The paper uses data from the Italian regional public accounts database to investigate and analyze the heterogeneity of the impact of public infrastructure across Italian regions, both on institutional and political grounds. Heterogeneity is addressed by linking the impact of infrastructure on Gross Domestic Product (GDP) with corruption by using a random coefficient panel data model. This paper considers a novel objective measure of corruption that consists of the difference between a measure of the physical quantities of public infrastructure and the cumulative price government pays for public capital stocks. Empirical analysis confirms the existence of parameter heterogeneity across Italian regions and is also consistent with the theoretical considerations that corruption negatively affects economic performance.

This paper empirically analyzes the economic impact of public spending on infrastructure (Barro, 1988 and Aschauer, 1989), as recorded under the Italian project Conti Pubblici Territoriali (CPT ) (i.e., Regional Public Accounts (RPA)) and corruption on Gross Domestic Product (GDP). The study differs from the existing literature in three main aspects:

First, it adopts a random coefficients model (RCM) approach in order to estimate the economic impact of public expenditure on infrastructure across Italian regions. The rationale for using an RCM is a drawback common to studies on infrastructure and productivity—they do not take into account the heterogeneity of parameters (Romp and de Haan, 2007). Within a regional context, differences involving the impact of infrastructure on economic performance might make little sense. However, in the Italian regional setting characterized by economic dualism, such kind of analysis might be of some interest in explaining differences in the economic performance systematically reported across northern and southern regions.

Second, it uses an `objective' proxy for corruption proposed by Golden and Picci (2005) as explanatory variable. Indeed, many theoretical considerations lead scholars on this field to think that, in general, social capital matters (Putnam, 1993). Public infrastructure spending, in particular, "are the classic locus of illegal monetary activities between public officials […] and businesses" (Golden and Picci, 2005). Moreover, there are reasons to assume that corruption and public investment vary together (Rose-Ackerman, 1999) or, more generally, that corruption modifies patterns of public spending (Mauro, 1998; Coppier, 2005; and Shaw, Katsaiti et al., 2007).

Infrastructure Journal, Multilevel Analysis, Public Infrastructure, Italian Regions, Gross Domestic Product, Economic Dualism, Economic Sectors, Government Interventions, Economic Infrastructure, Social Infrastructure, Geographical Latitude, Human Capital Infrastructures.