Determinants of growth in Italy. A time series analysis

This paper investigates the macro-determinants of growth in Italy in a time series framework, from 1950 till 2004. The analysis of economic growth, started with the Solows (1956) and Swans (1956) famous contributions, has developed rapidly since the mid 1980s. The empirical literature follows two main approaches: growth accounting and growth regressions. In this paper the empirical approach starts with a parsimonious specification of the growth equation and then it analyses extended models. The initial specification is consistent with the standard neoclassical model and includes human capital. The extensions involve the introduction of a set of policy and institutional factors potentially affecting the Italian economic performance. In relation to econometric techniques, we use the error correction model (ECM) representation: in a time series framework, it provides evidence on the existence of a stable long-run linear relationship between growth and its determinants. The main results are the following: investment is the key source of economic growth; time series properties of the variables of interest and regression analysis provide evidence in favour of endogenous growth models in Italy and the only variable that seems to be robustly correlated with growth, according to the Extreme Bound Analysis, is government consumption, which affects negatively the growth rate.