DISCUSSION PAPER NO. 1614 Urban Social Assistance: Emerging Insights from Three African Countries Vanessa Moreira and Ugo Gentilini December 2016 Urban Social Assistance: Emerging Insights from Three African Countries Vanessa Moreira and Ugo Gentilini 1 December 2016 Abstract: As countries implement social assistance (or safety net) programs, a range of technical hurdles can affect their implementation differently in rural and urban areas. In urban areas, the focus of this study, cost of living can be higher and more prone to economic slowdowns. Poverty can be more severe than in rural areas and accompanied by high malnutrition rates. Implementation challenges in most urban areas relate to the lack of proper identification, outreach, intake, and registration of potential beneficiaries. These are in part due to the lack of social cohesion and different channels of communication. Therefore, social workers are likely to play an even more fundamental role in program implementation and M&E processes. JEL Classification: O1, O2, O18, I38 Keywords: Social pension, non- contributory cash transfer program 1 The paper greatly benefited from advice and support by Ngoc-Dung T. Tran, Phillippe Leite, and Ruslan Yemtsov. Acronyms and Abbreviations CBT: Community-Based Targeting CTC: Community Targeting Committee M&E: Monitoring and Evaluation MIS: Management Information System NGO: Non-Governmental Organizations NSO: National Statistical Office PMT: Proxy Means Test PMU: Project Management Unit SAM: Severe Acute Malnutrition SSN: Social Safety Nets UNICEF: United Nations Children’s Fund BENIN CPS: Centre de Promotion Social INSAE: National Institute of Statistics and Economic Analysis of Benin MDGLAAT: The Ministry of Decentralization, Local Government, and Administration and Development of the Territory PSDCC: Projet de Services Décentralisés Conduits par les Communautés SSDCC: Secretariat for Decentralized Community Driven Services - Benin CONGO, Republic MASAHS: Ministry of Social Affairs of Humanitarian Action and Solidarity CAS: Social Welfare Center CCS: Targeting Community Committee GAC: Governance and Anti-Corruption INS: National Statistics Office i MALI OM: Orange Money BNDA: National Bank for Agricultural Development INSTAT: National Statistical Office SAM index: Severe Acute Malnutrition UTGFS: SSN UGTFS Technical Management Unit ii Table of Contents ABSTRACT:...................................................................................................................................I I. INTRODUCTION .................................................................................................................. 1 II. NATURE OF URBAN POVERTY ............................................................................................ 3 III. PROGRAM OBJECTIVES ...................................................................................................... 7 IV. KEY PROGRAM DESIGN FEATURES ..................................................................................... 7 4.1. PROGRAMS DESIGN SPECIFICITIES ..................................................................................... 8 V. KEY STEPS IN THE DELIVERY PROCESS ............................................................................. 10 5.1 OUTREACH & INTAKE ....................................................................................................... 10 5.2 SELECTION ........................................................................................................................ 13 5.3 TARGETING / REGISTRATION ........................................................................................... 16 5.4 PAYMENT / TRANSACTIONS ............................................................................................. 17 5.5 MONITORING &EVALUATION (M&E) ............................................................................... 19 5.6 GRIEVANCE & REDRESS .................................................................................................... 20 VI. KEY ISSUES EMERGED DURING PROGRAM IMPLEMENTATION IN URBAN AREAS .......... 21 VII. URBAN AREA CHALLENGES .............................................................................................. 23 VIII. EMERGING POLICY & OPERATIONAL OPPORTUNITIES .................................................... 24 IX. REFERENCES ..................................................................................................................... 26 ANNEX 1: BENEFIT LEVELS ...................................................................................................... 28 ANNEX 2: 3 COUNTRY COMPARISON ..................................................................................... 29 Figures Figure 2. 1: GDP, PPP and Urbanization trends ........................................................................ 6 Tables Table 2. 1: Key indicators .......................................................................................................... 6 Table A1. 1: Distribution of benefits in Benin ......................................................................... 28 Table A1. 2: Distribution of benefits in Mali ........................................................................... 28 Table A1. 3: Transfer amount by household composition ..................................................... 28 iii I. Introduction As countries implement social safety nets programs, a range of technical hurdles affects their implementation. Given the fluid expansion and contraction of urban informal settlements over time and their specific characteristics, the development of an effective outreach strategy to inform and attract urban poor for safety nets programs or to effectively communicate with prospective beneficiaries about available programs is challenging. One challenge faced by the poor population in most urban areas is related to the lack of proper identification (ID), therefore preventing access to public services and programs demanding ID for verifying identity. Obtaining IDs can be costly, incurring private and administration costs and, a lot of times, travels back to their villages of origin. Interestingly enough, the three countries in this note have not described problems in this regard. In Benin, the program created its own IDs as the national ID card system has limited coverage. In Congo, newborn children leave the hospital with a birth certificate (provided guardians have paid the fee of FCFA 5,000). Lastly, the outcome of recent elections in Mali may brighten the prospects for ensuring individual IDs. One key process that is somehow weak in the urban areas presented in this study is related to the outreach, intake, and registration of potential beneficiaries. The challenge of communicating with potential beneficiaries and reaching pre-identified families is problematic in urban areas because of the larger catchment areas and the multiple communication channels available, minimizing effectiveness of communication campaigns. The same occurs regarding payments messages. In contrast, in rural areas, the existence of social cohesion together with a single community radio station/ church/ community leaders enables to convey the messages quickly and efficiently. Social workers are a fundamental part of program implementation and monitoring and evaluation (M&E) processes. Due to larger catchment areas and the lack of social cohesion in urban areas with high density of people, families’ identification and monitoring can be 1 challenging and community committees are not very effective without participation of social workers. A strong and transparent system for identification and selection of beneficiaries is key to ascertain credibility to the program. In urban areas, this can only be achieved with strong participation of social workers who know the population of interest. Community committees composed of youth associations, churches, community leaders, etc. tend to identify participant families that are not necessarily the poorest, therefore increasing inclusion errors. The poorest are often marginalized and excluded from such committees. In urban areas, data collection and monitoring takes longer. Enumerators and M&E agents have difficulties returning to pre-selected households due to the bad quality of the address recording (houses disorderly numbered) and the differences in recorded names (household members known by a nickname), causing enumerators to perform multiple household visits in order to correctly identify families. GSM payments are relatively uncomplicated and inexpensive. However, electronic payments transactions that are not fully functional may adversely affect time to access benefits and accountability. Poverty can be more severe in urban areas. On top of the scarcity of basic services as sanitation, the urban poor have difficulties in accessing good schools and health centers because of their families’ income constraints. Urban areas have high malnutrition rates. Individuals are subject to precarious and temporary jobs as a way of generating income for meeting the end needs. Also, differently from rural areas, families cannot produce for subsistence. Therefore, food intake can be compromised. Living in urban areas is expensive and more vulnerable to an economic slowdown. Apart from having to pay for rent and some basic services that are otherwise free in rural areas (for example, health and education), the urban poor are subject to higher commodity prices and 2 face higher unemployment rates. When an economic slowdown happens, social unrest and violence increase mainly in the poorest urban areas. In this study, we will look into urban and rural social safety nets program specificities from three countries: Benin, Republic of Congo, and Mali. What kinds of strategies were used to make the social safety nets programs adapted to the urban context and more effective to fight urban poverty. In the next sections, we will present (i) nature of urban poverty; (ii) programs objectives; (iii) key program design features; (iv) key steps in the delivery process; (v) key issues emerged during program implementation in urban areas; (vi) emerging