Document(s) 6 of 10

“People living in poverty have the least access to power to shape policies — to shape their future. But they have the right to a voice. They must not be made to sit in silence as ‘development’ happens around them, at their own expense. True development is impossible without the participation of those concerned.”

— Nelson Mandela, Speech on receiving the Ambassador of Conscience Award 2006, Johannesburg, November 1, 2006

As shown in Parts 1 and 2, CBMS developed as a unique poverty-monitoring system, specifically designed to collect and analyze information at the local level for use principally by local authorities in diagnosing the extent of poverty, planning and budgeting development programs, and assessing these programs’ impact. CBMS emerged from the realization that finely targeted poverty-reduction interventions can best be achieved through understanding and applying individual and household information about the poor, involving communities with local governments, and ensuring political commitment for inclusive development. CBMS initiatives are therefore as much about improving local governance and empowering communities as they are about facilitating planning and resource allocation.

The examples that follow show that CBMS grew out of a process of interaction between local researchers, policymakers, and communities, adapted to specific on-the-ground conditions in different contexts, regions, and countries. It emerged as a localized data system for use in planning poverty reduction and other public sector programs at the local level, and not as simply a survey of the poor. Its community-level census approach to monitoring, however, also challenges its institutionalization as a national program, including the capacity-building required for local governments to scale-up the system and its integration with national poverty-monitoring systems.

The CBMS implemented in different countries (Table 2) share core elements. Each, however, reflects an adaptation to local needs and conditions as well as to the broader political, economic, and social environment.

ASIA Philippines: From cradle to national implementation

CBMS results led to the establishment of Task Force Clean and Green and a health patrol to address health and nutrition problems. A health centre was constructed; a supplemental feeding program for children was introduced; toilets were distributed to households to improve the sanitation problem.

In January 2005, the National Statistical Coordination Board officially endorsed CBMS as a tool to strengthen the statistical system at the local level and directed its technical staff to promote the adoption of CBMS by Local Government Units (LGUs). In early February 2006, the Philippines’ National Anti-Poverty Commission announced that it was adopting CBMS nation-wide. This welcome news followed similar directives by the Department of the Interior and Local Government that LGUs should adopt the CBMS core poverty indicators for poverty monitoring and planning.

By January 2009, CBMS was being implemented in 52 of the country’s 81 provinces, including 531 municipalities and 42 cities, totalling 13 498 barangays. The goal: 100% coverage by 2010, the target date for the national implementation of a core local poverty indicators monitoring system.

Starting in Palawan

The story of how CBMS evolved in the Philippines starts in the province of Palawan. One of the hurdles provincial officials faced when they began to plan the 1999 budget was a lack of detailed municipal, village, household, and individual level information. This led officials to the CBMS developed through the IDRC-funded MIMAP-Philippines project.

Following a joint assessment of the province’s data needs and availability by the CBMS research team and the Provincial Planning and Development Office (PPDO), provincial Governor Salvador Socrates issued an executive order for the creation of CBMS technical working groups within local governments, setting the stage for its institutionalization throughout the province. After pilot testing in two barangays of the Municipality of Taytay, the system went province-wide in 2000.

The strategy for implementing CBMS was based on partnership and resource-sharing, coordinated by provincial and municipal governments. Enumerators were recruited from the barangays and trained by the PPDO. After the household census was completed, PPDO staff returned to train community representatives in processing survey results. Municipal planning officers consolidated the data, while the CBMS team provided overall technical guidance and training for the provincial and municipal trainers, with support from IDRC. The cost for the first round of the survey amounted to only PhP 5.03 million (US$100 265), or US$1.42 per household. A substantial portion of this cost was for capacity building, the demand for which will decrease in subsequent rounds. The provincial government shouldered 13% of the cost; participating municipal governments provided the rest.

Improving the systems

The first round yielded valuable lessons about the types of questions asked and how they were phrased, which subsequent rounds helped to address. The validation process proved to be a particularly useful step in assessing the accuracy of the data as well as learning how communities understood the causes of poverty. CBMS implementation was most effective where municipal governments assumed a strong coordination role and where orientation, enumeration, and processing were carried out within a short time frame: delays reduced quantity and quality of the data and added to the cost.

Data processing was more difficult. Since many areas in Palawan had limited or no access to electricity, processing was done manually in the barangays and municipalities. The training was initially inadequate for those who were doing data processing for the first time and, for many, processing the local language questionnaire into English proved difficult. This led, in some cases, to processing having to be done by PPDO staff.

In spite of these difficulties, this first experience demonstrated the value of CBMS to the government. The provincial government used the CBMS data to prepare Palawan’s first Human Development Report (2001). Municipal governments used the data to prepare their annual investment plans and to prioritize projects for poverty reduction, including investments on rural infrastructure, water supply and sanitation, and health and education services. Furthermore, the very precise information helped local governments evaluate the impacts of some of their projects.

A notable innovation in Palawan was the integration of geographic information systems (GIS) to present CBMS data in the form of easy-to-understand colour-coded maps that show each household in the community (Figure 2). GIS proved extremely instrumental to CBMS not only in depicting the poverty situation graphically but also as a tool to communicate with communities and policymakers.

The poverty maps are colour-coded representations of the poverty indicators. Unlike poverty maps generated from small area estimates (such as estimates for municipalities from national sample surveys that are designed to generate estimates for higher levels of aggregation), the CBMS maps show the poverty situation starting from the household level. By aggregating the household level data, estimates of poverty indicators are also generated. Poverty maps are then produced, portraying the poverty situation of the villages, municipalities, or provinces. With the colour-coded maps, policymakers can prioritize areas and households when delivering programs.

Figure 2. Example of GIS-generated maps presenting CBMS data (colours have been altered for this publication). The software used to generate these maps was the Natural Resource Database (NRDB) freeware, developed by a British Voluntary Services Overseas co-operant (Richard Alexander), customized to suit CBMS data (www.nrdb.co.uk). It is now an integral component of CBMS.

