Measuring the Flemish competitivity through total factor productivity

The competitivity of the Flemish economy is a continuously ongoing priority of policymakers. Very often the focus is only on labour costs in comparison with the neighbouring countries (measured as labour cost per unit of output, for instance total labour cost divided by value added). A wider analysis also takes into account labour productivity (measured as value added divided by employment). A comprehensive measure of productivity is total factor productivity. Total factor productivity does not focus only on the production factor labour but also takes into account the other production factors, such as capital and energy. Total factor productivity is not available in the national accounts or other official databases. It is also not possible to obtain it via a simple operation of available series (such as a division of two series). Total factor productivity needs to be derived through econometric estimations (and by imposing hypotheses). For instance one could impose a Cobb-Douglas production function which needs to be estimated (after log-linearisation). The OLS estimator, however, is inadequate because of endogeneity problems. Using a 2SLS estimation method could be a solution. Other possibilities are panel data estimation and the method proposed by Olley and Pakes (1996). The goal of this paper is to derive sectoral total factor productivity for the Belgian regions and to show a graphical instrument for policy making purposes, as proposed by Goesaert and Reynaerts (2012). The graphical instrument allows to identify strong and weak sectoral branches.