Cambridge Journal of Economics 23:623-640 (1999)
Copyright © 1999 Cambridge Political Economy Society
Article |
A trivariate model of participation, fertility and wages: the Italian case
Department of Economics 'Cognetti De Martiis', University of Turin, Turin, Italy
Centre for Business Research, University of Cambridge, Cambridge, UK
z Correspondence address: ESRC Centre for Business Research, Austin Robinson Building, Sidgwick Avenue, Cambridge CB3 9DE, UK
E-mail: md236@econ.cam.ac.uk
Abstract
Italy has unusually low fertility by OECD standards, accompanied by unusually low female participation in paid work. This paper addresses the issue of the empirical relationship between fertility, female participation in the labour market and wages with these Italian 'peculiarities' as a backcloth. A trivariate model of participation, fertility and wages has been constructed and estimated using three pooled cross-sections of Italian micro data, allowing for the identification of cohort effects. This model follows a 'purist' approach: the participation and fertility decisions, as well as the wage equation, are modelled as completely joint. The cohort effects turn out to be significant: the point estimates do not appear to confirm actual trends, which are negative for fertility and positive for participation. The female wage is the most important variable influencing the propensity to have children and the propensity to participate in the labour market, casting doubt on suggestions that observed trends are the products of shifts in women's 'tastes'.
Key Words: Labour supply fertility discrete regression and qualitative choice models