Assessing food insecurity in Brazil: the issue of the non-symmetric effects

Abstract The main objective of this article is to study how a set of socioeconomics characteristics impacts the probability of occurrence of different levels of the food insecurity in Brazil. This is performed taking into consideration the informational complexity of the Brazilian Scale of Food Insecurity (EBIA). Our hypothesis, which are tested through a generalized ordered logistic model, is that the explanatory variables offer asymmetric effects regarding the different categories of food insecurity measured by the EBIA. The results confirm that hypothesis, pointing out that it is important to consider the complexity involved in the food insecurity phenomenon to discuss public policy.