Who Participates and Who Doesn’t? Lessons from a Community-based Development Project in Bangladesh

It is generally claimed that the second wave of participatory development, which received momentum in the 1990s, tends to be informed with broad-based social inclusion, social change and transformation. However, many have portrayed participatory development —the way it has been managed — as tyranny. This paper is a further attempt to explore some of these tyrannies of participation by focusing on a participatory development project in Bangladesh. This paper draws on a comparative case study on two project communities, and investigates: first, how socio-economic differentiation influences the extent of participation; and second, how this influence may be exaggerated by faulty project implementation guidelines. It is found that people’s ability to participate in community development project is greatly influenced by socio-economic status measured by, for example, household income, educational background and residency status. No additional effort from the project management side was made to include the most disadvantaged community people in the project activities or to make the weaker participants more active in group activities.
Main Article: 


During the 1990s, the growing interest in participatory development, stemming from the critiques against state-led top-down models of development, pushed various international development agencies, for example, World Bank (WB), United Nations Development Program (UNDP), Organization for Economic Co-operation and Development (OECD), and The Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), to implement various development projects utilizing participatory mechanisms. This second wave of participatory development, which gained momentum in the 1990s, tends to be informed by broad-based social inclusion, social change and transformation (Gaventa, 2003). However, participation in development has been criticized severely during the last two decades. Many have portrayed participatory development —and its management —as ‘tyranny’. Some such ‘tyrannies ‘were vividly highlighted by Cooke and Kothari (2001) in their book Participation: The new Tyranny? This paper is a further attempt to explore some of these tyrannies of participation —exclusion of the weaker community members, domination of the local elites in leadership positions and faulty project implementation strategies. The present paper focuses on a UNDP-sponsored development project in Bangladesh, namely the Local Partnership for Poverty Alleviation Project (LPUPAP). The paper draws on an empirical study on two project communities, and investigates: first, how socio-economic differentiation influences the extent of participation; and second, how this influence may be exaggerated by inappropriate project implementation guidelines. Comparing the research data and findings with those associated with other projects or development interventions is outside the scope of this paper.

Discussion of the Key Concepts and the Theoretical Underpinnings


LPUPAP has defined community as, “a body of persons sharing common problems, living in a physically identifiable area” (Government of the People's Republic of Bangladesh, 2001 p35). Thus, community in development practices is often perceived as synonymous with neighborhood, and as a ‘natural’ social unit. As Yuval-Davis (1993) observes: “the community at times is perceived as an organic whole which is ‘out there’ where one can either belong to or not” (cited in Akello, 1994 p14). This limited view of community might lead to failure of development interventions (Guijt and Shah, 1998). In reality, rural as well as urban communities, bounded by definite geographical area, are a complex arena of interaction, where various groups (men, women, youth, children, poor, rich, landowners, landless, permanent residents, renters etc.) cooperate or compete with each other as a strategy of their livelihood. Therefore, Smith (1990) notes, “the notion of a community is always something of a myth. A community implies a coherent entity with a clear identity and a commonality of purpose. The reality is that communities, more often than not, are made up of an agglomeration of factions and interest groups often locked in competitive relationships” (cited in Awortwi, 1998 p5). In this respect, the paper tries to understand the influence of these community level socio-economic factors in determining the extent of participation of the community members in the study area.

Community Participation

Generally, community participation is such a process where people should have access to decision-making, implementation and benefit sharing. But, the concept of community participation is contested, and there is no clear and commonly agreed definition. However, the confusion is more about ‘when’ and ‘how’ to achieve it, rather than why participation is necessary (Mukhopadhyay, 1993).

For Moser (1983), it is important to see whether the notion of participation is a ‘means’ to enhance efficiency, effectiveness and sustainability of a project by incorporating local knowledge and resources, or an ‘end’ where participation itself is counted as a goal, so that community could have greater control over resources and regulative institutions in a given social situation which ultimately requires redistribution of power (Abbott, 1996). Nonetheless, the practical application of this normative differentiation is rather difficult. Therefore, Moser (1983) notes, “In reality it is not the evaluation of participation either as a means or as an end which is important, but the identification of the process whereby participation as a means has the capacity to develop into participation as an end” (cited in Abbott, 1996 p36).

