Alternative approaches to modelling individual transitions: the labour force participation of the wives of unemployed men in the UK

Abstract

This paper argues that for many dynamic labour market issues of interest both fully structural and reduced form empirical approaches are inappropriate. The former often pose quite intractable computational burdens while the latter is unduly restrictive. We illustrate our arguments in the context of the common finding of the lack of an "added worker effect" for the wives of unemployed males using UK data. Our empirical strategy, which combines elements of both approaches, indicates that complementarily between the non-labour market time (leisure) of both partners drives this low participation rate. While having a potentially large disincentive effect on wives' labour supply, the effect of the benefit system is small relative to household characteristics, especially the presence of young children.