Convergence of firm-level productivity, globalisation and information technology: Evidence from France
Highlights
► The firm-level productivity convergence process in the 1990s and the 2000s in France has slowed down. ► This is explained by the acceleration of the productivity of firms on the technological frontier. ► Globalisation and information technology may have been of more benefit to the most productive firms.
Introduction
There is substantial dispersion of productivity across firms, even within narrowly defined industries (for example Baily et al., 1992, Oulton, 1998 and Oulton, 2000), which gives rise to a convergence process amongst firms (Oulton, 1998, Griffith et al., 2002). This article describes and explains this convergence process using French firm-level data. We find that the speed of convergence has fallen significantly since 1992. Evidence is presented that globalisation (via the acceleration of exports), and the spread of information and communication technologies (ICT) can explain this slowdown.
Section snippets
Data and estimation strategy
We mainly use the FiBEn database, a firm level database maintained by the Banque de France which includes fiscal returns of some 282,000 French firms between 1991 and 2004. 45,000 of those firms are observed during the whole period. Although we present our results on the whole set of firms, they hold on the sub-sample of surviving firms. FiBEn covers the whole of the market sector although industry is covered better than services. This database is used to compute total factor productivity,
Stylised facts about convergence
We find a significant degree of -convergence in TFP (Table 1). For a firm at the lowest productivity decile, half the gap with its “target” level of productivity is eliminated in less than two years. Estimating -convergence year by year (Eq. (1) in GMM and GLS and Spearman rank correlation) shows that the speed of convergence falls during the 1990s, i.e. beta becomes less negative in Chart 1. Dispersion is also significantly increasing throughout the period (standard deviation in TFP rose
Acknowledgements
The views expressed here are those of the authors and not those of their respective institutions. This paper has been published in an extended and preliminary version in Economie and Statistique (in French). The authors thank Philippe Askenazy, Gilbert Cette, Francesco Daveri, Sébastien Roux and Patrick Sevestre and an anonymous referee for their valuable comments on an earlier version of this paper and Nicolas Berman and Laurent Eymard for excellent research assistance. This research benefited
References (7)
- et al.
Another look at the instrumental variable estimation of error-component models
Journal of Econometrics
(1995) - et al.
Productivity dynamics in manufacturing plants
Brookings Papers on Economic Activity: Microeconomics
(1992) - et al.
Is learning by exporting important? micro-dynamic evidence from Colombia, Mexico, and Morocco
Quarterly Journal of Economics
(1998)