Similarly, Rifkin (1990) proposes three questions to determine participation: ‘Why participate?’, ‘Who participate?’, and ‘How do people participate?’ (refered by Akello, 1994 p16). These are the three crucial questions in assessing the process of participation. Community people should have incentives to participate. The individual goal behind participation could be complementary to the common well-being of the community, or it could be only for narrower self-interests. The next crucial question comes about participants—do they represent the whole community, or only the privileged groups? Finally, it is important to see the intensity of participation, which is described by Arnstein (1969) as ladder of participation. By considering the fact that for effective participation there is a need for building viable organization at the community level which will mobilize diverse sections of the community to put forward community needs, this paper primarily addresses the question: who participates and who does not in the community groups espoused by the LPUPAP intervention?

Socio-economic Differentiation and Community Participation: The Theoretical Underpinning

The general notion of modernization theories is that economic growth leads to democratic political changes. It is Lipset (1959), who in his paper ‘Some Social Requisites of Democracy: Economic Development and Political Legitimacy’, made a claim to this direction (Lipset, 1959). Follow-up studies as well as attempts at refutation have abounded (Lipset, 1966; Hadenius, 1992). However, reviewing much of this vast literature more than 30 years later, Larry Diamond finds that on the whole Lipset’s general claim has been vindicated, as well as many of his notions of the mechanisms that relate economic development to democracy (Diamond, 1992). Thus, economic growth (GNP per capita) drives political change in the democratic direction through the enhancement of the quality of life for large segments of the population. It seems to mean that the distribution of wealth also matters for democratic change in the political culture, often stimulated by higher levels of education and diffusion of new values through international contacts (especially at the elite level). Class structure changes because of economic growth. Therefore, at the macro level, various indicators of socio-economic modernization have been shown to drive democratization across nations by increasing citizens’ participation in political life.

Similar processes operate at the micro level. Many studies have revealed that certain socio-economic factors affect the level of community participation. For example, research by Milbraith (1965), Almond and Verba (1963), Hausknect (1962) and Lee (1998) indicate that low socio-economic status as measured by income, education and occupation is associated with low level of participation(refered by Awortwi, 1998 p17). Bamberger (1998) argues that focusing on the poor is relatively easy, but bringing them together to actively participate in a self-sustaining development process is rather difficult. Because there are differences in the ability of the most deprived groups in a society to participate even in a situation like level-ground field (referred by Awortwi, 1998 p4, p12). The differences in ability are mainly determined by differences in socio-economic status, for example, household income, educational qualification and residency status are important among many others.

As such, my main focus here is to explore how the community people’s accessibility and inaccessibility to the project groups were influenced by the community level socio-economic differentiation. Thus, the paper does not reflect that much on the effect of group membership on the participants’ socio-economic condition.

The Background Information of the Project

Bangladesh[i] is one of the most populous and densely populated countries in the world. Since the mid-1970s, the country has experienced rapid urbanization at an annual growth rate over 6% (Ghafur, 2000 pp262-63). The implications of such rapid urbanization are increasing concentration of poverty, inequality, high unemployment, underemployment and mushrooming of urban slums and squatters in urban centers. Urban poverty is generally associated with poor housing arrangement, inadequate sanitation facilities, unhygienic and filthy living condition and the absence of other services like health, education and so on.

In response to this state of affairs, UNDP supported LPUPAP from 2001 to 2007. It was one of the largest UNDP sponsored poverty alleviation projects in Bangladesh under the Ministry of Local Government, Rural Develop1ment & Cooperatives (MLGRD&C). Municipal/City Corporations, together with Local Government Engineering Department (LGED) were the sole implementing agencies of the project. The project was implemented in 11 towns and endeavored to improve the socio-economic and political status of the urban poor through empowering poor communities and building the capacity of local government bodies. Instead of the traditional supply-driven approach, the project adopted community based, bottom-up and demand-driven approach to achieve its development objectives: community groups were formed to identify and prioritize needs, communicate effectively with local government bodies, and participate in designing, implementing and evaluating community level development activities (GHK-International, 2006).

