Building The Gambias Social Registry Final Draft FRAMEWORK FOR THE DESIGN AND ROAD MAP LEADING TO IMPLEMENTATION PREPARED BY: MANUEL RODRIGUEZ PUMAROL AND JOHNSON MWEBAZE WITH GUIDANCE FROM PENNY WILLIAMS AND CORNELIA TESLIUC. Funding was provided by the Rapid Social Response trust fund supported by the Russian Federation, Norway, the United Kingdom, Australia, Sweden and the Global Facility for Disaster Risk Reduction. LIST OF ABBREVIATIONS BReST Building Resilience through Social Transfer for Nutrition Security CAPI Computer Assisted Personal Interviewing CT Cash Transfer CCT Conditional Cash Transfer DCD Department of Community Development DSW Department of Social Welfare GBoS Gambia Bureau of Statistics GDP Gross Domestic Product GPS Global Positioning System HH Households IT Information Technology MCNHRP Maternal and child Nutrition and Health Results Project M&E Monitoring and Evaluation MIS Management Information System NaNA National Nutrition Agency NSPP National Social Protection Policy PIC Public Information Campaign PLHIV People living with HIV/AIDS PMT Proxy Means Test SIM Card Subscriber Identity/Identification Module Card SP Social Protection SPIP Social Protection Implementation Plan SR Social Registry VPN Virtual Private Network WFP World Food Program TABLE OF CONTENTS LIST OF ABBREVIATIONS ---------------------------------------------------------------------------------------------------------- 2 TABLE OF CONTENTS -------------------------------------------------------------------------------------------------------------- 3 I. INTRODUCTION ------------------------------------------------------------------------------------------------------------------ 1 A. PURPOSE OF THIS REPORT -------------------------------------------------------------------------------------------- 2 B. METHODOLOGY --------------------------------------------------------------------------------------------------------- 3 II. CONCEPTUAL BACKGROUND ON SOCIAL REGISTRIES --------------------------------------------------------------- 3 III. KEY PROCESSES AND STEPS TO MOVE TOWARDS A SOCIAL REGISTRY ---------------------------------------- 7 I. DEFINE THE SCOPE --------------------------------------------------------------------------------------------------------------- 7 DECISION 1- SCOPE AND USE ------------------------------------------------------------------------------------------------- 8 2. DEFINE METHODS FOR HOUSEHOLD REGISTRATION ----------------------------------------------------------------- 9 DECISION 2 – HOUSEHOLD REGSITRATION ---------------------------------------------------------------------------- 11 3. PUBLIC INFORMATION CAMPAIGN (PIC) ------------------------------------------------------------------------------- 14 DECISION 3 – PUBLIC INFORMATION CAMPAIGN -------------------------------------------------------------------- 15 4. DEFINE THE HOUSEHOLD ASSESSMENT MECHANISM-------------------------------------------------------------- 15 DECISION 4 – HOUSEHOLD ASSESSMENT ------------------------------------------------------------------------------- 21 5. DESIGNING THE DATA INTAKE QUESTIONNAIRE -------------------------------------------------------------------- 21 DECISION 5 – DATA INTAKE QUESTIONNAIRE DESIGN --------------------------------------------------------------- 22 6. DEFINE A CASE MANAGEMENT & DATA UPDATE MECHANISM ------------------------------------------------- 23 DECISION 6 – CASE MANAGEMENT & DATA UPDATES ------------------------------------------------------------------ 25 7. DEFINE INSTITUTIONAL ARRANGEMENTS ---------------------------------------------------------------------------- 26 DECISION 7: INSTITUTIONAL ARRANGEMENTS ------------------------------------------------------------------------ 27 8. SOCIAL REGISTRY MANAGEMENT INFORMATION SYSTEMS ----------------------------------------------------- 30 DECISION 8: SOCIAL REGISTRY MANGEMENT INFORMATION SYSTEM ----------------------------------------- 32 V. POTENTIALCHALLENGES IN DEVELOPMENT OF THE SOCIAL REGISTRY --------------------------------------- 32 IV. SOCIAL REGISTRY IMPLEMENTATION ROAD MAP -------------------------------------------------------------- 33 ANNEX 1 - INFORMATION SYSTEMS AND HARDWARE INFRASTRUCTURE OF THE MAIN SOCIAL PROTECTION PROGRAMS IN THE GAMBIA-------------------------------------------------------------------------------- 36 A.1. NATIONAL NUTRITION AGENCY (NANA) ---------------------------------------------------------------------------------- 37 A.1.1. Maternal and child Nutrition and Health Results Project (MCNHRP) ------------------------------- 37 Findings And Challenges ------------------------------------------------------------------------------------------------ 38 A.1.2. Building Resilience through Social Transfer for Nutrition Security (BReST) ----------------------- 39 A.2. GAMBIA BUREAU OF STATISTICS (GBOS) --------------------------------------------------------------------------------- 41 I. INTRODUCTION 1. The Gambia is one of the smallest countries in Africa with an estimated population of 1.9 million1, a large proportion of youth (71 percent under the age of 30) and an average annual population growth rate of 3.1 percent.2 The incidence of poverty is 48.6 percent 3 of the population living under the national poverty line, almost unchanged since 2010 and an estimated real GDP per capita of US$488 (2017). The country faces important development challenges as reflected by the relatively low ranking in the HDI in 2015 (173 out of 188 countries) (UNDP 2017). 2. To mitigate these challenges, in February 2016 the Cabinet approved the National Social Protection Policy (NSPP). The policy defines a comprehensive and cross- cutting social protection (SP) agenda and proposes a set of priority actions to guide the gradual establishment of an integrated and equitable social protection system in The Gambia. The NSPP sets out in detail the Government’s vision and commitment to develop a modern and comprehensive social protection system. Additionally, it seeks to broaden coverage for those in need of support particularly, the poor and the vulnerable. 3. The NSPP is accompanied by a Social Protection Implementation Plan (SPIP), which defines a set of activities to guide the implementation of the NSPP across four results areas/goals: • Establish and strengthen the NSPP leadership, coordination and implementation mechanisms. • Increase coverage of social protection policies and programs to meet the NSPP objectives. • Strengthen the social protection system for effective planning, delivery and monitoring of social protection programs. • Develop a sustainable financing strategy and mechanism to fund the implementation of the NSPP and specific programs. 4. Many initiatives are underway to support the NSPP and SPIP, but they are fragmented. There are challenges in measuring the effectiveness of these initiatives in order to measure the real impact on the society in general and efficiency of individual programs in particular. The implementing agencies are operating in silos with little information being shared across the agencies and this lack of information, integration and consolidation causes challenges in effectively measuring progress of the programs. Moving forward, in an environment of fiscal constraints and budget realities, it is 1 Integrated Household Survey 2015/16 (IHS 2015/16) 2 As per UNDP Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat projections. Gambia Bureau of Statistics (GBoS) is currently working on providing revised 2013 population data. 3 IHS 2015/16. 1 important to develop strategies to improve programs’ design, including modalities for identifying and registering the intended beneficiaries. While the NSPP promotes universal coverage for various target populations, there is insufficient fiscal space to put this in practice and so there is a need for targeting of social protection programs to prioritize coverage of the poorest and most vulnerable in order to make social spending more efficient. 5. Based on the aforementioned challenges, the development of a Social Registry (SR) has been identified as a key priority as part of the third results area of the SPIP4. A SR can provide a gateway for inclusion in social protection initiatives and an improved information system on beneficiaries and benefits to promote better coordination and synergies among programs. A. PURPOSE OF THIS REPORT 6. The purpose of this report is to provide a framework for the design and eventual implementation of a SR as a gateway for inclusion in social programs and as a means of bringing efficiency and transparency in the system through better coordination. This document will review the concepts behind a SR and define key questions that would need to be answered to determine the most appropriate model for a SR in The Gambia. The focus will be on Decision-Making and Institutional Arrangements (see Figure 1 below). A subsequent phase of the work would elaborate the means of achieving the desired model, including Development of Processes and Procedures, Systems, Human Resources and Capacity Building. Thereafter, the implementation of the SR can be planned and tested. 4 Outcome 3.1: ‘Electronic single registry system that records all households and individuals enrolled in social protection programs in The Gambia will also be established’ 2 Figure 1: Key Processes and Steps to Move Towards a Social Registry B. METHODOLOGY 7. This report is based on analysis through a desktop review and consultations with different stakeholders during World Bank missions that visited The Gambia in November 2017, March and May 2018. Different stakeholders were interviewed including: Department of Social Welfare, National Nutrition Agency (NaNA), Gambia Bureau of Statistics, Department of Community Development, National Authorising Officer Support Unit (NAOSU), Ministry of Finance & Economic Affairs, Ministry of Information and Communication Infrastructure, Directorate of Health Promotion and Education, National Disaster Management Agency (NDMA), Ministry of Health and Social Welfare, UN Agencies and other development partners. This exercise provided insights into the role of each agency and its competence to contribute towards creation of a Social Registry. The report also builds on a recent World Bank Guidance Note and Assessment Tool for SRs.5 II. CONCEPTUAL BACKGROUND ON SOCIAL REGISTRIES 8. Most countries offer a series of social benefits and services to common target populations, such as the extreme poor, food insecure families, pregnant women, 5 Kathy Lindert, Phillippe Leite and Tina George (2017 ) Social Registries for Social Assistance and Beyond: A Guidance Note & Assessment Tool 3 children, youth, active age adults with vulnerable status, the disabled, and the elderly. The delivery of these benefits and services can result in fragmentation, gaps and duplication if there is not a common ‘gateway’ for assessing potential eligibility and no repository of information on actual beneficiaries. This can lead to duplication of processes and unequitable distribution of resources. Moving from isolated interventions to integrated systems can bring greater effectiveness and efficiency. 9. Most social programs pass through similar phases or “business processes� to reach out, collect information, and assess potential eligible beneficiaries and then to decide on their enrollment in a specific program. With multiple programs, this can get complex and lead to duplication of efforts. Parallel efforts in assessing eligibility both consumes scarce public resources and contributes to duplication and gaps in coverage, thus reducing the effectiveness of social protection spending. Figure 2: Business processes of multiple programs 10. A SR aims to consolidate the outreach, registration and assessment of needs and conditions of populations likely to be targeted by programs. The SR contains information not only on those benefiting from a program, but also those who may potentially be eligible and those who may have applied for a particular program but have not been enrolled in that program. In other words, it is a long list of intended populations with their characteristics recorded. Being included in the registry does not automatically guarantee a program enrollment - it is up to each individual program to make 4 enrollment determination based on the parameters of the program. Figure 3: Social Registries support the Outreach, Registration, and Assessment of Needs and Conditions 11. The name of these registries may vary from country to country: some countries call it a Social Registry, others a Single Registry or National Registry. Despite name discrepancies, in essence, the purpose is always the same: to consolidate into a single common database, structured and organized key information about current and potential beneficiaries of social programs in order to create a single-entry point for citizens to access the main social protection programs6. 