Spatial Dynamic Modelling of Tax Gap: the Case of Italy

Abstract

This paper analyses the determinants of regional tax gap in Italy testing if tax evasion is characterised by spatial persistence. The size of spatial correlation in regional tax gaps has been tested and the role of additional determinants of evasion over the period 2001–2011 has been estimated. Using a dynamic spatial panel model, it is shown that regional tax gap is determined by tax evasion in neighbouring regions and is characterised by spatial persistence. Results make it possible to draw a taxonomy of the determinants of regional tax gap: contextual factors and operational factors linked to the relative efficacy of tax evasion contrasting policies and geography.

Keywords

• determinants of tax gap
• spatial econometrics
• panel estimation