Mortality risk and housing wealth of the aged

This thesis analyzes and empirically estimates a life cycle model of consumption and housing choices of American elderly. Unlike the existing literatures, I assume that the elderly utility depends on non housing consumption and housing consumption, and I combine Hurd's (1989) extended life cycle model and Artle and Varaiya's (1978) model. Using panel data from the Retirement History Survey, I first estimated the parameters of the model and found that the marginal utility of bequests from housing wealth is small and moving either to rental units or to small houses is superior to staying in owned houses for the majority of low and medium wealth households. I then used the estimated parameters to simulate the elderly choices on consumption and housing. I found that only 23.62 percent of households would prefer staying in their homes until they die, 37.49 percent will downsize their houses and 38.89 percent would move to rental units. Those who are predicted to stay in their houses have higher initial liquid assets and higher valued houses than those who are predicted to move. These findings contrast with most existing literatures and claim that the elderly tend to move or downsize their houses as they age. Descriptive analysis in chapter 3 also shows that the housing values and housing equities of the movers were down, while those of the stayers were up. And the housing sizes measured by the number of rooms were clearly decreasing for owner-to-owner movers, but unchanged for stayers.