12. The scope of the SR may also vary between one country and another: the SR may be an opportunity to strengthen the targeting process to serve multiple social programs; to integrate operations and services with other sectors; to develop coordinated interventions; facilitate the planning process; and/or provide a combined M&E and grievance redress across programs. 13. Depending on the scope and the objectives to be achieved, a SR may offer multiples advantages. 14. From a policy perspective, the SR may: 6 In the context of this report SR do not refer to Registries of Beneficiaries of existing programs which exist in some countries. 5 a) Apply a more equitable approach to distributing resources to beneficiaries based on objective and comparable information; b) Increase responsiveness and inclusiveness of interventions that serve the chronic poor and populations structurally vulnerable to poverty and to respond more quickly to shocks (for example, natural disasters or conflict); c) Facilitate access to universal programs such as education and health services for those who lack information on the services, costs, procedures, eligibility criteria, or enrollment processes; d) Identify gaps in coverage – or duplications; e) Increase transparency and accountability since program information can be more easily shared and compared; f) Build a stronger link to institutional frameworks and wider social and economic policies and opportunities; and g) Increase knowledge on poverty and vulnerability based on access to the large amount of information available. 15. From an operational perspective, advantages include the ability to: a) Facilitate oversight of multiple schemes and report to policymakers responsible for social protection interventions; b) Avoid duplication of administrative effort (for example, with data collection activities for targeting programs); c) Establish common systems across all schemes, such as grievance redress or payment systems, increase efficiency and save money; d) Avoid duplication errors and better manage fraud; and e) Enable beneficiaries to transition between schemes as their circumstances change. 6 III. KEY PROCESSES AND STEPS TO MOVE TOWARDS A SOCIAL REGISTRY 16. This phase involves building consensus among all stakeholders. Here, the Government and development partners should agree on all the key design parameters related to: scope, household assessment mechanism, data to be collected, registration methods, data collection and sharing mechanisms. This will be the focus of this report; implementation is for future phases of the work once key decisions have been made. I. DEFINE THE SCOPE 17. Decisions should be taken regarding the role and functions of the SR. 18. One potential role is to provide a common ‘gateway’ for eligibility – that is a ready-made list of poor and vulnerable households from which potential beneficiaries meeting a program’s eligibility criteria can be selected (e.g. all children under two or al l disabled or all those who have inadequate dwellings or who live in a certain geographical area). When more program funds become available, there is a ready-made pool of potential beneficiaries in the next geographical area or the next poverty bracket. In times of crisis, such as natural disasters or conflict, having a SR can facilitate a rapid response by quickly providing a list of households in the affected area and their characteristics. This would help to facilitate the coordination of resources for a more efficient delivery of any intervention to the affected households. 19. Another potential role is to improve the effectiveness of programs e.g., by bundling or coordinating benefits and services, such that eligibility for one program can automatically confer eligibility for another. For example, anyone enrolled in an income support program targeted to the poor could become automatically eligible for a health insurance fee waiver or a subsidy or, it can facilitate referrals to other services, such as specialized disability care or for treatment of malnutrition. 20. In addition to supporting registration and eligibility functions, the data produced by a SR may also be used for other purposes. For example, validating information collected through other methods or sources, assessing potential demand for interventions, planning and costing interventions depending on projected coverage rates, monitoring and evaluation, or other analytical purposes. 21. A key decision to be taken is whether or not certain social protection programs should mine the data from the SR when determining who gets enrolled in their programs. Having a single repository of information on potential beneficiaries, supported by a Management Information System that can track actual beneficiaries to 7 different programs, can facilitate oversight, monitoring and evaluation. This ensures a more equitable approach to distributing limited resources by identifying duplications and gaps among the chronic poor or populations vulnerable to poverty. This can also reduce the burden of potential beneficiaries applying to different programs and reduce the burden of staff and agencies from collecting information for each program separately. The SR may help identify areas where common delivery systems could be used, such as payment systems or grievance redress. 22. Decisions should also be taken about the geographical coverage of the SR; will it cover the entire country or only some areas (if the latter, which Regions/Districts/Ward/Village; Rural/Urban Areas, etc.)? Which criteria will be used to select areas (e.g., geographical targeting based on poverty maps or hunger maps)? Generally, the decision will be conditioned by the purpose for which the SR will be used, financial availability and institutional capacity to carry out such processes. Countries usually start with a small SR which over time is converted into a nation-wide tool serving multiple programs, starting on the regions where key and bigger programs are already working or where a new program is being introduced. DECISION 1- SCOPE AND USE 23. The proposed name for the Social Registry is Gambia National Social Registry (GAMNSR) 24. The purpose of building the GAMNSR is to consolidate into a single common database, structured and organized key information about current and potential beneficiaries of social programs in order to create single entry point for the citizens to access the main social protection programs. 25. All social protection programs shall be able and requested to use GAMNSR for targeting. The programs may apply their own inclusion and exclusion criteria to enroll households. 26. The GAMNSR will be hosted in the SP Secretariat and will comply with the following key functionalities: • Provide robust data on key household characteristics to facilitate assessment and eligibility determination for individual programs. • Provide periodic updating of information on beneficiaries to ensure they remain eligible. • Facilitate rapid identification of beneficiaries for programs that are scaling up or trying to create synergies with complementary interventions to enhance their impact. 8 • Produce key monitoring reports required for decision making and management. • Capability to be linked to other SP beneficiary databases and administrative databases as when these databases become available. Potential (universal) databases are CVRS, National Identification Database. 27. To ensure that the GAMNSR can meet the aforementioned functionalities, the following key features are foreseen: • The database of the GAMNSR shall be composed of data on multiple characteristics of households, including their ranking according to their level of poverty (non-poor / poor / extremely poor); on that basis the different programs apply their own inclusion criteria to screen potential eligible beneficiaries. • Capacity for regular quality control to ensure data verification, validation and update of information. • Capacity to generate regular monitoring reports and descriptive statistics about households and SP coverage in the GAMNSR. • Capacity to continuously register households and receive feedback on household information from different Social Protection programs which facilitates information updating. 2. DEFINE METHODS FOR HOUSEHOLD REGISTRATION 28. International experience shows that the data to build a SR can be collected under two mechanisms used independently or combined: 9 ➢ En masse registration through census approach: entails implementing a house-to- house survey with the objective to visit every household and register households by collecting the information. ➢ On-demand approach: This can be done in two ways: a) Targeting Centers where people from the community go to complete the information questionnaire - the center stays for a certain number of days in a community and then moves to the next community; b) Registry Offices where households come to a local community development office and request to be interviewed, then an enumerator is sent to the household for the purpose of applying the information questionnaire. 29. The table below compares the different approaches: Relative advantages Disadvantages Best suited On-demand • Lower total costs due to • Poor may not participate • In areas with low or application self-selection of non- because they lack moderate approach eligible out of registry information, fear stigma poverty/eligibility process (interviewing and face other barriers to • In heterogeneous areas fewer non-eligible access (illiteracy, • When registry is well households) distance, disability etc.) known or well publicised • Dynamic, ongoing entry • Costs can be higher if (and outreach campaigns and easier to update social workers must verify encourage applications (including changes linked (via home visits) in poor areas) to life-cycle events) information provided • When people have • More democratic • Can be a slow process, higher education levels nationally — everyone has involving long queues and • Where a network of the right to be interviewed bureaucracy social protection offices at any time • Requires large network of is available at local level • Permanent process helps staff at local level or municipal staff are build and maintain • Unlikely for people to well trained to perform administrative and report positive changes the registration function logistical structures to household conditions (to minimise travel for • Does not allow for easy applicants) collection of household’s GIS geo-referenced data Census • Better chance to reach the • Periodic surveys can lead • In areas with high approach poorest and other to static/inflexible poverty rates (more than vulnerable groups, who registries — especially if 70%) and/or high are less informed and target population is poverty density more stigmatised (less linked to life-course • In homogeneous areas likely to apply) events (pregnancy, (rural areas and urban • Lower marginal registry children 0–3 etc.) slums) costs (per household • Members of eligible • In areas with relatively interviewed) due to households may not be stable poverty dynamics economies of scale with home or respond when • With new registries travel the survey is conducted (programs), particularly • If conducted often • Costly in areas with many when a large program enough, there is a higher non-eligible households needs to start quickly chance of capturing or where households are • For registries that also positive changes to very dispersed want to keep a record of household conditions (less • Re-registration very near-poor and non-poor likely to be reported) costly and often households (e.g. to be • House check conducted postponed beyond targeted in case of an during survey process (no recommended 2 years emergency or linked to misreporting assets etc.) social insurance schemes) 10 30. The decision about which method to employ will depend on the country context, the available funds and the implementation capacity. To decide on the best methodology or mix of methodologies, the following issues may also need to be considered: • Time Sensitive: The methodology selected should consider the time factor and goals to be met in terms of scaling up programs. • Budgetary Constraints and Efficiency: Financial resources are limited and thus cost-effectiveness of any approach needs to be considered. • Institutional Capacity: Not only is the institutional capacity of the implementing agency key for the implementation of a massive registration exercise, but also the availability of partner institutions that can carry out the survey (whether census or on-demand approach). The selection process can be lengthy, cumbersome and itself requires institutional capacity. Once selected, monitoring and supervision of partners is important and can be supported through third party firms contracted to carry out timely spot-checks and quality controls. • Inclusiveness: a methodology that will increase responsiveness and inclusiveness of interventions to serve the chronically poor, serve those who are structurally vulnerable to poverty, and respond to individual shocks (e.g. job loss, disability, childbearing or old age) or large crises (e.g. natural disaster or conflict) 31. In addition, regular spot checks (or operational audits) of the data collection process need to be carried out to ensure the quality of the process. 