Abstract
This study provides a nonparametric test for the parametric conditional distribution in the sample selection model. The nonparametric test statistic proposed by Fan et al . (Citation2006) is applied to the Korean housing demand model. Test results indicate that the conditional normality in the housing expenditure equation is rejected.
Acknowledgements
The author wishes to acknowledge the helpful comments of Qi Li and Inchul Kim. The author's research is partially supported by the Second Stage of BK (Brain Korea) 21 Project.
Notes
1 A shortcoming of their test is that it can detect Pitman local alternatives approaching the null at a rate slower than O(n −1/2) while the detecting rate of Andrew's test for Pitman type alternatives is O(n −1/2).