Participation and Hours of Work for Married Women in Great Britain *

THE estimation of labour supply functions for men and women from household survey data has been undertaken by several authors using cross-section data for various years in the U.S. (A collection of such work appears in Cain and Watts, 1974.) Recently some estimates for men in the U.K. have been obtained by Brown et al. (1976) using a small sample of data for 1971. However there are no previous estimates for British women which utilise individual data on participation and hours, although town participation rates have been analysed (Greenhalgh, 1977). In this paper we examine the supply responses of married women using data for British families surveyed in 1971 in the General Household Survey. This is a large, nationally representative sample providing detailed information of hours of work, earnings, family and personal characteristics for individuals and households in Great Britain. Details of data definitions and sample selection are given in the Appendix. In part I, the model is outlined and the estimation problems are discussed. Two procedures are suggested for obtaining estimates of wage and income elasticities, together with the effects of other variables which may be thought to affect preferences for work and leisure. In part II, the empirical estimates are presented and both methods yield positive wage elasticities (close to unity) and negative income elasticities (mlmus two-thirds) for total labour supply. The separate components relating to probability of participation and to hours of work of participants reveal smaller wage and income elasticities for the former. The effects of children, colour, qualifications and geographical location on participation and hours of work are also investigated in detail. In the concluding section, the implications of some of the findings for government wage and tax policy are presented.