policy and operational opportunities; and (vii) challenges. II. Nature of Urban Poverty Rural areas in all three countries are largely agriculture dependent and commonly affected by natural shocks such as droughts and floods. Therefore, poor people from rural areas move to urban areas to escape the difficulties. However, instead of finding relief, they are subject to worse conditions. For example, some public services, which are available for free in rural areas, are unavailable or fee-based in urban areas, such as schools and hospitals, creating additional costs for families. Poverty can be more severe in urban than in rural areas. On top of the scarcity of basic services as water, electricity, and sanitation; the poor in urban areas have greater difficulties in accessing good schools and health centers as their families’ income are constrained. The urban poor are subject to higher commodity prices and higher unemployment rates; and in any economy slowdown, social unrest and violence increase mainly in the poorest urban areas. Moreover, in urban areas, individuals are subject to precarious and temporary jobs as a way of generating income for subsistence. Food intake can be compromised. In contrast, in rural areas, most families may be less exposed to economic slowdown, with subsistence farming ensuring some basic access to food. 3 BENIN Benin’s economy relies mostly on the cotton trade, and agriculture is the main source of income for 70 percent of the country’s workforce, making the country very susceptible to natural shocks. In 2010, Benin experienced awful losses due to floods—thousands of people were forced to leave their houses, many schools and public buildings were damaged or destroyed, and most crops were lost, worsening the food insecurity situation. Difficulties in the rural life underline the continuous migration to urban areas, where migrants often settle in highly dense, underdeveloped areas. The urban poor often live in poorly built, rented accommodations without access to water, electricity, and sewage. Urban areas of Benin have numerous pockets of poverty. People living in those areas lack the access to basic services. Food insecurity is high as a consequence of irregular sources of income and the lack of a family production whereas in rural areas, families can produce something to eat. Violence is widespread; women are subject to prostitution and various sorts of sexual abuse. High indices of child labor are observed as children work to help increase the household income. Social unrest is observed as a reflection of diminishing jobs in the formal sector. CONGO Since the oil discovery in the early 1980s, oil revenues and considerable external assistance have motivated the economy in Congo. The formal economy was centralized by the state, and as a result, public spending and relatively high wages in the public sector attracted migration to Brazzaville and Pointe Noire—two important urban areas. Nonetheless, oil wealth was not shared properly and poverty is still persistent with the rate standing at 40.9 percent in 2011. However, almost 50 percent of the poor live in urban areas due to the high urbanization rates of the country. Gini inequality is high at 0.46. The sustained growth over 2000-2014 did not translate much into poverty and inequality reduction despite a slight decline in urban poverty. It should be noted also that urban mobility is high in certain areas (pockets of poverty) due to violence, lack of services, and the high cost of living (rental/ food prices). 4 Formal jobs, which are a leading factor of economic inclusion in Congo, are concentrated in the hands of a few. The Congolese labor market has a very small formal sector, dominated by the public sector and state-owned enterprises, leading the poor to informal, activities that are not sustainable, mainly in urban areas. The urban poverty rate in Congo is rather low. But the high concentration of people living in poor urban areas has negative impacts on their access to the already weak health and education services. For example, reflected by higher rates of children per teacher in public schools compromising education quality. MALI Mali’s political and social situation remained fairly stable until March 2012, when a military coup ousted the previously elected president. The presence of armed separatists and jihadist groups in the north caused massive internal displacements, mostly to urban areas in the south and neighboring countries. However, poverty is still predominantly rural where people depend on subsistence farming and are therefore vulnerable to natural disasters. Conflict in Mali nowadays is exacerbated by climate change and population growth. In urban areas, people’s livelihoods are under stress mainly due to limited opportunities (jobs and services) and persistently high levels of youth unemployment—an easy feeder into the informal economy. Therefore, profound challenges remain due to economic fragility— growing insecurity, rapid demographic growth, and the impact of climate change on agricultural productivity. 5 Figure 2. 1: GDP, PPP and Urbanization Trends Source: WDI, World Bank Table 2. 1: Key Indicators Indicator Country 2010 2011 2012 Benin 2.12 2.96 4.64 GDP growth (annual %) Congo, Rep. 8.75 3.42 3.80 Mali 5.81 2.72 0.03 Benin 732.95 799.04 807.69 GDP per capita (current US$) Congo, Rep. 2,953.19 3,453.22 3,191.16 Mali 610.79 680.83 641.79 Benin 1,710 1,750 1,810 GNI per capita, PPP Congo, Rep. 4,120 4,180 4,470 (current international $) Mali 1,440 1,460 1,430 Poverty headcount ratio at Benin .. 36.20 .. national poverty lines (% of Congo, Rep. .. 46.50 .. population) Mali .. .. .. Benin 9,509,798 9,779,391 10,049,792 Population, total Congo, Rep. 4,066,078 4,177,435 4,286,188 Mali 15,167,286 15,639,115 16,112,333 Benin 16.97 16.37 15.81 Population in the largest city Congo, Rep. 61.23 61.38 61.54 (% of urban population) Mali 35.41 35.42 35.47 Source: WDI, World Bank (last Updated: 05/02/2016) 6 III. Program Objectives Although main objectives are similar when implementing a social safety nets program in the three countries, additional objectives are specific to each country’s needs. While all three countries are interested in establishing an efficient social safety nets system to protect the poor and/or vulnerable while reducing current poverty and smoothing household consumption; in Benin, the program Projet de Services Décentralisés Conduits par les Communautés (PSDCC) aims to also improve households’ ability to cope with shocks; while in Congo, the Lisungi program also aims to promote children’s access to education and health services; and in Mali, through the Jigiséméjiri, the effort is also to increase the food security for households, due to particularly high rates of food insecurity in the country. Programs vary in size and implementation areas. Although Benin’s and Mali’s programs are relatively large in scale (especially when compared to previous programs existing in the countries), only a small part of the benefits is destined to poor households in urban areas (main urban areas of both countries are included in the respective programs; namely, Cotonou, Benin’s capital city; and Bamako, Republic of Congo’s capital city). On the other hand, Lisungi, which is still in its pilot phase, is mostly implemented in urban areas due to the country’s urban population density and the program’s desired conditionalities. To guarantee that households can comply with school attendance and regular health care visits, minimum infrastructure was required for implementation, which proved to be hard in rural areas. IV. Key Program Design Features Programs’ outcomes are directly related to implementation. Hence, design issues become extremely relevant when implementing a determined program in any country where regions/ areas have distinct characteristics. If such characteristics are not considered, the entire project and its relevance may be compromised (for not producing sufficiently good results). In general, community committees are set to facilitate the process but they are not working as expected in urban areas. Due to larger catchment areas and the lack of social cohesion in 7 areas with high density of people, identification of households and individuals in urban areas can be challenging. Therefore, social workers become a fundamental part of the identification process, as they are familiar with the families that ask for regular help. Ideally, every community committees would have local social workers as its members, as their knowledge of the families can help minimize both errors of inclusion and exclusion. But this is not always the case. Community committees are often composed of youth associations, churches, community leaders, etc. This structure works well in rural areas where families are close together, but not so much in urban areas. Identification of poor families in urban areas takes longer. In response to urban areas specificities, social workers revise the preliminary lists formulated by community committees. The limited number of social workers, however, prevents their participation in all urban community committees, with implications for program accountability and possible delays due to possible community claims and concerns of beneficiary selection. In Benin’s PSDCC, about 10 percent of the selected household are in urban areas (1,187 out of the total of 13,000). The program makes use of community targeting committees (CTC) composed of members with good knowledge of the neighborhood and background in social assistance, charity work or identification of those in need. Members are elected democratically in neighborhood assemblies and committees. In rural areas, community targeting is facilitated by social cohesion whereas in urban areas, CTCs coordinate with local service providers (like women/youth associations) to help identify the households. Exclusion errors are likely because the poorest are often marginalized and inclusion errors tend to be higher because of a tendency to select people who are more connected to the local structures. The program administration needs to work closely with social workers