This paper sets out a theoretical framework for models of the household production and labour supply decisions of families, and estimates empirical specifications on time use survey data containing information on labour supply, domestic work and pure leisure. The models are then compared with the standard household model estimated on household survey data containing information on time allocated to labour supply alone. The aim of the paper is to demonstrate, both theoretically and empirically, that the “income pooling hypothesis� of the standard model cannot be tested on household survey data of this kind. Due to missing information on individual consumptions, including pure leisure, these data allow only the implied aggregation restrictions of the model to be tested. Nor can individual preference parameters, or the parameters of a sharing rule, be identified unless the empirical specification imposes some essentially arbitrary assumption. The study goes on to examine the gender implications and implied child costs of various assumptions that have been introduced in the literature on modeling multi-person househol
