How much has labour taxation contributed to European structural unemployment?

Abstract

We analyse the effect of labour taxes on euro area unemployment. Making use of the Blanchard and Katz (1999) [Wage dynamics: reconciling theory and evidence. American Economic Review 89(2), 69-74.] NAIRU model, we represent structural unemployment as that made up of two elements: an observed part related to labour taxes plus a non-stationary unobservable component related to the reservation wage and to the wage mark-up. Contrary to empirical estimates available so far, this modelling explicitly acknowledges the existence of non-measurable factors responsible for the unemployment non-stationarity. For the period 1970-2004, we obtain a tax effect that lies in the middle of the estimates that can be found in the literature. This study gives support to the view that lowering labour taxes can help to reduce unemployment in continental Europe.

Labour taxes and unemployment evidence from a panel unobserved component model

“…Yet, extensive empirical research has not succeeded in providing robust evidence confirming the alleged positive relation between taxes and unemployment. Surveying the literature, the estimated elasticity of unemployment with respect to taxes ranges from zero (Bean et al, 1986;Layard et al, 2005;Nickell, 1997;Blanchard and Wolfers, 2000) over medium-sized (Elmeskov et al, 1998;Nickell and Layard, 1999;Nickell et al, 2005;Planas et al, 2007) up to large (Daveri and Tabellini, 2000). In our opinion, the reason for these conflicting results is threefold.…”

“…Building on the above literature, Planas et al (2007) extend the Blanchard and Katz (1999) NAIRU model to allow for labour taxes. They show that equilibrium unemployment can be seen as the sum of an unobserved component related to non-measurable factors and an observable component related to the effect of labour taxes, with the extent of the latter effect depending on the bargaining strength of workers, the indexation of the reservation wage to labour taxes, and the elasticity of wages with respect to the unemployment rate.…”

“…The finding in Berger and Everaert (2009), for instance, that unemployment does not cointegrate with a large set of labour market institutions might be caused by omitted non-stationary variables and, as such, does not imply that there is no long-run relation between unemployment and labour market institutions. To the best of our knowledge, Planas et al (2007)is the only empirical paper on this subject to account for the missing variables problem. They estimate the impact of labour taxes on unemployment within an unobserved component (UC) model in which the equilibrium rate of unemployment is made up of a non-stationary unobservable component capturing non-measurable factors plus an observed component that represents the effect of labour taxes.…”

“…Panel data unit root and cointegration tests show that unemployment and labour tax rates are non-stationary but not cointegrated. As this finding may be induced by missing non-stationary variables we set up a panel UC model similar to Planas et al (2007). The main difference is that instead of aggregating the data we add a cross-sectional dimension by pooling countries.…”

Financing Social Security by Taxing Capital Income - A Bad Idea?

“…Empirical evidence by Daveri and Tabellini (2000) and, more recently, by Planas et al (2007) suggests that a significant part of European unemployment can be traced back to a steady rise in the costs of labour. Consequently, there exists a direct link between wage-related social security contributions and unemployment.…”