TECHNICAL NOTE CÔTE D’IVOIRE EVALUATING DIGITAL PAYMENT SYSTEMS FOR EXPANDING SOCIAL PROTECTION COVERAGE DECEMBER 2019 CONTENTS ACKNOWLEDGEMENTS ............................................................................................................... 3 I. MOTIVATION AND OBJECTIVES ........................................................................................ 4 Choice of Social Protection and Labor Programs Included in this Note ....................................... 4 Structure of this Note ................................................................................................................. 5 II. ENABLING ENVIRONMENT FOR DIGITAL PAYMENTS ........................................................ 5 Institutional Framework .............................................................................................................. 5 PSP landscape in Côte d’Ivoire ................................................................................................... 5 KYC Requirements ....................................................................................................................... 6 Interoperability ............................................................................................................................ 6 Identification Systems ................................................................................................................. 6 III. EFFECTIVE MONITORING OF DIGITAL PAYMENTS ............................................................ 8 PSP Reporting ............................................................................................................................. 8 Specific Challenges of Monitoring Digital Payments ................................................................. 8 The Grievance Redress Mechanism ............................................................................................ 9 GRM Basics .................................................................................................................................. 9 GRM and M&E ........................................................................................................................... 10 Putting It All Together: A Monitoring System for Digital Payments ........................................... 10 IV. PERFORMANCE EVALUATION......................................................................................... 12 Evaluation Framework .............................................................................................................. 12 Detailed Findings ...................................................................................................................... 13 Accessibility ............................................................................................................................... 13 Integration ................................................................................................................................. 15 Lessons Learned from Global Experience .................................................................................. 16 V. RECOMMENDATIONS .................................................................................................... 18 Interoperability, choice and quality-based competition ............................................................ 18 Monitoring and Accountability.................................................................................................. 19 Grievance Redress .................................................................................................................... 19 Conclusion ................................................................................................................................ 19 VI. ANNEXES ........................................................................................................................ 22 A. Fundamental Performance Indicators for Payment Mechanisms ...................................... 22 B. Good Practice for M&E of Digital Payments....................................................................... 23 C. Good Practice for GRMs .................................................................................................... 25 D. Example Survey Instrument for Payment Monitoring – PFSP............................................. 28 E. Example Grievance Form – PFSP........................................................................................ 30 F. ISPA Payment Assessment Tool Overview ......................................................................... 32 2 ACKNOWLEDGEMENTS This technical note was prepared as part of the World Bank's technical assistance program1 (2015-2018) in Côte d'Ivoire on social protection and labor policy (contributory and non-contributory), "Building Productive and Resilient Social Protection and Labor Systems" and "Strengthening Social Protection and Labor Delivery Systems". The note complements the broader World Bank policy note, Modernizing Social Protection and Labor Policy for Inclusive Growth (2019). The present note is based on an adaptation of the Inter Agency Social Protection Assessments’ (ISPA) Payments Assessment Tool, described in Annex D. The program was implemented in partnership with the Intersectoral Social Protection Technical Committee under the Presidency, responsible for the operationalization of the National Social Protection Strategy and the National Employment Plan (2016-2020), as part of the National Development Plan 2016-2020. This work has accompanied investment projects on strengthening the social protection and labor system for vulnerable populations, funded with support by the World Bank/International Development Association (IDA), under the Productive Social Nets Project and the Social Protection and Economic Inclusion Project, managed by the Ministry of Employment and Social Protection and the Ministry of Solidarity, Social Cohesion and Poverty Reduction. The team appreciates the effort, time and views shared by public and civil society representatives as part of the preparation of this work. This technical note was prepared by a World Bank team including Chris Tullis (Consultant, Digital Development Specialist) and Heba Elgazzar (Senior Economist, Social Protection and Jobs), with contributions by Meritxell Martinez (Operations Officer, International Finance Corporation/IFC), Giulia Parzani (Consultant, Operations), Rita Oulai (Consultant, IFC) and Michael Muhanna (Consultant, Social Protection). The work benefited from helpful inputs and discussions with Robert Palacios (Lead Economist, Social Protection and Jobs), Tina Georges (Senior Public Sector Specialist, Social Protection and Jobs), Laura Rawlings (Lead Social Protection Specialist), and Kathy Lindert (Lead Economist, Social Protection and Jobs). The team thanks the peer reviewers of this work for their helpful comments and suggestions, including Giuseppe Zampaglione (Lead Specialist, Social Protection and Jobs), Silvia Baur-Yazbeck (Financial Analyst, International Finance Corporation) and John Van Dyck (Senior Specialist, Social Protection and Jobs). The team also thanks the broader Côte d'Ivoire country team, including Oumou Coulibaly (Program Assistant, Abidjan Office), Lydie Anne Billey (Program Assistant, Washington) and Dora Harris (Senior Program Assistant, Washington) for their helpful logistical and office assistance. The technical assistance program was mobilized under the general direction of Jehan Arulpragasam (Practice Manager, Social Protection and Jobs, Africa Region) and Pierre Laporte (Country Director, Côte d'Ivoire). 1This work has received generous financial support from the Rapid Social Response Trust Fund and these donors (Sweden, Norway, the United Kingdom, Australia and the Russian Federation), with sincere appreciation from the authors. No part of this report may be reproduced or disseminated in any form or by any means, electronic or mechanical, photocopying or recording, storage, or retrieval system, without the prior written permission of the World Bank. The observations, interpretations and conclusions expressed here are those of the authors and do not necessarily reflect the views of the Executive Directors of the World Bank or the countries that they represent, or the financial donors to the work. The World Bank does not guarantee the accuracy of the data cited in this report. 3 I. MOTIVATION AND OBJECTIVES Coverage of social protection and labor programs (SPL) in Côte d’Ivoire remains very limited, but emerging delivery platforms hold the promise of enhancing inclusion among the most vulnerable groups. As of 20142, labor market programs covered 0.3 percent; all social insurance, 7.9 percent; all social assistance covered 27.5 percent, including cash transfers at 0.2 percent. While this means that 33 percent of the population are covered in total, the bulk is in-kind food assistance, followed by pensions for formal sector workers. Given that 46 percent of the population is under the poverty line, determining the relative effectiveness of different SPL instruments can help maximize the value of spending for the poorest households. Côte d’Ivoire’s 2014 National Social Protection Strategy (NSPS) calls for a shift in public expenditures towards a more targeted, focused approach on the poor, and this note aims to shed light on innovative approaches to delivering services to the hard-to-reach in rural regions. Within the framework of the NSPS, the Ministry of Employment and Social Protection (MESP) plays an important role in coordinating key programs and has been gradually instituting a new delivery platform, with a focus on designing coherent payment, targeting, identification, and monitoring systems to support income and livelihoods programs across agencies. Prior to the creation of the NSPS steering committee in 2015, coordination across programs has been limited since financing, implementation, program databases, policies, targeting and delivery mechanisms remain largely independent. Concerns around error, fraud and corruption (EFC) and international experience with different payment modalities, discussed outside the scope of this note, have led to more effective and transparent mechanisms for delivering benefits, notably those in cash used by pension schemes, labor programs, cash transfers, and voucher schemes for education and health programs. The main objective of this note is to evaluate the institutional and implementation framework needed for strengthening payment systems as a core element of delivering SPL benefits to the poor, around which targeting, identification and monitoring systems operate. The assessment aims to support improving how new national programs and initiatives identify the poorest households, assess their needs and facilitate the delivery of investments and services to them by maximizing digital financial technology. Choice of Social Protection and Labor Programs Included in this Note The four SPL programs features in this note were chosen for several reasons. The first is their size. Côte d’Ivoire’s SPL system is still in its early stages, with relatively few programs that reach a substantial percentage of the population. The two pension funds—the Caisse Natioale de la Prévoyance Sociale (CNPS), for the private sector, and the Caisse Générale de la Retraite des Agents de l’Etat (CGRAE), which covers the public sector—and by far the two most mature and developed SPL programs in the country. While the PFSP is significantly smaller—currently in its pilot phase—its scope is set to expand manifold in the short and medium term. Payment and contribution mechanisms used for other programs, notably employment programs and health insurance, will form the basis of future