The Homicide Drop in Recife, Brazil

• DOI: 10.1177/1088767916634405
• Published date: 01 February 2017
• Author: Débora V. S. Pereira,Martin A. Andresen,Caroline M. M. Mota
• Date: 01 February 2017
• Subject Matter: Articles

Homicide Studies

2017, Vol. 21(1) 21 –38

© 2016 SAGE Publications

Reprints and permissions:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1088767916634405

journals.sagepub.com/home/hsx

Article

The Homicide Drop in

Recife, Brazil: A Study of

Crime Concentrations

and Spatial Patterns

Débora V. S. Pereira1,2, Caroline M. M. Mota1,

and Martin A. Andresen2

Abstract

Studies in crime concentrations have focused primarily on North America with a rather

restrictive set of crime types. In this article, we analyze the crime concentrations and

spatial patterns of homicide in Recife, Brazil. Brazil’s homicide rates have remained

stable but at high levels, approximately 30 homicides per 100,000. Some places have

experienced notable decreases in homicide: In Recife, the capital of Pernambuco,

there has been a drop in the homicide rate of 46.67%, 2000 to 2012. We analyzed

the decline of homicides finding that it continues to be highly concentrated, but the

decrease has not been uniform.

Keywords

homicides, crime concentrations, spatial point pattern test, Pact for Life Program

Introduction

In Brazil, between 1980 and 2010, more than one million people were murdered

(Waiselfisz, 2012). According to Julio Jacobo Waiselfisz, a researcher in Brazil, the

situation in this country can be considered an epidemic because its rate is greater than

10 homicides per 100,000 inhabitants, a threshold set by the World Health Organization

(United Nations Development Programme [UNDP], 2013). Brazil has a higher homi-

cide rate than the most-populous countries in the world such as China, India, the

1Federal University of Pernambuco, Recife, Brazil

2Simon Fraser University, Burnaby, British Columbia, Canada

Corresponding Author:

Martin A. Andresen, Institute for Canadian Urban Research Studies, School of Criminology, Simon Fraser

University, 8888 University Drive, Burnaby, British Columbia, Canada V5A 1S6.

Email: andresen@sfu.ca

634405HSXXXX10.1177/1088767916634405Homicide StudiesPereira et al.

research-article2016

22 Homicide Studies 21(1)

United States, Indonesia, Pakistan, Nigeria, Bangladesh, Russia, Japan, and Mexico

(Waiselfisz, 2014a), and is more similar to countries like South Africa with its levels

of violence (Breetzke, 2010).

Although Brazil had made many efforts to reduce crime in recent years, the number

of homicides is increasing. In 2002, the country had a homicide rate of 28.5 per

100,000 inhabitants, with a small decrease to 25.2 per 100,000 inhabitants in 2007.

After 2007, Brazil had an increase in its homicide rate, reaching 29 homicides per

100,000 inhabitants in 2012 (Waiselfisz, 2014b). Such a trend at the national level is

counter to the international crime drop that began in the 1990s for most crime types

and most countries that have been investigated (Farrell, Tilley, & Tseloni, 2015;

Farrell, Tseloni, Mailley, & Tilley, 2011; LaFree, 1999; Ouimet, 2002; Tseloni, Mailley,

Farrell, & Tilley, 2010; van Dijk, Tseloni, & Farrell, 2012).

The growth of these homicides is not a trend for all states. Some states have been

highlighted by the reduction in homicides, such as Pernambuco, São Paulo, and Rio de

Janeiro. In 2000, Pernambuco was considered the most violent state in Brazil, with 54

homicides per 100,000 inhabitants (Waiselfisz, 2012). According to the Secretariat for

Social Defense of Pernambuco (SSD; 2014), this homicide rate dropped to less than 32

homicides per 100,000 inhabitants in 2013. This is a large magnitude drop (approxi-

mately 40%) but is counter to much of the international crime drop literature that has

found decreases in crime rates at both the national and subnational levels, indicating that

similar processes are operating at different scales of analysis (Farrell et al., 2011; Farrell

et al., 2015; Ouimet, 2002). Clearly, Brazil and its states are different, and understanding

local conditions (local and micro places) may prove to be important for understanding

the decrease in homicides within Pernambuco and Recife, specifically. Such an under-

standing could prove to be instructive for not only Brazil but also the international com-

munity in general because of the implications for reducing violent crime.

Theoretical Background

Some of the research over the past 25 years within spatial criminology has shown that

crime is highly concentrated in a small number of places (Curman, Andresen, &

Brantingham, 2015; Sherman, Gartin, & Buerger, 1989; Weisburd & Amram, 2014).

These places are referred to as micro-spatial units of analysis, such as street segments

and intersections or actual addresses. This relatively small literature has been able to

replicate this finding in a number of cities across the United States, Canada, and Israel.

Because of the consistency of these results, David Weisburd, Groff, and Yang (2012)

have put forth the “law of crime concentration at places” and recognize the presence

of chronic street segments in parallel with the long-known fact that a few chronic

offenders commit the vast majority of crimes (Wolfgang, Figlio, & Sellin, 1972),

along the lines of the Pareto principle.

The first study to emerge in this literature considering an entire city was that of Sherman

and colleagues (1989) investigating robbery, motor vehicle theft, and rape in Minneapolis,

Minnesota. These researchers found that 3% of street segments in Minneapolis accounted

for 50% of these crimes: 2.2% for robbery, 2.7% for motor vehicle thefts, and 1.2% for