While there seems to be no end to estimates of housing tenure determinants, prior studies have not accounted for the simultaneity of tenure choice with household formation, labor supply or the marriage decision. Our estimates are superior to those in the literature both because we address these issues and because we better measure the cost of owning relative to renting. Accounting for simultaneity with the household formation and labor supply decisions matter. Using a household's predicted wage rate rather than its observed income doubles the response of tenure choice to the price of owning relative to renting. Including household formation selectivity correction variables cuts the response to tenure choice to the predicted wage by 25 percent. Moreover, the impact of variations in demographic variables on tenure choice is sharply reduced after correcting for selectivity bias.