to revise initial listing to minimize leakage and under-coverage. Districts and villages were selected based on an index that combines poverty and access to services while poor households are selected through the implementation of a proxy means test (PMT) to identify the poorest households. Selected households receives FCFA 3,500 monthly as an unconditional transfer, and some households with at least one adult, who is capable of working, can be selected for 8 labor intensive public works. The public works option is available for 12 days a month during four months at a daily rate of FCFA 800. Therefore, transfers are either FCFA 42,000 or 80,400 a year (US$ 84 or US$ 140), which represents in per capita terms about 1.5 or 2.4 percent of the per capita GDP. In Congo, six of the seven localities selected for the pilot of the program Lisungi are urban. As the program aims to promote children’s access to education and health, it was a requirement that selected areas had infrastructure facilities with supply of health care and primary education that is adequate for implementation. Moreover, the program is implemented in areas with well-established Social Welfare Centers (CAS). At the CAS, social workers are organized into small catchment areas (social sector) and work with the community committees to identify potential beneficiaries and organize local information campaigns. The CAS are also present in rural areas. Starting in 2011, the government invested in modernization of centers and CAS in urban areas. As of today about eight CAS are in place with the infrastructure needed for implementing the program nationwide. Due to the conditionalities imposed by the program, regions were selected based on the existing health and education infrastructure in order to guarantee the program implementation. Extreme poor households, with at least one child aged 0-14 or one elderly aged 60 or above, are identified through a combination of community-based targeting and PMT. Selected households receive FCFA 10,000 monthly as an unconditional transfer, and each child within the households (for a cap of three per household) is eligible for FCFA 5,000 conditional on visiting health centers and attending school, depending on their age. The elderly, defined as those aged 60 or above, are eligible for an unconditional monthly transfer of FCFA 10,000. Therefore, transfers per household range between FCFA 120,000 and 540,000 a year (US$ 240 - 1,080), which represents in per capita terms 1.5 – 6.7 percent of per capita GDP. In Mali, just one urban community of the capital city Bamako, Bamako III, is eligible for the program. The program is decentralized and operated by different levels with different attributions: National (Project management Unit – PMU), regional (focal points and M&E agents), circles (M&E agents), communities committees, and villages (rural areas) or 9 neighborhood (urban area) committees. Although CTC members are supposed to work as volunteers in both urban and rural areas, they are often demanding some sort of payment as a result of the intense work to identify households. In urban areas, the pressure to remunerate CTCs was more intense, and often CTCs were not supporting program implementation activities such as monitoring payment activities. Therefore, as a correction to the original design, Mali is introducing a unique payment scheme destined to support the work of CTCs. Notice that as one of the objectives of the program is to improve children’s outcomes in human capital (health, nutrition, hygiene, and family savings), a series of information sessions delivered by NGOs are organized, and all households become beneficiaries of the program and are expected to participate in the sessions. Regions were classified according to an index combining poverty 2 and severe malnutrition 3 levels. Within each region, districts were ranked based on a weighted severe acute malnutrition index that combines poverty and infrastructure indicators while communes are selected based on geographical targeting using a combination of poverty maps and infrastructure indicators indices, which identify the potential availability of agencies for implementation of the accompanying measures and payments. All villages within each commune have a quota of eligible households, based on the population size. Selected households receive FCFA 10,000 monthly as an unconditional transfer. The total annual transfer is FCFA 120,000 (US$ 240), which represents in per capita terms about 6 percent of per capita GDP. V. Key Steps in the Delivery Process 5.1 Outreach & Intake In the urban areas of all three countries, enumerators have difficulties in returning to the households identified by the preliminary household lists. There are issues with the household information transmitted to the firms responsible for additional data collection (NSO in Congo and Mali and a private company in Benin) due to errors in address recording (a lot of times, 2 Provided by the National Statistical Office. 3 UNICEF. 10 house numbering when available is done in a disorderly manner) and also because names recorded do not match household members’ names (it is common that people are known by their nicknames). Another problem in urban areas is the fact that household heads and/or reference adults are not found during the time of the survey because they were working or looking for a job. Therefore, enumerators need to visit households multiple times in order to collect the data, increasing the amount of time needed for data collection. To minimize this issue, social workers are once more fundamental for the correct identification of households and subsequent interview scheduling (to minimize the number of visits to each household). In Benin, as part of the outreach, sensitization of national and local authorities is done through the use of training modules explaining the approach and defining concepts of poverty, extreme poverty, and household. The program administrators at the central level organize the training sessions in the morning, followed immediately by the community committees’ field activities to identify potential beneficiaries. The lists of poor households produced by the committees are validated in public during neighborhood assemblies open to the public. Households on the final validated list are subjected to the PMT survey to verify their poverty status. As it is common in urban areas, families in Cotonou are not strongly associated as in smaller towns/ rural areas. Therefore, more time is required to identify poor households by the committees, and social workers from the centers for social promotion (CPS) are needed to revise the initial list to mitigate potential targeting errors, mainly in urban areas. Otherwise, no differences were observed between urban and rural areas. Data collection uses smartphones without the need of connection to the Internet and data transfer to the network happens when Wi-Fi network is available. In Congo, 4 community committees are responsible for identifying poor and vulnerable households, which are further selected through household social survey and data collection. 4 The share of beneficiaries in each zone is defined by quotas given the population and poverty rates in each zone by the PMU that later organizes departmental workshops to promote cash transfers awareness. Those sessions are designed to city councils, sub-prefects and departmental directors. In a later stage, the PMU organizes training sessions to the leaders of Social Welfare District (CAS), CCS chiefs, and basic education inspectors. 11 Identification is made by CAS and targeting community committees (CCS) on the basis of the existent indigence card (carte d’indigence), which contains information of population that often asks assistance at CAS. The outreach process is similar in both urban and rural areas but the challenge of communicating with potential beneficiaries and reaching the families pre- identified is problematic in urban areas because of the multiple communication channels available that minimize effectiveness of communication campaigns, the lack of good address to locate the household in case social workers cannot accompany surveyors, and low social cohesion. In Mali, identification of households is performed by the members of the village committees based on the poverty criteria explained during technical workshops organized by the SSN Technical Management Unit (UTGFS). Identification methods vary according to the size of villages: (i) in the districts of Bamako and in some rural towns, a sort of a census approach was used to review the measureable conditions of dwellings and make sure that all families would be identified. This work is done by the members of the village/ neighborhood committee, where district/ village areas were divided in sectors and assigned to each member/ team of two members; and (ii) in small villages where people are known to each other and have daily interactions, the village committee held meetings for exchange of information among its members, reviewed poorest households and established the preliminary beneficiaries list taking into account the quota allocated to the village/ neighborhood by the UTGFS. In small villages, the work of identification and selection of poor households took between two and three days. In rural towns, it took two to three weeks. In Bamako, efforts were made to achieve results after one month (the deadline was set by the program). It is important to note that field workers signaled this amount of weeks was inadequate for collection of data in large areas where census could not be exhaustive. Despite aiming for having a census approach to map all households in urban areas, the activity did not reach all households in large neighborhoods. Therefore, it is likely that not all poor households have been identified and pre-registered. In order to fix this problem, community committees had to revise the analysis of the preliminary lists to make sure that the selection 12 was not solely done by the neighborhood committees. Communities’ involvement is then required to improve outreach but at an additional costs not considered by the program (communications, travel, and meeting arrangements). 