For example, when poverty maps were presented during a provincial planning convention, the head of municipal planning and development office in Bataraza, Palawan commented that the data on access to sanitation facilities by households seemed to be very low, as the municipality had recently carried out a latrine-distribution program in partnership with the national Department of Health. But when the municipal planner returned to the town to verify the CBMS survey results, he discovered that the latrines were distributed but not installed: the households were supposed to install them and contribute the cement and labour. Unfortunately, this had not been done and the latrines were never used. As a result, the program was redesigned to ensure that public investments into sanitation were properly used.

The successful implementation of CBMS in Palawan led to successive rounds in 2002, 2005, and a fourth in 2008. The information allowed planners to assess whether policies, programs, and projects implemented to address the problems were yielding the intended results, and whether living conditions were improving. For example, the provincial government became aware of the low school participation rate. This prompted them to examine more closely the issue of access to basic education and the quality of education being provided. In June 2006, the Palawan Regional Development Council reported that “CBMS has greatly benefited the province of Palawan,” including helping to establish computerized databanks in almost all municipalities.

CBMS data has also been used for other planning purposes. In 2005, for example, the provincial government of Palawan prepared a poverty map of the entire province, in collaboration with the Peace and Equity Foundation and the Palawan Network of NGOs Inc. The National Red Cross in Puerto Princesa used the data to select sites for its Integrated Community Disaster Planning Program and to identify health and sanitation needs in communities.

Spread and evolution

From Palawan, CBMS spread to other provinces and is now in the process of being scaled-up up nation-wide. The costs of implementation have been largely borne by local governments, a clear indication that they see the value of the system. Other stake-holders have also contributed, which bodes well for sustainability.

The growth of CBMS in the Philippines has been accompanied by adaptations to local circumstances. The CBMS team has assisted local governments in identifying and developing indicators specific to their communities — on migration, environmental protection, business-related activities, and natural disasters, among others. Similarly, the data collection and processing tools and training modules (in English and Tagalog) have been customized, resulting in four sets of household profile questionnaires and computerized processing systems incorporating specific local government concerns.

The CBMS team also developed a computerized national repository system, installed at the National Anti-Poverty Commission and the League of Municipalities of the Philippines. This repository houses all the CBMS data collected by the local governments and assists the Commission in its mandate of coordinating poverty-reduction policies and programs. The League can now use municipal-level data to identify and address inter-municipal concerns.

As the system has spread, so have the benefits. In barangay Kalamunding, Labo, for example, CBMS results led to the establishment of Task Force Clean and Green and a health patrol to address health and nutrition problems. A health centre was constructed; a supplemental feeding program for children was introduced; and toilets were distributed to households to improve the sanitation problem. A scholarship program was also launched to enable deserving students to finish their education. As significant, the barangay’s village captain (Constancia Labios) reported that attitudes also changed: CBMS has “prodded the parents in the community to attend to the health needs of their children such that more children aged 0–5 years are no longer malnourished.”

The changes engendered by the information provided by CBMS to government officials and communities have further been reflected in policy and planning. As a 2004 review of the MIMAP program commissioned by IDRC noted, “Local officials have acknowledged that the community-based monitoring system made possible by MIMAP has helped depoliticize and strengthen the local government’s budget allocation process by providing an objective basis for budget prioritization” (Saumier et al. 2004).

Rural to urban migration

CBMS has also found a home in the Philippines’ crowded cities. Pasay City, for example, pilot tested CBMS in 2004 and became the first city in the national capital region to adopt it in its program planning and budgeting cycle. The indicators and questionnaires were customized to include data on the victims of crime and the number of persons with disabilities, as well as on the number of households with family members working overseas.

The city partnered with church-based NGOs to implement the CBMS “to preserve the integrity and transparency of the data collected,” says Mayor Wenceslao Trinidad. In 2005, CBMS was conducted throughout the city’s 201 barangays. Among the findings was a noticeable mismatch between labour supply and the types of employment available. A number of measures were subsequently taken to remedy the problem, including a public–private–civil society partnership to inventory skills and promote jobs and city-sponsored job fairs.

Spreading the word

The CBMS team based at the Angelo King Institute has been guiding the replication and adaptation throughout the country. As CBMS has grown, the team has been able to draw a number of lessons about what conditions need to exist for the system to root itself in the community, and how it can best be carried out. These are summarized in Part 4 of this book.

The Filipino experiences with developing and implementing CBMS have been well documented and disseminated. Advocacy is a critical component. Publications, digitized maps, databoards, a computerized database, an online database, meetings, and forums are just some of the tools used by the research team, collaborating agencies, and communities themselves.

Evidence-based program design, implementation, and monitoring have continued to be promoted through the CBMS Development Grant Program, launched in 2005. This program provides funds to local governments and NGOs to undertake interventions that address development needs identified through CBMS surveys. Funded by the United Nations Development Programme–Philippines and the Peace and Equity Foundation, the program has awarded 25 grants to date.

Vietnam: Focusing on basic needs in communes

Among those plans and programs have been vocational training in such areas as traditional crafts to generate employment, particularly for women, assistance to improve inadequate housing, and the provision of tools and agricultural inputs to boost food production.

Standing in the yard of her modest home on the edge of rice fields in Dai-Yen commune in Chuong-My district of Ha Tay province, a villager displays her poverty certificate. It attests that she is poor — or rather, that she is poorer than her neighbours in this farming community of close to 5 000 people on the outskirts of Hanoi. The card entitles this widowed mother of two to free health care, preferential loans, and other services. It is a valuable entitlement. Her right to the certificate was identified by other commune members — and by the CBMS piloted here.

This commune and others in the province serve as poverty observatories. Here, researchers have been working with the Managing Office of Vietnam’s National Program on Hunger Eradication and Poverty Reduction (HEPR) and the Provincial Department of Labour, Invalids and Social Affairs (DOLISA) to implement CBMS.

Vietnam, once one of the poorest countries in the world, has made impressive strides in reducing poverty. The country now has one of the fastest economic growth rates in the world, averaging around 8% annually, and has emerged as a global exporter of many commodities and manufactures. The impact of Vietnam’s economic policy reforms over the past decade and its rapid integration into the global economy has significantly improved living standards, bringing new wealth and uplifting the poor. According to the Vietnam Living Standard Surveys conducted by the General Statistics Office, the poverty rate dropped from 58.1% in 1993 to 20% in 2004.

