The choice concerning the method of acquisition of the property (either owned or rented) is closely linked to other individual choices of great importance not only for consumers but also for the country where these consumers live. The article “What are the determinants of the mode of housing tenure?” aims to make a comparison of the various types of approaches that have been used in the economic literature on the analysis of the decision of individual housing tenure . The models used in economics are different and vary from binary choice models to model duration. In this aspect, the determinants of tenure are unclear because depending on the definition of the dependent variable and the estimation method, the determinants of tenure vary.
The study conducted in this study examines the various models that have been used in economic theory to model correctly the analysis of housing tenure. In short, the list of models to consider are: the classical model of tenure (classical model), which describes the probability that an individual homeowner is, at any given time, the model estimated only for those individuals who have made a transition and therefore have already opted for a form of tenure (model recent migrants), sequential bivariate probit model, which estimates the tenure decision and the status of migrants together, the duration model, which reflects the decision to do a transition from an specific state tenure (rental or ownership) to another state at any given moment of time, and finally the transition model, which uses a sample based in the tenure decision, giving this more weight in the sample. The data for this analysis was obtained from the European Community Household Pamel (ECHP), in particular the periods from 1994 to 2001. The ECHP is a standardized survey that is conducted in the European Union. This database contains information about housing characteristics and its inhabitants. From these data, we conducted a sampling on Spanish owners who had been on the panel for more than one period. This estimation had data for about 6778 households, nearly 82% of which were proprietary. In addition, a 40.36% of those had perfomed at least one movement during the five periods observed. The definition of the dependent variable for all the models in which the dependent variable is the decision between the decision to rent or buy property is as follows: the status of tenure, (variable that has value 1 if the consumer owns or 0 if not), “move”, (in the equation used in the model selection of recent migrants), “stayer” (the second equation for the sequential model), transition (this variable takes values 1 for those individuals who make the transition of ownership to property, taking 0 in any other cases), income (estimated permanent income and transitory income based on the current monetary income’s owner of the of the household head), gender (This dichotomous variable takes values 1 if the person providing the main source of income is a man and 0 otherwise) and age, property size (how many people are in the house), marital status, work and education, relative price (this is the price relation-tenant) duration (for those owners who have made a transition) and finally the number of rooms.
The results of this model are obtained by grouping the various models analyzed in terms of the dependent variable and analyzing these based on your top 6 formats. The first group, in which the dependent variable is the state of holding together the neoclassical models, the model of recent migrants and the sequential model. The second group, in which the dependent variable is the transition from rental to home ownership, includes duration and transition models. Overall, the results provide evidence of a better performance in models in which the dependent variable is the transition between renting and home ownership. This higher performance is not only explained by the results observed in terms of goals but also the ability to provide predictive and marginal effects closer than previously predicted. In terms of predictive ability, both models (duration and transition models) show relatively high success rates. To summarize, we conclude that, if we want to describe the profile of consumers who are home owners, permanent income, age, marital status and level of education are the main explanatory variables. All these variables can be considered life cycle, with marital status with a greater impact. On the contrary, if we want to answer the question about what are the determinants of the transition to rental property, we conclude that, in addition to life-cycle variables, variables related to prices and transaction costs are also important determinants. In conclusion, the results underline that the model in which the dependent variable is the fact of owning or not a house underestimate the effects of prices and transaction costs.