Abstract
This article attempts to estimate household tap water demand function using the data from a survey conducted on households in the Korean metropolitan cities by employing a sample selection model. The results show that there exists sample selection bias in the sample and that failure to correct for sample selection bias distorts the mean estimate of water demand upward by 13.8%.
Acknowledgement
This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOEHRD) (KRF-2003-041-B00675).