5.2 Selection Geo-targeting and PMT are used for selection. In all three countries, a combination of geographical, poverty, and social characteristic measurement is used to identify beneficiaries. Benin Selection follows a three-level selection process: (i) municipalities, (ii) villages, and (iii) households. Municipality selection: The program covers 12 municipalities with highest extreme poverty levels, 5 about 15 percent of the country's municipalities. Selection was performed during a national workshop and based on data provided by the National Institute of Statistics and Economic Analysis of Benin (INSAE), according to an index on the access to basic infrastructure. Village selection is done in two stages: a) Use of poverty maps (based on census and household survey data) 6 to estimate the distribution of poverty in small geographic areas (such as counties and villages or city neighborhoods). The targeting criteria will also be based on the villages’ share of extreme poverty to the municipalities. Each commune will establish a broad list of eligible villages. 5 Share of the number of people living below the food poverty line in a municipality, relative to the country’s extreme poor population 6 2002 census data and 2011 Household Living Conditions survey (Enquête Modulaire Intégrée sur les Conditions de Vie des ménages - EMICoV). 13 b) From the eligible list of villages, villages/ communities will be selected by the municipal councils. Following the selection of municipalities and poverty map analysis, a total of 120 villages were selected among the established lists of villages. Ultimately, the final number of participating villages varies by municipalities with regards to the resources allocated by the program. Household selection: In October 2013 Benin adopted a unique and rigorous method for the identification of the poor for social protection programs. This approach recommends a pre- identification of the poorest by the communities and further validation through the implementation of PMT to estimate households’ poverty levels. The PMT takes into consideration the accessibility to health, education, and other services and was prepared by the national statistical office using 2011 household survey data. To account for differences between urban and rural areas, different models are in place and the eligibility cut-offs were defined differently, being higher in urban areas (adapted to the purchase power parity for urban areas). The index developed is to be updated every two years. Congo Selection follows a three-level process: (i) regions, (ii) communities, and (iii) households. Regions and communities were selected according to the adequacy of infrastructure facilities to provide access to education and health for children, as those are conditionalities imposed by the program. Furthermore, the program is implemented in areas with well-established Social Welfare Centers (CAS). The program covers three regions and seven communities: Brazzaville (Makélékélé, Bacongo, Moungali, Talangai), Pointe-Noire (Mvoumvou), and Cuvette (Makoua and Oyo). The distribution of the number of beneficiaries between the three regions and the CASs centers on quotas determined by the poverty rate and the population of each region. Household selection: During project preparation in Congo, the National Statistics Office (INS) worked with the PMU to assess the characteristics of poor and vulnerable households in the 14 country based on the data from the Congolese Household Economic Survey (ECOM). Based on this analysis, the INS validated a PMT model that allows the PMU to estimate the household per capita consumption for each household registered in the Unified Registry. The management information system (MIS) uses this linear combination to generate a list of potential beneficiaries per district by comparing the estimated welfare and food poverty line estimated by the proxy, which by the way have different models for urban and rural areas. Mali Selection follows a four-level selection process: (i) regions, (ii) circles and towns, (iii) municipalities, and (iv) households. Regions, circles and towns, and municipalities are classified according to synthetic indexes based on estimated indicators at the regional, circles and town and municipalities level. Three different indexes are used (one for each selection level) and, among other things, they comprise measures of poverty (provided by Mali’s National Statistical Office (INSTAT)), child severe acute malnutrition (SAM) 7 rates and access to services (water, electricity). The index ranges from 0 (rich) to 100 (very poor). All villages from selected municipalities are eligible to receive the program. The distribution of benefits is done at the community level. Each community will get a fixed budget to cover program expenses, based on the percentage of extreme poor provided by INSTAT. Quotas are calculated by the UTGFS team and evenly distributed in all villages, defined as the number of households divided by number of villages Household selection: Community/ village assemblies identify the poorest households at the village level. Each identified household will receive a score to allow household classification from the poorest to the least poor in an economic criterion (transparent and direct). Validation of the beneficiaries list is done during neighborhood/ village general assemblies under the supervision members of the municipal committee and the M&E program officer. 7 UNICEF - normalized between 0 and 100 15 The preliminary list is read publicly and, in some cases, heads of households are introduced. Then the assembly is asked to rule. If there are no objections, the list is approved by acclamation and is a definitive list. If there are objections, complaints are registered and followed by immediate debate regarding eligibility. Households that are considered less poor than others are excluded from the final list and replaced by households considered poor. But public protests during general assemblies are rare mainly because people are reluctant to speak in public on a sensitive issue. 5.3 Targeting/ Registration Generally, in urban areas, registration was facilitated by the fact that people have identification documents. When it happened that registration was lower in urban than in rural areas, it was due to the higher urban mobility. The electronic data collection 8 in the urban areas of Benin increased data quality while preventing data entry mistakes (for example, out-of-range values) and facilitated the process of identifying beneficiary though pictures of household heads. Pictures were taken by the same device used for data collection and were used to issue beneficiary ID cards. The country has an ID card system but coverage is limited. Therefore, the program has developed an alternate ID in coordination with the National Statistics and the National Health Insurance Program. Once the household selections trough the PMT is finalized, social workers deliver ID cards and compare with documents to guarantee delivery to the correct person for final registration in the program. Each Community Committee in Congo organizes an information campaign amongst pre- identified households. In Brazzaville, this process was more intense due to the lack of social cohesion, which affects outreach. In rural areas, local radios, community leaders, social assistance centers reach a larger number of people at once. Records team installed at 8 GPS location of the household, roster, assets and standard living measures used by the PMT. 16 registration centers request households to register the vouchers which were handed in during intake phase. At this stage, households have to present identity documents, birth certificates, or any other documents required by the census strategy. Then, enumerators collect the necessary additional information, take photos, and collect fingerprints of at least two members in the household. Field data is transferred to local and central servers by connecting laptops, used during registration, to local servers in each CAS. The processes are identical in both urban and rural areas but take more time in urban areas as a result of the lack of strong connections within families from the same community. INSTAT performed additional data collection on selected households if needed and subsequent program registration. INSTAT was responsible for instructing enumerators, training and fieldwork. During this step, household heads and a second adult member were photographed for purposes of identification and for provision of beneficiaries’ card to conclude the registration process. In Bamako, the percentage of households in the final beneficiaries list that were found to receive the registration card was slightly below than in other regions (97.7 percent against 99.8 percent and 99.3 percent in other regions). Replacement of households then were done with the help of village/ neighborhood committees. Household heads who were not found or absent were replaced by other members of the household. 