[This paper estimates all possible multidimensional interaction effects in a logit model of homeownership, assesses the relative importance of these interactions, and interprets the results in light of existing theories of housing consumption.]