Oskar Nordström Skans och Lena Hensvik, Institutet för arbetsmarknads- och utbildningspolitisk utvärdering
The Montgomery (1991) model of employee referrals suggests that it is optimal for firms to select new employees through referrals from their most productive workers, as these are likely to know others with high unobserved productivity. In this paper, we use a rich matched employer-employee data with cognitive and non-cognitive test scores to assess the model’s ability to explain why firms recruit former coworkers of incumbent employees. Our empirical results support key elements of the model: Incumbent workers of high aptitude are more likely to be linked to entering workers. In addition, firms acquire entrants with better unobserved abilities when hiring linked workers, and entering workers with links to incumbent employees receive higher entry wages than other entrants. Finally, entering workers exhibit wage returns from their linked employees abilities. The results thus support the notion that firms use referrals of productive employees in order to attract workers with better unobserved qualities as well as the notion that firms use the ability-density of social networks when setting entry wages.