The survey was conducted as a part of a Country Gender Assessment (CGA) for Bangladesh undertaken by the Social Development Unit in the South Asia Region (SDU-SAR) of the World Bank. The CGA was aimed at filling some key remaining or emerging gaps in knowledge on gender issues in Bangladesh and then presenting an identification and analytical assessment of the emerging gender issues in the country to national actors, bank operations and other donor activities related to Bangladesh for their perusal.
Mitra and Associates carried out the two-month survey with the main objectives to understand the dynamics of the marriage market, including the changing nature of the matrimonial contract, importance of dowry, its role in determining women’s employment, and insecurity and violence within and outside marriage, among others. The findings of the study were then incorporated into the GSA of the WB SDU-SAR.
The survey collected data from around 5,000 adults in a nationally representative sample of 2,000 households across the country, each of which had at least one woman member. The sample was drawn in two stages following a sample design similar to that of the Bangladesh Demographic and Health Survey (BDHS). At the first stage, 90 clusters were selected as the sample sites, where a cluster was a census village in case of rural areas or a census block in urban areas. They spread across 57 districts, with 64 rural and 26 urban clusters. Households were sampled at the second stage by selecting 22 to 23 households from a cluster to yield the total sample size of 2000 households.
Besides a retrospective component of the oldest female of the household, the core features of the survey instrument included ages of marriage, nature and amount of dowry or, for earliest memories, bride price, income, education and a substantial sub-module on the nature of marriage choices and arrangements. Considering the unconventional nature of questions of these issues, Mitra and Associates made a careful pre-testing before actually employing the instrument to collect date from the selected respondents.
A special smaller module to understand the perceptions on gender relations administered to local decision-makers was also a part of the survey.
The survey involved three broad categories of activities: preparatory activities, fieldwork, and data processing, including data cleaning.
The preparatory activities comprised:
The fieldwork consisted of:
The data processing phase included:
The survey was conducted in February–March 2006 and a cleaned set of survey data was submitted April 2, 2006 to the World Bank Bangladesh office that sponsored the study.