diagnostic work based on lessons learned from applying the approach for this assessment. The other factor is significance to Côte d’Ivoire’s SPL strategy. The SPL system is in the process of undergoing major reform. The current system is dominated by the two pension funds, which cover only formally employed workers, and represent 93 per cent of total SPL spending in Côte d’Ivoire. However, formal employment arrangements account for only 6-10 per cent of the active labor force. Recognizing that the current focus excludes significant portions of the population that could benefit from SPL 2 Most recent data, similar to 2015-2018 relative distribution. Data on how beneficiaries are distributed across quintiles are available from the National Household Living Standards Survey, which may not accurately reflect the number of beneficiaries found in administrative databases. Data available from administrative databases is used in discussing each individual program in this policy note. 4 programming, the current government strategy calls for this gap to be filled. The two most strategically significant programs are current the CNPS, which is currently in the process of implementing an ambitious program to extend contributory pensions to the entire informal sector, and the PFSP, which is set to offer non-contributory unconditional transfers to the bottom 20 percent of the population by income, regardless of employment status. Structure of this Note The note is structured as follows. Following the introduction in section I, section II looks at institutional framework, including the regulatory environment for financial services as well as the personal identification (ID) landscape in Côte d’Ivoire. Section III assesses the current payment systems in the four focus programs chosen. Section IV provides recommendations for improvement of the payment systems of the four focus programs in the short and long term, with an eye not only toward improving service delivery for the beneficiaries of individual programs, but also toward improving coordination between programs. The final section also discusses in detail monitoring and evaluation and grievance redress mechanisms and how they link to payments. The note concludes with a summary of key opportunities and challenges to coordinating payment systems between SPL programs. II. ENABLING ENVIRONMENT FOR DIGITAL PAYMENTS Institutional Framework PSP landscape in Côte d’Ivoire While Côte d’Ivoire has an established banking sector, its reach is limited to only 15 per cent of the population, and only 10 per cent of the rural population.3 Banks’ traditional service models, which rely on capital-intensive bank branches for service delivery, inherently limit banks’ ability to reach several population segments, including those in rural areas. In recent years, technology has allowed FSPs to innovate alternative FSP service-delivery models to better reach lower income clients. Some examples of such service-delivery models include agency banking, OTC transfer services, and mobile money. While many markets have a mix of different types of services, mobile money is clearly dominant in Côte d’Ivoire. The following chart shows the extent to which the network of access points of mobile money providers in Côte d’Ivoire dwarfs the competition (Figure 1). Figure 1 Number of financial service access points by PSP type Source: MIX Market (2015-16). There are several reasons for this, but a key reason is regulatory. While mobile money services have been given the regulatory space to flourish in WAEMU, the regulatory environment has not been hospitable to agency banking over the same period, providing a barrier to the development of this sector as an alternative to mobile money.4 From the perspective of SPL delivery systems, mobile money 3 Global Findex, 2014, latest available data. 4 CGAP study, p21 5 systems offer advantages to other PSPs as they are already familiar to most of the population, requiring less time to adopt particularly among rural populations. KYC Requirements Know Your Customer (KYC) requirements are a basic aspect of financial sector regulation that establishes minimum standards that FSPs must adhere to identify their customers. Traditionally, KYC regulation requires identity documents that much of the financially excluded population are unable to provide. Banks have the most stringent requirements, which in Côte d’Ivoire include a valid state-issued photo identification and a valid proof of address. Both documents can be costly to obtain, especially when accounting for informal and indirect costs, and remain inaccessible to certain vulnerable groups, including those whose birth is not registered in the civil registry. Recognizing the barriers to financial inclusion of overly stringent KYC requirements, the regional regulator (BCEAO) is working on risk-based KYC requirements that would decrease the rigor of required documentation for certain low-value accounts. A mobile money account can already by opened with an ID only, without proof of address. Additionally, new regulation provides for the opening of low value accounts (transacting less than 200,000 FCFA or $325 per month) without any proof of identification at all. While this is a very encouraging step, and a potential boon for SPL programs, especially those targeting low-income and informally employed populations, there is one caveat. Namely, despite regulations allowing low-value accounts not to be identified for several years, no mobile-money provider in Côte d’Ivoire has to date offered such a low-value account. Although it is unclear exactly why this is the case, the fact that identification is required to purchase a SIM card (sales of which are governed by a different regulator), which is a prerequisite for using mobile money, may be a principal barrier.5 Interoperability Some markets feature a robust environment of interoperability in payments systems, whereby money stored in an account at one provider can be transferred to an account at a competitor. This can benefit all participants in the SPL payments ecosystem, with governments find it easier to administer payments, beneficiaries being able to choose the PSP most suited to their needs, and PSPs themselves gaining additional revenue from payments in interoperable systems that they may not be able to achieve otherwise.6 Currently, sector-wide interoperability in Côte d’Ivoire is limited to the banking sector, where the regional SICA-UEMOA switch allows inter-bank wire transfers and the GIM-UEMOA network allows card withdrawals at competing banks’ ATMs. There is currently no interoperability or sector -wide payment switch for mobile money. Interoperability between mobile money and the banking sector (including some MFIs) does exist, but this functionality still immature and limited to bilateral partnerships between individual banks and mobile money providers. In the absence of sector-wide switches to ensure interoperability, there are certain payment aggregators on the market, which allow SPL programs make payments using multiple PSPs while only themselves dealing with a single actor for the purposes of making payment orders and reporting. Identification Systems In general, SPL program need to be able to both identify their beneficiaries initially and then also authenticate these identities at the time of service delivery. Historically, the main SPL programs in Côte d’Ivoire were contributory programs such as the public- and private-sector pension funds, both of which used the national ID card for identification purposes. Despite the limited coverage of this document, it has been possible to rely on it because the pension funds, themselves limited in coverage to the 10 per 5 CGAP study. 6 Benson and Loftesness 2012 6 cent of the adult population in formal employment, 7 had target populations that in practice were subset of the population covered by the national ID card. Only recently has the national SPL strategy called for expanding contributory coverage to the informally employed as well as offering non- contributory benefits, such as the means-tested social safety net (PFSP) and subsidized health insurance scheme (CNAM).8 The primary ID document in Côte d’Ivoire is the national ID card. In general, coverage is limited to Ivoirian nationals who have access to documentary evidence to prove their nationality. Despite efforts in recent years to expand access to this document, coverage remains at 64 percent of the adult population by one recent estimate.9 There are also other ID cards available in Côte d’Ivoire, but many of them share the same accessibility issues as the national ID card. The ID card for the CNAM’s was cause for high hopes for expanding coverage when it was introduced, given the CNAM’s mandate for universal coverage. In its current form however, this card suffers from many of the same accessibility limitations as the national ID due to the need for breeder documents (another state-issued ID card or a birth certificate) as a prerequisite for obtaining a CNAM ID that vulnerable populations may not have access to. Apart from the accessibility issues, this card is highly attractive, with a robust identifier whose uniqueness is guaranteed by fingerprint biometrics. The PFSP also issues a program ID card to one person designated as the primary transfer recipient in each beneficiary household. This card is non-biometric, and the identity of its bearer is not backed up by any breeder documents, but instead by community witness testimony given during the community validation of the program’s beneficiary lists. The decision to forego breeder documents was made by the program to ensure the card’s accessibility. Unfortunately, the lack of a legal basis for the community identification method used limits the card’s applicability outside of the PFSP program itself, and ultimately the utility of the card to the beneficiary. These limitations extend to receiving the PFSP benefit itself, as in the future the card will no longer be accepted by the mobile money provider who delivers the PFSP payments. Thus, PFSP beneficiaries will have to acquire an additional ID document to receive their payments (or designated a proxy recipient who has). Additionally, it should be noted that the identifier used for the PFSP card is a household-level identifier (due to the program’s beneficiaries being defined on a household level) and not an individual identifier, further limiting the card’s usefulness outside of the program. Further work will delve into linkages with the development of the proposed national biometric identification system. The characteristics of the cards issued by the SPL programs selected for this assessment are summarized in Table 1. Table 1 ID documents issued by SPL programs in Côte d’Ivoire PFSP CNAM CNPS CGRAE PEJEDEC Issues program- Yes Yes No No n/a specific ID? Number included on Yes (household-level identifier Yes n/a n/a n/a program ID? only, no individual identifier) How is uniqueness of Witness testimony during Biometrics n/a n/a n/a the identifier assured? community validation of (fingerprints) beneficiary lists Biometrics included in No Yes (fingerprints) n/a n/a n/a program ID? 7 Côte d’Ivoire Modernizing Social Protection and Labor Policy for Inclusive Growth, World Bank, forthcoming. 