32. Updating data in social registries is very critical especially if the registry is used for determining eligibility across programs. For a registry to be fully effective it should aim to offer: (i) dynamic inclusion of newcomers (e.g. migrants, newborns); (ii) dynamic exclusion of those who have passed away; (ii) dynamic management of transitory shocks (e.g. natural disaster, crop failure, unemployment, sickness, pregnancy). Any SR without data updates will likely face serious challenges in providing support to those most in need, especially when the data in the SR is seriously outdated. For example, in countries primarily collecting data based on a census survey approach scheduled updates are set far apart (e.g. two years in Costa Rica; three years in Colombia, Indonesia and Mexico). However, these deadlines for updating are often not met because of budget and logistical problems. DECISION 2 – HOUSEHOLD REGSITRATION 33. A census approach will be used to collect SR information from Gambian households. This would entail implementing a house-to-house survey with the 11 objective of visiting, interviewing and registering every household in a community/settlement. 34. Gambia Bureau of Statistics (GBoS) as the agency established by the Statistics Act of 2005 and charged with the responsibility for coordination all statistical activities in The Gambia will carry out field data collection for the Social Registry. GBoS will be responsible for the overall implementation, coordination, supervision for all Field Data Collection Activities. 35. There will be need to address the risks this arrangement entails, most importantly collecting data for social protection provision could undermine GBoS’ perceived independence. The government could consider learning from countries that have adopted a similar approach, (e.g., Indonesia) 36. Computer Assisted Interviewing (CAPI) tools with built-in quality controls shall be used as a means of collecting SR information. 37. A data collection software solution will be developed for the purpose of data collection. Among other functions, the solution will provide data validation, i.e., checking completeness of data, applying internal consistency checks, checking for duplication, ensuring contents and formats (such as conformity of names and ID numbers to the defined data dictionary) and ensuring the length of fields and content structure for aggregation and reporting is standardized. The objective is to obtain a clean and correct dataset 38. Process evaluation, spot checks and monitoring & evaluation will be carried out to review the performance of the data collection process and validate the accuracy of the information received by the SP Secretariat with respect to field work carried out by GBoS. 39. A Consultant/Firm shall be hired to carry out Spot Checks & Process Evaluation activities with the purpose of: • Assessing whether GBoS is implementing the Targeting Process following the methodology described in the Operations Manual. • Evaluating the quality of the fieldwork conducted by the GBoS, as well as their overall performance during the Targeting Process. • Validating the accuracy of the information received by the GBoS Implementation Unit with respect to original fieldwork carried out by data collection firm. • Assessing if the targeting methodology is producing the outputs expected from the targeting process. 12 • Reviewing the implementation approach recommended in the Operations Manual and provide evidence-based advice on which improvements may be incorporated in the GAMNSR targeting process 40. Process Evaluation activities will include the following: • Verify that the data collection firm/organization has appropriately conducted standardized training on the data collection process for all staff; • Verify that the Public Information Campaign follows the guidelines established by SP Secretariat; • Confirm that the design of the logistics plans is in line with the guidelines and in accordance with the agreed fieldwork methodology. • Hold regular meetings with SP Secretariat to provide timely feedback for appropriate changes and corrections to be made by the data collection and software solution firm. • Monitor and confirm that all targeting activities have been carried out in accordance with the roll-out procedures. • Submit an evaluation report at least one week after the data collection and software solution firm have concluded the process in a particular area. • Provide timely feedback to the SP Secretariat on observed activities and recommendations for corrective actions based on field observations. 41. The aim of spot checks is to verify the quality of information collected by the GBoS during the fieldwork. Specific tasks include the following: • Verify the information collected by GBoS through the PMT questionnaires in selected areas of a Region; • Collect information on a sample of households in a Region and calculate their score; • Make a comparison between scores resulting from the evaluated sample and the scores provided by the GBoS; • In case of any significant discrepancy in the two scores, inform the GBoS, through SP Secretariat Unit to improve its data collection; • Submit the analysis report to SP Secretariat in case discrepancies persist. 42. The SP Steering Committee shall appoint a Technical Committee for Monitoring Field Data Collection activities. Members of the technical committee must have experience in household surveys. The Technical committee shall produce weekly progress reports and shall establish modalities for timely resolution issues arising from monitoring activities. 13 3. PUBLIC INFORMATION CAMPAIGN (PIC) 43. A key activity for the success of any data collection mechanism is a Public Information Campaign (PIC) to do community outreach and inform the public about the registration process. 44. PIC through outreach is intended to inform the public about the registration process. Such outreach, prior to the beginning of the data collection process in a community is critical to ensure the success of any type of data collection approach. The PICs would be tailored to the literacy levels, cultural and ethnic differences, and accessibility to remote areas. The effectiveness of the PIC would be closely monitored and evaluated in a timely manner as to make necessary changes and adaptations along the way. The SP Secretariat with the support of an external firm/organization or public institution, will design the information strategy and will provide the PIC materials. 45. The campaign should be designed to address the following objectives: • Establish legitimacy of GAMNSR as a credible and efficient mechanism to categorize potential beneficiaries of social programs in an objective, homogeneous and equitable manner by: o Communicating the processes and procedures with the aim to improve understanding and requirements of the GAMNSR to the wider public; o Communicating the grievances redress mechanisms; o Seeking proactive involvement of the media o Sensitizing public sector on the benefits and positive outcomes of the GAMNSR; o Highlighting the importance and value of GAMNSR through a national level public information campaign; o Raising the public profile of the GAMNSR nationally and regionally with identified audiences; o Ensuring effective advocacy with critical stakeholders. • Develop conceptual understanding on GAMNSR amongst various audiences and share its achievements, successes and contributions by: o Determining the communication requirements that must be met to support the GAMNSR goals; o Identification and evaluation of the opinion, perception and expectations of the communities and stakeholders regarding the GAMNSR in particular, operations and community outreach activities; o Proposing the use of various types of media platforms including print and broadcast relating to the main themes of GAMNSR; 14 o Designing a national level mass media/marketing campaign on GAMNSR processes; o Seeking media’s proactive involvement; o Strengthening external communication campaigns; o Identifying how the activities can be improved and targeted population can benefit further; o Making available information on processes, qualification, criteria and roll-out schedules; o Establishing information-sharing protocols to facilitate research and results sharing with common clients/audience. • Ensure mass awareness about the implementation process, motivate communities to get interviewed and provide accurate household information; by o Designing a local level mass media/marketing campaign on GAMNSR processes; o Proposing the use of various types of media products, including print and broadcast, that can be related to the rollout themes of GAMNSR. o Making available information on processes, qualification, criteria, and roll-out schedules; o Contacting the representatives of local NGOs, community leaders and notables of the area to seek their support in terms of information dissemination about GAMNSR and the targeting implementation process. DECISION 3 – PUBLIC INFORMATION CAMPAIGN 46. The Ministry of Information & Communication Infrastructure shall lead the development and implementation of the PIC. 47. The Ministry may co-opt other institutions/organizations for the implementation of the PIC and/or use existing community structures to ensure information gets to all households. The following have been proposed for the implementation of PIC: • Gambia Bureau of Statistics • Directorate of Health Promotion, Ministry of Health & Social Welfare • Social Protection Platform-The Gambia • Department of Community Development • Department of Social Welfare 4. DEFINE THE HOUSEHOLD ASSESSMENT MECHANISM 15 48. Social Registries contain information on all registrants, whether or not they are deemed eligible for, or enrolled in, select social programs. Some of the information in the social registry can be transformed into measures of welfare (such as means-testing, proxy means test scores, etc.). These measures are used to determine poverty status (non-poor, poor, extreme poor) which is recorded in the SR. 49. It is important to note that having household characteristics recorded in the Social Registry is distinct from enrollment in to social program(s). Registrants are not always or automatically enrolled in a particular program due to budgetary limitations and/or additional criteria guiding enrollment decisions, beyond poverty status, such as geographic prioritization or other categorical risk factors. So, the measures of poverty in the SR combined with other enrollment criteria, would guide the social programs to select the vulnerable registrants that need social assistance. 50. Most social programs tend to concentrate resources on the poor, food-insecure and/or vulnerable and in doing so they use some form of targeting which permits the SR to put in place a harmonized targeting mechanism which could serve more than one program. Concentrating resources on the poor or vulnerable can increase the benefits that each program can achieve within a given budget or can increase the impact at a lower cost. For example, if all the benefits provided by a transfer program were targeted to the poorest quintile of the population rather than uniformly distributed across the whole population, the budget savings or the difference in impact for a fixed budget would be five to one. In practice, very few programs use an entirely universal approach and the full theoretical gain is not realized. Furthermore, targeting is never completely accurate, yet costs are associated with targeting. The size of targeting errors and costs will differ according to the setting and the types of targeting methods used which must be thoroughly assessed in any policy proposal. Ideally, targeting can focus resources on the poor to a moderate or high degree without incurring unacceptably high errors of exclusion and administrative, private, or incentive costs. 