The number of poor was more than halved during that same period but, says Vu Tuan Anh, Vice-Director of the Socio-Economic Development Centre, the country still has to overcome many challenges. There are significant regional disparities (notably between the cities and rural mountainous regions) and the gap between the haves and have-nots has increased. To meet Vietnam’s MDG of halving the number of people living below the poverty line by 2010 and reducing by 75% those living under the food poverty line, two national programs have been launched: the National Target Program of Poverty Reduction 2006–2010 and the Program for Socio-Economic Development in Ethnic Minority and Mountainous Regions. But, says Vu Tuan Anh, identifying the poor for these programs, and evaluating their progress and success, requires reliable information (Vu 2007).

Most of the information on the poor in Vietnam is generated by national level surveys, such as the biennial Household Living Standards Survey (HLSS). Separately, poverty assessments have been carried out through the Ministry of Labour, Invalids and Social Affairs (MOLISA) and its provincial departments (DOLISA) at the commune level, as well as through various social organizations. These assessments have been largely on the basis of income, have been fairly subjective, and have not been comparable.

While these surveys offer good information at an aggregate level for the central government, they are not appropriate for planning poverty interventions and allocating resources at subnational levels of government. And because of the lack of reliable poverty estimates at local levels, schemes have been formed by different communities and social organizations — such as the Women’s Union and Farmers’ Union — for targeting the poor in their different locales. These have been uncoordinated and unreliable, however. Moreover, says Vu Tuan Anh, frequent changes in personnel in commune administrations and social organizations have meant that records have not been systematically kept.

Complementing data sources

Obtaining accurate, detailed data is the goal of the community-based poverty-monitoring system that Vu Tuan Anh and his research team have been developing since 1996 with IDRC support. The research team developed and tested a poverty-monitoring method based on simple indicators, short questionnaires, and participatory techniques. Local staff carried out the survey, which combined quantitative and qualitative methods: the questionnaire and indicators were combined with group discussions and interviews with key informants. Results were analyzed using simple data-processing software.

Some of the initial results were surprising. In a commune of Lam Dong province, for instance, researchers found that only half of the poor households were receiving the credit to which they were entitled under the poverty-alleviation program, and that they were using it to meet basic consumption needs, such as food, and not for longer term production-oriented poverty reduction activities as intended (Asselin and Vu 2005).

The research concluded that poverty monitoring in Vietnam’s rural areas should incorporate measures on basic needs rather than only on income and expenditure measures, be understandable and useful to a wide range of local users, and be participatory. In fact, viewed in the context of Vietnam’s long tradition of community participation, the researchers found that CBMS created opportunities for local communities to define their poverty as well as to participate in reduction efforts.

The initial research led to CBMS being implemented to monitor a poverty-alleviation project in 30 communes of Thanh Hoa province. This project was supported by the Canadian International Development Agency (CIDA) and the government of Thanh Hoa province, and was carried out by CECI, based in Québec, Canada.

Subsequently, in cooperation with the research team, the Managing Office of the national HEPR program selected a set of poverty observatories in 20 communes of 12 provinces to pilot test the CBMS. The data enabled a detailed picture to be drawn of poor households. The results were also used to assess government poverty-reduction programs such as healthcare provision, education, housing, and credit. The Managing Office of the HEPR found that the CBMS baseline data would help assess the implementation of future poverty-reduction policies.

Increasing coverage

From this base, CBMS spread to different geographical and socioeconomic regions of Vietnam, including to 30 commune observatories in all 13 districts of Ha Tay province and 10 commune observatories in eight districts of Yen Bai province. Elsewhere, the Women’s Union of Ninh Binh province implemented CBMS in 27 communes of the poor, mountainous Nho Quan district. Five communes in Quang Nai, a province in Vietnam’s central coastal region, are implementing CBMS. And, in Lam Dong, a province of the southern Central highlands where most of the population belongs to ethnic minorities, seven communes are using CBMS.

Indicators in Vietnam cover the community situation, household living standards, and the implementation of poverty-reduction policies and programs. As Vu Tuan Anh points out, indicators can be modified to reflect local needs. In Ninh Binh province, for instance, indicators were added on education, employment, the participation of women in social activities, and household decision-making.

“Because of this census, we have information we have never had before,” says Le Thanh Trinh, chairman of the Gia Son Commune Administration in the very poor Nho Quan district. “Based on this data we can make plans.” Among those plans and programs has been vocational training to generate employment, particularly for women, assistance to improve inadequate housing, and the provision of tools and agricultural inputs to boost food production.

Vu Tuan Anh points out that provincial authorities responsible for implementing poverty-reduction programs have benefited greatly from the CBMS results. The collected data has been used to assess poverty, evaluate poverty reports received from communes, and monitor the implementation of policies and programs of the National Targeted Program of Poverty Reduction. Several provinces are now requesting technical assistance to implement CBMS. And recently, questionnaires and indicators were adjusted to facilitate reporting on the MDGs: the national indicators now also include child mortality, mothers’ health, and the incidence of HIV/AIDS, among other diseases, and draw on CBMS data sets.

Bangladesh: From research to practice

The information gathered helped to identify those who should benefit from public programs.

A low-income country, Bangladesh has about 144 million poor citizens, the third largest concentration in the world. While the economy has grown and livelihoods improved over the past decade, it has a considerable way to go to meet its objective of reducing poverty. The country’s poverty profile reveals areas of extreme poverty as well as growing inequalities. Its 2005 Poverty Reduction Strategy Paper (PRSP) called for a more detailed multidimensional monitoring system taking in the ground realities that are useful for policymakers and planners.

“Poverty has always been on the government’s agenda,” says Mustafa Mujeri, former leader of the IDRC MIMAP-Bangladesh project, launched in 1992 by the Bangladesh Institute of Development Studies. One objective of that project was to strengthen the country’s ability to monitor poverty. As Mujeri explained, poverty data in Bangladesh focused almost exclusively on income and consumption. Surveys were carried out infrequently — every 5 to 10 years. And more than 5 years could lapse between data collection and dissemination.

Working with the Bangladesh Bureau of Statistics (BBS) — the only national systematic data source in the country — the MIMAP research team met with various government departments, civil society, and other researchers to develop a multi-dimensional core set of 12 poverty indicators. Data was collected in 21 districts throughout the country. The indicators were refined and adjusted over the life of the project, including increased gender disaggregation.