8 République de Côte d’Ivoire (2013), “Stratégie Nationale de Protection Sociale.” http://www.coopami.org/fr/countries/countries/cote_ivoire/social_protection/pdf/social_protection07.pdf 9 Joseph Atick (2015), “Identification for Development (ID4D) Identification Systems Analysis: Côte d’Ivoire Country Assessment,” World Bank Group. https://openknowledge.worldbank.org/bitstream/handle/10986/25197/108189.pdf?sequence=4&isAllowed=y 7 PFSP CNAM CNPS CGRAE PEJEDEC Prerequisites for No breeder documents required, National ID card or n/a n/a n/a program ID only witness testimony from birth certificate community to the beneficiary’s identity Program ID enough to No, to open the required mobile Yes n/a n/a n/a fully benefit from money account, transfer recipient program? must have a valid state-issued ID Source: World Bank staff compilation. n/a: not available. III. EFFECTIVE MONITORING OF DIGITAL PAYMENTS Monitoring digital payments to SPL program beneficiaries presents unique challenges not faced when monitoring cash payments. The core of a monitoring and evaluation (M&E) system for digital payments consists of two fundamental building blocks: The Grievance Redress System and PSP reporting. Because the GRM also has an important role as a fiduciary safeguard in addition to serving as an M&E data source, this section will begin with a presentation of the GRM overall along with some examples of international good practice before detailing its role in payment monitoring specifically. Then the role of PSP reporting will be described, and finally it will be explained how GRM and PSP data can be combined to form a coherent monitoring system for digital social payments. PSP Reporting All providers of digital payments have the capacity to produce individual-level reports on bulk payments. Because this technical capacity is inherent to the digital payment platform, there is no reason that SPL program administrators should settle for aggregate reporting, which is frequently used in cash programming. This individual-level data is not only important because it allows transactions to be reconciled individually and thus helps reduce losses due to fraud and corruption on the part of PSP staff, but also because it provides an important input to the program’s M&E system. Specific Challenges of Monitoring Digital Payments Monitoring cash payments is conceptually a straightforward affair. Since program / PSP staff are physically present at the moment when beneficiaries receive cash in their hands, monitoring is often little more than keeping a checklist of who receives cash and who does not. Even if in practice this type of monitoring can be costly and open to corruption, it is theoretically simple to implement. When payments are digitized and deposited onto a transaction account on behalf on the beneficiary instead of being paid in cash, monitoring becomes more complex. Conceptually, it becomes necessary to divide the payment into two steps: first, there is the deposit onto the transaction account, and secondly, the beneficiary must access this store of value. It is important not to conflate the concept of access with a cash withdrawal. Funds can be accessed in a variety of ways, including not only withdrawal (in cash) but also by making a transfer or payment (within the digital ecosystem). In addition to these various active ways of accessing the payment, an SPL program beneficiary can also access her payment passively if she is aware that the deposit has been made onto her account but chooses to save it instead of using it immediately. From a monitoring perspective, one key advantage of digital payments is that they allow rigorous monitoring of the deposit onto the account. However, in environments where program beneficiaries are not accustomed to digital payments—and in many cases have just opened their first-ever financial account in the context of the program in question—having data on deposits is not enough to determined that beneficiaries were also able to access their payments. There are many reasons who a beneficiary might not have access to a deposit made in her name. For example, she might have misplaced the physical hardware (ATM card, mobile phone, etc.) needed to access her benefit. Or she may not understand the payment mechanism enough (for example, how to enter a PIN code) to access 8 her account. Or she may simple not be aware that the deposit had been made. In such cases, beneficiaries may find themselves in a situation where, although money is successfully being deposited on an account in their name, they are unable to take the additional steps needed to access or benefit from this transfer. It is also important to note that even in cases where programs can obtain access to the PSP’s data on beneficiary transactions (such as withdrawals), this data is still not enough to monitor the success of the payment. Withdrawal and transaction data are unable to differentiate between cases where beneficiaries who were unable to transact, for whatever reason, and those who chose not to—to save, for example. Because empowering beneficiaries to save is one of the key advantages of making SPL payments onto a digital account, it is not desirable to discourage this savings behavior to make monitoring more straightforward. Instead, the monitoring system needs to be adapted to complex array of choices that digital payments offer program beneficiaries. The Grievance Redress Mechanism GRM Basics A GRM is meant to given program-affected persons (PAPs) an avenue to express any concerns they may have concerning the project implementation. The type of grievance handled by the GRM will depend on the type of program. In the context of cash transfer programs, complaints will often be linked to the payment mechanism (non-receipt of transfer, problems with PIN, etc.). In programs that are targeted, complaints will likely relate to eligibility (improper inclusion or exclusion from the program). Other more general complaint types might include inefficient service delivery, bribery or malpractice, sexual harassment, etc. Once complaints are made by PAPs, they are recorded in the program’s complaints database, which ideally should automatically assign a unique tracking number to each complaint. Then the complaints are then referred to the appropriate actors—which may be program staff, but could also be staff of partner organizations, such as the PSP, or actors in the local justice system—for investigation. Once the appropriate action has been taken, the grievance is marked as resolved in the tracking system and the outcome is communicated clearly to the complainant. This process is summarized in Figure 2. Figure 2 The Life Cycle of a Complaint in the GRM Reception Recording Grievance channels: Investigation Phone hotline Data entry into Local program staff tracking system Resolution Local community Complaint tracking Complaint codified number assigned according to type leaders Feedback Other intermediaries Grievance Confirmation sent Case assigned to resolved according to complainant appropriate personnel to procedures in Complaintant (unless GRM manual informed of anonymous) Investigation resolution carried out (unless anonymous) Source: World Bank staff compilation. 9 A good GRM is accessible to and understood by all PAPs (not only program beneficiaries), giving them adequate means to express any grievances they may have. It also has efficient processes for resolving grievances and communicates clearly and proactively with PAPs about not only the resolution but also the process and any expected delays. Annex F gives a summary of good practice in GRM design SPL programs as well as some examples of programs that have effectively implemented GRMs with these features. GRM and M&E Although the grievance redress mechanism (GRM) is often considered more as a fiduciary safeguard, from an M&E perspective, the GRM has a core role in providing payment monitoring data when monitoring digital payments. When a GRM is well functioning—that is, when grievance channels are well understood by program beneficiaries and are accessible—then it not only allows grievances to be resolved in an efficient manner, but it also supplies a steady stream of qualitative and quantitative data on any problem areas in the program. It is the GRM’s quality as a data source that makes it central to the M&E system. Putting It All Together: A Monitoring System for Digital Payments Figure 3illustrates the logical framework needed to assess whether beneficiaries have meaningfully had access to their digital payment. To begin, it should be noted that knowing whether there has been a transaction on the beneficiary account is always insufficient to determine whether the beneficiary had access to their benefit. This is true both in cases where there was a transaction was and was not observed. For example, a beneficiary would not have accessed her benefit despite the program observe a withdrawal in the PSP data if that beneficiary was a victim of fraud (PIN stolen, etc.). Similarly, a beneficiary would have accessed her benefit despite not transacting on her account if she was aware of the deposit on her account but had chosen to save and withdraw later. It follows then that PSP transaction data, while necessary for monitoring digital payments, is always insufficient for monitoring the delivery of social payments deposited onto beneficiary accounts. This is due to the very nature of digital deposits, where in general—in contract to traditional cash payments— the expectation is that no program or PSP staff are physically present at the moment when the benefit is accessed. Indeed, in the case of saving, there is no event to observe apart from the beneficiary’s internal intention to save. To determine whether beneficiaries have accessed their benefits, the PSP data must thus be supplemented by data collected from beneficiaries themselves: either passively (in the case of the GRM) or actively (in the case of spot checks and surveys). 10 Figure 3 Logical framework for access of digital SPL payments, including suggested data sources Source: World Bank staff compilation. In general, due to the challenges of monitoring digital payments, a key to successful monitoring is to combine various data sources to assemble a complete picture of beneficiaries’ access to their transfers. In addition to data on deposits from the PSP (and, if possible, transfers and withdrawals), other data sources include surveys of beneficiaries, data produced by grievance redress systems (GRMs) or complaints hotlines, and mystery shopping exercises. Although administering a survey to each beneficiary following each payment is unlikely to be cost-effective in the context of most SPL, if surveys 11 are deemed necessary the costs can be kept to a minimum by extrapolating from small samples (spot checks), conducting surveys by phone, or both. Among these various sources of data, however, two should be considered as primary, namely the payment report from the PSP and the GRM. The primacy of the GRM over surveys and spot checks is due to the facts that a) it makes use of an existing system, b) the system gives data in real time, as opposed to the periodic nature of surveys, c) the system is passive, and thus is less costly to administer than surveys, which are usually the most expensive item in any M&E budget. In a well-functioning SPL program, data from the PSP and GRM alone is usually enough to conclude, on a day-to-day operational level, that payments are being made successfully and that beneficiaries have access to their transfers. In general, the monitoring system will have produced strong evidence that a beneficiary had access to her transfer if the following five conditions are met ( Table 2). Table 2 Conditions for effective monitoring of digital payments using data from GRM and PSP Condition Data source 1 The beneficiary is aware that they should expect a transfer and of the transfer amount Precondition 2 A deposit has been made onto the beneficiary’s account PSP reports 3 The beneficiary is aware of and can use the GRM Precondition 4 The GRM data is feeding into the M&E system Precondition 5 No complaint has been filed complaining of non-payment GRM data Source: World Bank staff compilation. It should