Relocation of Government Sector Jobs

Place-based policies have gained increasing interest the recent years. An often used place-based policy is relocation of government sector employment. This study estimates the local labor market effects of changes in government sector employment. To estimate the local labor market effects, I use a difference-in-difference two-way fixed effects approach with a direct projections specification to allow for areas with multiple treatments. I apply my empirical approach to a balanced panel data set spanning over ten years from 2010 to 2019, with observations of 2145 parishes in Denmark each year. The panel data set is created by combining several registers containing Danish administrative data on working places, geographical data, financial statement data, and educational data. I argue that by using the finest possible spatial scale, the parishes, I get around potential selection biases from non-random placement of place-based policies. I find a positive short-run job multiplier of around 1, meaning that a change in government sector employment of 1 changes the private sector employment with 1 in the same direction. However, the job multiplier decreases over time and is only 0.35, and not significantly different from zero, after three years. I further find that an increase of the government sector share of the total employment by one percentage point will increase the private sector wages by around 0.15 percent in the short run. However, changes in government sector employment do not seem to impact the local housing costs. The job multiplier effect is very local, as the neighboring parishes within the same municipality are not affected by a change in government sector employment. The employment effect is primarily due to changes in the non-tradable sector, while the tradable sector seems unaffected. However, the job multiplier appears to be non-linear as only negative changes in government sector employment seem to affect the private sector employment.