Marco Escobar

While studies of crime and the relationship between crime and socioeconomic variables are fairly common in wealthier countries, few researchers have investigated crime in less industrialized countries. Colombia, the country in which the city focus of this study is located, had in 2003 64 homicides per 100,000 people (Policia Nacional, 2004) compared to 2 for Canada (Statistics Canada, 2004) and 5.7 for the United States (FBI, 2003). Throughout the 1980s and early 1990s the City of Medellin (Colombia) had higher homicide rates than the Colombian average, even though the annual number of homicides steadily declined over the last fourteen years. This paper examines homicides that occurred in Medellin during 2004 and 2005. Homicides were geocoded and a kernel density surface was created to identify hotspots of homicide occurrences. Half of the hotspots were located in the north of the city in neighbourhoods in which the predominant socioeconomic strata are one, two, and in a smaller proportion three. The area with the lowest density of homicides was located in the southeast in neighbourhoods with the highest strata mostly 5 and 6, low percentages of unemployment and high average annual employment income. The homicide rate by neighbourhood shows significant spatial autocorrelation, with a cluster of high homicide rates in the downtown area and clusters of low homicide rates in the southeast and the central east side. A 2004 survey provided demographic and socioeconomic variables measured at the neighbourhood scale, which were used to build regression models with the homicide rate as the dependent variable. The regression results indicate that there is a weak relationship between the socioeconomic variables and the homicide rate by neighbourhood in Medellin during 2004 and 2005.