WP#2008-12

Abstract

Economic conditions and labor force participation vary significantly across the states of the Union.  Despite these marked differences, little is known about the reasons for such variations in retirement patterns.  Using the Current Population Survey for the period 1977-2007, this paper demonstrates that the differences in the labor force participation of men age 55-64 are related to the labor market conditions, the nature of employment, and the employee characteristics in each state as well as a pseudo replacement rate.  These variables explain more than one-third of the total variation.  Even moving to a fixed-effects model only cuts the explanatory power by half.  The question remains, however, whether these relationships reflect different populations or unique aspects of the state.  To answer that question we turn to the Health and Retirement Study (HRS).  We estimate equations for the probability of working and for the expected retirement for men in their late fifties and early sixties.  In each case, the first equation includes just the state-level variables and the second the state-level variables and the HRS demographic and economic information for each individual.  The results show that the state-level variables explain almost none of the variation in the probability of working or the expected retirement age, but most of the state-level variables are statistically significant both before and after the inclusion of the HRS information. 

For executive summary in PDF

For full paper in PDF

For related Issue in Brief