Law 488/92 is the principal instrument to promote private accumulation in the less developed areas of Italy. The selection procedure is based on 3 indicators: the share of owners' founds on total investment; the new job creation by unity of investment; the cut on the maximum possible capital aid accepted by the firm, that mimics an auction mechanism. Is this procedure efficient, reducing the amount of incentive by units of investment but nevertheless financing additional private capital formation and employment? The paper investigates the impact of incentives on employment in subsidized firm using a non-experimental design. We want to estimate if employment creation in subsidized firms is additional compared to the non-subsidized ones. Several different econometric approaches are used in order to tackle the selection bias: a diff-in-diff model with and without the Heckman correction for sample selection; a random growth model; a discontinuity design regression model. The results show a positive, statistically significant and robust impact of the incentive by 488/92 on employment: the subsidized firms present a employment dynamics from 3 to 15% points higher than in non subsidized ones.