The research team sought to find ways of ensuring that the data could be collected, processed, and delivered to policymakers in the shortest time and in forms they could easily understand. Training programs helped develop the capacity of BBS to collect process and verify data. Overall, it was found that the MIMAP poverty-monitoring system indicators provided deeper measures of poverty than those previously available. The effort paid off: MIMAP’s poverty-monitoring survey data was used to prepare Bangladesh’s PRSP. It also influenced budgetary allocations, increasing the percentage allocated to the social sector.

Involving communities

As communities and local governments became involved in monitoring poverty, the MIMAP poverty-monitoring system evolved into a community-based approach known as the Local-Level Poverty-Monitoring System (LLPMS). Pilot-tested in 2002–2003 in four villages of a Union Parishad, the system had three major components: participatory poverty and development monitoring; resource profile monitoring; and village development planning.

The pilot test was carried out by the Bangladesh Institute of Development Studies and the Bangladesh Academy for Rural Development, a national training and action-research institution under the Ministry of Local Government, Rural Development, and Cooperatives. Local institutions and the community actively participated in the process, and trained local youths collected the information from households.

Local officials and villagers collaborated with each other to identify and prioritize problems in an information book for each village, while the research team provided technical assistance. The information was shared with villagers and a ward information book was developed, setting out both village and household data and prioritizing identified problems. The villagers recognized the usefulness of the system and started using it to pressure higher levels of government to deliver better public services.

The trial enabled researchers to identify a number of challenges. First and foremost, these included mobilizing local people and obtaining local government support (a prerequisite for sustaining the process). To ensure the quality of data — and the collaboration of all households — the team found that enumerators needed to be of different professions, ages, and gender, and from different socio-economic groups. The questionnaire needed to be kept simple and cross-checked through focus group discussions, and carried out during the slack agricultural season to allow sufficient time for interviews (Mujeri and Guha 2006).

The success of the pilot test led to its replication the following year in all villages of Muhammadpur Union (West), under the Daudkandi Upazila of Comilla District. Local government representatives and villagers were trained to analyze the information and to identify and prioritize problems to develop a pragmatic plan. Ward meetings were held to disseminate findings.

Villagers were motivated by the information — new to them, they said — to organize to fight poverty. The result was that representatives from neighbouring Union Parishad asked that the LLPMS be implemented in their areas and representatives of the National Statistics Department be involved in the process.

Linking research to practice

The lessons from Bangladesh are consistent with those from the Philippines and Vietnam:

• Support and training need to be provided to local people in collecting and tabulating data.
  
  
• The systems’ sustainability rests on the involvement of governments.
  
  
• Governments at all levels, as well as other agencies, need to participate to ensure the wide use of the data.
  
  
• Linking the CBMS and national indicators helps sensitize policymakers.
  
  
• Indicators and questionnaires need to be simple and short.

The researchers also found that information dissemination by the local government officials was instrumental in mobilizing people. Says Ranjan Kumar Guha, the Academy’s assistant director and CBMS-Bangladesh project leader, “Incorporating some household data in the village information book helps to ensure the quality of the data because villagers go to check what is recorded about them.” Local authorities also noted that the information gathered helped to identify those who should benefit from public programs such as government-issued vulnerable group feeding cards (Guha 2006).

Academy researchers have been actively promoting the LLPMS through a series of workshops. But institutionalizing the system in Bangladesh is a long-term process requiring the commitment of union-level officials as well as NGOs and community representatives in partnership with the research community and higher levels of government.

Cambodia: Improving local statistics and governance

When the survey in Kbal Snoul village showed how many children did not attend school because none was accessible, a donor stepped forward to build a new school within walking distance.

Reducing poverty is a pressing need in Cambodia where more than 35% of the population lives under the poverty line of US$0.50 a day. The country is also committed to the process of decentralization to make the provision of public services more responsive and more effective to citizens’ needs. As part of this process, 1 621 communes were established as legal entities, governed by councils first elected in 2002, to be responsible for local development planning and implementation.

As Try Sothearith, deputy director of the National Institute of Statistics’ (NIS) Department of Demographic Statistics, Census and Survey, notes, “Commune councils need adequate information, generated in a systematic and reliable way, in order to effectively conduct their needs assessments, planning, monitoring, and evaluation of development projects” (Sothearith et al. 2006). At the time, commune databases existed, but the information they contained was largely garnered from administrative reports completed by the village chief without household visits. Moreover, the databases, set up under the SEILA program (a joint government–donor program to promote local planning and development in support of decentralization), were to close at the program’s end in 2006.

In 2003, the Cambodia Development Resource Institute, in collaboration with the National Institute and the Seila program, piloted a CBMS in six communes in two provinces. To meet the goal of creating a sustainable system to locally monitor poverty over time, the project emphasized institutional capacity building at the local level.

Drawing on the experience of other Southeast Asian countries where CBMS had been established, and with the assistance of the CBMS Network Coordinating Team from the Angelo King Institute in the Philippines, a core set of nine indicators was developed.

The CBMS developed in Cambodia involved all levels of government. School teachers and other knowledgeable villagers were recruited and trained as enumerators. Village chiefs worked with the enumerators in listing households and mapping villages. Commune councils supervised the survey teams and data processing, helping to ensure their buy-in. District and provincial statistical offices handled data cleaning and coding as well as data entry. Finally, the research team from the National Institute of Statistics analyzed the data, wrote reports, and disseminated the results.

The project yielded valuable results in terms of adequately describing the different facets of poverty in the pilot communes and in building the capacities of local authorities. According to CBMS project leader Sothearith, “The pilot project successfully promoted links between the commune, provincial, and national level planning processes.” And, as important, the project “built the capacities of local authorities to implement and take responsibility to upgrade CBMS in their localities.”

Broadening the reach

Results of the pilot project were widely shared and NIS carried out a second round survey of 12 communes in 2006. Because most of Cambodia’s rural areas lack electricity and processing had to be done manually, one district officer introduced the use of a second-hand computer powered by a car battery. This small innovation enabled commune officers to process data electronically in even the most remote areas.

