The employer size–wage effect: evidence from Italy

We use micro data to look at the firm-size wage differential in Italy. Given the remarkable importance of small firms in this country, this is a relevant issue. We have three major findings: (a) empirical estimates are very sensitive to the assumptions made in the modelling of the allocation of workers to firm sizes; (b) based upon our preferred specification, the estimated unconditional firm-size average log earnings differential is not significantly different from zero. Hence, differences in the returns to similar characteristics for individuals working in small and large private firms, who are randomly drawn from the population, are small and not significantly different from zero; (c) the raw average log earning differential is explained mainly by differences in observed individual characteristics and by selection effects.