At the community[ii] level, there were two types of community groups: the Primary Group (PG) and the Community Development Committee (CDC). A PG was composed of persons in the community that wanted to address a development issue of common interest. The maximum number of PGs in a community was 20, each with 15 to 20 members. If a PG started a savings and credit operation, it was known as Savings and Credit Group (SCG). Every PG had a Group Leader (GL) and Secretary elected by group members for two years. Group funds were operated jointly by the GL and Secretary. PG members were supposed to meet once a week to discuss various issues, and workout plans to address those issues. On the other hand, a CDC was comprised of the GLs and Secretaries of the PGs in a community. CDC chairpersons, vice-chairpersons, secretaries and the treasures were elected as office bearers[iii] for two years by the members of the CDC. Every CDC had a constitution of its own and was supposed to receive accreditation from the local authority (municipality/city corporation). CDCs were the focal point for implementation for all development activities and addressing common development issues faced by the entire community. Therefore, CDCs were entrusted with a list of functions, for instance, physical implementation, water and sanitation, environment and solid waste management, literacy campaigns and establishing schools, social and cultural activities, law and order maintenance etc. The project provided two types of funds directly to the CDCs: the Community Development Fund (CDF) for physical infrastructure development and the Poverty Alleviation Fund (PAF) for income generating activates and skill development.

Research Methodology and Data Sources

This study employed an exploratory and descriptive case study social research method. The case study method is appropriate when an in-depth investigation is necessary. The paper primarily draws on primary data supplemented by secondary data. Primary data was collected through field investigation in 2006 lasting about one month in two project communities in Rajshahi City Corporation (RCC) area, situated in the North-West part of Bangladesh. According to the official evaluation report, 2006 (GHK International, 2006) RCC area was one of the best performing LPUPAP project towns. One of the reasons for selecting RCC area was that most of the project components were already well-established at the time of the field study, and the project operation was matured enough to conduct field investigation. Furthermore, RCC area was also selected for the sake of manageability —a term which is often mentioned in social science research (Bhuiyan, 2004, Annex I). I knew the city to a great extent. Therefore, it was easy for me to use my experience and network during my field work. Data was mainly collected from 60 community people through structured interviews using questionnaires comprised of both quantitative (restricted) and qualitative type questions (open-ended). At the beginning of the field work, two LPUPAP communities, Ramchandrapur Baze Kazla (West) and Khulipara Mohaldarpara, were selected purposefully. The following criteria influenced the selection of the communities: first, in both communities the project had been working for about five years; second, all the project components were introduced into both communities; third, project officials helped me to identify these communities, where I expected interesting variations in performance. A total of 30 respondents (10 GLs, 10 general group members and 10 non-participant community members) from each community were selected randomly. In addition, interviews were conducted with the two concerned Ward Councilors (WCs), two concerned Community Development Workers (CDWs), and the local project coordinator. Secondary data was collected from various published books, journal articles and various LPUPAP documents.

A Brief Description of the Study Communities

Inhabitants and their Identities

Baze Kazla (West) was composed of two types of inhabitants holding two different identities. The eastern part of this community was mainly populated by the migrants who came from different parts of the country—normally known as Gristho or Bengali. The other part (western part) of this community was populated by the people who had originally migrated from Murshidabad, a District of India, were known as Mohaldar. The term Mohaldar means fishermen. These people were distinct from the Bengalis in terms of cultural practices. During an informal discussion with me regarding various community related issues, one of the inhabitants from Mohaldar portion divided up the community by giving stress on ‘we’ and ‘they’. Although identity difference between Bengalis and Mohaldar was difficult to trace at the surface level, it was powerful and dominant in determining local power structure at the community level. Unlike Ramchandrapur Baze Kazla, Khulipara Mohaldarpara (from this point in this paper ithas been termed as Khulipara) was mainly populated by Mohaldar people, who were living in this locality since long time. Therefore, the community was relatively homogeneous in terms of group identity.