The CBMS survey results were widely disseminated and used by commune councilors, local development partners, and by other line ministries and development organizations. For example, when the survey in Kbal Snoul village showed how many children did not attend school because none was accessible, a donor stepped forward to build a new school within walking distance. The information also led to other investments. Two roads were built to link the village to the main road, saving villagers a 45-minute walk on a narrow path. Discovering the high number of households that were landless led to a program to allocate them public land.

In fact, participating communes noted that the CBMS results were an excellent tool for attracting donor assistance as they provided an accurate picture of the community’s needs. “CBMS is a clear measurement tool,” says one village councillor. “No other instrument provides actual rates. No one can challenge this data.” Survey results have also enabled communities to better deal with their own problems. For instance, the relatively high number of cases of domestic violence in one village has led to education and deterrence programs for offenders. Bringing the problem out into the open has led to a dramatic reduction in cases of violence.

The success of CBMS in Cambodia has led to considerable demand from both policymakers and local government planners for scaling up. The NIS is planning for the next round during which the coverage will extend to all the communes in three more districts, with the intent to cover one entire province. This has also generated interest among other provincial governors and development organizations. The main drawback, however, is low absorptive capacity and a shortage of external technical and financial assistance.

In an effort to address these constraints, H.E. San Sy Than, NIS’ director-general, has proposed that a CBMS bureau be established within the Institute to promote CBMS throughout the country and to build capacity among statistical officers and local communes. This would greatly assist the implementation of Cambodia’s 2005 Statistics Law, which also proposes that an NIS planning officer be placed in every commune in the country.

Lao PDR: Bottom-up data collection and poverty targeting

The CBMS provided the most reliable base from which to start socioeconomic development planning and has been used to improve the targeting of poverty alleviation projects.

Nearly one-third of the Laotian population is estimated to live in poverty. Many of the poor live in isolated regions and comprise numerous distinct ethnic groups with special needs. In the Laotian context, poverty is understood differently by different cultural groups: understanding culture and social structure is therefore important to meet the needs of the poor. The Government of Lao PDR has thus put considerable emphasis on “people-centred” development and improving “livelihoods” in its national effort to meet its MDG objectives and move out of its “least developed country” status.

To reach the goal of achieving the MDGs by 2015 and becoming a lower middle income country by 2020, the Lao PDR government adopted a National Growth and Poverty Eradication Strategy in 2001. The National Statistics Centre (NSC) was charged with poverty monitoring and analysis in the strategy’s framework. Implementing and monitoring it, however, requires detailed information about the poor. The strategy indicated that further disaggregating poverty information and analysis would help identify pro-poor policy actions (see World Bank 2006). It further acknowledged the importance of building capacity.

The approach taken by NSC was bottom-up, drawing from community-collected data to bolster national databases. To obtain this information, NSC introduced a decentralized system of Village Book Statistics in 2004 — a community-level demographic and socio-economic statistical database — to complement the Lao

Expenditure and Consumption survey to monitor poverty at the national level.

In implementing the Village Book system, however, NSC realized that a crucial step was missing: villagers were asked to aggregate data into the book but did not have any tools to collect data at the household level. Statistical capacity in Lao PDR was also weak at all levels. To address this problem, NSC pilot-tested CBMS in four villages of two districts. By 2005 it was extended to 24 villages in two provinces in the country’s poorest districts. In addition to providing timely information, the project was intended to facilitate the country’s decentralization efforts by directly involving people in the design of programs that best address their needs.

Through the project, NSC was able to develop an improved set of indicators to support the Village Book and inform local development activities, as well as corresponding data collection and processing instruments. The village enumerators were trained to collect, compile, and validate the data. Training was also provided in data entry, processing, and tabulation. The low level of education and weak statistical and computer skills of enumerators proved challenging and necessitated longer training than anticipated. The data was shared at the local level and with all statisticians in the provinces. The results were also disseminated in national forums.

According to Phosy Keosiphandone, deputy director general, Department of Planning and Investment of Saravan Province, CBMS provided the most reliable base from which to start socioeconomic development planning and has been used to improve the targeting of poverty-alleviation projects and resource allocation. In addition to yielding fundamental information at the village level on the poor for planning purposes, CBMS was invaluable in other respects. It provided evidence for policymakers to better understand the ground realities of the different communities and thus helped to prioritize public sector programs. The system strengthened capacity at the local level, as well as within NSC, and the involvement of communities greatly enhanced their own understanding of development programs, giving them a sense of ownership. It also helped to improve coordination between the NSC and local officials in planning.

Private investors, NGOs, and donors in Lao PDR have found that the CBMS data-enriched Village Books reduced the need for additional surveys in planning their programs. As Keosiphandone (2007) notes, the data explained discrepancies between improving incomes and stagnant poverty levels. “Interestingly, we found in the first year that one village had a high increase in income, but at the end of the second year the income poverty level had not declined. They spent most of their income on alcohol and satellite TV.” This helped local planners better understand local behaviour and views of poverty.

Indonesia: Tailoring CBMS to multi-dimensional indicators

Ensuring local ownership of the system and data, this initiative is fully supported by local government officials, including the mayor and his deputy.

Reforms in Indonesia have had an important positive impact on the poor, and include its system of decentralized government, which has sought to bring services closer to the people. However, despite significant strides in reducing overall poverty, close to 42% of Indonesians still live between the US$1 and $2 a day poverty lines, underlining the high vulnerability of many to falling into deeper poverty and the need to closely monitor the situation. More attention also needs to be paid to non-income multi-dimensional measures, such as malnutrition rates, access to safe drinking water, sanitation, and so on. There are considerable regional disparities and growing inequalities — all of which point to the need for disaggregated data for planning.

Indonesia’s decentralized planning structure and regional autonomy have given local governments responsibilities to implement development activities for their constituencies. When local governments needed to identify the poor to better target poverty reduction programs, they turned to the National Family Planning Coordination Board, the only national agency that annually collects household level data. Because the Board’s data was intended solely to monitor family-planning programs, it proved inadequate for the task. Subsequent efforts by local governments proved costly and unsatisfactory, largely because of a weak methodology and training of personnel (Suryadarma et al. 2005).

