The study of the Italian social security system in '20s and '30s seems to be relevant both for its macroeconomic implications and for its possible hints to the current debate on pension reforms. We offer a quantitative description of the funds raised by the CNAS (later INPS: Italian Social Security Fund) and an analysis of their use in 1919-1939, by the riclassification of datas availables in balance sheets. We argue the important role played by the Institution in financing public debt and in setting up public Corporations involved in financial and industrial policy. Then we put forward a qualitative and quantitative analysis of the public expenditure in infrastructures and reclamations that the use of current surpluses of the pension scheme made possible. In particular we focus on its regional redistributional effects. Implications of the collection of contributions and the payment of different kinds of benefits are then discussed. We find that all these redistributive mechanisms were significant and more articulated then commonly believed. Regional distribution of asset allocation at the end of first and second decade shows an important, even though disomogeneous, contribution to the infrastructural endowment of the country.