This paper uses a lifetime framework to address questions about the progressivity of social security and proposed reforms. We use a large sample of diverse individuals from the PSID to calculate lifetime income, to classify individuals into income quintiles, and then to calculate the present value of taxes minus benefits for each person in each group. In our basic calculations, the current system is slightly progressive, overall, on a lifetime basis. Social Security would become slightly more progressive in one of the reform plans, and it would become slightly regressive in each of the other plans. The pattern of progressivity is affected by alternative assumptions, but it is