5.4 Payment/ Transactions Communication with families regarding payments in urban area is disadvantaged by the absence of an efficient communication method, while in rural areas, radio/ church/ community leaders convey the message quickly and efficiently. Therefore, social workers and community committees need to spend some days in the field, before payment, to spread payment dates. Benin and Mali opted to deliver payments in urban areas through the use of SIM cards. But in Mali, the process is not entirely automated so beneficiaries have to go, in person, to respective payment centers, which incurs some personal cost. As a developed structure for 17 mobile payments is not yet available in Mali, payments are entirely manual and can take up to three months. In the near future, with transactions changing to electronic, beneficiaries will not need to show up at payment centers in order to receive cash transfer benefits, and payments can be paid monthly as in Benin. In contrast, Congo uses the same payment agency in both urban and rural areas, the Banque Postale du Congo, but beneficiaries still have to go every three months to an agency instead of cashing the transfer monthly using ATM cards. Benin uses distinct methods for payments in urban and rural areas. In Cotonou, cash benefits are transferred through SIM card electronic payment from MTN Bénin Telecommunications Service Provider. SIM cards are provided by the GSM operator while beneficiaries’ ID cards by program administrators. The GSM operator has a copy of the beneficiary database allowing for the monthly payment of all beneficiaries. Payments are made at any MTN operator in the region, and beneficiaries knows that between the 24th and 31st of the month they can demand payment at any MTN operator agency. As the transaction is electronic, accountability is greater—within seven days program administrators have transaction information ready to be archived. In rural areas, payment is through branches of the Caisse Locale Credit Mutuel Agricol (CLCAM). Transactions are manual and can take up to three months for being processed. In all areas of Benin, heads of social promotion centers have a copy of the beneficiaries’ database in their catchment areas and handle any complaints about missed payments. Congo pays all beneficiaries through the same bank system. One to two weeks before a payment date, the PMU provides interested stakeholders (internal auditor, the World Bank, payment agencies, etc.) and the payment agency, Banque Postal do Congo (BPC), with the nominal lists of beneficiaries that include the transfer amount to each household. Beneficiary households are informed by social workers/ community committees to be physically present at the associated BPC agency. Since BPC has created a virtual bank account, beneficiaries can cash entire or partial benefits by keeping their virtual account as a savings account. In urban areas there are more branches available therefore facilitating payment. In rural areas, payment centers by BPC are created to reduce the time and distance to payment points for 18 beneficiaries. Once more, in rural areas payments are manual, taking more time than in urban areas. Mali also uses distinct methods for payments in different areas. Following a study that was conducted to identify all agencies with capability to implement money transfers in all regions targeted by the program (banks and microfinance institutions that include mobile operators and post offices), operators were pre-selected based on three criteria: (i) national coverage; (ii) capacity to do large-scale operations; and (iii) experience in cash transfers. Selected agencies differed in terms of their areas of operation. In the urban area (Bamako) the Orange Money mobile operator was selected whereas in all other areas, the tender winner was the National Bank for Agricultural Development (BNDA). In urban areas, Orange money created an account to all beneficiaries and provided free SIM cards. Even though SIM cards are used, payments are not electronically made yet. Consequently beneficiaries (or their substitutes) have to present themselves at payment centers every three months during payment time, which is informed by community committees, to withdraw the benefits upon presentation of their beneficiary card and ID card. The main difference between urban and rural areas is that BNDA, the payment agency in rural areas, has more experience with payment transactions since it is a bank. As a result, payment cards are developed by BNDA and include the recipient's name, ID code, date of birth, and a column for observations, and payments are made upon presentation of the beneficiary card and national ID card (of the beneficiary or an authorized replacement). Given that BNDA has the expertise in providing payments, accountability is done on a timely manner; whereas in the urban areas, OM is inexperienced in such transactions, resulting in increased time needed for accountability to the PMU. 5.5 Monitoring &Evaluation (M&E) M&E allows government and all partners to regularly monitor the project progress and to estimate the program impact. To ensure the program is operational and implemented in compliance with its objectives, M&E is adopted in various ways. In Benin and Congo, M&E agents visit households regularly while in Mali, agents work closely with community 19 committees to make sure program procedures are being followed, evaluate problems, and solve issues when they appear. M&E agents in Benin visit households regularly in both urban and rural areas but the dispersion in urbans areas makes the process longer. GPS data collected during identification phase aides in the process since a lot of households do not have addresses or are not well documented. In Congo, M&E agents and social workers are in charge of following M&E activities and visiting households. The coordination of health and education agents with social assistance is difficult due to the limited number of agents in each circumscription and the high number of people to be monitored. As a result, a partnership with UNICEF is being formulated to support M&E activities. This issue is more relevant in urban areas due to the large catchment areas. M&E agents in Mali work closely with the municipal and village committees (rural areas) or neighborhood committees (urban areas), the social development service, and the payment agency. M&E agents help the committees understand program procedures through regular monitoring work, evaluating problems, and solving the issues. They also help during village/ neighborhood sensitization sessions to inform beneficiaries about payment dates and locations and are there in person to monitor payment transactions, track problems if any, and above all to make sure beneficiaries have all documents needed to receive the payment. Moreover, in urban areas, monitoring is easily assisted by M&E agents and benefits from a greater presence of the civil societies and partner NGOs physically located in the urban areas. In rural areas, M&E agents are mostly alone to monitor the activities. 5.6 Grievance & Redress The three countries have distinctive approaches to address grievance and redress. While in Congo information is collected electronically; in Benin and Mali, social agents play an important part in the collection and transmission of the information to program administrators. 20 In Benin, CPS have computers that are not connected to the network, but have info of all beneficiaries. Social workers are present on-site to support and/or address complaints. Following the personalized service provided by social workers, demands and complaints are delivered to program administrators via e-mail or telephone. In rural areas, as CPSs are not present to centralize the collection of complaints, field agents collect complaints and demands and coordinate with program supervisors. Therefore, grievance and redress is less pronounced in rural areas. In Congo, grievance and redress is done through social workers and M&E agents located in each CAS. Information is digitalized in the system and processed automatically. There is no difference in the processes for urban and rural areas. In Mali, grievance and redress is not a well-established service. But the fact that NGOs are more present in urban areas facilitates collection of demands and complaints and also transmission of information to program administrators. Anyone can communicate with M&E agents, community committees, and community leaders. VI. Key Issues Emerged during Program Implementation in Urban Areas Program implementation needs to be adapted to the difficult and challenging environment in urban areas. When program administrators organize general assemblies to present the program or to discuss a new feature or even to talk about results and monitoring activities, participation of authorities and the public is modest when compared to similar assemblies organized in rural areas. Possible explanations are the larger catchment area and weaker social cohesion, meaning program administrators need to put more efforts and resources— both human and financial—to reach poor households for program participation. Inability to find and monitor families. Common characteristics of urban poverty are the lack of a proper household address and higher mobility of people. Given the uncoordinated growth in certain areas, households are often referred to/ identified by certain characteristics or “landmarks” instead of an address. Therefore, program workers and statistical office staff 21 without good knowledge of the area may not be able to find the families. To escape or mitigate poverty, families are more likely to move (sometimes to areas not covered by the program) but they don’t have the habit to inform social workers or program administrators about their plans. When this happens before registration to the program is finalized, administrators have difficulties in monitoring the families and offering the desired services. Lack of interaction between community committees and social workers. When defining the plans for identifying program beneficiaries or potential beneficiaries, program administrators make active use of community committees. However, in urban areas, if social workers are not part of the identification/ selection process, the outcomes of such committees could be biased, excluding potential beneficiaries from the program as the poorest are often marginalized within the communities. Due to the nature of their work, social workers have better knowledge of the population that often requires assistance. Therefore, enforcing participation of social workers in the committees is essential but costs are high. In the three countries presented in this report, availability of social workers to join the committees was limited. Communication is harder in urban areas. In rural areas, catchment areas are smaller and often communication channels are limited to one radio station or one meeting place (a religious site like a church or a mosque). In contrast, the number of communication channels available in urban areas requires a better outreach strategy in order to efficiently spread program messages like