Federal Planning Bureau
Kunstlaan/Avenue des Arts 47-49, 1000 Brussels
The determinants of industry-level total factor productivity in Belgium
Bernadette Biatour, email@example.com, Michel Dumont, firstname.lastname@example.org, Chantal Kegels, email@example.com
Abstract ‐ In this Working Paper the impact of potential determinants of total factor productiv‐ ity, i.e. the part of output that cannot be explained by the quantity of production factors, is es‐ timated for Belgium using industry‐level data for the period 1988‐2007. Abstract ‐ In deze Working Paper wordt de impact van mogelijke determinanten van totale fac‐ torproductiviteit, d.w.z. het gedeelte van de output dat niet verklaard kan worden door de ge‐ bruikte hoeveelheid van de productiefactoren, geschat voor België, op basis van bedrijfstakge‐ gevens voor de periode 1988‐2007.
Abstract ‐ Dans ce Working Paper, l’impact des déterminants potentiels de la productivité tota‐ le des facteurs, c’est‐à‐dire de la part de l’output qui ne peut pas être expliquée par la quantité de facteurs de production, est estimé pour la Belgique en utilisant des données sectorielles cou‐ vrant la période 1988‐2007.
Jel Classification ‐ C82, D24, F43, O47
Keywords ‐ Total factor productivity, R&D, human capital, competition
With acknowledgement of the source, reproduction of all or part of the publication is authorized, except for commercial purposes.
Legal deposit ‐ D/2011/7433/12
Responsible publisher ‐ Henri Bogaert
WORKING PAPER 7-11
The share of the output of a company, industry or country that cannot be explained by the amount of capital, labour and other factors used for production is called total factor productiv‐ ity (TFP). TFP growth is considered as a proxy for disembodied technological change, which, following the contributions to neoclassical economic growth theory is believed to be the pre‐ dominant explanation of economic growth in developed countries. As total factor productivity is a residual, it is likely to be a biased indicator of technological efficiency due to measurement errors, omitted variables, aggregation and misspecification. If the heterogeneity in efficiency of capital goods and workers is taken into account, the TFP growth residual is often reduced dra‐ matically, even to the extent that it suggests that almost all technological change is actually em‐ bodied in production factors (e.g. ICT in capital goods) and that there is hardly any disembod‐ ied technological change. In this paper, we dwell on the issues that make the measurement of TFP and its interpretation as an indicator of technological efficiency troublesome. Despite the large list of problems involved in measuring total factor productivity, it is generally considered as a proxy of technological efficiency and, more in general, a major determinant of welfare. This warrants an analysis of the factors that may determine TFP. Research and Devel‐ opment (R&D) activities of firms, universities and research institutes are generally considered as the main determinant. Not only own R&D activities may affect innovation and productivity growth, but firms may also benefit from R&D performed by other domestic or foreign firms or research organisations. Empirical studies have shown the importance of these so‐called spill‐ overs for TFP and provide an argument for governments to stimulate R&D, e.g. through subsi‐ dies or fiscal support, as spillovers also imply that firms cannot fully appropriate all benefits resulting from their own R&D and ...