Why Are Older Workers Moving Less While Working Longer?

Abstract

Older workers’ labor force participation (LFP) and migration across state lines have been trending in opposite directions, counter to conventional economic wisdom. This paper investigates what might explain this puzzle using data from the Current Population Survey (CPS) and Health and Retirement Study (HRS). Descriptive analysis identifies several factors that may explain the decline in migration, including greater housing price dispersion, fewer opportunities for wage arbitrage, and greater geographical sorting. I employ a series of empirical tests to examine how older workers’ LFP, retirement, and migration decisions respond to income and housing wealth losses by exploiting job losses to identify individual income shocks, and shocks to specific labor markets to identify housing wealth losses using an import competition shock that began in 2001 after Congress ratified permanent normalized trade relations with China in October 2000. The puzzle appears to be driven by composition effects. In response to a housing wealth shock, non-college educated homeowners (the largest subgroup of older workers) reduce their two-year migration rate by 54% but only slightly reduce their labor supply, while college-educated renters (the smallest subgroup) increase their labor supply by 13% but only weakly increase their propensity to move.