Home     Research

The Geography of Assortative Matching

[ Draft ]

This paper investigates why assortative matching between workers and firms is stronger in large cities than in small cities. I develop a search and matching model with heterogeneous workers and firms to analyze how worker composition and labor market frictions affect equilibrium sorting. I calibrate the model to match salient moments of matched employer-employee data from Germany. I find that matching efficiency plays a major role in explaining differences in assortative matching across cities. Moreover, the effect is amplified by a more disperse workers productivity distribution, since there are higher returns from matching with similar types for both workers and firms. Using the calibrated model, I show that around 5% of GDP gap observed between large and small cities can be explained by differences in assortative matching. Overall, the paper stresses the importance of studying local labor market frictions and workers productivity distribution together to understand why the allocation between heterogeneous workers and firms vary between cities, and the resulting implications for spatial inequality.

Presented at: QMUL Macro Internal Seminar (2024), NSE 3rd PhD and Post-Doctoral Workshop in Economics and Finance (2024), 22nd edition Brucchi Luchino Workshop in Labor Economics (2024) (selected)

---

Task-Biased Technologies Adoption Across Countries with Paula Cesana

Draft Available Soon!

In this project we study how the task content of occupations differs across countries. Using data from PIAAC, we propose a measure of “task intensity” to quantify the relative importance of each task within occupation, across countries. We find there is a negative relation between the routine task intensity and GDP per capita: the richer the country, the less intense the use of routine tasks, within occupation. In order to rationalize these findings, we propose a simple production model in which technology is task-biased. The allocation of tasks within occupation depends on tasks productivity: when tasks are complements in production, an increase in the relative productivity of one task reduces its intensity with respect to the others. This is reflected in all occupations, since tasks are common to all of them. In the cross-country context, we interpret a relatively high intensity in routine tasks as a low routine specific productivity relatively to the other tasks. In the quantitative exercise we calibrate the model by matching moments obtained from the survey and from aggregate data. Eliminating dispersion in tasks productivity reduces GDP per capita variation by 10%, significantly benefiting low-income countries.

Presented at: QMUL Macro Internal Seminar (2022)