payment dates or information on conditionalities. Three other related issues are (i) Unawareness of program administrators regarding alternates to receive payments in the absence of the main beneficiaries. The weak interaction between beneficiaries, who do not inform who the alternates are, and administrators prevent the alternates to receive transfers; (ii) Loss of SIM cards and/or program ID cards, without which payments cannot be made. Beneficiaries do not inform community committees or social workers about 22 their lost documents. Therefore, on payment day, if either a social worker or a member of community committee member is not present at the payment location to provide a replacement card, beneficiaries cannot receive the benefits; and (iii) Grievance mechanism are still under development; therefore, complaints can be left unanswered by program administrators. At times, social workers and community committees try to solve the issues by themselves, without communicating to program administrators. VII. Urban Area Challenges Strengthen the ID system for security and to increase the program’s credibility. Urban areas would benefit from better access to ID systems. Challenges arise from the limited coverage of national ID systems already in place. Countries in this study worked around this issue by establishing a program ID system (like in Benin) or by making investments prior to the implementation of the programs (like in Congo and Mali). Falling motivation of local actors for outreach and follow-up. Volunteer work to help program implementation in urban areas is unappealing because of the larger catchment areas and population dispersion. Therefore, social workers engage in supporting program implementation besides their other duties. As the task proved to be very labor-intensive, their motivation has decreased. One way to increase the willingness to help could be the establishment of TORs, describing specific activities to be carried out in exchange for some sort of payment. Limited knowledge of the household poverty state. Weaker social cohesion in urban areas implies that families known by committees and social workers are those who more frequently make demands and they might not be the poorest families. Poverty indicators as poverty headcount and poverty gap are not often available in small geographic areas as communities and most of poverty indicators used by communities can be very subjective and different because of different perceptions of poverty between community committees in urban areas. 23 When households have similar characteristics, it is difficult for the committees to determine if one household is poorer than the other. Data collection takes longer. Household intake and registration is an issue stemming from: (i) larger catchment areas; (ii) the lack of trust on data collecting agencies/agents, (iii) the bad quality of addresses recorded, and (iv) the use of nicknames to identify household members. Monitoring is hindered by higher mobility. To escape poverty, violence, lack of services, and the high cost of leaving, families in urban areas move around more frequently, sometimes before registration process is finished or even to an area without the program coverage. Higher operation costs as enumerators perform multiple visits to households in order to correctly identify families and to follow-up. In Benin, although GPS data is collected during the identification phase, the information is not shared with social workers for further household visits. Sharing GPS data with social workers would improve effectiveness of registration and M&E activities. Manual payments’ accountability takes too long. It can take up to three months for program administrators to process payments, leading beneficiaries to frequently visit the centers in order to receive their cash. VIII. Emerging Issues By looking at the experiences in Benin, Republic of Congo, and Mali, this study has identified five emerging policy and operational issues. 1. Household identification (location) can be improved through the use of GPS data. GPS location data is available in cell phones, while tablets are often used by urban program administrators as well. Recorded coordinates could be used for improving identification and targeting accuracy, diminishing the problem of lack of a proper street address often observed in urban areas. It can also be used to facilitate M&E activities, including 24 improved measures against fraud (as a way to ascertain the enumerator has visited the household). 2. A greater number of social workers is needed as well as a work program with solid terms of reference and remuneration to motivate neighborhood and community committees. It is also important to create/ strengthen partnerships with other agencies so as to train social workers and a few community members in order to improve the quality of identification and M&E capacity. 3. Creation of communication strategies adapted to urban areas. Improving communication between program administrators and beneficiaries is emerging as a core constraint in cities. More flexible arrangements could be ironed out, such as allowing alternate household members designated by beneficiaries to attend information and communication meetings, like in Mexico. 4. Implementation of automated/ electronic systems of payment will reduce time, enhance accountability, and possibly provide more regular payments. Given the greater development of financial institutions in urban areas, programs should make use of the existing sizeable presence of payment points to establish monthly electronic payment and open bank accounts to beneficiaries in order to increase financial literacy. When well- designed, an automated payment system could help minimize delays, processing time and the risk of fraud. 5. Adjustments in accordance with underlying differences and provision of multiple programs (bundles) to better protect and serve the population. Programs’ design should take in consideration the different costs of living in urban and rural areas and also differences regarding household composition. Urban poverty is more severe not only in terms of monetary poverty but also other vulnerabilities (such as violence, weak health and education services, and limited job offers) that should all be taken into consideration in order to better achieve programs’ objectives. 25 IX. References Gentilini, U. 2015. Entering the city: emerging evidence and practices with safety nets in urban areas. Social protection discussion paper n.1504. Washington, DC: World Bank. http://pubdocs.worldbank.org/en/5841448382581833/Entering-the-City- Emerging-Evidence-and-Practices-with-Safety-Nets-in-Urban-Areas.pdf World Bank. 2014a. Evaluation de la pauvreté au Benin. Washington, DC ; World Bank Group. http://documents.worldbank.org/curated/en/2014/03/23797599/assessment -poverty-benin-evaluation-de-la-pauvrete-au-benin World Bank. 2014b. Benin - Decentralized Community Driven Services Project: additional financing. Washington, DC; World Bank Group. http://documents.worldbank.org/curated/en/2014/01/18925310/benin- decentralized-community-driven-services-project-additional-financing World Bank. 2014c. Benin - Executive summary: poverty assessment. Washington, DC: World Bank Group. http://documents.worldbank.org/curated/en/2014/04/23791532/benin- executive-summary-poverty-assessment World Bank. 2013a. Congo, Republic of - LISUNGI Safety Nets Project. Washington, DC; World Bank Group. http://documents.worldbank.org/curated/en/2013/12/18794003/congo- republic-lisungi-safety-nets-project World Bank. 2013b. PAD - Mali Emergency Safety Nets project. Washington DC: World Bank. http://wbdocs.worldbank.org/wbdocs/viewer/docViewer/indexEx.jsp?objectId =090224b081e5883c&respositoryId=WBDocs&standalone=false World Bank. 2011. Mali - Filets sociaux. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/2011/01/14231097/mali-social- safety-nets-mali-filets-sociaux Programs’ websites: 26 Benin’s PSDCC: http://www.psdcc.org/ Rep. of Congo’s Lisungi: http://www.masahs-gouv.net/lisungi.php Mali’s Jigiséméjiri: http://jigisemejiri.org/index.php/en/ 27 Annex 1: Benefit Levels Table A1. 1: Distribution of Benefits in Benin Adults Base Amount / household/ month Amount / household/ year transfer Amount (FCFA) (FCFA) Number (FCFA) … 3,500 3,500 42,000 FCFA 9,600, for 4 3,500, for 8 months 28,000 3,500 1 80,400 months 13,100, for 4 months 52,400 Table A1. 2: Distribution of Benefits in Mali Group 1 Amount / month / household Group 2 Core villages (FCFA) Interventions: Interventions: Yes CT CT 10,000 62,000 hhs AM Cash transfer AM NP beneficiaries Interventions: Interventions: No AM … AM NP Note1: AM, accompanying measures are not targeted; all households in the village are expected to participate in the sessions covering approximately 283,000 households. Note2: NP, nutritional package, is expected to cover 140,000 children in total from CT and non-CT beneficiaries Table A1. 