Aggregate Matching Efficiency : A Stochastic Production Frontier Approach, France 1990-1994

We evaluate the determinants of aggregate matching efficiency changes through a stochastic translog production frontier model. The efficiency coefficient is represented by a stochastic function of variables meant to capture workers and firms characteristics. The model is estimated on French data covering twenty-two regions from March 1990 till February 1995. Our estimates suggest that aggregate matching efficiency has decreased steadily in the early nineties. There are also wide cross-regional differences. On average, about 30% of the variations of efficiency observed across time and regions can be related to changes in the explanatory variables used in the model. The most important explanatory variables are the proportion of youngsters, females and immigrants in the stock of job seekers. Long-term unemployment has a significant negative effect, population density a significant positive one. The huge decline in the proportion of permanent job offers has apparently little effect on matching efficiency.