Socio-economic Profile of the Communities

The total number of household in Baze Kazla (West) was about 500 and the total population was 1968 (543 were children). On the other hand, the total number of households in Khulipara was about 370 with 1968 members (469 were children) and was less densely populated than Baze Kazla (West). The overall socio-economic condition of the inhabitants in both communities was almost the same, and most of them represented the poorest of the poor. However, the overall socio-economic condition in Khulipara was found better in comparison to Baze Kazla (West) due to diversified job opportunities, since the locality was very close to the city center (about one kilometer). The entire community of Baze Kazla was under the constant threat of river bank erosion as it was outside the city protection dam. During the field research, inhabitants informed that just fifteen years ago the river bank was almost one and half kilometer away from where it was at the time of the field study. Besides river bank erosion, the community had to also face the plight of floods at regular intervals. On the contrary, Khulipara had relatively better infrastructural facilities. The community was under the City Corporation’s water supply coverage. However, the community had severe water logging problem mainly during the rainy season.

From the collected data, it was found that 74% of the total residents in Baze Kazla (West) lived below the subsistence level as their monthly household income was less than TK. 3000[iv], with the average household size of 4.73. Although economic condition was relatively better in Khulipara, the income of the majority of the households (53%) waslower than BDT 3000 per month with the average household size of 4.8. In both communities, generally men were the main wage earners and women were mainly involved in reproductive activities with a few exceptions. Most of the people were engaged in informal sector activities, like rickshaw pulling, vegetable vending, petty trading, retail fish trading etc. Many of such jobs were seasonal and irregular. However, in Khulipara, the majority of the people were somehow related with fish mongering since it was their traditional occupation, and their income was relatively more stable than in Baze Kazla (West), and it was found that 60% of the total households in Baze Kazla (West) did not have regular income .

The overall literacy rates were 77% and 83% in Baze Kazla (West) and Khulipara respectively. However, about 27% among them could only sign their names in Baze Kazla (West) and the figure was 17% for Khulipara. The lower socio-economic status of the both communities was also visible by observing the household characteristics. In Baze Kazla (West), about 47% houses were Kancha[v] and 47% were Semi-Pucca. In contrast, only 7% houses were Pucca. Furthermore, about 60% of the total inhabitants of this community lived in rental houses. On the contrary, the relative higher economic condition was also visible by seeing the house conditions in Khulipara. About 17% houses were Pucca, and 53% were semi-Pucca. The majority of the residents (70%) was permanent and lived in their own houses.

The study data show that the residents of Khulipara enjoyed relatively better socio-economic status in comparison with the residents of Baze Kazla in terms of population density, job opportunity, infrastructural facilities, average household income, literacy rate and housing condition.

Participation and the Development Outcomes in the Study Communities

In both communities, the project started working in 2001. At the time of the field visit, the total number of PGs in Baze Kazla (West) was 14 (4 fromGristho portion and 10 from Mohaldar) comprising 260 members with total savings of BDT 212,800[vi]. Although Mohaldar were the majority, CDC leadership was mainly occupied by Gristho people —3 out of 4 office bearers were from them. In Khulipara the number of PGs was 13 with 255 members and the total savings was BDT 344,460.

Baze Kazla (West) CDC completed two CDF contracts till that time. Under these contracts, a total of 105 latrines, 6 tube-wells and 639.65 meter footpath had been constructed. Furthermore, under the PAF, 19 (18 males and 1 females) community people received on-the-job training. In Khulipara, a total of 98 latrines, 13 tube-wells, and 261 meters drain had been constructed under two CDF contracts. Furthermore, 33 community people (31 males and 2 females) received on-the-job training under PAF.

Interviews with project officials revealed that both Baze Kazla (West) and Khulipara were officially recognized as well-performing project communities for many reasons: (i) the majority of the households living in those communities participated in community groups; (ii) group meetings were relatively regular; (iii) attendance rates at PG and CDC meetings were relatively higher; (iv) both CDCs successfully completed two CDF and one PAF contracts; (v) both CDCs had relatively higher amount of savings; (vi) both were running relatively smooth savings and credit program; (vii) there were some really successful cases of income generation through the PAF and credit operation; and (viii) in general, both enjoyed, better sanitation and walkway facilities than before.

