Modelo baseado em agentes para avaliação de políticas públicas de segurança

Abstract

Currently, there is a feeling of public insecurity in the world, as tranquility and order have been stolen by terrorist attacks, gender violence and other crimes against people and property. In Brazil, the Federal Government, in an attempt to restore the feeling of security, launched a public policy aimed at bringing the police closer to the community, the National Program for Security with Citizenship (PRONASCI), which conditions the transfer of resources to the change in the format of action of the police, and must now seek greater proximity to the community and respect for human rights. And, in 1997, the Federal Government decided to take on the role of promoting public security policies, creating the National Secretary for Public Security (SENASP/MJ) within the structure of the Ministry of Justice. Thus, the need to evaluate public security policies in the country became evident, but the evaluation models that already exist and are applied to evaluate other types of public policies, do not contemplate the peculiarities of public security problems, which emerge in a “war” format. urban” with constant growth in crime and violence rates. In this study, we seek to offer a public policy evaluation model that addresses the specific needs of the field of public security. In this study, theoretical references and support models were taken into account for the construction of something suitable for the public problem of homicidal violence, and, as security was characterized as a complex system, whose components interact with each other and with the environment, it is presented here in support of the Agent-Based Model (MBA). As there are multiple agents and variables of public security, for the choice of agents to be included in the model and computational simulation, the statistical method of Pearson’s correlation index is applied, thus identifying the degree of correlation of agents with the public problem of homicidal violence, using data from the Public Policy Pacto Pela Vida in Bahia (PPV). This enables an improvement in the decision-making process and greater effectiveness in evaluating the public policy being implemented. In simulations, applying the model to three different scenarios, limitations to the model were identified, since the field of public security is a recent study and has outlines of several theoretical, ideological and technical matrices that have not yet been sufficiently explored. Finally, the proposed model proved to be adequate, presenting different employment possibilities to evaluate public policies not only in the field of security, but any program that deals with the most diverse public problems, whose public policies are being implemented and lack evaluation and monitoring.