be noted that since the three preconditions in the above table must be met for the PSP and GRM data to allow adequate payment monitoring. As such, these preconditions themselves should also be monitored periodically in order ensure the reliability of such a payment-monitoring system. Additionally, although the above system provides for some quality assurance of the PSP data some initial monitoring of potential fraud on the part of the PSP, at intervals it will also be necessary to perform additional quality assurance where beneficiaries are systematically surveyed to ensure that they have received their transfers. For example, an SPL program’s M&E plan could provide for periodic spot checks where beneficiaries are explicitly asked whether they received and how they used their transfer, to complete the original picture offered by the PSP data and GRM. The survey instrument that is used for such post-payment surveys in the PFSP is provided as an example in Annex C. Additionally, an example of a PSP reporting framework that is adapted to the specificities of mobile money providers’ ICT platforms and can be integrated into contracts with mobile money PSPs is provided in Annex B. IV. PERFORMANCE EVALUATION Evaluation Framework The assessment approach used in this note is adapted from standardized methodology developed by the Inter Agency Social Protection Assessments (ISPA) group, housed at the World Bank. The ISPA payment assessment tool, whose detailed methodology can be found in Annex E, specifically looks at the quality of SPL program delivery, assessing three criteria, namely accessibility, robustness and integration, ranging on a scale of 1 (lowest) to 5 (highest) in terms of effectiveness and performance. The tool was specifically designed to review the payment mechanism used by an individual program or those payment mechanisms used across the whole SPL sector, providing a common framework for comparing multiple payment delivery mechanisms ( 12 Table 3). An existing SPL program’s payment mechanism can be measured against these criteria to provide a snapshot of its stage of development as well as tracking its evolution over time. This assessment approach is designed to help to identify areas that may benefit from strengthening, in line with best practice, as well as ensuring that trade-offs between criteria are made explicit to policy makers. Table 3 Summary of Assessment Scores for Selected SPL Programs PFSP CNPS10 CGRAE PEJEDEC Cost of Access 3.33 TBC TBC TBC Accessibility Appropriateness 2 3 Rights and Dignity 2.33 Reliability 1.4 Robustness Governance 3 Security 2.33 Integration Financial Inclusion 2.67 3.5 Coordination 1 OVERALL SCORE 2.25 Source: World Bank staff compilation. The scores of the selected payment systems assessed in this note are summarized in Table 3. The following sections will give details of these assessments by sub-score. The discussion in the following section focuses on the case of PFSP and discusses the following key areas in detail in this note given the availability of data to date: cost of access, appropriateness, rights and dignity, security and financial inclusion. Preliminary scores are given for reliability (based on a nearly 90 percent success rate in payment delivery as of 2017 based on the first 5,000 beneficiaries, with an additional 30,000 beneficiaries to follow), governance and security and a more detailed discussion of these dimensions will follow through future work. Detailed Findings Accessibility The PFSP payment system is fully digitized, based on mobile money technology. This choice was made possible by the relatively advanced development of the Ivoirian mobile money market, which boasted 53 per cent of the transaction volume of the entire seven-country WAEMU region.11 At the same time, the market is not yet so mature that providers’ agent networks extend into beneficiaries’ villages. Indeed, during the current pilot phase of the program, only 3 of the 56 pilot villages had a permanent mobile money agent present in the village itself (although in many cases there was an agent in a nearby market town). In order the overcome this limitation in the next phase of the program, the PFSP staff have begun working with mobile money providers to help them expend their agent networks to cover program villages. Until this is accomplished, the payments are made by roving agents who come to villages on a fixed schedule after payments are made and facilitate on-site withdrawals. • Cost of access The PFSP program management has clearly thought about cost of access and has put several measures into place to minimize the cost to beneficiaries. During the program’s pilot phase, the cost of not only the SIM card but also the phone was subsidized to 100 per cent. Additionally, PFSP program officers 10 Note: The assessment for CNPS, CGRAE and PEJEDEC will depend on data availability; the note mainly focuses on the case of PSNP as a case study representing the main challenges and opportunities faced across SPL programs. 11 BCEAO (2015), “Situation des services financiers via la téléphonie mobile dans l’UEMOA à fin septembre 2015.” http://www.bceao.int/IMG/pdf/situation_des_services_financiers.pdf 13 came to beneficiary villages to collect the documents necessary to open their mobile money accounts and opened accounts on their behalf. The program also gave all beneficiaries a free program-specific ID card that they could use to authenticate the first few payments. Cash out was also subsidized, with the market rate withdrawing the full transfer amount at one time added to the amount of the original transfer (the PSP backend software offers no way to eliminate cash-out fees directly for transfers or accounts). Additionally, for those villages where the PSP did not have a permanent presence, the PSP sent itinerant agents to beneficiaries’ villages on a set schedule, once after each transfer, lowering beneficiaries’ transport costs to zero. Because the villages where small, wait time were also low. There are however some costs that still get passed on to beneficiaries. For example, if beneficiaries wish to withdraw at another time or wish to withdraw in multiple tranches, then they must cover their own transport costs to regular mobile money agent, which in many cases can be far away. Additionally, the cost of obtaining a national ID card, which will be required to continue receiving electronic payments after the pilot period, is not subsidized. This will likely imply significant expenses for certain groups of beneficiaries, including migrants and internally-displaced persons, who may have to travel to a far-away birthplace to acquire a birth certificate. There is some evidence that some beneficiaries may be excluded from the program and/or prevented from receiving their benefit because they are being discriminated against by local leaders, on whom the program depends for communication with beneficiaries. The program plans to review its communications strategy to make more use of above-the-line communications, including posters and radio announcements in local languages, in addition to expanding below-the-line communication, primarily through the line ministry’s national network of social workers. Although Orange money accepts a variety of ID documents for account opening, 12 what these documents have in common is that in almost all cases they require the account holder to have access to a birth certificate, not only for themselves but in many cases also for their parents. This is problematic for non-citizens, IDPs, and those whose births were registered in civils registers that have been lost, either due to the 2002–2011 political crisis or due to poor archival conditions. In the short term, Orange Money has agreed to open accounts for these beneficiaries and pay them based on the program- specific ID document that the PFSP has issued to beneficiaries, while waiting for them to obtain a national ID document such as the CNI. It is not yet clear what percentage of PFSP beneficiaries will be able to obtain the necessary ID documents by the time that Orange Money stops accepting the project- specific ID. At this point, it is likely that at least some beneficiary households will lose access to the current payment mechanism. Training on the payment channel for PFSP beneficiaries has been inadequate to allow them to autonomously make withdrawals and perform basic functions on their accounts (change PIN, etc.). Training on the payment mechanism needs to be reinforced significantly to minimize fraud as well as to maximize the financial-inclusion benefits of the accounts. • Appropriateness The social workers are not well placed to offer supplementary training during these on-site payments because of time pressure to complete all the payments within the allotted time (usually one-half day). Beneficiaries general understand the frequency of the payments, which are quarterly, but they only learn the actual date that the roving agent will pass through their village with one- or two-days’ notice. They are communicated this information mainly by social workers, which is problematic, as some evidence suggests that they are not reliably transmitting PFSP messages on time due to a variety of factors often outside of their control, but it will be important to monitor whether and why information is being withheld. 12The list of acceptable identification documents includes the national ID card, passport, driver’s license, voter registration card, and birth certificate for citizens, as well as a foreign national ID card, consular card, or passport for non-citizens. 14 Currently, beneficiaries have no choice of PSP. The PFSP has an exclusive contract with Orange Money for the first two transfers, after which a competitive-bidding process will be used to select another PSP, possible Orange Money again, for another exclusive contract. There is also the question of the inherent appropriateness of mobile money to populations such as the one targeted by the PFSP, since some beneficiaries, due to illiteracy or other factors, may be ill equipped to use payment channel that requires manipulating a phone, and may thus be at heightened risk for fraud. • Rights and Dignity There is a complaints mechanism, but it is still in the preliminary stages of development. Evidence from the first payment shows that the complaints mechanism is highly utilized, suggesting that it is well publicized and accessible. Its chief manifestation at the moment is through a helpline, whose number is printed on the back of the program ID card that was issued to each transfer recipient. Although due to the lack of PSP agents in PFSP localities, most withdrawals are currently performed using special itinerant PSP agents that are dispatched according to the PFSP timetable. It is, however, not necessary that beneficiaries use this program-specific infrastructure if they happen to live in one of the few localities with a PSP agent or they travel to an area where there is one. • Security Training is currently poor, and most beneficiaries do not know how to use their phones independently to perform basic functions, such as changing their PIN code, or making a withdrawal. Current means by which payroll lists are communicated is not secure and open to fraud. No systematic review of fraud risks has been conducted and no systematic effort has been made to establish protocols for mitigating fraud risks. Integration • Financial Inclusion The current method for subsidizing beneficiaries’ withdrawal fees penalizes beneficiaries who choose to withdraw their benefit in multiple smaller tranches—for example, if they want to save or use their account to help with budgeting—which incentivizes beneficiaries to cash out their entire benefit at one time and normalizes