El desempleo regional en el Ecuador. El rol de la heterogeneidad espacial a nivel cantonal en el periodo 2010-2022.

Abstract

The study aims to analyze regional unemployment in Ecuador and the role of spatial heterogeneity at the cantonal level in the period 2010-2022, identifying spatial patterns in the distribution of unemployment rates between cantons and determining the factors that explain these rates. The methodology uses a quantitative approach with a non-experimental design, being developed in two stages: first, a spatial autocorrelation analysis using spatial weight matrices and the Moran I statistic to evaluate the spatial dependence between cantons; second, the application of spatial econometric models, including the General Nesting Spatial Model. The results, based on data from the 2010 and 2022 censuses with 217 cantons, reveal that construction employment affects the unemployment rate only in 2010, while population density and urbanization are relevant in 2022. The spatial models demonstrate significant correlations not captured by traditional models, showing that unemployment in a canton is correlated with neighboring cantons. Using the Akaike and Bayesian information criteria, it was determined that the Autoregressive Spatial Model provides the best fit for the data, effectively capturing the spatial dependence between cantons and its direct and indirect effects on unemployment rates.