| Summary: | 碩士 === 國立成功大學 === 都市計劃學系碩博士班 === 93 === Since there is a cost associated with ownership (the transactions cost of selling a house) and the various cost of renting a dwelling exceeds the various cost of owning an equivalent unit, and related researches show that the portal factor in determining the least costly tenure alternative is the length of time the household stays at a given location, thus the longer a household stays at a given location the more likely it is that the household will choose to own a house. Thus, the probability that a household makes a given tenure choice and expects to move should be estimated as a simultaneous system of equations.
However, related researches on the topic are scant, some of them didn’t consider the two decisions are causal chain relation and estimated the relationship by using the multinomial Logit model, some researches considered two decisions are determined in order and estimated by using the nested multinomial Logit model. This dissertation is to test the hypothesis that they are simultaneously determined by using a simultaneous Logit model. To improve the multinomial Logit model in dealing with the problem that the independence from irrelevant alternatives, this simultaneous Logit model is estimated by a two-stage procedure.
This study uses the Panel Study of Family Dynamics (PSFD) to establish the simultaneous choice Logit model on tenure choice and prospective moving decision. The model will estimate by using maximum likelihood method of multinomial Logit first. Second, the two bivariate recursive models will be established. A simultaneous discriminative Logit model is finally established in order to test the causal relation between two decisions.
In the empirical analysis it was found that the coefficient on the endogenous variables (tenure choice and prospective moving decision) is negative and significant. The effects of a vector of independent variables on the household’s tenure choice and prospective moving decision are causal and simultaneously determined. In addition, we note that the effects of age and marital status are significant and positive on the probability of ownership, and the effects of the size of family is significant and negative on the probability of prospective moving.
|
|---|
