Análise de dados em painel das variáveis que afetam a criminalidade no Brasil

Abstract

This work analyzes the main variables as determinants of crime and verifies the relationship between them. For this purpose several variables considered relevant as unemployment , social inequality , income, poverty , amount invested in public safety and also crime which had as proxies unemployment rate , GINI coefficient , average family income , rate of people living below were used poverty line , the amount that each state invests in public safety and the number of homicides per 100,000 inhabitants, respectively. Thus, samples were collected between 1995 and 2009 periods for each state of Brazil and the Federal District and was made an estimation of panel data with fixed national level to analyze the impact each variable has on crime effects. With the estimation made, proved to be significant variables representatives of unemployment, inequality and poverty. While income and government spending on public safety variables were not significant at 5% (though the latter presented significant at 10%). Regarding signs presented by the estimated parameters, social inequality had a positive behavior towards variable used in the study to measure crime, so the greater the inequality, the greater tends to be the number of homicides per 100,000 inhabitants. The variables representatives of unemployment and poverty had negative signs, contrary to having an index used to measure crime behavior. While this result is not surprising for the unemployment rate, it may surprise some people with regard to poverty variable, since the intuitive idea that most people have is that the lower the rate should be less poverty crime but this can be explained by socioeconomic theory.