NGOs in Bangladesh: Activities, Resources, and Governance 2003
This survey is one of the first large, nationally representative surveys of non-governmental organizations (NGOs) in a developing country. The NGO sector in Bangladesh is highly organized and relatively homogeneous. Most NGOs utilize a branch and headquarters structure in which branches have limited autonomy from headquarters. At the branch level, most NGOs in the country, whether big or small, focus on credit services, derive more of their income from fees for services than from grants, rely on salaried rather than voluntary staff, keep detailed financial accounts that are externally audited, and hire middle-class college educated men as managers. The convergence to a modal institutional form probably is the result of the persuasive power of ideas, sociological pressures toward acculturation and conformity, as well as material incentives.
Kind of Data
Sample survey data
Unit of Analysis
Anonymized version for public dissemination: name of NGO and NGO manager were removed, as well as name of mother NGO and address.
Unit of Analysis
What counts as an NGO is an important question for a survey like this one. No answer is self-evident, and the definition chosen depends in large part on the purpose of the research. Because the objective of this research project was to analyze NGOs that are involved in what the donors and government call ''development,'' organizations were deemed NGOs if they appeared on any of the lists used to synthesize the sample frame, or, if the NGOs were discovered during the field enumeration, reports from informed observers suggested that the NGOs were engaged in activities that are generally considered development work in Bangladesh, such as the provision of safe drinking water, sanitation, health care, adult and child education, agricultural training, roads construction, transport, skills training, land rights/tenure, credit, arsenic reduction, environmental work, employment generation, poverty reduction, advocacy in any of these areas, or other related topics. Although the survey questionnaire itself focused on the attributes of NGOs as service providers, they did not have to be engaged in service provision to be included in the sample frame; NGOs engaged in ''consciousness raising'' were also included. In fact, as the results below show, most NGOs were involved in both service provision and consciousness raising. Registration with the NAB is a precondition for obtaining foreign funding in Bangladesh. Because that list constituted an important element of the initial sample frame, the NGOs that were surveyed were typically those that had either applied for foreign funding or that might have considered themselves in a position to do so. The Department of Social Services keeps a separate list of ''social welfare'' or non-profit organizations in the country. That list is much broader than the list kept by the NAB, includes organizations as diverse as mosques, trade unions, tea clubs, and local sports leagues, and includes over 23,000 organizations. That list was not used for the construction of the sample frame because the activities of those organizations, while important for understanding the broader context of civil society organizations in Bangladesh, were not the subject of this research.
Two final points should be emphasized. First, the site to which the questionnaires were applied were the field branches of NGOs (and the headquarters offices of smaller NGOs if the headquarters offices were also engaged in direct development work with communities). Most NGOs in Bangladesh, even the smaller ones, had both field offices and headquarters. Because central headquarters often did not have comprehensive information about the activities and resources of field offices, the survey was applied to the branches themselves. A detailed analysis of the resources and governance of the branches from the point of view of the headquarters staff, while important both to validate the findings at the branch level and to obtain a fuller portrait of resources, the flow of funds, and governance, was beyond the scope of this study.
Second, the data presented here were provided by the NGO managers, and there may be discrepancies between self-reporting and what a thorough review of accounts, based on an external audit, might have revealed.
Producers and sponsors
Authoring entity/Primary investigators
Varun Gauri and Julia Galef
World Bank and Columbia University
The selection of NGO relied on a two-stage sampling design. The initial sample frame was constructed by merging a list of known NGOs from the government's NGO Affairs Bureau (NAB), a list of members of the Association of Development Agencies in Bangladesh (ADAB), and lists of the branches of the larger NGOs. On the basis of that synthesized list, 35 political-administrative units, thanas, were selected according to a procedure in which the probability of selection was proportional to estimated size, where size was understood as the share of registered NGOs in the thana. Before selecting NGOs to interview in each thana, survey interviewers conducted a field enumeration of the actual number of NGOs operating in the selected thanas. Employing that updated list based on field enumeration, NGOs were randomly selected in each thana on the basis of a stratified sampling procedure that assured statistically useful sub-samples among NGO types: ''big'' NGOs that had over 1,000 branches across the country (BRAC and ASA), ''big'' NGOs that conducted field operations in more centralized branches (Proshika and Caritas), and other typically ''small'' NGOs.
The decision to obtain representative subsamples of the large and small NGOs was based on the hypothesis that small NGOs might exhibit different governance structures, have access to fewer resources, and engage in different activities than the large NGOs. There is obviously a continuum between large and small NGOs; the decision was made to obtain statistically representative sub-samples of the four large NGOs named above because they are generally considered to have a relatively large ''market share'' in the sector, and because the government's Bangladesh Bureau of Statistics (BBS) made a similar distinction in its 2000 Household Income and Expenditure Survey (HIES), which distinguished between the existence of those four NGOs and other NGOs.
Although the community questionnaire of the HIES also inquired about Grameen Bank in its section on NGOs, branches of Grameen Bank were not included in the present survey of NGOs because Grameen Bank is not actually incorporated as an NGO.
THE SAMPLE FRAME AND THE SAMPLE
The initial list consisted of 6,559 NGOs and NGO branches in Bangladesh. The number of NGOs in a given thana ranged from a minimum of one (in 17 thanas) to a maximum of 192 (in the thana of Mohammadpur in metropolitan Dhaka).
The mean number of NGOs per thana was 15.1, and the median was 12. Of the 6,559 NGO offices, some 1,456 consisted of BRAC branches, 1,186 were ASA branches, 125 were Proshika branches, and 98 were Caritas branches. A large percentage (79%) of the small NGOs were also organized into headquarters and branch offices. Among the small NGOs with a large number of branches were Buro (56 sites), Family Planning Services and Training (51 sites), Dhaka Ashania Mission (41 sites), and Esdo (41 sites).
On average, the number of NGOs based on the field enumeration exceeded the initial list by 21%. One question with both policy and theoretical interest is the extent to which Bangladeshi NGOs are serving the needs of the poorest or most needy in the country. One way to address the question (though it has limitations because many NGOs serve clients outside the communities in which they are located) is to examine the concentration of NGOs in the poorest or most needy areas of the country. Income data from the preliminary returns of the 2001 census permitted an analysis of NGO concentration in a given thana as a function of average household income in that thana.
NGO concentrations per capita were lowest in the poorest thanas in the sample, suggesting that NGOs are not locating in the poorest regions of the country. It is also possible, of course, NGOs were targeting not poor thanas per se but poor areas within the thanas. To analyze that possibility, we analyzed NGO concentration per capita by community wealth quintile, where community wealth was constructed using the first principal component of three community wealth measures asked of participants in the focus group discussions (percentage of households in the community with televisions, cement floors, and a resident in paid employment). The concentration of NGOs per capita was relatively high in the second lowest wealth quintile but lowest in communities in the poorest quintile.
Although the results here should be interpreted with caution because they are from a random sample of only 35 thanas, they are consistent with other findings that NGO location decisions in Bangladesh are not targeting the poorest communities.
As noted above, some of the small NGOs in the sample were themselves relatively large, and included organizations with a headquarters and branch structure. Using appropriate sample weights that adjust for the frequencies of NGOs in the sampled thanas as compared to their thana frequencies in the initial synthesized list, we were able to estimate the relative percentages of NGO types in the country. Note that even among the stratum of small NGOs, a branch-headquarters structure was far more prevalent than stand-alone organizations, and that branches of international NGOs were relatively rare in the country.
Dates of Data Collection (YYYY/MM/DD)
Mode of data collection
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.