this behavior. The fact that the PFSP authentication and monitoring protocols do not require beneficiaries to use program-specific payment infrastructure is a key advantage of the program, as it allows beneficiaries to progressively become more and more integrated in public financial services infrastructure as this infrastructure becomes more developed in their communities. However, the state of PSPs’ current agent networks means that few beneficiaries will be able to take advantages of the savings and budgeting possibilities offered by their accounts unless these agent networks are brought closer to beneficiaries’ villages. (For details of the PFSP strategy on agent network expansion, see box X.) Financial literacy training is also lacking, to allow beneficiaries to take full advantage of their accounts and help them see the advantages of not withdrawing their entire benefit immediately in cash. This will become more and more relevant as the program matures and agents become more and more accessible to beneficiaries. (For details of the PFSP financial inclusion strategy, see box Y.) • GRM The PFSP GRM was operational before the first transfer was made to the pilot group of 5000 households. Over the three-month period following the first payment, 2072 official complaints were filed, of which 97 per cent were successfully resolved. All the received complaints pertained to the payment mechanism, with delays in opening of mobile money accounts (59 percent), delays in receiving the transfer to the account (31 percent), and blocked SIM cards due to incorrect PIN entry (9 percent) being the principle reasons for filing a complaint. Most of these difficulties were due to logistical problems stemming from setting up an entirely new payment mechanism, including some delays on 15 the PSP side. The fact that all complaints related to the payment mechanism and not, for example, to targeting, is likely to do the fact that the GRM had not yet been adequately publicized within the wider community, and so only program beneficiaries were aware of the possibility to file a grievance. Concerning the channel used to make complaints, most official grievances were recorded though the dedicated complaints hotline (whose number was printed on the back of the program ID card). However, many beneficiaries, especially those who do not speak French, did not place the call themselves, instead asking a community leader or local social worker to make the call on their behalf. This experience shows that the GRM is fully operational. The large volume of complaints shows that a large proportion of households were able to express themselves through one of the various available grievance channels, and the 71 per cent resolution rate within four weeks shows that the system worked efficiently to redress many of beneficiaries’ grievances. Compared to international experience, where GRMs have often been added to SPL programs as an afterthought if they are included at all, this be a very positive performance for a pilot GRM of a greenfield SPL program. Lessons Learned from Global Experience International experience shows the positive effects of promoting financial inclusion, where having a financial account increases individual savings, female empowerment, consumption and productive investment of entrepreneurs, and income for the poor, particularly among women.13 “Mobile finance” is an umbrella term that describes the how mobile phones can be used to perform tasks that have traditionally only been able to be done through personal accounts in the formal banking sector account or through services such as Western Union. Some of these tasks include mobile cash transfers, direct- to-vendor payments, as well as a broad range of banking services, such as account-based savings, credit or transaction products. Kenya has used digital mechanisms to extend social protection coverage to the informal sector at scale. The case of Kenya is interesting because of its reliance on a new platform to collect contributions and promote savings through mobile phones. In March 2007, M-PESA was launched by Kenya’s largest cellular network operator, Safaricom, which is partly owned by London-based Vodafone. The project was largely funded by donors from the financial services sector. Since its inception, M-PESA has witnessed an incredibly fast adoption rate and now boasts near 50 percent market penetration. As of the end of December 2016, M-PESA served almost 29.5 million active customers through a network of more than 287,400 agents. It is important to mention that a large portion of M-PESA users have previously been excluded from the formal financial service sector. Through partnerships with local cellular providers, M-PESA has grown to other markets internationally including as Tanzania, Afghanistan, South Africa, India, Romania, Albania, Mozambique, Lesotho, and Egypt. Also, in a recent development, through a new partnership, Kenyan users of M-PESA will be able to transfer money directly to any Western Union location worldwide. In terms of services, the platform can fill the needs satisfied by many traditional banking services. Using their mobile phone, users can benefit from many services, including send money; withdrawing cash at licensed M-PESA agents (many of whom are retailers who also sell SIM cards, airtime and other products); buying airtime; buying goods and services through licensed vendors; paying bills, including paying contributions to NSSF; saving money and requesting loans, through licensed products with partnering banking institutions. All individuals who use M-PESA must have register for the service from a registered agent using their National ID (card or passport) and mobile device. To add cash into their mobile wallet, users deposit cash at an agent shop. The same is to true for withdrawals. The organization business model operates on transaction fees that vary based on amount of transaction and type of transfer (to M-PESA user, to 13Deléchat, et al. 2018. “What Is Driving Women’s Financial Inclusion Across Countries?” IMF Working Paper WP/18/38, International Monetary Fund, Washington, DC. See also: Aportela 1999; Ashraf and others, 2010; Dupas and Robinson 2009; Swamy 2013. 16 unregistered users, and cash withdrawals). Registered businesses who allow users to pay bills through M-PESA cover the full transaction fee (1.5 percent per transaction). Recently the NSSF in Kenya has partnered with a digital provider to fully adopt a cashless payment system, both for contributions and withdrawals, erasing the need to visit an NSSF office. Going forward, employers with more than 20 employees will now be able to make their staff contributions through M- PESA. It is worth noting, while M-PESA dominates the market of mobile money subscriptions, (66 percent of the total number of subscribers as of 2017), there are at least three other smaller mobile money platforms with over ten million users collectively. In addition, the NHIF (National Hospital Insurance Fund) has partnered with a leading telecommunication company in Kenya to provide the informal sector an easier way to pay contributions. The fund stopped cash transactions aiming to reduce utilization rates to NHIF services. It is important to mention that since NHIF started using mobile money to collect contributions, travel time and long lines to be registered/to pay contributions has been reduced given that, one of the main problems that the informal sector had, were the penalty fees (5 times the monthly contribution 14) payed once a member default. Members no longer worry about not being able to pay contributions on time, in fact M-PESA allows individuals to contribute on behalf of others too. It is important to note that, although NHIF in Kenya is mandatory for the formal sector, NHIF is voluntary for the informal sector. Therefore, healthcare coverage is still low in general, but due to M-PESA the informal sector grew significantly between 2009 and 2017. Mobile payments made contributions easier, safer, and quicker. It also reduced workload on the NHIF staff. A study15 shows that out of 60 emergency bills, 36 are paid though M-PESA, fast access to cash leads to fast access to healthcare. These reported bills were not only linked directly to hospital bills, but to the transportation an individual need to get to the health facility. In addition to that, patients can access preventive care through NHIF easier using M-PESA and by going to preventive care patients would seek emergency care less. In addition to that, it is interesting to explore the central connectivity that using M-PESA offer. NHIF’s network interconnection gives access to the members, customers, and stakeholder to the system, no matter how far. A member could contribute while out of town and the transaction would be displayed in all NHIF branches. An interesting part is not only in the display of the transaction itself, but in the information a member submits to the system. Information is transferred then to a data center this is how it becomes accessible to all branches. Consequently, NHIF witnessed important growth over the last five years. Overall membership has grown from 2.8M in 2010 to 5.2M in 2015. Membership in the informal sector has reached 1.7M as of 2015, from around 530,000 in 2010 to 1.7 M, and as of 2015, with an increase in monthly revenues to NHIF from M-PESA. Overall, Mobile Money has effectively supported expanding coverage to the informal sector. Its key advantages include its ability to lower opportunity costs in registration and payment (time saved, ease of services), improve access to health care, and in the case of health insurance, reduce out-of-pocket payments for vulnerable workers. 14 Case Study: Kenya National Hospital Insurance Fund (NHIF) Premium Collection for the Informal Sector- USAID https://www.hfgproject.org/wp-content/uploads/2014/06/14-1022-HFG-Mobile-Money-Case-Study-NHIF-mpesa.pdf 15 Stuart, G. Ph.D., Cohen. M., Ph.D. - Cash In, Cash Out Kenya: The Role of M-PESA in the Lives of Low-Income People http://www.fsassessment.umd.edu/publications/pdfs/cash_in_cash_out_kenya.pdf 17 V. RECOMMENDATIONS This note demonstrates that Côte d’Ivoire has successfully gradually scale-up of digital financial inclusion through mobile money accounts for social protection beneficiaries through an innovative public-private partnership.16 This approach has supported the Government’s efforts in developing a new productive social safety net national program, including livelihoods support and tools for supporting social protection service delivery as a system since 201517 18. As of March 2018, 50,000 beneficiary households (approximately 600,000 individuals) have been receiving targeted cash transfers as of March 2018, exceeding the original target of 35,000 households, and reaching three percent of the poor. The program has also supported a new mobile money account platform for delivering cash transfers and economic inclusion activities to boost earnings among beneficiaries, largely subsistence farmers. An important aspect of the program is its reliance on a social registry information system, housing socioeconomic information on 164,000 households (800,000 individuals, or three percent of the national population) located in 16 regions across a total of over 882 villages, coordinated with the national pension funds and universal health insurance program. This system includes an identification mechanism, which will be enhanced through coordination with the unique biometric identification system project launched in 201919 with the National Identification Office. In the future, given the pension system capitalizes on multiple payment channels, the productive safety nets program likewise aims to develop a competitive process for