A stochastic model to analyze the dynamics of poverty and social mobility in Mexico

Abstract

Objective: To propose a model based on non-homogeneous Markov chains to analyze social mobility and the transitions between the categories of poverty defined by the National Council for the Evaluation of Social Development Policy in México. Methodology: A non-homogeneous Markov chain approach is employed to project the overall proportions of the population across categories of poverty and vulnerability. Transition probabilities are adjusted using observed data to ensure an accurate match between the model’s projections and actual proportions. Results: The projections reveal significant disparities between regions in México. Northern states show a higher proportion of non-poor and non-vulnerable populations, while southern states exhibit the highest levels of extreme and moderate poverty. Limitations: The model may be constrained by the quality and availability of the observed data used to adjust the transition probabilities, as well as by unaccounted variations in underlying social and economic dynamics. Originality: This work introduces a novel approach by applying non-homogeneous Markov chains to analyze social mobility and poverty transitions in the context of México. It also incorporates data-driven adjustments to transition probabilities to ensure precise projections. Conclusions: The projections highlight the regional heterogeneity of social mobility in México, emphasizing the need for targeted public policies that address the specific needs of each region, particularly in the most disadvantaged areas.