Conflict, parenting, and early childhood mental health in conflict-affected settings: Evidence from Colombia

We examine the effect of the exposure (or reduction in the exposure) to conflictrelated violence on parental mental health, caregivers' parenting stress and responsive caregiving, and early childhood mental health. We use data collected from the impact evaluation of a psychosocial group intervention in Tumaco, Colombia, a community chronically affected by the armed conflict. Using an Instrumental Variables approach, we use the program's randomization into cohorts and staggered design of the data collection to exploit a natural experiment in which the armed groups in the municipality agreed to a Truce and municipal violence rates dropped between data collections. We find that the exposure to recent conflict-related violence had negative effects across the four main dimensions: increased parental mental health problems in 0.68 sd (SE=0.342 ; q-value=0.074), increased parenting stress in 0.76 sd (SE=0.389 ; q-value=0.074), reduced responsive caregiving in -1.023 sd (SE=0.434 ; q-value=0.074) and increased child mental health problems in 0.556 sd (SE=0.343 ; q-value=0.074). By providing causal evidence on the direct effect of conflict-related violence on parenting outcomes, we conduct an exploratory mediation analysis to assess whether parental mental health, parenting stress and responsive caregiving behaviors may partially account for the association between violence and child mental health. Yet, correlational links between violence and parental mental health and parenting behavior outcomes as mediators suggest important associations for understanding children's mental health vulnerability in conflict settings.