Estimating the impact of added jobs : a local multipliers approach

Abstract

Indirect job creation is often an argument in favor of several policies. In this paper I seek to identify the impact an additional tradable job has on non tradable employment. I find that the addition of a tradable job in Argentina leads to the creation of 4 jobs in the non tradable sector in a given city. I identify this effect using a shift-share instrument and two datasets that register employment. The first is the national household survey that allows me to measure employment in different industries up to city level. The second is a random sample from an administrative dataset that allows to follow workers over time. I exploit this to estimate effects on hirings and separations. My findings suggest that the main mechanism is increased hirings.