Abstract:
Empirical residential location choice models have occasionally reported estimated coefficients of dwelling-unit price that are small, statistically insignificant, or even positive. This would imply that households are insensitive to changes in dwelling-unit prices and taxes; this not only is counterintuitive but makes the models useless for policy analysis. One explanation for this result is price endogeneity, which means that the included dwelling unit’s price variable is correlated with the model’s error term. In discrete choice models this problem is caused principally by the omission of attributes correlated with price. An application of the control-function method is presented as the most suitable to correct for endogeneity in choice models of residential location. The method is tested for robustness to different error structures by using two Monte Carlo experiments, and it is also applied to real data from Santiago, Chile. The results show that price endogeneity is significant in choice models of residential location and that the control function method is suitable to account for it.
