Community Mobilization, Individual Agency and Maternal Survival: Evidence from a Randomized Controlled Trial in Rural India

Rationale: It is widely accepted that maternal survival rates can be improved by positive health practises and effective treatment seeking behaviour. When women have little individual agency however, their control over health practises and care seeking may be significantly curtailed. This analysis is nested within a broader randomised controlled trial of womens' groups to improve maternal and child survival in Rural India. Objectives: This paper aims firstly to test whether participation in womens' groups improves individual agency and secondly, to measure the extent to which any change in individual agency affects care seeking during pregnancy. Methodology: The trial within which the study is nested sought to evaluate the impact of participatory community groups on maternal mortality, neonatal mortality and maternal depression using a cluster randomised controlled trial. The trial was carried out in 36 clusters in three contiguous Districts in Jharkhand and Orissa. The total population of the 36 clusters in 2005 was 228,186. Within each district, clusters were randomly allocated to control or intervention status, with 6 intervention clusters and 6 control clusters in each district. All 18 intervention clusters had women’s groups and all 36 clusters received health service strengthening. Monitoring and evaluation data, including data on individual agency, was collected annually from participating women in all clusters over a period of three years. The measure of agency used in this analysis is a multi-attribute index generated using principle components analysis. A two-stage least squares system of equations is used to measure a) any impact of the women's group intervention on individual agency and b) any impact of individual agency on care seeking at key stages in a woman's pregnancy. Findings & Conclusions: We will be able to measure the extent to which women's groups impact individual agency as well as the extent to which any changes in agency correspond to changes in care seeking during key stages in pregnancy (antenatal, delivery and postnatal). These findings will enable us to discuss whether increased agency is one of the pathways through which community mobilisation effectively improves care seeking behaviour and health outcomes