Labour and Capital Adjustment Dynamics Using Bivariate Random Probit

Abstract

This paper provides empirical evidence on the interrelationship between employment and capital adjustment decisions using a sample of Italian firms during the period 1989-1997. A dynamic bivariate probit model is estimated using a short-cut in the spirit of Heckman estimator. Unobserved heterogeneity and state dependence are found to play an important role for both hiring and investment equations. The estimates of the state dependence are significant and positive for both capital and labour implying that the convex components of adjustment costs are important for the adjustment process of capital and labour. The significant positive correlation between random effects and errors is a strong evidence of the simultaneous interrelationship between factor demand adjustment processes. A positive relation is found only between the occurrence of hiring spikes in one year and the investment spikes in the next year (and not vice-versa) which can be due to the fact that firms need anticipation of skilled labour and training in order to appropriately challenge investment strategies.