51. A number of different methods are available for assessing the poverty status of households in the SR. Some demand some sort of assessment of eligibility for each applicant (individual or household); others assess broad categories of people, for instance, all those residing in certain areas (geographical) or all those of a certain age (demographic). 52. For purposes of emphasis, been categorized as extreme poor or vulnerable in the SR does not imply any decision regarding eligibility for individual programs – that decision is taken by each individual program’s management, but drawing from a long list of individuals already identified as poor, food-insecure or vulnerable whose details 16 have been recorded in a SR. 53. This section defines a number of the common household assessment methods and summarizes some of the main advantages and disadvantages of each7. In practice, a combination of methods is almost invariably used. MEANS TESTS 54. A verified means test seeks to collect (nearly) complete information on households’ income and/or wealth and verifies the information collected against independent sources. Where suitable databases exist, and interagency cooperation can be obtained, information may be verified by cross-linking the registries of, say, the welfare agency, property registrars, tax authorities, social security agencies and the like. When this is not possible, households may be asked to submit copies or records of transactions, such as pay stubs, utility bills, or tax payments. Simple means tests with no independent verification of income are not uncommon. Sometimes verification is completely nonexistent in that a program intake worker simply records what an applicant says. Sometimes a social worker will visit the household to verify in a qualitative way that visible standards of living (which reflect income or wealth) are consistent with the figures reported. Alternatively, the social worker’s assessment may be wholly qualitative, considering many factors about the household’s needs and means without necessarily quantifying them. 55. Means tests work best in situations of high levels of literacy and documentation of economic transactions. They are administratively demanding when combined with meaningful attempts at verification. Means testing is also the form of targeting most likely to discourage work effort because eligibility is linked directly to current income. Means testing is most appropriate where declared income is verifiable, where some form of self-selection limits applications by nontarget groups, where administrative capacity is high or where benefit levels are large enough to justify the costs of administering a means test. PROXY MEANS TESTS (PMT) 56. The PMT aims to predict household incomes by measuring household characteristics. The PMT methodology uses the national household survey as its basis. The first step in developing a PMT is to undertake analysis of the household survey to identify proxies that have some correlation with household consumption. These are usually based on demographics (such as age, number of people in the household, etc.), human capital (such as level of education of the household head), type of housing (such 7 This section is drawn largely from ‘For Protection and Promotion’ Grosh, del Ninno, Tesliuc, Oue rghi (2008) 17 as the type of roof, walls, floor and toilet), durable goods (such as whether a household has a radio, refrigerator or television) and productive assets (such as whether a household owns animals). A set of multiple proxies with the best correlations, which can supposedly be easily measured and observed, are chosen and a scorecard is created. 57. Once the scorecard is developed, the PMT can be implemented by undertaking a survey of households to determine their poverty scores. Often, this is done as a census with enumerators visiting as many households as possible in a country or a particular region. However, sometimes households are asked to apply individually. The enumerators are expected to verify the answers to the questions to reduce the chances of fraud. 58. The advantage of proxy means testing for household assessment is that it requires less information than true means testing yet it is objective. Moreover, because it does not actually measure income, it may discourage work effort less than a means test would. Proxy means testing also has some drawbacks. Administering it requires a large body of literate and probably computer trained staff and moderate to high levels of information and technology. It also implies an inherent inaccuracy at the household level, as the PMT score is only a prediction, although on average, good results can be observed. The scorecard usually relies on indicators that are fairly stable and may distinguish chronic poverty well but can be insensitive to quick changes in household welfare or disposable income which may be frequent and large when an economy is suffering from a large downturn. Additionally, the formula and results may seem mysterious or arbitrary to some households and communities – as such, the PMT is often combined with community validation. 59. Proxy means tests are most appropriately used where a country has reasonable administrative capacity and where they are used to identify a large pool of potential beneficiaries or to provide the targeting mechanism for several programs so as to maximize the return for a fixed overhead. COMMUNITY-BASED HOUSEHOLD ASSESSEMENT 60. Community-based household assessment in this report refers to the use a group of community members or leaders or a village development committee to decide which individuals or households in the community are extreme poor. The main argument put forward in favour of community-based household assessment is that “communities know best� who is most in need within their communities. Village Development Committees or a group of village elders or special committees composed of community members or a mix of community members and local officials may be specially formed to assess households. 18 61. Community based household assessment may encounter several possible challenges. Local actors may have other incentives besides good targeting. For example, the granting or denial of benefits to different members of the community may continue or exacerbate any existing patterns of social exclusion. In addition, if local definitions of welfare are used, evaluating how well community-based household assessment works becomes more difficult and ambiguous. Welfare may be considered differently by different villages or regions. 62. Another challenge with community-based household assessment is that the wellbeing of communities varies greatly so that the poorest ten percent in a relatively well-off community may have higher standards of living than the more affluent in a poor community. So, while community-based household assessment may be the best option available for a small-scale program working in a few communities, it would be less suitable for a national social registry. 63. Community-based household assessment may be most appropriate where local communities are clearly defined and cohesive for programs that are just starting or plan to include just a small portion of the population and for temporary initiatives that cannot support administrative structures of their own. It can be coupled with other measures to enhance accuracy. TARGETING IN GAMBIA 64. Considering the restrictions and limitations of the different methods of targeting, a good instrument for targeting should meet at least two conditions8: first, it must be able to discriminate statistically among the population to identify the poorest or those intended as recipients of social benefits, and second, it should consider verifiable measures to avoid misrepresentation of a household’s living conditions. 65. The report “Moving Towards an Integrated and Equitable Social Protection in The Gambia�9 has been used to review some of the targeting methods used currently by social protection programs but does not go into details about the efficiency or effectiveness of these methods in The Gambian context. Three main methods have been identified: income-based, categorical and geographical targeting. A combination of categorical and geographical targeting however, appears to be the most common approach to selecting beneficiaries currently. The NSPP identifies the need to ‘assess the scope for adopting a standardized and harmonized system of categorizing, identifying and targeting eligible individuals and households. 8 Herrera, Larrañaga, and Telias; 2010 9 This report was prepared by Ms. Maja Gavrilovic with Mr. Yusupha Dibba. 19 66. Within the income-based targeting approaches, most projects adopt either a simple means test and/or community-based targeting. Simple means testing was used in social assistance schemes run by the Department of Social Welfare (DSW) which selected beneficiaries through an interview and visit to the household by a DSW social worker to verify in a qualitative/visible way, the level of poverty. 67. Community-based targeting was adopted in the Ministry of Basic and Secondary Education Conditional Cash Transfer. Through community meetings involving mothers’ clubs and school management committees, the most vulnerable girls were identified for support. To avoid leakage of benefits, the CCT combined community-based targeting with a means-testing method to identify the poorest girls. 68. Categorical targeting is also commonly employed in social assistance and livelihood promotion projects defining eligibility in terms of individual or household characteristics that are fairly easy to observe and are correlated with poverty. Categorical groups that have been defined as vulnerable and deserving of support in The Gambia include: children (Orphans and Vulnerable Children and Almudos (street children) in particular); pregnant or lactating women; young unemployed people, women, people with disabilities, the elderly and PLHIV. There are other categories that do not often feature in social protection programming, such as migrants, single parents, widows, and child-headed households. Not many programs disaggregate between transient, extreme and chronic poor in their design. 69. Geographical targeting has been used in projects that address food insecurity and livelihood promotion. World Food Program (WFP) food security assistance is based on an assessment of households’ food and poverty gaps with the most food -insecure households prioritized for assistance. While geographical targeting is very important in directing resources to spatially excluded and disadvantaged regions, the focus on rural poverty in The Gambia may neglect small pockets of the poor emerging in urban locations. 70. A combination of methods can be observed in programs such as: Maternal and Child Nutrition and Health Results Project (MCNHRP) and Building Resilience through Social Transfer for Nutrition Security (BReST) which use a combination of Geographical and Categorical targeting. These methods have been used for simplicity and are effective when programs have limited reach. 71. A social registry should be designed to cater for all common targeting methods: by ensuring that there is 100% coverage of the entire population and the data collected and stored in the social registry can be used to user programs to determine eligibility of 20 households. One this is achieved within the social registry; social programs would use any criteria to select eligible households/members for inclusion into their programs. For example, a universal program which targets everyone above 60 years would request households from the SR with household members above the age of 60 for enrollment. DECISION 4 – HOUSEHOLD ASSESSMENT 72. PMT shall be used for household assessment/categorization. 73. The fact that the GAMNSR will provide a categorised list of households does not mean all social programs have to be ‘poverty-targeted’. Programs may use any criteria for selecting potential eligible households form the social registry. 