Household Demand for Housing in Kenya

Previous studies have established that demand pattern for housing identifies with that of a necessary good. The nature of housing goods consumed is highly heterogeneous in dimensions like ownership, size, location and tenure type. In this study I model the probability of household tenure types in Kenya using Household Budget Survey data. Using a Multinomial Logit Model (MNLM) formulation three tenure, household size and age of the household head emerge as significant predictors of household housing choices. Based on the results, the government and firms can rely on household size and age of the household head to approximate demand for housing services in the various tenure types. Furthermore, the ability to interpolate these variables based on occasional population surveys makes it easier to design mechanisms of matching demand and supply for housing services