3: Transfer Amount by Household Composition Children and/or pregnant woman Elderly Transfer amount / household Base / month transfer Amount Amount Number Number (FCFA) (FCFA) (FCFA) 1 5,000 15,000 2 10,000 20,000 3 or + 15,000 25,000 1 10,000 20,000 2 20,000 30,000 3 or + 30,000 40,000 1 5,000 1 10,000 25,000 FCFA 1 5,000 2 20,000 35,000 10,000 1 5,000 3 or + 30,000 45,000 2 10,000 1 10,000 30,000 2 10,000 2 20,000 40,000 2 10,000 3 or + 30,000 45,000 3 or + 15,000 1 10,000 35,000 3 or + 15,000 2 20,000 45,000 3 or + 15,000 3 or + 30,000 45,000 28 Annex 2: Three Countries Comparison Benin Congo, Republic Mali Program PSDCC Lisunji Jegiséméjiri Duration 24 months 2014-2018 2014-2017 Poverty Poor: 36.2% of population Poor: 46.5% of population Poor: 43.6% of population (2009) Assessment Extreme poor: 15% of population Extreme poor: 25% of population 25% of population suffers from chronic food insecurity Objectives 1. Establish an efficient SSN system to 1. Establish an efficient SSN system to V.I Establish an efficient SSN system to protect the poor and vulnerable while protect the poor and vulnerable while protect the poor and vulnerable while reducing poverty and smoothing reducing poverty and smoothing reducing poverty and smoothing consumption. consumption consumption 2. Improve household’s ability to cope with 2. Promote access of children to education V.II Protect food insecure households shocks and health and facilitate access of vulnerable population to productive inclusion measures Size 1,187 HHs in the main urban area: Cotonou 6,000 HHs in periurbans areas of Pointe- 916 HHs in the urban area of Bamako (13,000 total) Noire, Brazzaville and Cuvette) (62,000 HHs total) Nature of Urban Lack of economic diversification (heavily Lack of economic diversification (heavily Occupation on the North by separatists and Poverty dependent on the cotton trade, and dependent on oil) stimulating jihadists causing massive internal agriculture is the main source of income displacements to urban areas in search of displacement, mainly to urban areas. More for 70 percent of the country’s). jobs recently, climate change and population growth. Program 1. Basic unconditional cash transfer to all 1. Establishment of admin pillars for a 1. Cash transfers to chronically poor and Components targeted HHs national SN program with laying the vulnerable households 2. Labor intensive public works for the foundations of a unified registry, MIS, and 2. Accompanying measure to improve same HHs to provide extra transfer during strengthening local capacities; children’s outcomes in human capital agriculture lean season 2. Implementation of a cash transfer (health, nutrition, hygiene and family program to improve access to health and savings). education services 3. Project management and M&E 29 Benin Congo, Republic Mali Key program design features Eligibility criteria Households are eligible if deemed extreme HH are eligible if extreme poor, with at Households are eligible if deemed extreme poor by the community committee and least on child 0-14y.o. or an elderly aged 60 poor by the community committee below defined PMT threshold or more and below defined PMT threshold Conditionalities None Health and education for HHs with children None, but all households from selected (regular visits to health centers and 80% villages are expected to participate in the minimum attendance to school) accompaniment measures sessions Institutional The Ministry of Decentralization, Local The program has its own Project The program has its own Project arrangements Government, and Administration and Management Unit (PMU), the Lisungi team, Management Unit (PMU), the UTGFS team, Development of the Territory (MDGLAAT) under the Ministry of Social Affairs of under the Ministry of Economy and will be responsible for the overall Humanitarian Action and Solidarity finance. supervision and coordination of the (MASAHS). The UTGFS team manages all project Project. The Lisungi team manages all project activities, being responsible for all A new Secretariat for Decentralized activities, being responsible for all disbursements and procurement Community Driven Services (SSDCC) disbursements and procurement procedures supported by the project. created to take responsibility for procedures supported by the project The project coordinator oversees all implementing Components Two and Three The project coordinator oversees all activities related to the implementation of and for the fiduciary oversight of the activities related to the implementation of the project. Project. the project. Response to PMT threshold is higher, adapted to Targeted communities were selected Targeted community of Bamako was urban purchase power parity for urban areas according to health and education care selected according to high density that specificities offered (to comply with program leads to significant number of poor despite implementation needs) and infrastructure low poverty rates. level (to implement accompanying Social workers revise the preliminary lists measures). formulated by community committees. Social workers revise the preliminary lists formulated by community committees Urban poverty line used. 30 Benin Congo, Republic Mali Beneficiary The number of beneficiary households in Quotas for each social welfare sector, The share of beneficiaries in each zone is quotas each community follow a simple system of proportionally to poverty estimates at the defined in quotas given the population and fixed quota per community defined by the department level poverty rates in each zone. Project preparation team. The number of households in each community that will receive the basic cash transfer is estimated to be 100. Outreach & Sensitization of national and local Community committees responsible for Identification of households in the Intake authorities is done through the use of identification of poor and vulnerable preliminary beneficiaries list was training modules explaining the approach households which will be selected through performed by the members of the village and defining concepts of poverty, extreme household social survey and data committees based on the poverty criteria poverty, and household. collection. explained during technical workshop training organized by the SSN Technical Management Unit (UTGFS). Selection Household selection is done through the The NSO has validated a PMT model that Validation of beneficiaries list is always implementation of a proxy means test allows the PMU to estimate households’ done in neighborhood / village general (PMT) which takes in consideration access per capita consumption for each household assemblies under the supervision members to health, education and other services. To registered in the Unified Registry. The of the municipal committee and the M&E account for differences between urban and management information system (MIS) program officer. rural areas, different models are in place uses this linear combination to generate a and the eligibility cut-offs were defined list of potential beneficiaries per district by differently, being higher in urban areas. comparing the estimated welfare and food The index developed is to be updated poverty line estimated by the proxy. every two years. 31 Benin Congo, Republic Mali Targeting / Electronic data collection in the urban The processes are identical in both urban Mali’s National Statistical Office (INSTAT) Registration areas increasing data quality and and rural areas but take more time in performed additional data collection on preventing data entry mistakes Pictures urban areas as a result of the lack of strong selected households if needed and were taken by the same device used for connections within families from the same subsequent program registration. data collection and were used to issue community. Each Community Committee Household heads and a second adult beneficiary ID cards. Program has in Congo organizes an information member were photographed for purposes developed an alternate ID in coordination campaign amongst pre-identified of identification and for provision of with the National Statistics and the households. Records team, installed at beneficiaries’ card to conclude the National Health Insurance program. Once registration centers request households to registration process. In Bamako, the the household selections trough the PMT is register the vouchers which were handed percentage of households in the final finalized, the social worked delivers ID in during intake phase. Enumerators collect beneficiaries list that were found to cards and compare with documents to the necessary additional information, take receive the registration card was slightly guarantee delivery to the correct person photos and collect fingerprints of at least below than in other regions. Replacement for final registration in the program. two members in the household. of households were then done with the help of village / neighborhood committees. Household heads that were not found or absent were replaced by other members of household. Payment / In Cotonou, cash benefits are transferred All beneficiaries are paid through the same In the urban area (Bamako) the Orange Transactions through SIM card electronic payment from bank system. Beneficiary households are Money mobile operator was selected MTN Bénin Telecommunications Service informed by social workers, community whereas in all other areas, the tender Provider. SIM cards are provided by the committees to be physically present in the winner was the National Bank for GSM operator while beneficiaries’ ID cards associated BPC agency. In urban areas Agricultural Development (BNDA). by program administrators. As the there are more branches available Given that BNDA has the expertise in transaction is electronic, accountability is therefore facilitating payment. In rural providing payments, accountability has fast, within 7 days program administrators areas, payment centers by BPC are created been on time whereas in the urban area, do have the accountability ready to be to reduce the time and distance to OM is inexperienced in such transactions archived. In rural areas, payment is payment points for beneficiaries. Once increasing the time needed for through branches of the Caisse Locale more, in rural areas accountability are accountability to the PMU. Credit Mutuel Agricol (CLCAM). manual, taking more time than in urban Transaction are manual therefore areas. accountability can last up to three months. 