Purpose The purpose of this paper is to investigate the impacts of left Brazilian political parties and partisan disruption on the homicide rate in Brazil. Design/methodology/approach The authors use panel data for the states between the years 1980 and 2011. The database used is an unbalanced panel covering a sample of 27 Brazilian states over 32 years, 1980-2011, totaling about 855 observations. Findings It is estimated that these two political factors are sources that have connection to the increased level of violence in Brazil. These analyses provide several important results. First, partisan disruption is associated with a higher homicide rate, compared to non-partisan disruption. The results from the panel also suggest that left-parties in government have a positive impact on homicide, compared to non-left-parties. Research limitations/implications Information regarding premeditated homicides (CID-BR-9 database) is available for all Brazilian states, and may be tabulated from the same micro-data at any level of aggregation. Some of the well-known problems regarding the choice of this variable are as follows. First, deaths resulting from wounds are sometimes included in the statistics whether wounds were intentionally inflicted or not. In addition, some incidents end up not being registered because certain deaths are not reported. This tends to occur more frequently in rural areas. Fortunately, this second problem does not appear to be too significant, as under-registry of deaths due to external causes is much lower than the amount resulting from natural causes (see, e.g. Cano and Santos, 2000). In addition, this problem may be controlled if under-registry remains stable over time by applying fixed effects to the panel data. Practical implications The main Brazilian political parties diverge on the causes of crime and how criminals should be punished. For example, in Brazil, the minimum age for one individual to be punished with imprisonment is 18 years old. Practices crimes for young people between 12 and 18 implies only in socio-educational measures. Given the high level of violence in Brazil, there is a bill being debated in the parliament that proposes to reduce the age to 16 years. Based on the research, 90 percent of the population approves the reduction of age to 16 years. However, the majority of parliament is opposed to changing the law. In general, the more conservative parties are favorable to changing the law. Social implications These divergent postures can be associated with the ideological essence or to belief system of each political party. Political parties have the potential capacity of changing crime trends through economic and social policies as well as by applying stronger sanctions against crime. Given the law enforcement system, the cycle of crime in Brazil may be related to the profile of the political party elected. Originality/value The authors assume the hypothesis that the current Brazilian multi-partisan system has an incentive system in which politicians do not respond adequately to the basic wishes of voters. Among such desires, the authors emphasize public safety. This paper evaluates the empirical effect of partisan disruption on homicide rate.