This paper investigates the impact of corruption on economic growth in the Italian Regions. We estimate a dynamic growth model for the period 1980-2004 addressing both the potential bias of the measures of corruption and the endogeneity between corruption and economic development. We find strong evidence of a negative correlation between corruption and growth. Moreover, since government intervention has been traditionally used to reduce income differentials between the Northern and the Southern regions, we also analyze the interaction between corruption and government expenditure. Our results indicate that corruption undermines the positive impact that public expenditures have on economic growth.

We analyze the effect of different legislature size on per capita regional expenditure in Italy. According to the theory, legislature size has an indefinite effect on government spending because logrolling and transaction costs may have canceling effects. We find a large and significantly positive effect of the number of legislators. We use these findings to forecast the effects of the increase in the number of legislators that is taking place in some regions: a 10% increase in legislature size commands on average a 12% increase in per capita regional expenditure. Copyright Springer Science+Business Media, LLC 2007Legislature size, Regional expenditure,

This paper tests the explanatory power of alternative theories on the determinants of judiciary independence using annual and decision-based data on the Italian Constitutional Court. The estimates show that structural measures of judiciary independence, such as the share of constitutional judges elected by the ranks of the professional judiciary and the age of justices, are positively correlated with independent behavior of the Court. Contrary to previous studies on a similar sample, we find that the Court counteracts the greater cohesion of the other government branches with more independent behavior, improving the effectiveness of the system of checks and balances.

We argue that the decision to vote in European Parliament (EP) elections is made at the intersection of three political dimensions: one related to the attitude of citizens towards the European Union (EU), one to the characteristics of the national political system and one associated with socioeconomic variables observed by voters at the local level. This article investigates this intersection by analysing the last four EP elections in the EU‐14, for 164 regions. We tested a multilevel model. The results indicate that compulsory voting, domestic political cleavages, labour market conditions and trust in the EU play a significant role. No evidence was found that GDP per capita affects turnout. Finally, the oldest segment of population seems more likely to vote than the youngest.

Are countries characterized by more decentralized fiscal and spending powers less corrupt? Or is a higher degree of government fragmentation a more effective way to reduce and deter corruption? And finally, is there any evidence that these alternative ways to enhance government accountability reinforce or hinder each other? These are the three questions that the empirical analysis presented in this chapter will try to answer. The idea that centralization brings about high levels of rent-seeking, corruption and lack of accountability of government officials is behind the demand for greater decentralization that has characterized many OECD and non OECD countries since the 1990s (Rodden et al., 2003; Arzaghi and Henderson, 2005; Bardhan and Mookherjee, 2005). These decentralization reforms have taken two main directions: the devolution of tax and spending power to sub-national governments, i.e. fiscal decentralization, and/or the increase in the number of sub-national units of government, i.e. government fragmentation. There is still much debate over which of the two facets of decentralization shows a greater effectiveness in deterring corruption and improving the quality of governance. So far neither has the theoretical literature drawn unambiguous predictions nor has the empirical one produced conclusive evidence