Please use this identifier to cite or link to this item:
Miniaci, Raffaele
Panteghini, Paolo
Rivolta, Giulia
Year of Publication: 
Series/Report no.: 
CESifo Working Paper No. 6928
Tax competition has long been studied using panel models. According to this approach, each country’s tax rate is assumed to depend on a weighted average of the tax rates applied in the rest of the world, where weights are exogenous. As a consequence, the estimated reaction functions of countries throughout the world have the same sign. This means that all tax rates are either strategic complements or strategic substitutes. Moreover, the intensity of a country’s reaction depends on certain exogenous weights, with a unique proportional factor common across all countries. Our article departs from this standard approach and proposes a VAR model as an alternative estimation strategy. Accordingly, weights are no longer determined exogenously but rather endogenously. As such, we compare and explore the implications of the panel versus the VAR model in terms of structural contemporaneous parameters and impulse response functions. We show that results obtained with a VAR model differ from those obtained from a panel approach. In particular, we find that strategic complementarity between certain countries (with a positive slope of reaction functions) may co-exist with strategic substitutability between other countries (negative slope). Given these results, we can say that a standard panel approach is relatively restrictive and therefore can lead to unreliable estimates, and fail to provide helpful policy recommendations.
tax competition
VAR models
Bayesian methods
Document Type: 
Working Paper

Files in This Item:

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.