Microcredit and women's empowerment in Bangladesh: a structural equation model for categorical observed variables

This paper employs the Bangladesh Demographic and Health Survey (2004) to explore the relationship between participation in microcredit programs and women's empowerment using a structural equation model with categorical observed variables. A MCMC-based Bayesian approach is adopted for estimation. Along with participation in microcredit, we consider a variety of socio-cultural aspects as potential predictors of empowerment in the Bangladeshi context including men's perceptions about women's status. We conclude that gender community norms are strongly rooted in women's minds regardless of the partners' perceptions of women's status, and microcredit interventions may actually contribute to change gender beliefs and social attitudes.