Inteligencia artificial y planeación presupuestaria en México: promesas y retos en América Latina para la asignación del gasto público

Abstract

Public budgeting is a complex process that depends on multiple factors, including the consideration of scarce resources and the quality of available information. Artificial intelligence (AI) has the potential to discover models that help explain complex social phenomena, such as the budget planning. However, currently there is little work exploring the application of AI in this context. This paper aims to propose an algorithmic perspective for the allocation of public budget in Mexico by exploring the potential of multilayer perceptron and multi-objective genetic algorithms. The guiding research question is: what is the potential of the multilayer perceptron and multi-objective genetic algorithms in decision making regarding the allocation of the Mexican public budget? The study analyzes open data from the World Bank related to Mexico’s expenditure budget and poverty measurement data provided by the National Council for the Evaluation of Social Development Policy for the period 1990-2020. Additionally, some limitations of this work derived from the inherent complexity of the public expenditure allocation are identified. The model showed indicators that allow us to understand the importance Mexican public budget expenditures in promoting economic development, as well as potentially reducing inflation and inequality. From the analysis, based on AI techniques, it was found that the most important aspects to generate an effective and efficient public budget should focus mainly on fighting poverty, investing in the agricultural sector, encouraging industry, generating policies for the improvement of the health sector, as well as promoting research and development.