including more than one payment service provider, where possible. Quality improvement will focus on areas such as payment reconciliation (that is, reporting arrangements and quality assurance regarding payments credited to beneficiary accounts versus withdrawn), coordinating with other government agencies involved in e- government-to-person payments, notably the two national pension funds and tax administration. More specifically, detailed recommendations for improving social protection payment mechanisms, notably of the productive safety net program, in the short, medium and long term are summarized in Table 4. Interoperability, choice and quality-based competition In the short term, if SPL programs wish to give beneficiaries different options for their PSP, programs will have to either process bulk payments separately for each PSP or go through an aggregator. The eventual emergence of a sector-wide interoperable switch will simplify this process for SPL programs and eventually allow them to be truly PSP agnostic. However, it is not clear that SPL programs in Côte d’Ivoire should aspire to offer this kind of choice. Giving SPL program beneficiaries a wide array of choice of PSP is considered best practice, and is common in more mature SPL environments, with that rationale that this choice both allows beneficiaries to choose the service that is best suited to their needs while also encouraging competition among PSPs. In less-mature payment environment such as Côte d’Ivoire, however—where financial infrastructure is not fully developed, especially in rural areas—this “choice” may be a false one if there is no appropriate provider in beneficiaries’ localities for them to choose from. 16 IFC-World Bank. 2018. “Granting Access: Leveraging Social Payments to Expand Digital Financial Inclusion in Côte D’ivoire.” Field Note No. 11, World Bank Group, Washington, DC. 17 World Ban (2015). Project Appraisal Document for Côte d’Ivoire Productive Social Safety Net Project. Washington, DC: World Bank. 18 World Bank (2019). Project Appraisal Document for Côte d’Ivoire Social Protection and Economic Inclusion Project (Additional Financing to Productive Social Safety Net Project). Washington, DC: World Bank. 19World Bank (2019). Project Appraisal Document: West Africa Unique Identification for Regional Integration and Inclusion (WURI) Program (including Côte d’Ivoire), P161329. Washington, DC: World Bank. 18 Monitoring and Accountability It is important that contracts with PSPs offering digital payments should explicitly require a) individual reporting on transfers, and b) that all transactions be individually reconciled before the PSP’s fees are paid. Annex B contains a sample reporting framework to be annexed to a contract with a PSP. Grievance Redress While the overall design of the PFSP GRM is in line with international best practices, the program will want to continue to make improvements to ensure that the GRM is efficient enough to keep pace with the growth of the program during scale-up. In particular, the PFPS should formalize the roles of local leaders in the GRM and training relative actors in their responsibilities (village committees, local political leaders, local social workers, etc.); make grievance processing more efficient (digitization of complaints database, coding of complaints to allow for priority treatment of certain complaint types, standardization of communication with beneficiaries on expected delays and on complaint resolution); continued awareness raising with beneficiaries on possible grievance channels; opening the complaints system to non-beneficiaries of the PFSP (allowing complaints on program targeting, etc.); and reinforcing the capacity of the GRM call center, including implementation of a toll-free complaints hotline, and support for the most common local languages. Conclusion Overall, it may be wise in the short and medium term for SPL programs to focus on developing partnerships with a small number of high-potential PSPs, to collaborate with them on the development of the financial infrastructure—networks of agents and cash points—that will pay dividends for SPL program delivery down the road. 19 Table 4 Recommendations for Evolution of the PFSP Payment Delivery Mechanism Short term Medium term Long term Subsidize the cost of and assist beneficiaries in acquiring a Work with ONI on a system to give PFSP beneficiaries national ID cards en Liaise with ONI during rollout of the RNPP such at by 2020 all PFSP Cost of Access national ID card (for those who do not have one). masse. beneficiaries are systematically and automatically enrolled in the Work with PSP to facilitate provision of a low-value, relaxed- Liaise with relevant stakeholders to clarify regulation on low-value accounts at national and civil registries as an integral part of PFSP enrolment. KYC account to PFSP beneficiaries. regulatory level and seek a potential program-level exception to KYC requirements for SIM cards. Work with current PSP to deploy additional itinerant field Begin working with other two mobile money operators, giving beneficiaries a Branch out to banks, MFIs and other relevant PSPs for those agents to reduce wait times. choice between multiple services. beneficiaries who prefer them. Work with current PSP to encourage it to expand its agent Work with all PSPs to encourage them to expand their agent networks to Permanent PSP agent networks in beneficiaries’ localities allow networks to beneficiary localities. (For those localities that beneficiary localities. (For those localities that do not immediately offer agents on-demand withdrawals, reducing wait times to negligible levels. do not immediately offer agents a viable business case, a viable business case, continued reliance on itinerant agents may be necessary Stay abreast of developments in the Ivoirian DFS market and continued reliance on itinerant agents may be necessary in in the short/medium term.) integrate mobile-enabled savings-and-loans and insurance the short/medium term.) Depending on the conclusions of the evaluation of the pilot phase, consider products and other new innovations that are relevant to PFSP Improve training of beneficiaries on use of payment channel expanding to PSPs not based on mobile, which maybe be more appropriate for beneficiaries as they become available. and ensure that this training is adapted to the specific needs certain types of beneficiaries (illiterate, elderly, etc.). Coordinate with relevant stakeholders/donors in the of the PFSP population. development of a national payment switch that will allow Improve communication with beneficiaries on the seamless integration of multiple PSPs, delivering maximum Appropriateness parameters of the PFSP. It is important to diversify beneficiary choice. Accessibility communications channels (posters, radio, SMS, robo-calling, through community leaders, through social workers, etc.) to mitigate the risks of communicating through any one channel. Improve communication about GRM and encourage Use feedback from pilot to review GRM processes and make improvements in Give local communities and meaningful and clearly defined role Rights and Dignity beneficiaries to voice complaints. case-management processes. in the GRM. Foster linkage to integrate the GRM with the broader Monitor exclusion closely in M&E design and take corrective Work with PSP on joint case-management process for payment-related justice system. Publish comprehensive GRM manual publicly. action as necessary. If any beneficiary households are complaints, including data sharing on complaints. unable to access their benefits due to the payment channel Conduct consultations to ask local communities what role they wish to have in (no access, unable to use), then corrective action is the GRM and integrate them as appropriate. necessary. Work with PSP to identify inefficient processes leading to Conduct systematic review of risks leading to payment delays and develop Ensure regular updating and adaptation of the risk review and payment delays. mitigating measures to manage these risks. mitigating measures. Develop M&E tools to allow payment delays to be tracked Work directly with PSPs to better understand how PFSP payments can be Working with other stakeholders, work toward build a mobile- Governance Reliability reliably at the transaction level. leveraged to ensure a viable business case the maximal number of PSP agents. money ecosystem in beneficiary villages, to ensure a clear and For localities where no viable business case exists, contemplate a subsidy to the sustainable business case for PSP agents. PSP in the short/medium term. Work with PSP to improve quality and timeliness of Codify staff roles and responsibilities clearly in a comprehensive operational Integrate PFSP and PSP MIS such that reporting is automatic and Robustness transaction-level reporting. manual and ensure that all staff, including field staff, understand their role. in real time. As program matures, increase independence of internal M&E to improve data reliability. 20 Short term Medium term Long term Develop M&E tools to automatically flag potentially When developing PFSP MIS, provide for functionality to automate creation of Integrate PFDP and PSP MIS to enable straight-through payment fraudulent activity from payment reports and GRM data and payroll lists, with checks to ensure data quality, and generate a secure file for processing. ensure that PSP agents are also monitored for fraudulent transmission to the PSP. Regularly update review of fraud risk and update fraud-control activity. Conduct systematic review of risks leading to fraud and develop any fraud- measures as appropriate. Reinforce training for beneficiaries on PIN codes, including control measures identified as lacking. on how (and when) to change them. Engage with ONI as a stakeholder to ensure that the new national registry Security facilitates PFSP’s fraud-prevention needs (ability to remove duplicate beneficiaries based on biometric data, etc.). Work with PSPs to expand network of permanent agents to Work with all PSPs to encourage them to expand their agent networks to Permanent PSP agent networks in beneficiaries’ localities allow beneficiaries’ localities, reducing program reliance on beneficiary localities reducing program reliance on itinerant agents. on-demand withdrawals, reducing wait times to negligible levels. itinerant agents. Develop financial-literacy training training for beneficiaries on using their As additional financial services become accessible through the Work with PSPs to devise a fee structure whereby accounts as a tool for managing their household budgets, with a special PSPs used by the PFSP, expand the nature of the financial-literacy Financial Inclusion beneficiaries are not penalized for making multiple, smaller emphasis on medium-term savings. Use regular household visits from social training to include borrowing, investing, insurance, etc. withdrawals. workers to systematically reinforce key messages from this training as well as Stay abreast of developments in the Ivoirian DFS market and provide individualized coaching on household budgeting. integrate mobile-enabled savings-and-loans or insurance Work with PSPs to offer earmarked- and commitment-savings sub-accounts to products and other new innovations that are relevant to PFSP beneficiaries and other relevant budgeting tools. Incorporate these into the beneficiaries as they become available. training. Subsidize the cost of and assist beneficiaries in acquiring a Work with ONI on a system to give PFSP beneficiaries national ID cards en Liaise with ONI during rollout of the national registry such at by national ID card (for those who do not have one). masse. 