Household labour supply in Great Britain: can policy-makers rely on neoclassical models?

This thesis empirically examines whether the neoclassical economic model provides an adequate framework to analyse a couple’s labour supply behaviour in Britain using recent data from the British Household Panel Survey. The thesis comprises three empirical chapters. The first chapter uses the instrumental variable (IV) estimation procedure to model the hours of work of married couples. This approach allows us to test whether some of the assumptions of the neoclassical model (e.g., income pooling and Slutsky properties) are satisfied by the data. In addition, further variables that have been identified as distribution factors in the literature are introduced to the empirical model to assess whether they play a role in explaining a couple’s hours of work. The first chapter only considers couples in which both spouses work. In the second chapter, the sample is amended to include all couples (i.e., those that work and those that do not) and the analysis conducted models a couple’s labour market participation decisions rather than their hours of work. After testing for income pooling and the impact of distribution factors, a further variable, the wife’s mother-in-law work status when the male spouse was aged 14, is introduced into the model. This is done to determine the effect of 'cultural' variables on labour market decisions. In the last chapter, this issue is explored further by explicitly modelling attitudes to a woman’s role in the labour market. This approach uses a bivariate ordered probit model given the ordinal nature of responses to the attitudinal questions and again restricts the analysis to couples only. Finally, gender-role attitudes are introduced to the labour supply framework used in the second chapter in order to evaluate whether beliefs regarding women’s role impact on a couple’s labour market decisions.