Industrial specialization and employment growth in Mexico: evidence from a spatial panel data model

This paper examines the relationship between industrial specialization and employment growth across Mexican municipalities over the period 1986–2019. The analysis is framed within the debate on Marshall–Arrow–Romer (MAR), Porter, and Jacobs externalities. Using the Krugman Specialization Index (KSI) and a spatial autoregressive (SAR) panel data model with municipality and time fixed effects, we estimate the impact of specialization on total and sectoral employment. The results indicate that higher specialization is generally associated with lower employment growth at the municipal level. However, this relationship varies across sectors: the negative effect is strongest in commerce and services, not statistically significant in manufacturing, and positive in mining. In addition, we find strong evidence of positive spatial spillovers, suggesting that interactions across neighboring municipalities shape employment dynamics. These findings suggest that specialization may enhance productivity without necessarily generating employment gains. More broadly, the results highlight the role of productive structure, sectoral heterogeneity, and spatial interdependence in shaping regional labor outcomes. Policy should therefore emphasize balanced and context-specific development strategies rather than relying on specialization alone.