2020 all PFSP beneficiaries are systematically and automatically Identify other programs that deliver payments to similar Establish dialogue between all country SPL programs on a shared emergency- enrolled in the national and civil registries as an integral part of target populations and begin exploring synergies to payment system. Even if payments are outsourced to PSPs, emergency PFSP enrolment. coordinating payment delivery. payments may have to be handled internally by the government, leading to All country SPL programs are fully integrated into the national Coordinate on any internal processes used, for example to significant economies of scale to establishing common protocols across population registry, using its unique identifier and biometric data open accounts or process payments, and share tools programs. for all identification purposes (including account opening and between programs. Leverage different SPL programs to improve local digital payment ecosystems. payment authentication). Coordination If multiple programs can pay into the same account—including perhaps non- Review SPL procurement strategy for PSPs and try to combine Integration SPL payments such as government salaries—then this will only improve PSP multiple programs’ procurements into a single RFP to decrease business models, lowering prices of payments for the government and costs. improving the quality of agent networks. Source: World Bank staff compilation. 21 VI. ANNEXES A. Fundamental Performance Indicators for Payment Mechanisms INDICATEUR PAR COMPOSANTE Mode d'obtention/definition Unité Dénombrement sur declaration du ménage et 1-Nombre moyenne de paiement reçu par le Nombre vérification des portable (message de ménage (moyenne) notification) ménage Nombre de transfers effectués pendant le 2-Pourcentage de paiements ayant été effectué dernier paiement avec succès (selon le % avec succès rapport de l'operateur de paiement) divisé par le nombre de bénéficiaires à payer Pourcentage de ménages pour lequel la fiche 3-Etat d'avancement du suivi des paiements de suivi de paiement a bien été soumise par % les TS après le dernier paiement Le temps mis par l'opérateur de paiement entre le temps du dépôt du transfert sur le 4-Delai de transmission des rapport de l'opérateur compte du bénéficiaire et la transmission des jours de paiement rapport, en jours (Etat de paiement/ PAIEMENT Reporting) (dernier paiement) Pourcentage de ménages bénéficiaires inclus 5-Integralité du rapport de l'opérateur de dans le rapport de paiement soumis par % paiement l'operateur de paiement (dernier paiement) Pourcentage d'accord entre de deux bases de 6-Accord entre les données fournies par Orange suivi des paiements (fournie par Orange et % money et le suivi effectué par l'UGP collecté par l'UGP) (dernier paiement) Pourcentage de ceux qui n'ont pas retiré leur 7-Motif de non retrait du transfert % dernier paiement (a) Volonté de faire une épargne (b) Cas de maladie / invalidité / voyage (c) Manque de liquidité au PDV OM (d) Non informé de la récéption du transfert (e) Perte du matériel (portable, puce, etc.) (f) Oublie du code sécret (g) Manque de la carte d'identité Pourcentage de recipiendaires qui arrivent à CONDITIONS PREALABLES 8-Pourcentage de recipiendaires sachant effectuer une transaction simple de manière % manipuler leur portable de manière satisfaisante independante Il s'agit de l'effectif des bénéficiaires ayant des documents d'identité (Carte Nationale 9-Pourcentage de ménage ayant des pièces d'Identité ou attestation d'identité pour les % d'identité valides nationaux, carte consulaire ou passeport ou titre de séjour pour les non nationaux) dont la date de validité n'est pas encore arrivée Nombre de ménages ayant porté plainte (au 10-Pourcentage des ménages bénéficiaires ayant moins une fois) divisé par le nombre total de % porté plainte ménages ayant réçu au moins un paiement Nombre de plaintes pour lesquelles une PLAINTES solution satisfaisante a été proposée au 11-Pourcentage des plaintes ayant été resolu % bénéficiaire dans le delai prevu divisé par le Nombre de plaintes réçus 12-Les types de plainte recurrente/ fréquente Pourcentage de ceux qui one porté plainte % (a) Modalité 1 (b) Modalité 2 (c) … Source: Republic of Côte d’Ivoire, Productive Social Safety Net / Social Protection and Economic Inclusion Project, 2015 to present. 22 B. Good Practice for M&E of Digital Payments Deliverable Report must include… Reporting Required remediation measures frequency 5000 project • Extract from database of Once, at latest If <99.5% of project beneficiaries are beneficiaries OM clients (excel/csv), 3 weeks registered for OM service, then OM must successfully showing customer names before first register remaining beneficiaries and registered to OM and phone numbers for all scheduled resubmit report service project beneficiaries payment • For any beneficiaries that were not registered successfully, an explanation of what went wrong in each case Agents are alerted • Details of communication No later than 3 If not all agents identified by the UGP as to expect higher with agents, including date days before participating in the PFSP program have than normal of communication each been communicated with or not all demand for cash • Summary of the agent’s scheduled agents have submitted a liquidity- withdrawals and strategy to manage payment management strategy, then both must have developed a liquidity in face of occur and the report must be credible plan to increased demand for cash resubmitted. manage their withdrawals liquidity Transfers made • Transaction statement of No later than 5 If transfer failure rate >0.5%, then the successfully and the PFSP OM account, days following appropriate remedial actions must be on time into listing all transactions in each payment taken to ensure delivery of transfer, and beneficiary and out of this account the report must be resubmitted. accounts (including all transfers into the beneficiary accounts, including timestamps, beneficiary name, and beneficiary ID), including any and all failed transfers along with their error codes • For any failed transfers, an explanation of what went wrong in each case Beneficiaries who EITHER Monthly, If the PFSP UGP, in analyzing beneficiary wish to make cash starting no complaints delivered through its withdrawals from • Individual customer later than 30 Grievance Redress Mechanism, transaction statements their accounts are days following identifies geographic zones and/or (excel/csv) for the able to, on the first specific agents who are consistently individual customer payment into unable to meet beneficiaries’ demand demand account of each project the first for cash withdrawals, then it will notify beneficiary, listing all beneficiary OM of the problem agent. In response, transactions (deposits, account OM must take at least one of the withdrawals and transfers, following remedial measures within 10 and including timestamps, days of being notified by the PFSP UGP of and beneficiary ID) in and the problem agent: out of the beneficiary’s account • Assist the existing agent to OR manage liquidity such that the agent is able to satisfy • List of all transactions demand for withdrawals (excel/csv) listing all • Dispatch a mobile agent to go transactions (deposits, to each beneficiary village in withdrawals and transfers, this zone, at a fixed time and and including timestamps, date, at least once per month, and beneficiary ID) in and until the immobile agent is out of the beneficiary’s able to satisfy demand for account withdrawals 23 • Deploy a new agent in this zone who is better able to address beneficiary demand than the existing agent OM will be required to submit a report on all such remedial actions taken within 15 days of receiving notification of the problem agent by the UGP. Remedial measures shall only be considered successful if the number of complaints delivered through the GRM drop to within 1 standard deviation of the overall mean of all agents for the project, as calculated by the UGP. If not considered successful, then additional remedial measures will be required. Source: World Bank staff compilation. 24 C. Good Practice for GRMs Good Practice Description Example Programs20 Multiple entry The GRM should not limit complainants to a • Philippines points single point of entry for expressing their • India—Tamil Nadu Empowerment and Poverty grievances. As an example, a GRM might Reduction Project (TNEPRP) allow complainants to file grievances either • Upper Cisokan Pumped Hydroelectrical Power with local project staff or via a telephone Project hotline. This approach helps ensure that PAPs • (Indonesia) have recourse to a grievance channel that is • Tanzania: Zanzibar Urban Services Project appropriate to their needs. It also helps avoid • Lake Victoria Phase II, APL 2 conflicts of interest in cases of corruption or • Indonesia: Upper Cisokan Pumped Storage malfeasance: if the corrupt actors are • China: Hubei Yiba Highway Project implicated in the GRM, then they may have an • Vietnam: Trung Son Hydro interest in preventing the complaint from • Azerbaijan: Highway 3 being registered. • Kazakhstan: South West Roads Project • Kyrgyz Republic: National Road Rehabilitation (Osh transport) • Colombia: Rio Bogota Environment Infrastructure • Southern West Bank Solid Waste Management • India: Vishnugad Hydropower • India: National Ganga River Basin Project Make use of Various programs have tried to create • India—Tamil Nadu Empowerment and Poverty existing social project-specific complaint channels, for Reduction Project (TNEPRP) structures example using SMS, only to discover that • Hubei Yiba Highway Project (China) PAPs preferred to complain directly to the • Tanzania: Zanzibar Urban Services Project same local community leadership structures • Lake Victoria Phase II, APL 2 that they were used to communicating with. Programs should make an effort to incorporate these local leaders in the program GRM. Allow for In certain cases, a PAP might wish to file a • Kyrgyz Republic: National Road Rehabilitation anonymous grievance anonymously, and the GRM should (Osh transport) complaints accommodate this wish. This is especially • Kazakhstan SW Road Project relevant in sensitive programs where the PAP • Azerbaijan Highway 3 fear reprisal or retaliation for complaining, for • China: Hubei Yiba Highway Project example if there are ethnic tensions at play. • India—Tamil Nadu Empowerment and Poverty Reduction Project (TNEPRP) • India: Second Madhya Pradesh District Poverty Initiatives Ability to file It is important to ensure that PAPs can • Pakistan–Flood Emergency Cash Transfer Project complaints in express grievances in the language they are • Kazakhstan SW Road Project local languages most comfortable with. If the local context makes central management of complaints in all local languages impractical, then local leaders and project staff can often be employed as intermediaries. Accessible Since the vulnerable groups targeted by many • India—Tamil Nadu Empowerment and Poverty communication SP programs may not be fully literate or Reduction Project (TNEPRP) conversant in the national language, it is • Uganda: Electricity Sector Development Project important that the communication strategy • Indonesia: Upper Cisokan Pumped Storage for the GRM and the options for filing a • China: Hubei Yiba Highway Project complaint take into account this limitation. • Vietnam: Trung Son Hydro For example, text-based communication can • Azerbaijan: Highway 3 be eschewed in favor of illustrations or local- • Kazakhstan: South West Roads Project language radio broadcasts. • Kyrgyz Republic: National Road Rehabilitation (Osh transport) • Southern West Bank Solid Waste Management • India: Vishnugad Hydropower 20 Marie Brown (2014), “Global Review of Grievance Redress Mechanisms in World Bank Projects,” World Bank Group. 