Accordingly, research data suggest that the extent of participation was higher in Khulipara with 68% of the total households participating in PGs, while the figure was 53% for Baze Kazla (West). But, one of the noteworthy findings is that, with some degree of variation, a large segment of the ultra-poor[vii] was excluded from being participants and, thus,beneficiaries of the project. Furthermore, the group leadership positions were largely occupied by the dominant section of those communities.

Socio-economic Differentiation, Project Rules and the Extent of Participation in the Study Communities

One of the noteworthy findings is that there was a great variation in household income between participant and non-participant community members -participants were comparatively better-off than the non-participants. Economic ability to pay the required BDT 10 as group savings per-week restricted a huge number of poor households to participate in groups in both communities. Therefore, the project merely excluded the poorest of the poor. This phenomenon was also well-documented in the Report of the Evaluation Mission (2006) by saying that the project was not explicitly targeted to the ultra-poor[viii](GHK-International, 2006, p59). Furthermore, within groups there was variation in household income between office bearers and general group members in both CDCs. For instance, in the case of Baze Kazla (West), no one among the office bearers had a household income below BDT 4000 per month; whereas among the general group members no household’s income exceeded BDT 5000 per month, with 80% of them in the range of BDT 1001 to 4000. Such income difference was also evident in the case of Khulipara CDC and even more acute where all office bearers’ household income exceeded BDT 5000 per month.

There were also differences in educational qualification among the participants and non-participants.From the educational profile of the two CDCs, firstly, it can be easily seen that participants had higher educational qualification than the non-participant community members in both communities. Many were illiterate among the non-participant community members (40% in Baze Kazla (West) and 30% in Khulipara). On the other hand, the majority of the group members had at least primary education (45% in Baze Kazla (West) and 50% in Khulipara). Besides the difference between the participants and the non-participants, a considerable extent of differentiation did also exist between the office bearers and the general group members. In both communities, the minimum level of education of the office bearers was primary education. In contrast, 60% of the general group members in both communities were either illiterate or could only sign their names. However, the project fund was not available for literacy campaign or for adult-basic education.

Apart from differences in household income and educational qualifications, residency status also came out as an influential determinant as to decide who participated and who did not. In both communities, 100% of the group members were permanent residents having their own houses. On the contrary, a considerable number of the non-participant community members did not have their own housesand most of them lived in rental houses (60% in Baze Kazla (West) and 30% in Khulipara), who were considered as temporary residents. When the office bearers were asked about this issue, they indicated that there was an implicit instruction from the project office to avoid non-permanent members to be included in the groups because many of the project benefits were directly related to having your own house and land. In addition, group leaders also ignored non-permanent residents since giving credit to them is a bit risky.

Therefore, the research findings clearly show the interplay between community level socio-economic differentiation and the extent of participation in community groups. In accordance with the theoretical assumptions of the modernization theories, it comes out that lower level of socio-economic attainment was associated with lower level/non-participation in the project groups. Furthermore, some of the project rules regarding group membership, for example, weekly savings and residency status restricted some of the community members from being group members. Besides these, reluctance of the project management to initiate additional efforts with the intention to make disadvantaged members more active in groups paved the way for concentration of power in the hands of those with higher socio-economic status.


The paper shows that socio-economic status measured by family income, educational attainment and the type of residency shaped the extent of participation in both the study communities. Community groups were largely represented by the relatively well-off family members. In both communities, the group leadership positions were occupied by those who were economically more solvent and had relatively higher educational attainment. At the same time, temporary residents of the both communities did not have access to group membership. As a whole, the relative higher socio-economic status of the community people in Khulipara influenced in achieving higher level of participation (68%) than in Baze Kazla (West) where 53% of the total residents participated in project groups. Thus, despite the general claim that the second wave of participatory development tends to be informed with broad based social inclusion and social change, this case study clearly shows that LPUPAP was far of attaining such broad-based social inclusion. The project implementation guidelines were biased towards well-off members of the communities. Moreover, there was no additional effort from the project management side to make the disadvantaged members —ultra-poor, illiterate or less educated and non-permanent residents —more active in groups.


