The near-retirement households among all working-age groups in the United States experienced larger drops in consumption and greater increases in labor force participation during the financial crisis of 2007-2009. Also, the retirement probability for near-retirement homeowners (but not for renters) decreases more in those areas where housing prices also decline more. This paper argues that the wealth effect of housing prices on retirement can account for those issues. It creates an incomplete-market life-cycle partial-equilibrium model with risky housing assets and endogenous retirement and verifies that the joint response of retirement and non-durable consumption implied by the structural model is consistent with the empirical findings using data from the Health and Retirement Study 1992-2012.It then shows that, after an unexpected 28 percent housing price decline, near-retirement homeowners ages 55-64 will reduce their non-durable consumption by 4.6 percent and increase their labor force participation by 1 percentage point immediately and delay their retirement by 2.8 months in the long run. The model also quantifies endogenous retirement as self-insurance for elderly homeowners against housing price risk.
Keywords: housing wealth effect, endogenous retirement, self-insurance
JEL: E21, E24, J26
Download the full text E2017008.pdf