Wage Rigidities in a Quantitative Spatial Economy: Commuting and Local Unemployment

In this paper we build a quantitative spatial general equilibrium model to study the geographical variation in unemployment rates in the presence of wage rigidities and when workers are allowed to commute from residence to workplace. Calibrating the model on Belgian district data, we find that, were workers' location choice driven only by preferences for amenities, workers would relocate away from the center of the country, generating a less concentrated spatial distribution of economic activity. We also explore the role of unemployment insurance in determining the location choices of workers. We find that when the risk of unemployment is fully insured, workers relocate to districts with initially high unemployment rates, therefore accentuating the spatial misallocation of labor. Removing unemployment insurance would instead not generate significant changes in the spatial distribution of workers. To gauge the magnitude of wage distortions, we compare the observed gross wage levels with the counterfactual market-clearing wages. Removing wage rigidities would generate significant gains in local and total GDP (+3%) and modest gains in the average real net labor income per resident (+1%). Lastly, we determine the level of the employers' social contribution rate that would allow to achieve full employment in all districts. We find that the optimal social contribution rate should be 24%, 12 percentage points lower than the observed rate, while at the same time it would increase fiscal revenue by 1.5%.