%0 Trabajo de grado (Bachelor Thesis) %A Quintero Escobar, Luis Eduardo %D 2014 %G Desconocido (Unknown) %T Essays on specification and estimation of models of markets for heterogeneous housing = Ensayos sobre la especificación y estimación de modelos de mercados de vivienda heterogéneos %U http://babel.banrepcultural.org/cdm/ref/collection/p17054coll23/id/352 %X My dissertation proposes new ways to study decisions in the housing market with structural models that deliver a consistent connection between house values and rents, incorporating investment and consumption motives for housing consumption, and allowing for preference heterogeneity. The models also take into account the difficulty of measuring complete housing quality in estimation by treating it as latent. The dissertation includes applications where I analyze issues like the size of agglomeration benefits of larger cities (New York City vs Chicago), and the dynamics of rent and prices as a reaction to household financial and demographic conditions during the re- cent housing crisis (in Miami). I find that New York has relatively lower qualities of housing and higher prices than Chicago, which implies positive compensating variations required by households moving from the latter to the former; and that there was a disconnect between the rent and ownership markets that suggest that the crisis mainly affected the asset and not the real market, respectively. Also, I find that heterogeneity in tastes plays an important role in the behavior of households in the housing market. I also present models that introduce heterogeneity in tastes to analyze the interaction between households with different demographics. Going beyond a single type of household presents additional challenges for estimation and identification that are addressed. I define types as households with and without children, and asses the effect of the presence of children on how sensitive households are to changes in the market. I obtain the robust result that the presence of children reduces the sensitivity of demand for housing quality to changes in the market. Additionally, I also use machine learning clustering methods to categorize house- holds into different demographic types based on age, number of children and expenditure in the housing market. With these types, I per- form analyses that contrast the implied demands for housing quality in the different obtained clusters. The methods developed in the different chapters are of interest to policy makers and agents who want to follow the evolution of rental rates and prices at different levels of the quality distribution, as well as for those interested in making welfare evaluations of policies that affect local housing markets differently for households with different demographics. Also, they give insight on agglomeration benefits of larger cities and compensations required when moving a household from one area to another.