74. Community validation will be carried out by the user programs with support from the SP Secretariat. Programs shall be expected to use local structures/committees for the community validation exercise. The proposed structure for the rural communities is: • Village Head (Alkalo) • Community Development Assistant (CDA) • Village Development Committee (VDC) • Child Protection Committee 75. For Urban communities, the Sub-Ward committee is proposed to carry out the community validation 76. For poverty targeted programs, exclusion and inclusion cases identified by the community validation process shall be re-interviewed by re-applying the PMT questionnaire to the household. 5. DESIGNING THE DATA INTAKE QUESTIONNAIRE a) The SR usually collects the kind of information often required by the social programs. Such information typically includes: • Identifying information and household composition, such as name, birth date, gender; relationship with household head; marital status; unique identifiers such as a national or biometric information; • Geospatial identifying information on the location of the household, including 21 address and other contact information, as well as GPS geo-coding information (where possible); • Socio-economic status such as self-reported and/or verified information on incomes for each household member; education, and employment status of each individual; • Information on housing and assets, such as housing characteristics (e.g. type of housing material, connection to water, electricity, and so forth), self-reported and/or verified information on assets (e.g. vehicle, land, livestock, etc.); and • Other program eligibility information and addition community information to assists with planning and designing SP Programs. (such as disability status, access to services, health, food security status, registration with employment agencies, health insurance, unemployment benefits etc.). 77. To collect the information needed for a Social Registry, a Household Questionnaire should be developed and for this purpose the Government should create a multidisciplinary team composed of representatives of the different SP Programs and any other stakeholders that are going to be users of the collected information. The purpose of this team is to agree on the kind of information to collect and how to develop questions to be understood in the context of The Gambia. 78. Any SR database must always have a mechanism to identify uniquely each household and its members. Unique identification of each household and/or member is key for data sharing, tracking benefits, to avoid duplications, etc. The recommended strategy for unique identification is to use the National ID, which provides a unique code for individuals. However, coverage of the National ID may not be sufficient and is likely to be lower among the extreme poor. Apart from a National ID, other methods, e.g. use of biometrics, can be used to uniquely identify a household and/or member. DECISION 5 – DATA INTAKE QUESTIONNAIRE DESIGN 79. A multi-disciplinary committee has been proposed to develop the data intake questionnaire. The multi-disciplinary committee will be composed of the following • SP Secretariat (not yet established) • Gambia Bureau of Statistics • Representatives from Development Partners • Ministry of Health and Social Welfare (Department of Social Welfare) • Ministry of Lands and Regional Government (Department of Community Development) • Ministry of Finance & Economic Affairs • National Nutrition Agency (NaNA) 22 • The Association of Non-Governmental Organizations (TANGO) • Ministry of Basic and Secondary Education • Ministry of Interior (Department of Immigration) 80. The following general categories have been proposed for data collection; • Identifying information, including photos of all household members • Geolocation of households • Demographic characteristics of the household members • Education of Households members aged 3 and above • Health of household members • Employment & income sources • Information on housing and assets • Agricultural activities • Impact of shocks and coping strategies • Additional community level information to assist in planning and design of SP programs, e.g., cell phone coverage, distance to Social facilities, like schools, health centers, markets 81. The integrated household survey (IHS 2015/16) can provide the information required for proxies of poverty. 82. The enumeration could also be combined with the planned civil registration and vital statistics data collection process. Both should be assessed to determine the usability of the National Census data and the practically involved with combining data collection with the CRVS exercise. 83. In the absence of a national ID, the government will collect fingerprint biometrics for household heads and/or caretakers above the age of 5. Given the costs involved with collecting & managing biometrics, the government should bring together the Ministry of Interior, Ministry of Health & Social Welfare to ensure harmonization in the implementation of biometrics to reduce on the cost of running parallel activities. 84. Clear data sharing policy to be developed as part of the social registry development process. The data sharing policy should define the rules and define the nature of data to be released, their intended use, protection and the conditions of their release. 6. DEFINE A CASE MANAGEMENT & DATA UPDATE MECHANISM 85. The NSPP refers to the need for ‘an independent appeal process’ for the Registry. A case management and data update mechanism, here-in referred to as Case 23 Management, allows citizens and any other stakeholder to (i) file grievances or complaints concerning the targeting, quality of service provided, and other aspects related to the GAMNSR; and (ii) request Information Data Updates: 86. The main grievances that might arise include: a. Exclusion appeal: Households which were declared non-poor and therefore unqualified for many social protection interventions but who consider themselves eligible can appeal through the GRM. b. Inclusion appeal: community members may lodge a complaint against households categorized as poor but have been identified as a non-poor household. c. Missed out households: households which were not interviewed by the data collection firm or were missed out due to any reason, such as: declined to be interviewed; were not at the community at the time of the information gathering process, etc. can also appeal through this system. d. Complaints about Quality of Services: Occurs when a person complains about the general quality of the service provided by social registry unit or any partner organizations. 87. Updating of information and reassessment of households is another key function of social registries. Outdated information on socio-economic status of households can lead to inaccuracies in the determination of eligibility of social programs. The socio- economic situations of individuals and households can change in many ways in terms of: their demographics (births, deaths, marriages or divorces); addresses and location (moving residence, internal or external migration, displacement); economic status, such as incomes (lost wages, promotions, changes in pension or other benefits, changes in unearned income), employment status (job loss, newly employed, changes in employment, seasonal work); educational status (new degree or professional certification, school enrollment); health events, conditions, and expenses; housing and assets; and other factors. 88. Updating of information is usually through the following: a. self-reported information by the households through on-demand-centers established as part of the social registry or case management system; b. Self-reported information via en masse registration systems, updating and reassessment usually depends on the next round of “census sweep� usually referred to as a “recertification� process. c. Information is sourced from other administrative systems; d. Data updates provided by programs implementing social programs in the communities. 89. Different access channels for concerned individuals to submit their cases and 24 updates need to be designed. These could include the use of the Regional Social Welfare offices or Community Development offices, if established, to act as gateways for all requests made on the field. In situations where the households find it difficult to go to the Regional Social Welfare/Community Development office, Community Committees may be set up or Village Development Committees or other local structures used to serve as receptors of requests and complaints from citizens, or concerned citizens submit their requests and complaints through a “contact center� of the Social Registry. DECISION 6 – CASE MANAGEMENT & DATA UPDATES 90. A case management mechanism shall be incorporated to be part of the GAMNSR. 91. The existing Community Development offices shall be gateway for receiving cases and updating requests in the Wards & Regions. The Community Development staff will take on an additional role of receiving and resolving or referring appeals and updates. 92. The proposed structure for Case Management is as follows: • At National Level, the Cases and Updates Officer shall be the Focal Person who shall handle complaints with support from the SP Secretariat • At Regional level, the Regional Social Worker will be the Focal Person at the Region. Cases at this level shall be dealt with by SP Sub Committee of the Technical Advisory Committee (TAC). • At Community Level, a Social Protection Community Committee (SPCC) a sub- committee of the Village Development Committee shall be established. The SPCC will include representation from the Child Protection Committee, representation from the Village Support Group and may coopt Multi-Disciplinary Facilitation Team (MDFT) on a case-by-case basis. The Chair of the SPCC shall be Community Focal Person (CFP) for the Village. The CFP is responsible for interacting with the Households for identification of cases and bring them to the SPCC for resolve. Women should have representation on the committee. 93. The Secretariat will provide multiple channels by which project beneficiaries and community members can file complaints. These include; • In person: The Community Focal Person (Or any member of Social Protection Community Committee) shall gateway for receiving complaints at the community level. • By Phone: Individuals shall submit their complaints through a “call center� of the SP Secretariat 25 • Webpage: Individuals will also have the possibility to submit their requests and complaints through the SP Secretariat’s Webpage that will have a section for case management. 94. A Case Management System will be developed that will provide a secure login where approved users will be able to: (a) enter new cases into the system; (b) view cases, (c) classify case, (d) update what action has been or is being taken, and (e) update the status of the case, for example whether it is “open�, “closed�, etc. 95. In the event that the Community Focal Person receives a complaint, s/he should transfer the completed form to the nearest Community Development offices to be processed. If the household request is filled in at a Community Development office, the officer should process the form. The office will do the data entry directly to the case management system, if the office does not have access, the officer should send the form to the SP Secretariat for processing or referral of difficult cases. 96. Data updates will be supported as follows: • Dynamic inclusion of new births and death by integration with the CVRS Database • Directly from households: any change to household conditions can be communicated directly through the case management mechanism • From a next round of “census sweep� usually referred to as a “recertification� process. To try to establish an ideal timing for the recertification, a study to determine the length of validity of variables used for targeting may need to be carried out based on the country’s context. 7. DEFINE INSTITUTIONAL ARRANGEMENTS 97. The institutional arrangements for managing and operating SRs vary significantly across countries. Several models are observed10: a) Hosted, managed, and operated by a “central social agency,� such as a social protection ministry. Country examples of this model include: Azerbaijan, Chile, Djibouti, Georgia, Macedonia, Mauritius, Mexico, the Philippines, Senegal, Sierra Leone, Turkey, and Yemen. These central agencies also share information from the SR with other “user programs� and partner agencies via data sharing agreements. b) Central Social Agency� hosts and manages (“custodian�) the SR but outsources implementation to an operating agent. Examples include Brazil, Mali, and 10 Social Assistance and Beyond: A Guidance Note & Assessment Tool. Phillippe Leite, Tina George, Changqing Sun, Theresa Jones and Kathy Lindert. 