Dinarte, Lelys Ileana (1); Baires, Wilber (2) (2017). 'Losing grams due to Gangs: The effect of Salvadoran Maras Violence on birth outcomes.' Paper presented at the annual conference of the HDCA, Cape Town 2017.


Children’s health at birth is crucial in the development of their later well-being, both in the medium and long term. Almond, Chay and Lee (2005) and Almond (2006) obtain evidence of the effect of low birth weight on different health and education outcomes in adult life, such as brain problems, lower IQ, behavioral issues, blood pressure, and cognitive under-development. These health and learning problems have an impact on their productivity and, therefore, on their ability to generate income in the future, limiting their probability of emerging from poverty and intergenerational mobility (Duflo, 2001).

 The objective of this research is to measure the impact of in-utero exposure to gang violence shocks on at birth outcomes during the period 2000-2015 in El Salvador. We will focus on shocks that occur during the first trimester of pregnancy, because medical evidence indicates that the greatest impact of these stress shocks occurs during the first few months of gestation (Glynn et al, 2001). Additionally, we will restrict our analysis to shocks related to violent acts committed by gangs in El Salvador, which include homicides, curfews and massacres; separating these effects by their intensity. The advantage of focusing on gang violence is that it is not only relevant for El Salvador, but for all countries in the northern of Central America.

 Evidence of the impact of violence shocks at birth health is still scarce, and has been developed using different types of exposure to violence (Aizer, 2010; Fergusson, Horwood and Ridder, 2005; Curry and Harvey, 1998; Lipsky et al, 2003; Brown, 2008; Camacho, 2008; Mansour and Rees, 2011; and Koppensteiner and Manacorda, 2015). The two main mechanisms identified in the literature are exposure to stress by the mother and economic conditions faced by households, such as loss of resources and less availability of nutrients.



The database to be used in this research will be unique and constructed from three main sources. Data on children's health variables such as birth weight and height; and maternal health, such as the mother's pregnancy period, preterm birth, among others; will be obtained from the Statistics of Births of the Ministry of Health.

 Violence shocks data will be obtained from two sources. First, data on homicides committed by gangs in certain municipalities and the exact date they occurred will be get from the Information Office of the National Police. A second source will be the main newspapers in El Salvador. From these sources we will obtain information related to curfews, massacres of gang members, clandestine cemeteries, etc. This is an innovative feature of our research, since this database does not exist and will not only be used to measure the effects of shocks of violence on children's health outcomes at birth, but also for other types of economic outcomes that are relevant In terms of public policy.

Estimation methodology.

To identify causal effects, we propose to use Differences in Differences. This will be based on differential changes in the level of violence to which mothers are exposed, as measured by a variable indicating whether the pregnant mother was living in a specific geographic area during a violent shock.

scroll to top