Abbott, J. (1996), Sharing the City, Community Participation in Urban Management, London: Earthscan Publication Ltd.

Akello, M. (1994), Community and Women Participation in Rural Water and Sanitation East Uganda Programme (RUWASA), Hague: Institute of Social Studies.

Arnstein, S. R. (1969), "A Ladder of Citizen Participation." Journal of the American Institute of Planners 35, (4) pp216-224.

Awortwi, N. (1998), The Riddle of Community Development: Factors influencing organization, Participation and Self-Management in 29 African and Latin American Communities, Institute of Social Studies.

Bhuiyan, M. S. H. (2004), Unveiling the Face of Social Capital: Evidence from Community-Based Solid Waste Management Initiatives in Urban Bangladesh, a PhD Thesis Bonn: The Centre for Development Research, University of Bonn.

Cooke, B. and U. Kothari (2001), Participation: The New Tyranny?, London: Zed Books.

Diamond, L. (1992), "Economic Development and Democracy Reconsidered", American Behavioral Scientist, Vol. 35 pp450-499.

Gaventa, J. (2003), "Towards Participatory Local Governance: Assessing the Transformative Possibilities", Participation: From Tyranny to Transformaiton,: Manchester, UK.

Ghafur, S. (2000), "Entitlement to Patronage: Social Construction of Household Claims on Slum Improvement Project, Bangladesh." Habitat International, Vol. 24 pp261-278.

GHK-International (2006), "Report of the Evaluation Mission", Local Partnership for Urban Poverty Alleviation, UK: GHK International.

Government of the People's Republic of Bangladesh (2001), Implementation Guideline, Local Parternship for Poverty Alleviation Project, Dhaka: Ministry of Local Government, Rural Developement and Cooperatives.

Guijt, I. and M. K. Shah (1998), "Waking Up to Power, Conflict and Process", The Myth of Community: Gender Issues in Participatory Development, in I. Guijt and M. K. Shah (eds) London: Technology Publications.

Hadenius, A. (1992), Democracy and Development, Cambridge:Cambridge University Press.

Haldar, S. R. and P. Mosley (2004), "Working with the Ultra-Poor: Learning form BRAC Experiences." Journal of International Development, Vol. 16 pp387-406.

Lipset, S. M. (1959), "Some Social Requisites of Democracy: Economic Development and Political Legitimacy", The American Political Science Review. Vol. 53(1) pp69-105.

Lipset, S. M. (1966), Political Man: The Social Bases of Politics, London: Mercury Books.

Mukhopadhyay, A. (1993), "Community Participation in Urban Development", Urbanisation in Developing Countries- Basic Services and Community Participation. in B. Mohanty (Ed.), New Delhi: Institute of Social Science, pp331-344.

End Notes

[i] Bangladesh is a South Asian county emerges as an independent state in 1971.

[ii] The idea of community in the project was very much close to the idea of neighborhood. The project identifies communities mainly on the basis of geographical boundaries. The minimum number of households in a LPUPAP community was 30.

[iii] Either the chairperson or vice-chairperson must be a female member.

[iv] 1 US dollar = 70 BDT. According to the Bangladesh Institute for Labour Services (BILS, 2006), the minimum cost of living of a family of four, is to be Tk. 4,800 per month (Clean Clothes Campaign, 2006).

[v]Pucca: the structure which has its roof and wall made of bricks and mortar; Semi-pucca: it is a structure of normal height and has walls made of bricks. The roof is made of any material other than cement/concrete; kacha: it has a ceiling which is low and is made of very cheap construction materials like straw, bamboo, chhan (grass), golpata (leaves), polythene sheets, gunny bags etc.

[vi] 1 US dollar = 70 BDT.

[viii] According to a World Bank Study, Ultra-Poor are those who are physically fit for work but still very poor and consume less than 1805 kilo calories per person per day Haldar, S. R. and P. Mosley (2004). "Working with the Ultra-Poor: Learning form BRAC Experiences." Journal of International Development16: 387-406.