IN recent years, an extensive literature has developed on the theoretical and empirical modelling of labour supply decisions, both in the UK, and the US. Interest in this area ranges from a desire to quantify the likely effects of government taxation measures, to the analysis and understanding of the social re-structuring of work, and the division between domestic and waged labour within the family. Empirical analyses include cross section studies both at the aggregate and micro levels (for example Greenhalgh (1977) (1979) (1980a), Layard, Barton and Zabalza (1980), Zabalza (1983), Ashenfelter and Heckman (1974)), longitudinal studies of the behaviour of individuals and households over a period of time (Heckman (1980), Elias and Main (1982)), and pure time-series studies of aggregate data (Corry and Roberts (1970)). A particular problem that has dogged such empirical studies lies in the proper identification of supply responses from demand side influences, and the adequate integration of both sides of the market within a unified structural model. Thus, within the context of the current economic recession, the observed patterns of employment, unemployment, and labour supply undoubtedly reflect complex interactions and adjustments in supply and demand. Traditionally such adjustments would, in theory, operate through movements in the real wage, serving to equilibrate the market. Clearly, however, the market for labour in the UK cannot, in practice, be assumed to be in a state of continuous equilibrium, whilst wages evidently respond to a number of influences, and only rather slowly to the downward pressure of excess labour supply. The modelling of such behavioural systems becomes necessarily rather complex (see for example Rosen and Quandt (1978)), and one important implication stemming from the possibility of constraints on labour supply lies in the possibility of adjustments in commodity demands in response to labour market disequilibrium (Ashenfelter (1980), Blundell and Walker (1982)). In this study a quarterly time-series model is developed, by constructing averages of certain variables relevant to supply decisions from the Family Expenditure Survey (FES), a continuous rotating sample of household behaviour. The sample is restricted to all households where both husband
Read full abstract