alone or 11 percent for the combined programs. The OASI program in the model, however, does not improve individual well-being. This might be due to several factors. First, similar to other models in the literature, this model assumes that the accidental bequest of an individual is redistributed equally among living individuals, which effectively mimics an actuarially fair annuity market with some progressive redistribution. This undermines the annuity role of the OASI program. Second, the model disregards the survivors’ benefits components of the OASI program, which would increase the welfare of an individual from an increase in welfare of his survivors. The paper finds that the optimal replacement rate to be zero percent for the OASI program without the survivors’ benefit component and with equal sharing of unintended bequest, and the optimal replacement rate for the DI program to be 90 percent.