The life cycle hypothesis has become the dominant mode used to analyze the effects of a social security system on private saving, the labor/leisure choice, and social welfare. As both Barro and Samuelson indicate, a fully funded Social Security program (in a world of certainty) would drive out an equivalent amount of private saving. If the interest rate is r, the effects of a payment of a dollar into the social security pool while young would just offset* the effects of receiving (1+r) dollars as a transfer when retired. Papers by Diamond, Hi», and Samuelson, among others, have examined the effects of non-fully funded Social Security schemes in a growing economy. A non-fully funded program can be used to alter the private sector's saving rate and, hence, the capital/labor ratio. Social Security, then, can be used as a policy tool for achieving the (or some variant of the) golden rule growth path.
This paper was published in International Economic Review, Vol. 23, No.3 (Oct., 1982), pp. 647-658.
Enders, Walter and Lapan, Harvey E., "Social Security Taxation and Intergenerational Risk Sharing" (1979). Economic Staff Paper Series. 91.