25 Good Practice Description Example Programs20 • India: National Ganga River Basin Project • India: Eastern Dedicated Freight Corridor • Pakistan: Floods Emergency Cash Transfer Project Issuance of a Each complaint should be assigned a case • Philippines case number number that can be used by both the program • India—Tamil Nadu Empowerment and Poverty staff and the complainant to track progress Reduction Project (TNEPRP) toward resolution. • Tanzania: Zanzibar Urban Services Project • Indonesia: Upper Cisokan Pumped Storage • Lao PDR: Nam Theun 2 Hydropower • China: Hubei Yiba Highway Project • Vietnam: Trung Son Hydro • Azerbaijan: Highway 3 • Kazakhstan: South West Roads Project • Kyrgyz Republic: National Road Rehabilitation (Osh transport) • Colombia: Rio Bogota Environment Infrastructure • Southern West Bank Solid Waste Management • India: Mumbai Urban Transport Project (MUTP) • India: Vishnugad Hydropower • India: National Ganga River Basin Project • India: Tamil Nadu Empowerment Project • India: Eastern Dedicated Freight Corridor • Pakistan: Floods Emergency Cash Transfer Project Categorization Complaints should be categorized with • Kazakhstan SW Road Project and routing of routing codes in order to enable more • Pakistan Flood Emergency Cash Transfer Project grievances efficient processing. For example, complaints • Uganda: Electricity Sector Development Project that can be resolved quickly by low-level • Indonesia: Upper Cisokan Pumped Storage program staff could be in one category, while • Vietnam: Trung Son Hydro complaints that require input from high-level • India: National Ganga River Basin Project program staff or take more time to process • India: Vishnugad Hydropower can be considered separately. Categorizing • India: Eastern Dedicated Freight Corridor complaints not only enables more efficient • Pakistan: Floods Emergency Cash Transfer Project processing, but can also permit better communication with PAPs about wait times, as different types of complaints likely take different amounts of time to process. Written All PAPs should receive a written confirmation • Kazakhstan: South West Roads Project confirmations that their complaint has been received. This • China: Hubei Yiba Highway Project written record important even for illiterate • India: National Ganga River Basin Project PAPs. This confirmation should include the • India: Vishnugad Hydropower case number as well as the anticipated • India: Eastern Dedicated Freight Corridor treatment delays. For example, an automated • Pakistan: Floods Emergency Cash Transfer Project SMS could be sent to the PAP once the complaint had been recorded into the complaint tracking system. Once the case is closed, its resolution should also be communicated in writing (in addition to other communication mechanisms that are more suitable to illiterate PAPs). Transparency PAPs should understand the anticipated wait • Philippines about expected time before their complaint resolved. • India—Tamil Nadu Empowerment and Poverty delays Categorizing complaints according to wait Reduction Project (TNEPRP) times at entry into the GRM can help facilitate • India: National Ganga River Basin Project this communication. • India: Vishnugad Hydropower • Kazakhstan: South West Roads Project • Kyrgyz Republic: National Road Rehabilitation (Osh transport) • Southern West Bank Solid Waste Management • Colombia: Rio Bogota Environment Infrastructure • Turkey: ECSEE APL#6 • Azerbaijan: Highway 3 26 Good Practice Description Example Programs20 • China: Hubei Yiba Highway Project • Vietnam: Trung Son Hydro • Indonesia: Upper Cisokan Pumped Storage • Tanzania: Zanzibar Urban Services Project Training of all Everyone implicated in the GRM should be • Nigeria actors trained in their role and how it fits into the • Philippines overall GRM. This training should not be limited to program staff, but also any contractors involved in the GRM (such as the PSP) as well as other actors (community leaders, etc.). Electronic data In order to allow complaints to be tracked and • Philippines management processed effectively, they should be • India—Tamil Nadu Empowerment and Poverty managed—at least once they reach a certain Reduction Project (TNEPRP) level—in a GRM tracking database. Data entry • Tanzania: Zanzibar Urban Services Project can be performed either centrally or in the • Indonesia: Upper Cisokan Pumped Storage field, using tablets equipped with an • Lao PDR: Nam Theun 2 Hydropower electronic complaint form. • China: Hubei Yiba Highway Project • Vietnam: Trung Son Hydro • Azerbaijan: Highway 3 • Kazakhstan: South West Roads Project • Kyrgyz Republic: National Road Rehabilitation (Osh transport) • Colombia: Rio Bogota Environment Infrastructure • Southern West Bank Solid Waste Management • India: Mumbai Urban Transport Project (MUTP) • India: Vishnugad Hydropower • India: National Ganga River Basin Project • India: Tamil Nadu Empowerment Project • India: Eastern Dedicated Freight Corridor • Pakistan: Floods Emergency Cash Transfer Project Source: World Bank staff compilation. 27 D. Example Survey Instrument for Payment Monitoring – PFSP A. Information générale Rapport à Intitulé Modalité (réponse) Identifiant de l’enquêteur Enquêteur Nom, prénoms de l’enquêteur Date de l’entretien Région Village Ménage Identifiant du ménage Nom, prénoms du chef de ménage Nom, prénoms du récipiendaire Récipiendaire Numéro de téléphone (sur lequel il reçoit le transfert) Genre du récipiendaire Masculin |__| Féminin|__| B. FICHE DE SUIVI DES PAIEMENTS Intitulé Modalité (réponse) 1-Statut du bénéficiaire Transfert reçu|__| Transfert non reçu|__| 1.1 Si transfert non reçu, pour quel motif ? 1-Transfert non effectué 2-Carte SIM égarée ou bloquée 3-Pas de réception de notification 4-Autre Si autre, à préciser 1.2 Si transfert reçu, montant notifié au récipiendaire 1.3 Si transfert reçu, date de réception de la notification 1.4 Si transfert reçu, le récipiendaire a-t-il effectué un retrait ? Oui |__| Non|__| 1.4.2. Si oui : Combien le récipiendaire a-t-il souhaité retirer ? 1.4.3. Si oui : Le récipiendaire a-t-il pu retirer le montant souhaité ? Oui |__| Non|__| 1.4.1 Si oui : Date de retrait 1.4.4 Si non : Quels en sont les motifs ? 1-Volonté de faire une épargne 2-Incapacité à effectuer les retraits 2.1 cas de maladie/Invalidité/voyage 2.2 Indisponibilité de l’agent au point de vente (liquidité) 2.3 Service de mauvaise qualité au point de vente 2.4 Problème de réseau 2.5 Problème d’identification/authentification 2.6 Récipiendaire non informé 2.7 Perte du matériel 28 2.8 Non maitrise du service Orange money 2.9 Difficulté avec le code secret 2.10 Ne veut pas payer des frais de transaction élevés 2.11 Autre problème avec Orange money. Précisez :…………………………………. 1.5.1 Le récipiendaire a-t-il consulté son solde avant l’opération de Oui |__| Non|__| retrait ? 1.5.2 Le récipiendaire a-t-il consulté son solde après l’opération de Oui |__| Non|__| retrait ? 2.1. A quelle date le récipiendaire pense-t-il recevoir le paiement suivant ? 2.2. Quel doit être le montant de ce paiement ? 2.3. Le récipiendaire aurait-il eu besoin de l’appui d’une tierce personne Oui |__| Non|__| pour effectuer l’utilisation du service Orange money 2.4. Le récipiendaire maitrise-t-il l’utilisation du portable (administration Oui |__| Non|__| de teste de capacité de manipulation) 3.1 Avez-vous connaissance de la helpline/mécanisme de gestion des réclamations ? 3.2 Savez-vous comment l’utiliser ? 3.3. Souhaiteriez-vous enregistrer une réclamation maintenant ? Oui |__| Non|__| Si oui : Ouvrez la fiche d’enregistrement des plaintes Source: Republic of Côte d’Ivoire, Productive Social Safety Net / Social Protection and Economic Inclusion Project. 29 E. Example Grievance Form – PFSP FICHE D’ENREGISTREMENT DES PLAINTES A- REFERENCE GEOGRAPHIQUE PLAINTE N°………du ………………20……… REGION :…………………………………………………………………………………………………… DEPARTEMENT :…………………………………………………………………………………………… SOUS-PREFECTURE : …………………………… LOCALITE / VILLAGE : ……………................. NIVEAU DE TRAITEMENT : □ Comité Villageois (CV) □Travailleur Social □UGP B- IDENTIFICATION DU PLAIGNANT NOM : ………………………………………. PRENOMS : ………………………………………………………. SEXE : □ Masculin □ Feminin LIEU DE RESIDENCE : ………………………………………………………………………………………… CONTACT (Tél) : ……………………………………………………………………………………………………. STATUT □ Bénéficiaire □ Non bénéficiaire Si bénéficiaire, N° d’Identifiant : ………………………………………………….. Si Non bénéficiaire : □ Lien avec un bénéficiaire □ Pas de lien avec un bénéficiaire Si « NON » préciser Nom et Prénoms du bénéficiaire : ………………………………………………….. MOTIF DE LA PLAINTE : CIBLAGE DES MENAGES IDENTIFICATION DE MENAGES ELIGIBLES □ Exclusion d’un ménage dans le processus □ Exclusion du ménage dans le processus □ Inclusion d’un ménage dans le processus □ Inclusion d’un ménage dans le processus □ Omission d’un ménage dans le processus □ Omission du ménage dans le processus □ Autre à préciser :…………………………………… □ Omission du récipiendaire du ménage ………………………………......................................... □ Omission du suppléant dans le processus …………………………………………………………….. □ Autre à préciser :…………………………………… ………………………………......................................... …………………………………………………………….. (continued) 30 PAIEMENT DU BENEFICIAIRE MESURES D’ACCOMPAGNEMENT □ Exclusion du bénéficiaire □ Stigmatisation □ Inclusion du bénéficiaire □ Absence des agents de mise en œuvre des MA □ Omission du bénéficiaire □ Manque d’intrants-matériel-équipement □ Corruption □ Retard des agents de mise en œuvre des MA □ Non perception du transféré □ Non-respect du calendrier proposé □ Non perception du montant total à transférer □ Manque de courtoisie des agents □ Non reversement du moins perçu du moi précèdent □ Non-respect des programmes et modules □ Retard dans le transfert □ Manque de suivi des bénéficiaires -MA □ Mode de paiement non adapté □ Lourdeur dans le processus de traitement des plaintes □ Absence de point de paiement /de structure de paiement □ Autre à préciser :…………………………………… AUTRES ………………………………......................................... □ Retard de paiement du point de vente □ Stigmatisation …………………………………………………………….. □ Retard des agents de paiement □ Fraude sur l’identité □ Eloignement du point de paiement par rapport à la localité □ Conflit intercommunautaire □ Refus de paiement par le point de vente □ Conflit entre bénéficiaires □ Espace non propice au paiement □ Disfonctionnement du CV □ Disfonctionnement du Comité Communal □ Intimidation des agents de paiement □ Disfonctionnement du Comité Régional □ perte/vol de la carte SIM □ Inégalement de traitement des bénéficiaires □ Oublie/perte du mot de pass □ Manque de courtoisie des agents de paiement /MA □ Décès du récipiendaire du ménage □ Négligence constatée dans le traitement des plaintes □ Décès du suppléant du ménage □ Lourdeur dans le processus de traitement des plaintes □ Difficulté à manipuler le terminal/potable □ Autre à préciser :…………………………………… □ Manque de couverture réseau mobile du prestataire ………………………………......................................... …………………………………………………………….. □ Autre à préciser :…………………………………… ………………………………......................................... …………………………………………………………….. Description sommaire de la plainte ……………………………………………………………………………………………………………………………………………. ……………………………………………………………………………………………………………………………………………. ……………………………………………………………………………………………………………………………………………. ……………………………………………………………………………………………………………………………………………. Source: Republic of Côte d’Ivoire, Productive Social Safety Net / Social Protection and Economic Inclusion Project, 2015 to present. 31 F. ISPA Payment Assessment Tool Overview Source: Interagency Social Protection Assessment partner organizations, World Bank; http://www.ISPATools.org. 32 33