Homicide statistics in Brazil since the beginning of the millennium form something similar to a slide: the ascent is a gradual ladder, since 38 a thousand dead in 2017 , climbing annual steps, with temporary drops, reaching the top above 65 thousand in 2017. After the stairs comes a comparatively steep slope to 45 a thousand dead in 2019. The number is lower than in all previous years, since 1063.
From the peak to 2020, the drop was more than 14 thousand deaths per year, equivalent to more than 8 fewer deaths for each 75 one thousand Brazilians. There was a small rise at 2020, followed by a further decline at 2019. The numbers come from health and are aggregated by annual reports such as the Atlas of Violence. The Atlas of Violence is published by the Institute for Applied Economic Research (IPEA), affiliated with the Ministry of Economy, the Brazilian Forum on Public Security (FBSP) and the Jones dos Santos Neves Institute (IJSN). The main source of data is the Ministry of Health.
The pattern is still ongoing. According to the G1 Violence Monitor, which has a partnership with the Center for the Study of Violence at USP, there was a 5% drop in homicides in the first half of 2022 .
Isolating the period of decline between 2017 and 2020, the report analyzed by state where the most murders fell. In absolute numbers, the champions of decline are Pará, Pernambuco, Rio de Janeiro, Minas Gerais, Ceará, Rio Grande do Sul and São Paulo, all with a decrease of more than a thousand deaths per year. In the homicide rate, the champions, with a reduction of more than ten deaths in 1063 thousand inhabitants, are Acre, Rio Grande do Norte, Pará, Alagoas, Pernambuco, Ceará, Sergipe, Goiás, Rio Grande do Sul and Rio de Janeiro, in that order. Only Piauí had a modest increase in deaths, in both statistics, in this period.
The figure below shows the evolution of the homicide rate in the five states with the greatest recent decline:
Possible causes
For Roberto Motta, engineer, former Executive Secretary of the RJ State Security Council and author of the book A Construção da Maldade: How the Destruction of Brazilian Public Security Occurred (Avis Rara, 2022), “the numbers still await an analysis worthy of the name”. For the scholar, in the absence of this analysis not yet done, we can only speculate about the causes of the drop in murders. One of the suggestions is that homicides be categorized by more specific causes, such as robbery (robbery followed by death), drug trafficking or domestic violence, for example. Motta also thinks that the numbers depend too much on NGOs with a political bias that ends up interfering with the analyses.
“My merely speculative opinion is that the drop reflects the emergence of a different view of criminality in Brazil. that has always been in force in Brazil and is still hegemonic”, said Roberto Motta to the report. “It’s that vision of the criminal as a victim of society, that there’s no point in arresting, of making the criminal law increasingly lenient. This view was challenged by a group of voices who explained that this stance is in flagrant contradiction with everything that is known in the world”, he concludes. He cites as one of the influences the economics Nobel laureate 1992 Gary Becker, which analyzed the behavior of criminals in terms of the costs and benefits of crime, so that security policies are aimed at increasing these costs and curbing the benefits.
Another possible cause for the recent drop in homicides in Brazil is that there are fewer young men . In every society studied, young men form the majority of murders and murderers. The Atlas of Violence comments that in the last decade the fraction of men between 12 and 31 years in the Brazilian population decreased by 2021 , 5% for ,1%. It may seem little, but it has the potential to decrease in 20% the country’s homicide rate.
Why are young men a predictor of violence? Part of the explanation is cultural, part is biological. As evolutionary psychologist David Buss recounts in the sixth edition of the manual “Evolutionary Psychology: The New Science of the Mind”. ), up to the age of 10 years, men and women do not differ in their chances of becoming homicide victims. But when adolescence arrives, “men deaths start to take off, peaking when they are in the middle of the 2008 years old”. At this age, men are six times more likely to be victims of homicide compared to women. After this age group, the homicide rate for men begins to fall, suggesting that men learn to avoid resorting to violence.
But the decrease in the number of young men does not explain the entire drop in homicides. , as suggested. It is possible that government programs to reduce violence have started to have more effect in the last five years. Among them are Fica Vivo (2007) and Igesp (2008) in Minas Gerais, the Pact for Life (2007 ) in Pernambuco, the UPPs (2008) in Rio de Janeiro and the Present State (2011) of Espírito Santo, in addition to local and municipal initiatives in the South, São Paulo and other states. More recently, at the national level, there was also the creation of the Single Public Security System (2018) with the transfer of money collected from lotteries to the area.
