Public debt management and Maastricht treaty: an empirical analysis of deficit smoothing for Italy

1. Introduction The aim of this work is to provide empirical evidence of the debt management theory in order to smooth deficit over time. The coexistence of public policies exigencies directed to reduce inequalities, together with the EU public finance requirements calls into question the necessity to address the problems of defi-cit smoothing empirically even more. This question is also exacerbated by the recent policy shocks, such as the current pandemic and the Ukraine-Russia con-flict. The most important objective for an efficient debt management is to finance both the government’s needs and the expiring debt; policy makers choose among different ad hoc debt instruments by combining their risk and cost char-acteristics. Debt management literature developed many interesting theories related to three different economic objectives such as tax smoothing, deficit smoothing and minimizing risk and debt cost (see Aiyagari et al. (2002), Barro (1979), Lucas and Stokey (1983), Ramsey, F. P. (1927)). Interest-cost minimization is important especially in a country, like Italy, where the level of debt is high and interest payments absorb a large share of the budget penalizing relevant social policies. By issuing different debt instruments, policy makers may control interest pay-ments on public debt according to a certain degree of uncertainty. In fact, the optimal debt composition depends on the stochastic structure of the economy, and government may issue inflation-indexed, short or long-term debt, to sup-port tax smoothing and deficit smoothing. Under the constraint of the 3% defi-cit limit, the optimal debt composition is derived by minimizing the impact on the cost of debt instruments, and accounting for the effects of strategic varia-bles such as output, inflation and interest rate. We move from the Missale (2001) framework and develop for the Italian case a SVAR analysis to calculate the impact of the shocks deriving from the three stra-tegic variables mentioned above. While the impacts of these shocks have been computed in the past literature with partial correlation coefficients, in this work we use the SVAR methodology which is specifically suited for this problem be-cause it considers the impulse from exogenous uncorrelated shocks of the en-dogenous variables (Amisano and Giannini (1997)). Moreover, we apply the methodology developed by Ginebri et al. (2005) for calculating the final effect, in terms of semi-elasticities, of output and inflation on the public budget items, consisting in finding co-integration relations when possible. In this way we can better account for the two effects of real output and inflation (composition ef-fect) different from past literature (OECD (2009), Van Den Noord (2000), ECB (2001), EC (2002, 2012) and more recently (specifically referred to fiscal multi-pliers) El Mostafa Bentour (2022). Nonetheless, this method takes into account all the items of the public budget to analyze these two effects, while the above mentioned literature has focused mostly on the unemployment benefits. This empirical analysis was conducted during the pre-financial crisis of 2007. This way, it is possible to evaluate how to face interest rates and inflation in the absence of world crises with unexpected shocks for output and if when joining EU there were appropriate public debt management conditions. We find that, when the public debt to GDP ratio increases and the budget ex-penses account for an increasing number of items, the share of long-term fixed interest rate bonds should increase. However, on the contrary, both the shares of indexed and short-term bonds decrease. Furthermore, consistent with the literature, at the end of our sample period, our findings showed that the actual indexed bond is too small compared with the optimal share while the opposite is true for the actual short-term share. The article is organized as follows. Section 2 deals with the theoretical model. Section 3 explains the SVAR approach adopted. Section 4 presents the empirical results. Section 5 concludes and suggests some future research.