26 Montenegro. In the cases of Brazil’s Cadastro Unico and Montenegro’s SWIS, the outsourcing is done via a performance contract. Again, data sharing arrangements govern the use of SR information for the purposes of use by other programs managed by other agencies. c) “Central Agency� that manages and operates the SR but is not otherwise involved in implementing social programs. Examples include Colombia’s SISBEN which is managed by the National Planning Department, Dominican Republic’s SIUBEN which is managed in a specialized unit under the Social Cabinet within the Vice Presidency, and Indonesia’s UDB, which has been managed and operated by TNP2K within the Vice Presidency. d) Specific social program agency to manage and operate the SR but share information with other social programs and institutions. Example case: Pakistan’s NSER, which is managed under a unit within the BISP program agency. e) Managed Centrally, Implemented in Decentralized Manner. Example cases: China Dibao Registry: Ministry of Civil Affairs and local governments’ Civil Affairs Offices manages, but system is implemented by local governments. DECISION 7: INSTITUTIONAL ARRANGEMENTS 98. The SPS will manage, operate and ensure proper implementation of the social registry. The detailed functions of the SPS include the following: • Service the Social Protection Steering Committee and any sub-committees and technical working groups; • Support the review (and development if necessary) of social protection policy and legislation; • Establish and maintain a Social Registry; • Establish and maintain a national social protection M&E framework and system; • Support harmonisation and alignment in the design and delivery of social protection programmes, projects and services; • Issue information, publicity and communications and hold events to promote wider understanding of social protection; • Establish a repository of information and knowledge relevant to social protection in The Gambia and respond to requests for information. 99. Several institutions with their roles were identified that key in the implementation of the social registry. The roles will be formalized through MoUs with carefully designed incentives and mutually agreed terms of reference. 100. The Gambia Bureau of Statistics: will be responsible for carrying out the data 27 collection operational processes on the field. This will provide an opportunity for the SP Secretariat to benefit from knowledge and experience of the GBoS in the design and implementation of massive data collection effort. The Specific roles of GBOS shall be to: • Ensure all logistics requirements (assignment of human and material resources) to undertake the field data collection. • Develop operational, recruitment, and supervision plans • Recruit, contract, and pay the necessary staff for the collection of data and ensure timely and high-quality training. The SRU will define the criteria for the personnel selection. • Prepare staff training curriculum, materials and guidelines for field staff: The SRU will define the contents of the training. • Setup and provision of logistical support (transport, per-diem, etc.) to field staff while data are being collected. • Distribute Public Information Campaign materials within their assigned geographic area. • Collect household information by using the CAPI software designed for those purposes, as well as fill out the forms designed to control and monitor the process. • Ensure the security and proper use of all equipment and hardware • When necessary, adjust the process execution procedures according to findings of the process evaluation and spot checks. • Implementation of checks and controls designed in together with the SRU to ensure the quality of the information being collected during the implementation process and a protocol for returning to the field if necessary when errors for a particular questionnaire reach a certain threshold. • Ensure that data are maintained and stored in a manner that is fully confidential so that no external individuals or institution can identify any specific personal or private in the data. Names, biometrics, pictures and any other direct or indirect identifier should only be made available to the SRU. 101. The Ministry of Health & Social Welfare (Social Welfare Department) and Ministry of Lands and Regional Government (Department of Community Development) will play key roles in the SR communication, validating the results of the household categorization process and empowering citizens by ensuring their complaints or concerns are heard by GAMNSR implementers. 102. The Regional Social Welfare and Community Development Officers shall represent the GAMNSR at regional level and be the link between the Region and The Secretariat. The roles and responsibilities of the Community Development and Social 28 Welfare Departments with regard to the GAMNSR implementation are to: • Oversee GAMNSR’s operations in the Regions • Coordinate training and sensitize the community on the SR implementation process. • Act as liaison between the community members and SP Secretariat. • Participate on the community/regional social protection committees • Mobilize community for data collection process and validation of households’ categorization exercises. • Facilitate the formation of the Community Committees, where they don’t exist and facilitate the selection of Focal Persons in each selected community • Assist in addressing case management issues. • Conduct home visits to track and refer case management cases. • Follow up complaints brought to them by community members. • Help community members fill out case management forms. • Guide through the routes and assist enumerators in household identification inside the community, if needed. • Report GAMNSR activities at the Community level 103. The Ministry of Information and Communication Infrastructure will lead the implementation of the Public Information Campaign and provide ICT technical assistance where possible. ICT support may include the following; development of equipment specifications; equipment evaluation; hosting the SR Management Information System and recruitment of IT staff of the Secretariat. 104. The Department of Immigration will provide guidance on the development of a unique-identifier for household members. The unique identifier shall be used across social protection program databases for purposes of data sharing and to prevent double dipping. The secretariat will also establish the feasibility of a joint-data collection exercise with Department of Immigration which is yet to commence on a massive data collection exercise for the issuance of national IDs 105. The Department of NGO Affairs, Ministry of Lands and Regional Government will be responsible for ensuring that NGOs implementing any social assistance or social protection programs shall have registered and signed an MoU with the SP Secretariat on the potential use of the Social Registry for targeting and also to ensure NGOs provide updates and feedback to the Social Registry on beneficiaries enrolled on their social protection programs. 106. The Ministry of Health & Social Welfare registers births and deaths for the citizens of The Gambia and shall be able to indicate any such updates of beneficiary 29 records as part of the Social Registry Updates. The interoperability of the CVRS Database and Social Registry Database will be key to ensure synchronization arrangements for continual data exchange and/or updates between the two databases. 8. SOCIAL REGISTRY MANAGEMENT INFORMATION SYSTEMS 107. Operationally, Social Registries are information systems. Information is the core input and output of Social Registries. The main “inputs� to these systems include various types of information needed to determine potential eligibility for social programs and the main “outputs� of Social Registries are data that have been transformed into standardized formats that permit assessment of needs. 108. Management Information Systems for Social Registries provide functionality for Field Data Collection and an interface for Social Registry staff to transform and manage household data for eligibility assessment. It enables them to: (a) view data on individuals and households: unique ID, members, demographics, socio-economic, housing and assets, health, education, employment, utilities, all programs registered, etc., (b) manage data: data cleaning for validity, accuracy, completeness, consistency, uniformity by parsing, deduplication, transformation, statistical methods, etc.; data filtering, data matching & data archiving, audit trails, (c) assess eligibility using policy/criteria, (d) generate an eligibility list, (e) exchange data: (i) push eligibility data to beneficiary systems, (ii) extract, transform and load data sourced from other administrative systems, (f) update/rectify data based on cross checks, using protocols. 109. ICT Infrastructure on the other hand is the composite hardware, software, network resources and services required for the existence, operation and management of an organization’s IT environment. It can be as simple as setting up IT equipment (servers, network, storage, power supply and cooling) in a room onsite, or as complex as commissioning a data center in a warehouse-style building. Typically, a data center to support sophisticated operations include the following components of ICT infrastructure: • Facilities – including electrical power utility grids, UPS (uninterrupted power supply), back-up generators, power distribution units, automatic transfer switches, and cooling equipment. • Server equipment – including servers mounted on racks and cabinets (physical and virtualized). • Networking equipment – including routers, switches, load balancers and link technology (copper/fiber cabling). 30 • MIS hosting and Storage – including options for local hosting in a data center and storage or cloud hosting and storage 110. In order to manage the time and cost of procurement, investment and operations and to achieve economies of scale for government, as a whole, a shared data center is highly recommended. Fragmentation of programs have resulted in duplicate investments in software applications, databases and ICT infrastructure across and within government agencies. 111. Whereas using a data center is current practice, some programs opt for a cloud- based approach, to minimize procurement, investment and operations costs, and to take advantage of potentially unlimited computing power. It is worth understanding the benefits the cloud can provide, including but not limited to • Security in the cloud is extremely high—very hard to do yourself. • “Bring your own keys� means that data is encrypted and that the provider cannot read it. • Reliability is excellent because of the nature of the infrastructure. • Flexibility is virtually unlimited; new environments, new applications, and higher loads are all handled with a click and only when needed. • Costs are low because scale is massive. • Low-end servers cost from 2 cents per hour and medium level servers from 10 cents per hour. • There are many server options, so the right performance can be obtained at the right price. • 1 TB of storage costs from US$100 to US$1,200 per year depending on access and performance. • Autoscaling can be used to handle peaks and provide lower base costs. • There are many options for cold standby systems, active-pass or active-active configurations, backups, and disaster recovery • Many other services are available at low prices—intrusion detection, identity, monitoring, analytics. 112. Governments and businesses around the world are moving to the cloud, with many governments adopting a “cloud first� ICT strategy. Applications can be designed and deployed to be resilient enough to handle interruptions and degraded service levels. Common practice is to use the power and flexibility of the cloud to ensure that nonfunctional service-level agreement requirements are met. There is general acceptance of what cloud providers are doing and how they are doing it. This means that there is little need to be concerned about possible traps because these have been addressed as competitive pressures make cloud services more transparent and complete. 