32 Benin Congo, Republic Mali M&E M&E agents visit households regularly in M&E agents and social workers are in M&E agents in Mali work closely with the both urban and rural areas but the charge of following M&E activities and municipal and village (rural areas) or dispersion in urbans areas makes the visiting households. The cooperation of neighborhood (urban areas) committees, process longer. GPS data collected during health and education agents with social the social development service and the identification phase aides in the process assistance is difficult due to the limited payment agency. The M&E agent helps since a lot of households do not have number of agents in each circumscription committees to understand program addresses or are not well documented. and the high number of people to be procedures through regular monitoring monitored. As a result a partnership with work, evaluating problems and solving the UNICEF is being formulated to support issues. The M&E agents also help during M&E activities. This issue is more relevant village/neighborhood sensitization sessions in urban areas due to the large catchment to inform beneficiaries about payment areas. dates and place and are present during payment to monitor payment transactions, track problems if and when they appear and above all to make sure beneficiaries have all documents needed for payment. Grievance & CPS have computers that are not Done through social worker and M&E Not a well stablished services. But the fact Redress connected to the network but have info of agents located in each CAS. Information is that NGOs are more present in urban all beneficiaries but social workers are digitalized in the system and processed areas, facilitates collection of demands and present to receive those who come to the automatically. There is no difference complaints and also transmission of CPS for demands and/or complaints. regarding the processes for urban and rural information to program administrators. Following the personalized service areas. Anyone can communicate with M&E provided by social workers, demands and agents, community committees and complaints are delivered to program community leaders. administrators via e-mail or telephone. In rural areas, as CPS are not present to centralize the collection of complaints, field agents collect complaints and demands and coordinate with program supervisors. Therefore, grievance and redress I less pronounced in rural areas. 33 Social Protection & Labor Discussion Paper Series Titles 2014-2016 No. Title 1614 Urban Social Assistance: Emerging Insights from Three African Countries by Vanessa Moreira and Ugo Gentilini, December 2016 1613 Issues for Civil Service Pension Reform in Sub-Saharan Africa by Anita M. Schwarz and Miglena Abels, November 2016 1612 How to Target Households in Adaptive Social Protection Systems? Relative Efficiency of Proxy Means Test and Household Economy Analysis in Niger by Pascale Schnitzer, October 2016 1611 Pensions for Public-Sector Employees: Lessons from OECD Countries’ Experience by Edward Whitehouse, October 2016 1610 Pension Systems in Sub-Saharan Africa: Brief Review of Design Parameters and Key Performance Indicators by Miglena Abels and Melis U. Guven, October 2016 1609 Household Enterprises in Fragile and Conflict-Affected States: Results from a Qualitative Toolkit Piloted in Liberia, Volume 2 – Annexes by Emily Weedon and Gwendolyn Heaner, August 2016 1608 Household Enterprises in Fragile and Conflict-Affected States: Results from a Qualitative Toolkit Piloted in Liberia, Volume 1 – Report by Emily Weedon and Gwendolyn Heaner, August 2016 1607 Benefits and Costs of Social Pensions in Sub-Saharan Africa by Melis U. Guven and Phillippe G. Leite, June 2016 1606 Assessing Benefit Portability for International Migrant Workers: A Review of the Germany- Turkey Bilateral Social Security Agreement by Robert Holzmann, Michael Fuchs, Seçil Paçacı Elitok and Pamela Dale, May 2016 1605 Do Bilateral Social Security Agreements Deliver on the Portability of Pensions and Health Care Benefits? A Summary Policy Paper on Four Migration Corridors Between EU and Non-EU Member States by Robert Holzmann, May 2016 1604 Assessing Benefit Portability for International Migrant Workers: A Review of the France-Morocco Bilateral Social Security Agreement by Robert Holzmann, Florence Legro and Pamela Dale, May 2016 1603 Assessing Benefit Portability for International Migrant Workers: A Review of the Belgium-Morocco Bilateral Social Security Agreement by Robert Holzmann, Jacques Wels and Pamela Dale, May 2016 1602 Assessing Benefit Portability for International Migrant Workers: A Review of the Austria-Turkey Bilateral Social Security Agreement by Robert Holzmann, Michael Fuchs, Seçil Paçaci Elitok and Pamela Dale, May 2016 1601 The Greek Pension Reform Strategy 2010-2016 by Georgios Symeonidis, July 2016 1507 Integrating Disaster Response and Climate Resilience in Social Protection Programs in the Pacific Island Countries by Cecilia Costella and Oleksiy Ivaschenko, September 2015 1506 Effectiveness of Targeting Mechanisms Utilized in Social Protection Programs in Bolivia by Ignacio Apella and Gastón Blanco, September 2015 1505 Kyrgyz Republic: Social Sectors at a Glance by João Pedro Azevedo, Paula Calvo, Minh Nguyen and Josefina Posadas, August 2015 1504 Entering the City: Emerging Evidence and Practices with Safety Nets in Urban Areas by Ugo Gentilini, July 2015 1503 Pension Patterns in Sub-Saharan Africa by Mark Dorfman, July 2015 1502 Social Protection in Fragile and Conflict-Affected Countries: Trends and Challenges by Mirey Ovadiya, Adea Kryeziu, Syeda Masood and Eric Zapatero, April 2015 1501 Defining, Measuring, and Benchmarking Administrative Expenditures of Mandatory Social Security Programs by Oleksiy Sluchynsky, February 2015 1425 Old-Age Financial Protection in Malaysia: Challenges and Options by Robert Holzmann, November 2014 1424 Profiling the Unemployed: A Review of OECD Experiences and Implications for Emerging Economies by Artan Loxha and Matteo Morgandi, August 2014 1423 Any Guarantees? An Analysis of China’s Rural Minimum Living Standard Guarantee Program by Jennifer Golan, Terry Sicular and Nithin Umapathi, August 2014 1422 World Bank Support for Social Safety Nets 2007-2013: A Review of Financing, Knowledge Services and Results by Colin Andrews, Adea Kryeziu and Dahye Seo, June 2014 1421 STEP Skills Measurement Surveys: Innovative Tools for Assessing Skills by Gaëlle Pierre, Maria Laura Sanchez Puerta, Alexandria Valerio and Tania Rajadel, July 2014 1420 Our Daily Bread: What is the Evidence on Comparing Cash versus Food Transfers? by Ugo Gentilini, July 2014 1419 Rwanda: Social Safety Net Assessment by Alex Kamurase, Emily Wylde, Stephen Hitimana and Anka Kitunzi, July 2012 1418 Niger: Food Security and Safety Nets by Jenny C. Aker, Carlo del Ninno, Paul A. Dorosh, Menno Mulder-Sibanda and Setareh Razmara, February 2009 1417 Benin: Les Filets Sociaux au Bénin Outil de Réduction de la Pauvreté par Andrea Borgarello et Damien Mededji, Mai 2011 1416 Madagascar Three Years into the Crisis: An Assessment of Vulnerability and Social Policies and Prospects for the Future by Philippe Auffret, May 2012 1415 Sudan Social Safety Net Assessment by Annika Kjellgren, Christina Jones-Pauly, Hadyiat El-Tayeb Alyn, Endashaw Tadesse and Andrea Vermehren, May 2014 1414 Tanzania Poverty, Growth, and Public Transfers: Options for a National Productive Safety Net Program by W. James Smith, September 2011 1413 Zambia: Using Social Safety Nets to Accelerate Poverty Reduction and Share Prosperity by Cornelia Tesliuc, W. James Smith and Musonda Rosemary Sunkutu, March 2013 1412 Mali Social Safety Nets by Cécile Cherrier, Carlo del Ninno and Setareh Razmara, January 2011 1411 Swaziland: Using Public Transfers to Reduce Extreme Poverty by Lorraine Blank, Emma Mistiaen and Jeanine Braithwaite, November 2012 1410 Togo: Towards a National Social Protection Policy and Strategy by Julie van Domelen, June 2012 1409 Lesotho: A Safety Net to End Extreme Poverty by W. James Smith, Emma Mistiaen, Melis Guven and Morabo Morojele, June 2013 1408 Mozambique Social Protection Assessment: Review of Social Assistance Programs and Social Protection Expenditures by Jose Silveiro Marques, October 2012 1407 Liberia: A Diagnostic of Social Protection by Andrea Borgarello, Laura Figazzolo and Emily Weedon, December 2011 1406 Sierra Leone Social Protection Assessment by José Silvério Marques, John Van Dyck, Suleiman Namara, Rita Costa and Sybil Bailor, June 2013 1405 Botswana Social Protection by Cornelia Tesliuc, José Silvério Marques, Lillian Mookodi, Jeanine Braithwaite, Siddarth Sharma and Dolly Ntseane, December 2013 1404 Cameroon Social Safety Nets by Carlo del Ninno and Kaleb Tamiru, June 2012 1403 Burkina Faso Social Safety Nets by Cécile Cherrier, Carlo del Ninno and Setareh Razmara, January 2011 1402 Social Insurance Reform in Jordan: Awareness and Perceptions of Employment Opportunities for Women by Stefanie Brodmann, Irene Jillson and Nahla Hassan, June 2014 1401 Social Assistance and Labor Market Programs in Latin America: Methodology and Key Findings from the Social Protection Database by Paula Cerutti, Anna Fruttero, Margaret Grosh, Silvana Kostenbaum, Maria Laura Oliveri, Claudia Rodriguez-Alas, Victoria Strokova, June 2014 To view Social Protection & Labor Discussion papers published prior to 2013, please visit www.worldbank.org/spl. Abstract As countries implement social assistance (or safety net) programs, a range of technical hurdles can affect their implementation differently in rural and urban areas. In urban areas, the focus of this study, cost of living can be higher and more prone to economic slowdowns. Poverty can be more severe than in rural areas and accompanied by high malnutrition rates. Implementation challenges in most urban areas relate to the lack of proper identification, outreach, intake, and registration of potential beneficiaries. These are in part due to the lack of social cohesion and different channels of communication. Therefore, social workers are likely to play an even more fundamental role in program implementation and M&E processes. About this series Social Protection & Labor Discussion Papers are published to communicate the results of The World Bank’s work to the development community with the least possible delay. This paper therefore has not been prepared in accordance with the procedures appropriate for formally edited texts. The findings, interpretations, and conclusions expressed herein are those of the author(s), and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. For more information, please contact the Social Protection Advisory Service, The World Bank, 1818 H Street, N.W., Room G7-803, Washington, DC 20433 USA. Telephone: (202) 458-5267, Fax: (202) 614-0471, E-mail: socialprotection@worldbank.org or visit us on-line at www.worldbank.org/spl. © 2016 International Bank for Reconstruction and Development / The World Bank