Abstract

Means-tested social assistance programs have acquired, in the last two decades, an increasing role in the majority of industrialised countries. Although the issue of whether social assistance should be targeted or universally granted remains a main subject of dispute (Harding et al. 1994; Mitchell 1995; Smolensky et al. 1995; Atkinson 1995, 1998), the shift towards targeting is in practice quite widespread, mainly reflecting the wish to curb social security budgets in a general context of fiscal restraint. The issue of targeting welfare assumes a particular importance in Italy, whose social assistance programs resemble more the rudimentary regimes typical of Southern Europe than the evolved systems of other EU countries. Indeed, the Italian welfare system is characterised by a number of categorical schemes limited to aged and disabled people, a discretionary relief (mainly in kind) provided by regional and local governments, and the absence of a national minimum income scheme as a universal safety net. To tackle these severe shortcomings, an articulated plan of reform proposals has been put forward in a report issued in 1997 by a special governmental commission on social expenditure. Some of its proposals have recently found two important applications, the reform of means-testing criteria and the experimental introduction of a national safety net scheme, an absolute novelty for the Italian system. The strategy behind the reform is to increase assistance for those most in need and to ensure that public expenditure is better targeted. The paper deals with the distributive implications associated with this reform plan, and in particular with the new entitlement rules which have recently designed to establish the access to welfare expenditure. Section 2 critically reviews the Italian social assistance system and briefly describes the institutional design of the new targeting system, which will gradually substitute the tests of means currently applied. Using the micro-data contained in the most recent sample survey of household income and wealth conducted by the Bank of Italy1 and the tax-benefit microsimulation model Dirimod95, Section 2 also examines the sense in which the shortcomings mentioned above substantially undermine the effectiveness and efficiency of income support programs in alleviating poverty among Italian households. Section 3 is devoted to study how different economic and demographic groups of the population are likely to change their relative position if households are ranked first by a variable representing old means-testing criteria, and then by the new test of means; this analysis aims to clarify how various groups are likely to change their relative probabilities to access social services. Section 4 evaluates the distributive impact of the new spending programs introduced with the budget law for 1999 and of some reforms of the whole assistance system which, although highly speculative, can shed some light on the likely direction towards which the Italian welfare state is evolving. Finally, Section 5 concludes.