Matteo FRAGETTA | Progetti

ON THE SOURCES OF INFLATION DIFFERENTIALS IN THE EURO AREA

As documented in the previous section, the empirical literature has so far studied either single or a few determinants of inflation differentials in isolation or has employed factor models to identify area-wide and idiosyncratic components. What the literature lacks is an integrated analysis that simultaneously look at all, or at least many, potential determinants suggested by theory, so as to disentangle not only their effects but also their relative weight on inflation differentials over time.In order to fill in this gap, we plan to adopt a multivariate dynamic panel data model along the lines of Ciccarelli, Maddaloni and Peydró (2011), i.e. a panel vector-autoregressive model including variables at a country level, except for the measure of monetary policy that is common across countries. We will identify the structural shocks via the imposition of uncontroversial sign restrictions on the impulse responses dictated by economic theory.The planned set of endogenous variables, observed at a quarterly frequency, besides cumulated inflation differentials (which are a measure of intra-euro-area real exchange rate), includes output gap, inflation, a measure of monetary policy, the government budget balance, a measure of capital movements, and the nominal exchange rate. Even if the specification has to be parsimonious – as the inclusion of any additional variable implies that an additional shock needs to be identified – we believe this is wide enough to encompass a broad range of shocks that may affect inflation differentials. In fact, the specification captures aggregate demand and supply, monetary and fiscal policy, financial and exchange rate shocks that may affect inflation differently across countries depending on their degree of openess.The sign, magnitude and dynamics of the effects of the various shocks on inflation differentials is to be studied through the analysis of impulse response functions, while we will make use of an historical variance decomposition to compute the dynamic contribution of the various shocks. The latter allows us to determine, quarter by quarter, the proportion of variation in inflation differentials to be attributed to each shock (see Canova and Ciccarelli, 2012, for an application).We will also look at subsamples of countries (e.g. peripheral versus core). In addition, by means of recursive estimation, we will also provide a full account of any changes of the transmission of shocks before and at various stages of the financial crisis. We believe that a thorough understing of the causes and relative weight of intra-euro-area real exchange rate misalignments has very relevant policy implications. In fact, given the fixed exchange rate regime implied by the currency area, only internal devaluation may correct such imbalances. The latter, however, is costly in terms of unemployment and output losses, as the recent European experience is highlighting. Therefore, avoiding misalignments in the future is key to guaranteeing higher levels of social welfare and this can only be achieved if policy actions are targeted towards the core causes of the problem.