The Macroeconomic Effects of Tax Changes: Evidence from Southern European Regions

An original panel dataset of exogenous fiscal shocks for some Mediterranean countries is used along with the so-called "narrative" approach introduced by Romer and Romer (2010) to estimate the tax multiplier. Exogenous shocks comprise only tax changes motivated by the willingness to improve long term growth, reduce government debt as well as achieving redistributive goals. Endogenous shocks are excluded from the data since they are likely to be correlated with output. The aggregate tax multiplier on output is much smaller than the one found in Romer and Romer (2010): -1, instead of -3 at a three years perspective. When considering the effect of tax changes on unemployment, a multiplier of +2.8 is found on a two years horizon; a similar coefficient, with a negative sign, is found relating government tax liabilities with private investment, on the same temporal horizon. Splitting the sample between pre and post-crisis period, a multiplier non significantly different than zero is obtained for the first partition, while significant and equal to -0.75 for the second subsample.