Social security and saving: A comment

We check the robustness of Feldstein's finding that Social Security reduces private saving to two specifications not considered before in the literatu re: (1) constraining the coefficients on Social Security and household net worth to be equal; and (2) allowing the effect of wealth on consumption to be nonlinear or a function of the age distribution of the population. We re ject Feldstein's implicit assumption of linear but unequal effects of wealt h on consumption, finding instead that the effects of both forms of wealth on consumption have risen over the postwar era. Only in recent years has th e estimated effect of Social Security wealth on consumption for some of the se specifications been close to Feldstein's estimate of 2.9 cents per dolla r of wealth.