A Perspective on Real Exchange Rate Determination in Italy During the Convergence Process Towards the EMU : The 'Traded Non-Traded' Model

Though inflation rates in EMS countries have significantly converged over the past decade, exchange-rate adjusted price levels have sharply diverged and continue to do so, albeit at a decreasing rate. This violation of purchasing power parity (PPP) may be viewed as a stylised fact in international finance.The failure of PPP to hold continuously is well documented empirically. However, there is growing evidence that it does hold as a long run phenomenon. A broad body of evidence suggests that real exchange rate is not a random walk, and that shocks to real exchange rate damp out over time albeit very slowly.Consensus estimates and studies of purchasing power parity, surveyed in Froot and Rogoff (1995), put the half-life of deviations from PPP at about four years for exchange rates among major industrialised countries. Recent attempts to formulate and test the structural models of deviations from PPP, [Asea and Mendoza (1994), Marston (1987), De Gregorio, Giovannini and Wolf (1994)] stress the different behaviour across countries of the prices of manufactured goods relative to that of service. In fact, several recent studies [for example Engel (1996)] that employ unbundled data, have investigated the extent to which departures from PPP are caused mainly by the presence of non traded goods versus deviations from the law of one price in traded goods. In particular in the EMS area, a sizeable fraction of changes in Intra-European real exchange rates may have been driven by changes in domestic relative prices between traded and non traded goods, rather than by changes in the prices of tradable goods across countries. In fact a marked divergence in inflation rates between the tradables and non tradables sectors has been a feature of some EC economies, especially those of Italy, Spain and UK, since the early 1980s. Sectoral productivity differences, international competitive pressure and Government demand for non tradables are possible reasons.This paper investigates whether, the traded non traded model can be used to provide a measure of the deviation from equilibrium of the Italian Lira ECU real exchange rate