The distortion of the labor/leisure choice by social security during the period the earnings test is in effect is well known. This paper, using a life-cycle asset maximization approach to social security acceptance, shows that the earnings test is not a sufficient cause for such a distortion in the constrained period or over the life cycle. We use time-series analysis to test the net empirical importance of the substitution and wealth effects associated with social security on the market work of younger men and find that hours worked per week would have fallen from 2 to 3 hours since 1936 without the present social security system. Such findings suggest that large savings effects associated with social security are over-estimates.

Charles R. Plott, Vernon L. Smith; An Experimental Examination of Two Exchange Institutions, The Review of Economic Studies, Volume 45, Issue 1, 1 February 1978

This paper presents the results of an evaluation of the role of real cash balances as a factor input for 11 two-digit SIC code industries over the period 1952-73. Using a four-factor translog cost function for each industry along with duality theory, it was possible to estimate the partial elasticities of substitution and the elasticities of demand for all factors. The substitution elasticities between real cash balances and production labor as well as with capital were found to be significantly different from zero. The interest elasticity of demand for each varies with industry and ranges from -.22 to -.41. The overall findings suggest that the neoclassical model offers considerable promise for modeling the firm's demand for money.

The distortion of the labor/leisure choice by social security during the period the earnings test is in effect is well known. This paper, using a life-cycle asset maximization approach to social security acceptance, shows that the earnings test is not a sufficient cause for such a distortion in the constrained period or over the life cycle. We use time-series analysis to test the net empirical importance of the substitution and wealth effects associated with social security on the market work of younger men and find that hours worked per week would have fallen from 2 to 3 hours since 1936 without the present social security system. Such findings suggest that large savings effects associated with social security are over-estimates.