This paper studies gender wage differentials in Italy using first-order predictions of monopsony-search models. It compares empirical predictions of these models against other competing ones of wage determination in non-competitive settings.

The paper looks at the empirical relevance of the model in terms of third degree wage discrimination among men and women by estimating the labour supply elasticity to the individual firm. It also tests the monopsony model using a "natural" experiment. Italian administrative longitudinal data from INPS are used. 

Women have lower elasticity of labour supply yo the individual firm: employer size regressions indicate larger effects (and consequently lower elasticity) for women as predicted by the monopsony model. Using the theoretical dynamic monopsony-search model of Burdett and Mortensen (1998), wage elasticity of separations and recruits confirm this result. Using relative men/women employment effects resulting from institutional changes in wage indexation mechanism (Scala mobile), it is found that relative male employment responded differently in the two periods to the exogenous relative increase in the wage differential, as predicted by the monopsony model. Search frictions explain about 50% of the gender differential.