The distribution of wage inequality across municipalities in Mexico: a spatial quantile regression approach

Research background: According to classical labor economics, wage differences among regions of a country that has free-factor mobility should eventually vanish. However, the level of wage inequality among Mexican territories is increasing. The nature and causes of this discrepancy are worth identifying.Purpose of the article: To identify the spatial relationship of wage inequality that existed in the Mexican metropolitan system during the years 2010 and 2015.Methods: We develop a model of wages that considers the interaction between spatial units within a region. Then, we specify a spatial autoregressive model with the average wage per municipality as a dependent variable. This variable is spatially lagged along with other controls such as productivity, schooling, and migration. We combine data from population and economic censuses. Then, we perform a quantile regression to estimate the spatial effect of wage in a region upon quartiles of the wage distribution.Findings & value added: Wage inequality increases within a given region when the average wage increases in one of said region's municipalities. This phenomenon occurs because in municipalities that are neighbors of the one that enjoys a wage increase, the average wage tends to decrease. The impact is larger in those municipalities whose average wage is in the lower range of the regional wage distribution. Wage inequality is also increased by internal migration and increased productivity. These latter findings are some of the first for Mexico at this aggregation level. A novel aspect of our study is its use of territory as an observation unit for which statistics from population and economic censuses are combined to draw inferences about spatial inequality.