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

We evaluate the determinants of aggregate matching efficiencychanges through a stochastic translog production frontiermodel. The efficiency coefficient is represented by a stochasticfunction of variables meant to capture workers and firmscharacteristics. The model is estimated on French data coveringtwenty-two regions from March 1990 till February 1995. Ourestimates suggest that aggregate matching efficiency has decreasedsteadily in the early nineties. There are also wide cross-regionaldifferences. On average, about 30% of the variations ofefficiency observed across time and regions can be related tochanges in the explanatory variables used in the model. The mostimportant explanatory variables are the proportion of youngsters,females and immigrants in the stock of job seekers. Long-termunemployment has a significant negative effect, population densitya significant positive one. The huge decline in the proportion ofpermanent job offers has apparently little effect on matchingefficiency.