Fiscal Decentralisation: The Swiss Case

Abstract

Switzerland provides a potential laboratory for testing various hypotheses connected with tax competition because of its extremely decentralized fiscal system. Twenty-six cantons (some of them extremely small) have retained the ultimate power of deciding tax questions, and hence not only to limit the Federal level of the State to about 1/3 of total public expenditure, but also to retain absolute power to set their own levels of taxation. Furthermore, citizens can launch referenda on tax issues at any level of government.Fiscal competition between cantons should therefore be easy to spot, resulting in lower tax cantons enjoying higher economic growth (the classical hypothesis), or in a race to the bottom with lower tax cantons offering a lower level of public services (the harmful competition hypothesis). In fact, neither of these hypotheses is confirmed. This leads the author to suggest a third hypothesis based on public choice theory (politicians in high-income cantons are led to raise high taxes). The author suggests that the classical and the public choice hypotheses cancel each other out, leaving a mixed picture. The article ends with some reflexions on the implications of the Swiss experience for fiscal harmonization at a European level.La Suisse offre un veritable laboratoire pour tester plusieurs hypotheses de concurrence fiscale en raison de son systeme fiscal extremement decentralise. Vingt six cantons (dont quelques uns tres petits) ont conserve le pouvoir supreme de decider les questions fiscales et, donc, non seulement de limiter le niveau federal de lEtat a 1/3 des depenses publiques totales, mais aussi de conserver le pouvoir absolu pour fixer leurs propres niveaux de taxation. De plus, les habitants peuvent lancer des referendums en matiere fiscale a tout niveau du gouvernement.La concurrence fiscale entre les cantons devrait etre par consequent plus facile a reperer, les cantons les moins imposes ayant une croissance economique superieure (hypothese classique), ou en un « nivellement par le bas » face aux cantons les moins imposes offrant un niveau inferieur de services publics (hypothese de concurrence nefaste). En fait, aucune de ces hypotheses nest confirmee. Cela conduit lauteur a suggerer une troisieme hypothese basee sur la theorie des decisions publiques (les politiciens des cantons a revenus eleves sont conduits a lever des impots eleves). Lauteur suggere que les hypotheses classiques et celles des decisions publiques se neutralisent lune lautre, laissant place a un modele mixte. Larticle se termine par quelques reflexions quant aux implications de lexperience Suisse pour une harmonisation fiscale au niveau europeen.