Modelo de datos panel para la evasión del sistema de protección social en las principales áreas metropolitanas del país: estimación frecuentista y bayesiana.

Abstract

In Colombia the evasion of the social protection system, mainly in health and pension, is an issue that is being addressed in greater depth, since it has some influence on the country's economy, involving several factors in diferent contexts such as social, cultural, political among others, for the foregoing it is important to look for a way to reduce the percentages of this problem, one of the alternatives is to capture that unobservable heterogeneity of economic agents in the temporal dimension, because the analyzes of time series and cross-sectional analysis fail to control that heterogeneity, obtaining biased results, it is decided to opt for panel data models. In the present work we propose to compare the behavior of the evaders of the main metropolitan areas of Colombia through frequentist and Bayesian estimation, identify the type of model to be used, choose which areas and explanatory variables have a signi cant efect on the response variable.