113. Alternatively, although a data center is easily understandable and under direct control of the government, there are considerable costs involved, suitably trained staff 31 are needed to operate a data center, and equipment will need to be upgraded regularly. 114. Based on the review of the Information and Communications Act, 2009 and National Information Communication Infrastructure Policy and Plans for the Gambia, it seems there is no legislative impediment to keeping data in the cloud. DECISION 8: SOCIAL REGISTRY MANGEMENT INFORMATION SYSTEM 115. A software development firm will be hired to design, develop and implement the data collection software and social registry management information system (SR MIS). 116. The data collection software will provide the following functionality • Collect all data as defined in the PMT Questionnaire. • Provide a secure data connectivity solution for transferring data from the CAPI devices to owner-controlled storage servers Perform data quality control, data aggregation, and enumerator management • Perform data synchronization, data de-duplication, and adjudication. 117. The SR MIS will include the proposed functionality: • Information processing (e.g., Data storage, transformation, verification, validation, update, and exchange); and • Business process management (e.g. registering & updating household information, determining potential eligibility and processing grievances); • Analytical support (e.g. queries, monitoring and reporting). 118. A compressive assessment of the existing data centers and use of the cloud will be carried to decide on the MIS/database hosting and backup strategy. 119. A review of the existing laws shows that, there is no explicit statement preventing registry data from being held outside of the country. Given the benefits of the cloud hosting, the Government may need consider the use of the Cloud for MIS & Data hosting with a local copy of the data held at a government data center as a backup. V. POTENTIAL CHALLENGES IN DEVELOPMENT OF THE SOCIAL REGISTRY 120. The introduction of a SR may represent a substantial change in the established way of developing processes inside the SP Programs, since the SR is more than software and hardware; and represents a new way of operation of the social protection system. 32 Despite all the good things that this system would bring such as increased accuracy, efficiencies, productivity gains, and quality improvements, numerous challenges may arise for the successful implementation and the SP Secretariat should be prepared to handle them. 121. The feasibility of the development of the SR will be directly linked on how to address the challenges that typically arise during the development of any information system. The following challenges should be considered as part of the process of design and implementation: • Managing expectations: developing a SR does not mean that it has to have all desirable functions from the beginning but should have the potential to be developed further. • Defining clear operational processes is key: more important than the technical design of these systems, is the correct redefinition and update of operational processes and procedures of each social program. The systems are simply the automated reflection of the processes reflected in the operation manuals of the programs. • Organizational changes may be needed: efforts to develop the SR may also uncover the need for organizational changes, which may be even more unsettling than the procedural changes necessary to implement the system. • The support of the National Social Protection Steering Committee is necessary to ensure success of the implementation process. Responsible managers must be willing to devote sufficient time and effort to fully understand the general concepts and objectives of the SR. • Education Through Participation. The preparation of manuals of procedures and other explanatory materials is a necessary part of the educational process. The preparation of these materials is not the most important part of the process, however. Management at all levels must be convinced that the new system, in fact, is going to be used and that it will help them do a better job. One of the best ways to do it is to have managers teach managers - that is, top management should discuss the new system with subordinates, who then carry the message to their subordinates, and so on. Since the teachers become more fully trained, this process aids in the education of all those involved. • An automated SR is not a substitute for decision-making. It is important not to oversell the potential of the new system. The system can provide certain decision parameters, but it cannot make decisions for example on eligibility for programs. IV. SOCIAL REGISTRY IMPLEMENTATION ROAD MAP 33 122. The proposed activities leading to the development of the GAMNSR is provided in the table below. A detailed schedule with timelines is provided in the implementation plan. Areas of Project Activities Responsible Intervention Workshops for consultations with various WB stakeholders Identify the information needs of various WB stakeholders Review of targeting procedures and forms Questionnaire Design SR Design Phase for individual programs Committee Development of the PMT Questionnaire Consultancy Development GAMNSR Operations Manual Consultancy Development of Training Materials New Consultancy Staff Hiring of key staff of the SR – SR Manager SP Steering Committee Strengthening of the Hiring of second batch of SR Staff (Database SPS Institutional Administrator, MIS officer Complaints and SR Manager/Director Capacity Update Assistant) Training of Staff SPS/WB Development Strategic Partnership with SPS GBoS for Data Collection Development Strategic Partnership with Ministry of Information and Communication SPS for the Public information Campaign Development Strategic Partnership with Strategic Social Welfare and Community SPS Partnerships Development Development Strategic Partnership with SPS other institutions/Organizations Consensus building Workshop SPS Development of the Social Registry Management Information System Business Consultancy Model Documentation Design of the Social Registry Management Consultancy Information System Development of the Social Registry Management Integrated Information System Implementation of the Social Registry Consultancy Management Information System Goods MIS Infrastructure Procurement of Computing Equipment for Development SPS Development of the SPS Website Consultancy Implementation Third Party Data Center, MoICI/SPS backup and Hosting Services Design of the Public Communication Directorate of Health Campaign Promotion (DHP) 34 Hiring the Spot checks & Process Evaluation SPS Consultant Implementation of the Communication Ministry of Information and Campaign Communication Community Entry Activities e.g.., Social Welfare, Community Workshops, Consultations, meetings with Development, SPS, GBoS, DHP Regional, Community Committees, VDSs Formation of Community Committees and Social Welfare / Community CFPs Development / SPS Organize and conduct the training of the Social Welfare / Community Community Committees and Focal Persons Development/SPS/GBoS Structure Numbering and Updating / Community Focal Persons Preparation of Layout Map Selection of Field Staff GBoS Building the Training of all Field Staff GBoS GAMNSR Development and Testing of Data Collection SPS/GBoS Software and Tools Field-testing of the instruments & Pilot data SPS/GBoS collection Social Welfare, Community Process Evaluation Workshop Development, SPS, GBoS, DHP Adjust Processes SPS Data Collection, Quality Assurance, Data GBoS Processing, Household Categorization Spot Checks, & Monitoring and Evaluation SPS/Firm Social Welfare, Community Community Validation Development, SPS, GBoS, DHP Grievances and Case Management SPS 35 ANNEX 1 - INFORMATION SYSTEMS AND HARDWARE INFRASTRUCTURE OF THE MAIN SOCIAL PROTECTION PROGRAMS IN THE GAMBIA 123. This chapter provides preliminary analysis of the status of information systems and hardware infrastructure, presented in the form of findings and challenges, identified within two key stakeholders: The National Nutrition Agency and Gambia Bureau of Statistics. Even though the Department of Social Welfare in the Ministry of Health and Social Welfare and the Department of Community Development in the Ministry of Local Government play a key role in the planning, design and operationalization of SP interventions in The Gambia, there won’t be direct reference to them in this chapter since they do not possess technological infrastructure and functional information systems at present. The findings are based on the interviews, meetings, and information collected with the different stakeholders involved in the process. Further the analysis is used to establish the basis for making suggestions and recommendation leading towards a harmonious integration of data for an effective SR to support the Government’s Social Protection programs and policies. 124. The parameters used for analyzing and assessing IT capacity of main implementing agencies are: • Processes and procedures o how well information processes are defined, documented, used and updated with relevant business rules. • Software o how much the software is sustainable, useable and maintainable. • Database o use of relational database11 for storing, processing and archiving data. • Hardware and networking infrastructure o to ascertain the capacity, sustainability and reliability of computing resources used to build a Social Registry. • Security mechanisms o to understand and ascertain the preventive measures and safeguards implemented against security breaches and threats. • IT Staff o The assessment of IT Staff in their strength, size and capacity will help the organizations to identify the areas of improvements through training and development. 11 A relational database is a collection of data items organized as a set of formally-described tables from which data can be accessed or reassembled in many different ways without having to reorganize the database tables. 36 A.1. NATIONAL NUTRITION AGENCY (NANA) 125. The NaNA was established in 2000 and became a legal entity by an Act of the National Assembly in 2005. Placed under the Office of the Vice President, NaNA is responsible for coordinating all nutrition and nutrition related activities in the country; facilitating inter-sectoral collaboration in the area of nutrition and implementing the National Nutrition Policy. 126. The overall mission of the (NaNA) is to improve the nutritional status thus reducing malnutrition, morbidity and mortality among the general population, especially the most vulnerable groups; pregnant and lactating women and children under five years of age, thereby contributing to the productivity of The Gambian population and the socioeconomic development of the country. 127. The NaNA has been responsible for the implementation of key programs linked to nutrition and social protection, such as: Baby Friendly Community Initiative (BFCI), Baby Friendly Hospital Initiative (BFHI), Micro-nutrients Deficiency Control Program, Nutritional Education, MCNHRP and Building Resilience through Social Transfer for Nutrition Security (BReST). For the purposes of this report, the focus of interest will be the last two. A.1.1. MATERNAL AND CHILD NUTRITION AND HEALTH RESULTS PROJECT (MCNHRP) 128. The MCNHRP is a project supported by the World Bank for which the implementing partners are: NaNA as fund holder and verifier, and the Ministry of Health and Social Welfare as service provider, purchaser and regulator. 129. The project uses a mix of targeting approaches through categorical targeting (pregnant women) and geographical targeting (implemented in 5 regions: North Bank West, Central River, Upper River, Lower River and North Bank East). The main objective of the program is to encourage the demand and provision of maternal and child health services at the first level of care, through monetary transfers to the beneficiary women, conditioned on prenatal and neonatal check-ups, and performance-based financing with health centers. The payments are done manually, although a pilot test is on run to use mobile money through a private service provider (Q cell). 