In the quest to introduce a better poverty monitoring tool in Indonesia, the Lembaga Penelitian SMERU Research Institute proposed pilot-testing a CBMS in collaboration with the Board in four villages in Java. A set of locally specific proxy indicators of welfare was developed, which included items such as asset ownership, health characteristics, political participation, and access to information.

These were used to score families’ welfare. To check the scores’ robustness, the richest and poorest families were compared. The comparison showed that there was a wide gap between the two groups in almost all the indicators, in all villages. The test also found asset ownership variables to be the most significant welfare indicators. Education, health, and consumption patterns were also important, but the importance of each varied from village to village. This convinced researchers that local specificity in indicators is crucial in Indonesia given its large population, size, and heterogeneity.

The pilot test also showed that, with training and supervision, Board cadres and educated villagers could be effective enumerators — in fact, although their education level was important in ensuring accuracy, their level of enthusiasm was just as crucial a factor. Interestingly, and contrary to CBMS implementation elsewhere, the Indonesia team considers that “village officials should not be involved in order to ensure that data is not tampered with.... Village officials are more prone to making mistakes in data collection.”

After successfully completing the pilot phase, SMERU started to train local governments interested in applying CBMS in their areas, with the expectation that the system can be institutionalized by local governments. The local government of the City of Pekalongan in Central Java was the first to plan to implement CBMS, starting in 2008, to support their poverty-reduction programs. With assistance from SMERU, it is collecting data on the socio-economic status of the community and will use it to better plan and budget. This project covers the entire city, consisting of 4 kecamatan (sub-districts) and 46 kelurahan (villages), with a population of 320 000 divided into some 80 000 families. The coverage will effectively institutionalize CBMS in the city.

It is expected that the City of Pekalongan will be a model for CBMS institutionalization for other local governments in Indonesia. Ensuring local ownership of the system and data, this initiative is fully supported by local government officials, including the mayor and his deputy, who are serving as supervisors, and a technical team consisting of officials of the Regional Development Planning Board, Planning and Evaluation Division, and the Administrative Division.

Other CBMS initiatives in Asia

Pakistan

In the context of the Poverty Reduction Strategy Paper process, Pakistan introduced far-reaching policy commitments that stress decentralization and devolution of powers to local governments. The local government ordinances, which established new accountability arrangements giving local government responsibility for service delivery and the establishment of community citizens boards, created new demands for local-level statistics and monitoring of public programs.

CBMS has been tested in Pakistan to provide an empirical base for budget allocation and planning within the local government. Researchers at the Pakistan Institute for Development Economics designed a core set of indicators to monitor and evaluate development policies and programs and presented it to national stakeholders. A customized set of data collection, processing, and validation tools was also developed and pilot-tested. Work focused on incorporating into CBMS as many indicators as possible from Pakistan’s National Reconstruction Bureau’s Information Management System.

While the administrative, political, and fiscal reforms supporting devolution proved conducive for CBMS to assist local governments in their planning and allocative responsibilities, political tensions and concomitant uncertainties and risks with partnerships between the research community, civil society, and various local government administrations and institutions have limited the development of CBMS.

Nepal

In Nepal, decentralization allowed local initiatives and development interventions to be conceived, designed, and implemented by Village Development Committees, the lowest government level. This created demand for local level information that was not provided by the Nepal Living Standards Survey, conducted every 5 years. Under a MIMAP project, the research team from the National Labour Academy launched an information-gathering system at the local level in 1997. Unlike other CBMS initiatives, however, the data was collected through focus group discussions rather than through household surveys. CBMS started to make some headway in Nepal’s system of local government and decentralized planning, but work in many of the rural areas was interrupted by political conflict and was discontinued.

Sri Lanka

Sri Lanka set a course for devolution with the creation of provincial councils and an elected local government structure consisting of village councils (Pradeshiya Sabhas), and municipal and urban councils. Local government institutions, however, were slow to develop and were hindered by political problems and conflict. They were also constrained by a lack of effective planning and development coordination. Nevertheless, the need for good data collection at the village level to support poverty reduction efforts remained. In response, the Social Policy Analysis and Research Centre at the University of Colombo undertook to pilot-test a CBMS in three locations in 2003.

The project yielded a detailed picture of the communities and confirmed the importance of locally relevant multi-dimensional indicators. The researchers concluded that although CBMS could become a valuable tool in Sri Lanka, it “can materialize only if a concerted effort is made to change the status quo” with respect to the lack of capability and empowerment within local governments (Hettige 2005). As CBMS project leader Siripala Hettige noted, “The persisting marginalisation of local government institutions within the Sri Lankan political system has been a hindrance to institutionalising CBMS within the local government framework in Sri Lanka” (Hettige 2007).

AFRICA

In October 2007, of the 41 countries eligible for relief under the Heavily Indebted Poor Countries (HIPC) initiative, 32 were in sub-Saharan Africa. Among them were Benin, Burkina Faso, Ghana, Senegal, Tanzania, and Zambia, countries where CBMS has been introduced. IDRC’s MIMAP program supported teams in Benin, Burkina Faso, Ghana, and Senegal, starting in the late 1990s as part of an effort to rethink approaches to combating poverty. This work led some IDRC-supported researchers to collaborate in preparing their countries’ Poverty Reduction

Strategy Papers (PRSPs). A key element in implementing the PRSP was the development of a monitoring and evaluation framework to track progress.

Burkina Faso: Empowering the poor

Since the survey, villagers have built retaining structures to capture rainwater for agriculture, thus increasing crops and reducing hunger.

The UNDP’s 2007 Human Development Report ranks Burkina Faso number 176 out of 177 countries in its Human Development Index, an unenviable position. To provide impetus to the country’s poverty-reduction efforts, Burkina Faso approved its first PRSP in 2000 and its second in 2004, the country’s reference point for defining development programs and activities. Among the PRSP’s main objectives are to build the capabilities of the poor and to implement strong partnerships with communities in poverty-reduction programs. Building local capacity to promote development at the community level and to reinforce local governance is also an integral element of Burkina Faso’s decentralization program.

