The Introduction of a Private Wealth Module in CAPP_DYN: an Overview

Household saving rate in Italy declined over the last two decades.This trend still persists despite three pension reforms have been enacted since the beginning of the nineties. In this paper we search further evidence of general macroeconomic effects through the analysis of households behaviour. In the first part of the paper we use data from five surveys of the Bank of Italy Surveys of Household Income and Wealth (SHIW) to estimate the lifetime profiles of saving and wealth accumulation. Estimates show that the age profile of the propensity to save has been influenced more by cohort effects than by general trend effects; whereas the age profile of the ratios of financial assets to disposable income has been subject to relevant trend effects. In the second part of the paper we analyse the effects of pension reforms on saving behaviour of Italian Households. Firstly we use a difference-in-difference estimator in order to test whether the groups more severely hit by the reforms actually increased their saving rate relative to the other groups. Then we estimate the Social Security Net Wealth (SSWN) for each individual in the SHIW in the analysed period (1989-2000). Finally we estimate the substitution coefficient between SSWN and private wealth taking into account that the reaction of saving to a change in SSWN depends also on age of the individual. Our results show that the reduction of SSWN is unequally distributed across individuals. The cut is stronger for self employed, young workers and women. Most of the groups more severely hit by the reforms did not increase their saving rate relative to the control group: younger households, in particular, did not increase the saving rate. On the whole a reduction of one Euro in SSWN seems to induce, on the average, a compensating increase in private wealth by about fifty cents. The substitution coefficient between private and social security wealth is higher for the richest and oldest part of the sample. Finally when we split the sample observations by year we find that the more dramatised is the impact of the reform, the higher is the substitution coefficient.