We evaluate the effectiveness of a post-conflict development programme on maternal
health-care utilization in the Chittagong Hill Tracts of Bangladesh. Our work varies from
conventional impact evaluation studies because of the inclusion of two post-conflict
psychosocial risks: the household’s actual experience of violence, and subjective perceptions
about violence, as key determinants of programme effectiveness. Following the difference-indifference
estimator, and propensity score matching method this study establishes that the postconflict
development programme undertaken by Chittagong Hill Tracts Development Facility of
the United Nations Development Programme is successful in improving maternal health-care
utilization. Despite this, forced settlement by outsiders, household experiences of conflict, and
perceptions of insecurity lower maternal health-care utilization. The effectiveness of the
programme would have been greater in the absence of conflict, although the programme may
have mitigated some experiences of past conflict. The intervention fails to significantly narrow
the inter-ethnic gap in terms of health-care utilization, chiefly attributable to the adverse effects
of the forced settlement of non-indigenous peoples in the region.