Burkina Faso became part of the MIMAP network in the mid-1990s. The research was carried out by the Centre d’Étude, de Documentation et de Recherche Économique et Sociale (CEDRES) of the University of Ouagadougou and the Institut National de la Statistique et de la Démographie (INSD). In the wake of similar projects in Asia, CEDRES and INSD undertook in 1997 to pilot test a CBMS in partnership with a Canadian NGO, the Centre d’étude et de coopération internationale. The goal: to develop a methodology and select indicators.

Decentralization, and later the PRSP, set the stage for the development of the CBMS and for bridging the gaps and deficiencies in effectiveness or efficiency of the country’s development programs.

Those gaps include a lack of current, disaggregated data, of involvement in development planning, of participation in decision-making, of coordination between various programs, and of local capacity, particularly in rural areas where 90% of the population is illiterate.

The need for revisions

A pilot test was carried out in 110 households in three areas — rural, semi-urban, and urban — of Passore province. Drawing from the CBMS indicators developed in the Philippines, the Burkina Faso team selected the most relevant for local circumstances. Because of large seasonal changes in the mainly agricultural population’s living conditions, it was determined that data should be collected twice a year for some indicators, such as the use of healthcare facilities, once yearly for others.

The initial research showed that data-collection capacity in villages was weak. The existing tools and training manuals needed revision and more time had to be allocated to training. These were reflected in a second phase of research, launched in 2000. Among the new measures:

• community sensitization and training before the survey;
  
  
• the incorporation of a gender and development strategy to increase women’s participation in designing and implementing the CBMS; and
  
  
• creating greater synergy with local development financing programs so that needs identified by the communities could be met.

In 2002, the system was tested in five villages and one urban sector in the department (administrative division) of Yako. In 2003, this was extended to cover the department’s more than 73 000 people in 8 454 households. A controller supervised the work of locally selected and trained enumerators, while a central team ensured quality control. Data processing was done manually to respect the CBMS principle of community control and use. The research team helped aggregate and analyze the data electronically.

Because of the high level of illiteracy in the country, communicating the CBMS results back to the community posed a particular challenge. The solution: translate the data into easy-to-interpret drawings on paper and blackboards, one for each indicator — population size, health and nutrition, sanitation, education — posted at the village assembly offices (Figure 3). These drawings have also been used to illustrate handbooks and the information has been translated into Moore, a local language.

Figure 3. This poster illustrates an adult literacy rate of 11.1% in the village of Kabo.

The resulting detailed picture of poverty held up to the communities spurred some to action. In Lilbouré, for instance, the results galvanized the community into building retaining structures to capture rainwater for agriculture. As well, school enrolment more than doubled.

According to CBMS team leader Prosper Somda of CEDRES, CBMS is the only data collection tool that communities can use to develop evidence-based plans. Lilbouré wants to repeat the survey every 2 years but, says Somda, it does not have the resources or the expertise to do so on its own. And the computer skills needed for data analysis do not exist in the village.

In 2006, the team extended the CBMS to two new sites — the Departments of Diébougou and Koper, the latter at the request of Africa’s Sustainable Development Council, a consortium of development organizations. In Diébougou, the success of a pilot project in 5 localities led to the decision to survey the entire commune close to 40 000 people in 23 villages to prove that CBMS was feasible and to convince national authorities of its value to complement the national statistical system. As in Yako, the survey results painted a sad picture: high rates of child mortality and low levels of medical care, a severe shortage of sanitary facilities, low levels of school attendance, particularly for girls, and precarious living conditions.

In Koper, the findings showed that “For all indicators, the Department of Koper suffers from numerous deficits that constitute great challenges for local authorities and other development actors” (Konate et al. 2007).

In late April 2008 the mayor of Yako, reported that the CBMS results had enabled him to negotiate funding for priority development projects. “Whoever has information is wealthy,” he says. “I needed reliable data. I was working in the dark which made it difficult to obtain financing. The CBMS results are already being used to improve people’s welfare.”

Senegal: Supporting the PRSP

When village officials noted that schools were concentrated in the north of the village — and that children there were higher school achievers — they decided to build three new schools in underserved areas.

Two events were early stimulants for developing the CBMS program in Senegal. One was the passing in March 1996 of the decentralization law, which devolved economic and social development functions to local governments. The second event was the launch in 2001 of Senegal’s PRSP process.

As part of the MIMAP Senegal project, Senegalese researchers worked closely with the government on the design of the PRSP. The team included researchers from the Centre de recherches économiques appliquées and from various units of the Ministry of Economy and Finance. However, Senegal did not have an adequate system to monitor households’ living conditions. In October 2002, the Planning and Statistics Directorate led the development of an integrated system to monitor poverty, living conditions, and human development. A national observatory was set up to collect and analyze data and disseminate results that would be useful for local planning.

The CBMS Senegal team was tasked with developing and testing this system in three locations: Guédiawayé, a suburban Dakar commune; a semi-rural commune in Thiès region; and Ndangalma, a rural community in Diourbel. Local authorities in Guédiawayé confirmed their readiness to set up the system and to make available local personnel. In the rural commune of Ndangalma, inadequate access to services and markets had led to “a group dynamic among the local populations of participation in grass-roots leadership organizations for joint thinking and action to promote sustainable local development,” says Momar Ballé Sylla, the CBMS Senegal coordinator.

A vibrant civil society also existed in Tivaouane, Thiès. And, as in Guédiawaya, “the people have decided to take charge of their future,” says Sylla. In addition — and perhaps not coincidentally for the selection of Tivaouane as a pilot site — the then newly elected mayor, El Hadj Malick Diop, was a statistician formerly employed by the Planning and Statistics Directorate.

Both Tivaouane and Guédiawaye passed administrative acts to institutionalize the monitoring systems. No decree was passed in Ndangalma, but the community president made the community secretaries available to the research team.

Two questionnaires were administered: one to communities and one to households. The first covered such indicators as demographic characteristics, education and literacy, health and nutrition, community organizations, economy, and infrastructure. The household questionnaire gathered data on household composition, education, health, employment, and international migration, in addition to living conditions. In 2003, a small sample of households was interviewed in each locality because, says Sylla, it would be too expensive to survey all households. The sample was selected in collaboration with the district commissioner or village leaders. All district and village leaders were also surveyed.

