Educational outcomes of young people living in Italian public housing are dramatically worse than those of similar youngsters not living in public housing. The existing literature about the effect of living in public housing on the educational and behavioral outcomes of its young tenants is mostly based on U.S. relocation programs and it is not conclusive. This paper is the first contribution on the Italian case. It investigates the effect of living in public housing on the probability of grade repetition, school evasion, early school drop-out and on years of education. The identification strategy exploits the history of the Italian interregional migration. The second part of the paper investigates the existence of heterogeneous effects among projects of different dimensions. We find a strong unfavourable effect of living in public housing. For female school evasion we can support the idea of a ghetto effect of public housing. The analysis does not find strong heterogeneous effects of projects of different sizes, although we find that size has a different impact on boys and on girls.