37 FINDINGS AND CHALLENGES A. INFORMATION SYSTEM PROCESSES • There is very little focus on documenting, defining and describing information system processes that could help in better understanding of the workflow, procedures and protocols • All operations including payment list generation, payment calculation, data exchange, reconciliation is all done manually using excel files. • Update and changes to beneficiary list are managed manually in the excel files. B. SOFTWARE • MCNHRP has no specific software customized to their business and system requirements. All events of beneficiary data processing including payments, updates, and reconciliation is managed manually using an Excel file that is highly prone to error and insecure. • The health performance indicators for the project are tracked using District Health Information Software (DHIS) - a free and open source health management data platform used by multiple organizations and governments worldwide. DHIS is used for aggregate statistical data collection, validation, analysis, management, and presentation. This data analytics and management platform is a completely modular web-based software package built with free and open source Java frameworks and operates under a liberal BSD license12. Clients can get DHIS as software-as-a-service, which includes system backups with safe storage at a remote server, SSL (HTTPS/encryption) use for data security, and stable, high-speed Internet connectivity. The service provider for DHIS software-as-a-service in the cloud is BAO Systems. DHIS is easily interoperable with third-party clients, including Web portals, Android apps, and other information systems. The source code of DHIS is hosted on GitHub. C. DATABASE • There is no automated or standardized database for managing beneficiary data, payments data and reconciliation data. Ensuring and governing data integrity, consistency, and reliability in the absence of a database is unmanageable. Also, as there is no archive and historical data tractability, auditability, and verifiability of data is very subjective and not guaranteed. • There are no backup policies for backing MCNHRP data. 12 BSD licenses are a family of permissive free software licenses, imposing minimal restrictions on the use and redistribution of covered software. This is in contrast to copy left licenses, which have reciprocity share- alike requirements. The original BSD license was used for its namesake, the Berkeley Software Distribution (BSD), a Unix-like operating system. 38 D. HARDWARE AND NETWORKING INFRASTRUCTURE • MCNHRP performs all its operations using desktop and laptops computers. • There is no functional LAN and Data Center facilities at NaNA. • Data sharing takes place through e-mails with no protocols. E. SECURITY MECHANISM • The Excel files containing sensitive beneficiary information and payment information only could be secured using password protection. This security mechanism is very basic and cannot be relied upon with such sensitive information. • There are no information security policies followed and neither there is any information security assessment done. F. IT STAFF • Under the project structure there is no dedicated, functional IT unit, so there is a limited specialized skills and resources. A.1.2. BUILDING RESILIENCE THROUGH SOCIAL TRANSFER FOR NUTRITION SECURITY (BREST) 130. The BReST is a cash transfer program funded by an European Union Grant, and implemented through a UNICEF partnership with the NaNA, as a major implementing agency and accountable for overall implementation of the program partnering with the Department of Social Welfare (DSW) of the Ministry of Health and Social Welfare given its important role in promoting social protection policy. 131. The key objective of the BREsT is resilience building and improving nutrition status of lactating women and children under 12 months. The project uses a mix of targeting approaches through categorical targeting (lactating women and children under 12 months) and geographical targeting (implemented in 3 regions: Upper River, Central River and North Bank Region). The program carried out the first payment on April 2017, and to this day has 6,176 beneficiaries. In addition to the monetary transfer component, the program supports nutritional improvements through dietary diversification, promotion of optimal infant and young child feeding, as well as community counselling and capacity building of service providers. 39 FINDINGS AND CHALLENGES A. INFORMATION SYSTEM PROCESSES • The BREsT has defined and documented their information system processes with great detail. • All processes and procedures involved in the operations have been well documented in an operational manual and user manual. B. SOFTWARE • For implementing the BREsT, the SCOPE Platform from the World Food Program has been used. SCOPE is a flexible and powerful multi-modular cloud-based digital solution that offers the capabilities of managing: beneficiary registration, beneficiary and program management, conditionality, assistance disbursement via multiple channels (financial service providers, food, digital vouchers), closed loop offline payment solution, and a business intelligence solution that allows the huge amount of transactional data that build up on the platform. C. DATABASE • BREsT uses the cloud capabilities of the SCOPE to host their database at WFP HQ in Rome. • Since the project uses SCOPE comprehensive database backup policy is in place. • The important areas in the database are encrypted and only authorized database administrators have access to these areas. D. HARDWARE AND NETWORK INFRASTRUCTURE • Given that SCOPE is a cloud-based platform, the project is managed using desktop, laptops computers and tablets with internet access through a private service provider. E. SECURITY MECHANISM • SCOPE Platform has employed contemporary security mechanisms including firewall, IPS/IDS, encryption technologies and antivirus systems. • SCOPE Platform runs on SSL with digital certificates encrypting data on web/network layer. • SCOPE Platform uses a “privacy and personal data protection� policy. 40 F. IT STAFF • The program, through the UNICEF support, has been able to have a qualified staff to manage the system, even though the project does not have a specialized IT unit. 132. Although the use of SCOPE has given the project enough stability and efficiency, there are opportunities for improvement, for example in areas such as identification of beneficiaries. One of the key challenges the project faces is the lack of proper identification documents of the beneficiaries that are eligible and registered but because of the lack of documentation, they cannot receive payments. The use of biometrics would be a feasible solution for it. A.2. GAMBIA BUREAU OF STATISTICS (GBOS) 133. The GBoS was established under the Statistics Act of 2005. The Act also established for the bureau the National Statistics Council to carry out the functions and manage the affairs of the bureau. 134. The GBoS is a semi-autonomous statistic agency under the Ministry of Finance and Economic Affairs. The bureau is the supervisory authority for the national statistical system and is the only body in the country entrusted to provide official statistics and is responsible for the collection, production and dissemination of integrated, relevant, reliable and timely statistical information. 135. GBoS has an elaborate infrastructure for data collection such as Statistical Officers and trained enumerators. The bureau is also the custodian of all official statistical information and for this it archives information of all national surveys including National Population and Housing Censuses. It does not implement social programs. FINDINGS AND CHALLENGES A. INFORMATION SYSTEM PROCESSES • There is no focus on documenting requirements, business rules, validation criteria, functional specifications, design specifications, test cases, quality assurance, user manuals and change management. • Lack of formal documentation results in many challenges like poor assessment of systems, difficult to test and validate them and with time these systems become unmanageable and difficult to trouble shoot. 41 B. SOFTWARE • Most of the software used at GBoS are for the exclusive use of statistics management, such as: CS Pro, STATA and IBM SPSS Statistics • Although GBoS has used CAPI software for data collection, the practice is new at the institution having used it only in some small-scale surveys, like: Multiple Indicator Cluster Survey and Malaria Indicator Survey. In addition, the software used and the tablets for the collection of data used in these processes were not the property of the GBoS, therefore at present they do not possess either of the two. • On the CAPI software used, there were no quality inspection/review modules in the system through which quality of the already entered data could be assessed. • There were no validation checks to identify duplicate records or applications in the system. • There is no practice of producing technical documentation, specification documents, test cases, design documents, deployment instruction and user manuals for the software used so far for data collection, which makes it difficult to fully understanding the software itself and its functionalities • There is no source code management and version controlling tool that is used for storing and managing the software’s codes and releases. C. DATABASE • Referential integrity that should be included in every database is missing in some of the databases. It requires every value of an attribute of a table must also exist as a related value of an attribute in another table. The referential integrity conditions ensure that the related tables are consistent, and that related data is not deleted or changed, causing integrity challenges. The non-existence of these conditions in all databases can lead to ambiguities in the data. • No technical documentation exists of the data model used, nor of the tables, columns, and objects in the database, which could make it difficult to completely comprehend all databases. • Some of the databases have very few constraints on their tables that implies that the validation has to be perfected by any software used for those purposes. However, this is not a good practice and proper validation and constraints should be applied on columns where possible. • There are no archive or history tables of all the databases. Their existence implies that updates to database are logged and archived for future audit or historic data analysis. • There are no clear backup policies for backing all databases. • There is no use of advance database features like partitioning, table spaces, materialized views and others to enhance database efficiency. 42 D. HARDWARE AND NETWORK INFRASTRUCTURE • The GBoS has some hardware and networking infrastructure, however the server room is not a purpose-built facility offering standardized facilities of a professional data center. The server room has limited availability of redundant power, specialized cooling, fire suppression, environmental control, infrastructure management, structured cabling, raised flooring and others required facilities for hosting servers and computing equipment. • There is no disaster recovery and/or business continuity infrastructure or plan. The only remedy is to restore the backup of data on another server that will result in loss of data and it will cause downtime. • There are no specialized backup devices used and backup is kept on external devices and multiple copies are kept on local PC’s. This practice possesses a security risk as well as it is not a standard way of keeping backups. Also, there is no backup policy that is followed for backing up data. • Each hardware component comes with its support and warranty so that in case of any problem, the concerned vendor can be engaged for immediate resolution. However, in case of the GBoS, there are no support and maintenance contracts with hardware vendors. E. SECURITY MECHANISM • The GBoS has used firewall and antivirus server managing clients and updates, however there is no VPN, IPS and IDS solutions in place for ensuring security of next level. • There are no encryption technologies used and data in the databases is kept in plain text including the password which is a serious security risk. • There are no information security policies nor any information as to whether a security audit is carried out. F. IT STAFF • The IT Staff is capable and has good knowledge however the team needs specialized resources for database administration, system administration, storage administration, quality assurance, software architects and technical documentation experts. • Further additional training and courses in information security, database administration and relevant courses would certainly be advantageous in strengthening and enriching the skills of existing resources. 43