The surveyors were recruited from the community and chosen by the mayor or rural commune president. The training program included instruction, testing, and translation of parts of the questionnaires from French into local languages, and field observation. The process led to a revision of the training manual. Data processing was done electronically by locally recruited and trained personnel.

A case for community validation

The research team provided the compiled results to city officials who shared them with the community through workshops. While community responses were generally positive, some noted the absence of indicators on income and maternal and infant mortality. The accuracy of some figures, such as the estimated population in Ndangalma, was also questioned, forcing researchers to return to the survey files.

Some of the results astounded the residents of Tivaouane. For instance, they discovered that more than 52% of the population was unmarried, an anomaly in a very religious community. But, says Mayor Diop, this is a manifestation of poverty. “With very high unemployment and crowding of two, three, and even four people in each bedroom, life as a couple is virtually impossible,” he says. When village officials noted that schools were concentrated in the north of the village — and that children there were higher school achievers — they decided to build three new schools in underserved areas.

The mayors of Tivaouane and Guédiawaye committed themselves to finance the collection of resources needed for a second phase and to expand coverage to other districts. Among others who expressed their intention to use the CBMS data were UNDP’s poverty program and the NGO Enda Tiers Monde.

Although the value of the CBMS was evident through the pilots, local officials were unsure if communities’ financial resources were adequate to cover the costs of future surveys. They also noted that additional training was needed for those charged with data analysis. Among other lessons learned was the need to sensitize households before the survey, to translate the questionnaires into local languages, to better adapt questionnaires to the locality, and to further validate the data collected.

CECI’s Louis-Marie Asselin points out that the Senegal CBMS differs from other CBMS projects in that it has a very complex sampling design at the community level and is highly computerized, using three software programs to process and analyze data. The local government would need to have advanced capability in sampling and computer skills to adopt this system. Asselin further notes that the relatively small sample may not be sufficient to meet the needs of local development planning. The links with the national statistical system offer high potential for policy impact, however.

Sylla notes a number of challenges in the way of full-scale implementation of CBMS in Senegal:

• obtaining the commitment and support of communities to introduce and finance their own CBMS;
  
  
• ensuring the continued use of the data generated to evaluate the impact of development projects; and
  
  
• maintaining relationships between stakeholders, particularly the Ministry of Social Development, local authorities, and villagers.

Benin: Strong municipal support

The mayor’s office of the town of Cotonou announced six priority actions to improve living conditions.… All the measures proposed were identified through a CBMS survey.

In 2007, the mayor’s office of the capital city of Cotonou announced six priority actions to improve living conditions in the city’s 13th district. Topping the list: the provision of sewage, sanitary, and health facilities; the extension of electrical services and safe water supplies; and mosquito-control operations to combat malaria. All the measures proposed were identified through a CBMS survey carried out the previous year.

With a per capita income of US$540 in 2006 and a ranking of 163 out of 177 countries according to the 2007 UNDP Human Development Report, Benin remains a very poor country. Reducing poverty and accelerating economic growth is the government’s priority, articulated in its PRSP. In addition, the government is pursuing a policy to improve public accountability and governance. But, as the Benin CBMS team notes, doing so requires the participation of communities at every step of implementation and monitoring.

In 2005, a team from the Université d’Abomey Calavi, supported by the Institut national de la statistique et de l’analyse économique, piloted a CBMS in the 13th district of Cotonou. In 2006, CBMS was extended to two other districts, Covè and Adjarra. To ensure that the data collected could be compared with national data sources, the team adopted the EMICoV (Enquête Modulaire Integrée sur les Conditions de Vie — integrated modular survey of household living conditions) questionnaire. The national EMICoV survey had been carried out in 2005. In total, 16 300 households were surveyed.

The CBMS aims to provide disaggregated data at the local level to support the country’s decentralization process, which started in December 2004. Benin’s Observatory of Social Change only monitors poverty and measures its impact at the national level.

Project leader Marie Odile Attanasso notes that the census highlighted great disparities in the communities surveyed. “The results urge the local authorities and the NGOs and development associations to come to the aid of the populations in order to improve their living standard and their living conditions,” she writes (Attanasso 2007). This, Cotonou’s municipal council took to heart. “This survey made it possible for the town council to give this district a real face,” said Mayor Nicéphore Dieudonné Soglo (CBMS 2008).

The mayor of Adjarra concurs: “This tool is very suitable to know the level of poverty of the town and to define the best strategies for poverty alleviation,” he said. “So I promise to mobilize funds to expand the CBMS through the town after this first phase.”

A number of challenges lie ahead, however, including a lack of human and financial resources and of statistical services at the local level, the low level of education of local authorities, and the unavailability of data-storage facilities.

From West to East Africa

In 2006, Kenya, Tanzania, and Zimbabwe joined the CBMS family. In Kenya, the system is intended to become an early warning system to monitor drought and ethnic conflict and to understand the relationship between the two. In Tanzania, the project complements the national government’s efforts to achieve participatory planning and monitoring in the context of decentralization. In Zimbabwe, although constrained by political and economic circumstances, the system is building on the community monitoring program launched in 2003 and seeks to strengthen the links between community monitoring and community research. Zambia joined the network in 2007 in an effort to link data from local-level monitoring with decision-making for poverty reduction .

Conclusion

As these project descriptions demonstrate, CBMS has progressively made inroads throughout both Asia and Africa. By 2008, CBMS programs had spread to nine Asian countries. The extent of CBMS work is growing both in breadth and depth as more local governments adopt this approach to refine policies and programs to address the needs of the poor. Interest is also growing as efforts to “localize” MDG monitoring gain prominence.

Clearly, everywhere, “CBMS has not only permitted the acquisition of a richer body of information and data on the welfare of the poor in developing countries. It has also led to the almost serendipitous result of offering a way to empower poor local communities in asserting their needs to their local and national governments and in influencing budgetary allocations.… As such, CBMS has become much more than a research tool, turning into a direct instrument for empowerment and actual poverty reduction” (Habito et al. 2004).

The lessons learned from these experiences about the need for CBMS, its uses and benefits, and about the ways it can and should best be implemented, are presented in Part 4. These experiences have also shown the transformative power of a community-based process. CBMS is proving a good incubator for change agents. It is also an effective bridge between government and the people it serves.

Document(s) 6 of 10