ABSTRACT Cerina et al. (2023) affirm greater polarization of occupations between high‐ and low‐skilled labor in large cities relative to small cities. The authors attribute differential rates of employment growth by occupation types to skill‐biased technological change alongside consumption spillovers. They find a smaller role for differential total factor productivity gains in tradable sectors and extreme‐skill complementarity in production. We computationally reproduce their results, reconstruct the authors' data cleaning and harmonization procedures, and test the broader replicability of the results. We find the results support for the internal validity of Cerina et al. (2023)'s results and broaden the external validity of the analysis.
Kouchekinia et al. (2026) studied this question.