Children's time in household work: Estimation issues

Abstract

The purpose of this article is to analyze children's time in household production using tobit analysis to adjust for nonpartlcipation and to compare the results to a regression analysis. In general, more variables are significant in each of the AnMyses based upon gender and birth order of the children. More importantly, the additional variables go beyond whether the day was a schoolday and the age of the child. The past failure of children's time spent in household work to change with differences in their families' characteristics appears to have been an artifact of not accounting for nonparticipation when estimating m~r~d'nA1 effects.