Human Development in Hostile Environments: Impact of Urban Violence on the Academic Performance of Colombian Children

• In Event: 1-014 - Indirect Effects of Violence On Human Development: Lessons from Diverse Latin American Samples

Integrative Statement

Exposure to adverse experiences during childhood and adolescence limits human development and reduces the probability of a person reaching his or her maximum potential. In fact, exposure to a violent event, either directly or indirectly, negatively affects human development (Sharkey, 2018). This issue is of great importance for Colombia, a country transitioning out of the longest-running civil conflict in the western hemisphere. While significant progress in terms of security has been made over the last decade, the number of homicides is still extremely high. In 2016, approximately 12,000 intentional homicides were reported nationwide, representing a rate of 24.4 homicides per 100,000 inhabitants. In this paper, we posit that these events not only affect the social and emotional development of children (Molano et al., 2018), but they also their cognitive functioning.
This study test this hypothesis analyzing the performance of Colombian fifth and ninth graders on the State’s national standardized test. We focus on assessing how the psychological consequences of acute exposure to violent events affect children and adolescents’ performance on standardized math and language components of the test. Our empirical strategy seeks to overcome the limitations of previous studies. Following a strategy similar to the one proposed by Sharkey et al. (2012), we use both time frames and geographic proximity to identify the impact of homicide exposure on test performance in Medellín and Cali, two of Colombia's four major cities (which together represent 13.4% of the country’s urban population). To do this, we develop a measure of exposure to violence which we deem exogenous, that is, uncorrelated with other variables that could potentially explain both test performance and exposure to violence (such as household, school or neighborhood quality). Using time and geographically stamped datasets, we identified the number of homicides (i.e. violent deaths) that occurred in each day within a 500-meter perimeter around schools.
Our results show a negative causal effect between the number of homicides that occur around schools and students’ results on the standardized academic tests. In particular, we find that a homicide in the vicinity of the school (i.e. less than 500 meters away) occurring four days before the test generated a decrease of 0.18 to 0.24 standard deviations in math scores for fifth and ninth graders and 0.15 standard deviations in language scores for fifth graders (see Table 1). This average effect represents the difference between a student's score falling at the median and a student at the 43rd percentile. In addition, we identified a gradient in performance relative to exposure time; the impact on academic performance becomes more negative as homicides occur closer to the test date. However, we find that the effect of the exposure to violence on the academic performance of fifth graders does not disappear even when the adverse event happens more than 5 months before the academic assessment. This result is particularly important, since it shows that young students are not able adapt in order to be completely resilient to violent contexts.

Authors

• Andres E Molano, Universidad de los Andes
  Presenting Author

• Arturo Harker, Universidad de los Andes
  Non-Presenting Author