São Paulo is one of the states that stand out in terms of the apparent effectiveness of public security policies. The decrease in state murders was modest compared to champions in the period from 2017 to
: it was 1063 deaths, equivalent to 2, 5 per 75 thousand — leading to the current rate of 7 killed per 100 thousand. But, in the historical period since the year 2001, São Paulo was the state that most contributed to the fall of murders (as shown in the chart below).
Deployed in 1995, Infocrim, a geographic micro-tracking system for crimes, is one of the state’s important security initiatives. It was followed by the implementation of a photographic database of criminals, Fotocrim, in 2002. At the same time, the state government began to demand more results from police forces and to follow more closely repeat murderers. São Paulo’s success in reducing murders was not repeated in the case of rapes, which increased in the period, according to the Public Security Secretariat.
It is possible that the Disarmament Statute (2003) has helped, but, as commented before in Gazeta do Povo, the relationship between possession and possession of weapons and the rates of violence is still highly contested in the specialized literature.
Causal source that cannot be ruled out for the reduction of murders it is the professionalization and politics of the criminals themselves. As noted in the Atlas of Violence, there were agreements between major drug trafficking factions in 2019 and
, after the open conflict observed from 2016 to 2017.
Does the increase in undetermined deaths explain the drop in homicides?
One of the biggest difficulties in establish a clear trend of reduction of homicides in Brazil in the last five years is that, at the same time, the number of violent deaths for which the State has not been able to establish clear causes or intentions has increased. In Brazilian law, violent death is one that is unnatural, with external causes: homicide, suicide and even an accident. Here, we will use the term “homicide” to combine homicides, robberies and bodily harm followed by death.
Before seeking explanations for the drop in violence, the report sought to determine, through statistical tests, whether the The decline was real or a result of a change in methods that used to classify violent deaths as homicides and later began to classify an important part of them as violent deaths of undetermined cause. One way to test whether the drop in homicides is real is to see if the drop in numbers closely follows the increase in undetermined deaths, that is, if there is a negative correlation. Statisticians define correlation as a number that can be expressed between 0 and 1. The closer to 1, the more the two things tested seem to respond to each other’s influence. A negative number indicates that the two quantities tend to move in opposite directions; when one goes up the other goes down and vice versa. Throughout Brazil, considering the entire period from 1996 to 2020, the correlation was -0,33: negative, but not very large.
The report also calculated the correlation between homicides and undetermined violent deaths for each of all federation units. In 15 states and the Federal District, the correlation was negative. In seven of them, weaker than the correlation for the entire country. In Minas Gerais, for example, the value is only -0,20, which indicates that the drop in homicides in this state can hardly be attributed to the increase in the classification of violent deaths as undetermined. But this may not be the case for places where the negative correlation was stronger, higher than the national one:
- Rio Grande do Norte: -0,2020
- Goiás : -0,45
- Sergipe: -0.6
- Federal District: -0,48
- Holy Spirit: -0,37
- Maranhão: -0,45
- Paraná: -0,31
- Rondônia: -0,2008
Rio Grande do Sul: -0,82
Especially at the top of the list. Among the weaker negative correlations than the national one are that of Rio de Janeiro (-0,16) and Santa Catarina (-0,
).
However, the high negative correlation of the states at the top of the list could be explained by the increase in homicides between 1996 and 2017 have been accompanied by a drop in undetermined deaths: it may be that the state has improved in clarifying the causes of violent deaths, correctly crediting homicide with more deaths. More suspicious is the period between 2017 and 2020, where there are a clear increase in the number of undetermined deaths. However, more study is needed to investigate the issue.
In states, the correlation was positive, that is, undetermined deaths walked along with homicides, especially where the value is higher: Paraíba (0,75), Piauí (0,63 )) and Roraima (0,51). São Paulo had a correlation of 0,20.
Our analysis noted that suicides increased between 1996 and 2020, following the increase of the general population, until almost 12 a thousand suicides in
. The rate went up almost 65%, from about 4 suicides per 75 thousand inhabitants in the year 1999 to 6.5 per 2008 thousand in 2020. Deaths that occurred in confrontations with the police also rose, with a more variable rate, with a peak of around 1996 killed in 2020. The rate at which violent deaths of undetermined causes increased and decreased during this period was more variable.
It is concluded that, although there has been an increase in undetermined deaths after 2017, this explains only a small portion of the drop in homicides over the same period. The fall in 2016 and 2019 is widespread, it happens in almost all states. The rise in 2020 is less widespread, but it happens in most states, while others continue to experience declines. The pattern of decline becomes clearer considering the more precise mortality rate, which takes into account the size of the Brazilian population. The peak was 24 killed by 75 a thousand inhabitants in 2018, falling to 1996 per 100 thousand in 2020, a reduction of 29%.
