Understanding How Consumer Risks in Digital Social Payments Can Erode Their Financial Inclusion Potential

Low-income recipients of cash transfers—whether government to person (G2P) or donor to person (D2P), and whether conditional or unconditional—increasingly receive their payments digitally. This digitization trend is expected to continue. The value of electronic transfers that are delivered into store-of-value accounts and accessible via debit cards or mobile money wallets, referred to here as “digital social payments,” is estimated to more than triple between 2010 and 2017 to over US$194 billion (Riecke 2014).

Applying the consumer risk lens outlined in McKee, Kaffenberger, and Zimmerman (2015), this Brief focuses on the particular experiences of low-income and vulnerable recipients. It reviews existing evidence from DSP programs in 12 emerging markets to identify the most common and consequential consumer risks and challenges experienced in their design and deployment. 4 These consumer risks include (i) inability to transact due to network or service unreliability; (ii) insufficient agent or automatic teller machine (ATM) liquidity; (iii) complex user interfaces and payment processes; (iv) poor or no recourse mechanism; and (v) fraud that targets the recipient. Drawing on promising practices by governments, development agencies, and providers, it then recommends that DSP programs and their payment service providers should consider three critical basics-reliability, communication, and monitoring-to better mitigate these risks. While these three elements of "doing digital finance right" are basics that apply to any DFS offering, getting them right in a DSPs context will provide a solid foundation for ensuring that these programs achieve both their social protection goals and their financial inclusion potential.

Five Most Common and Consequential Consumer Risks
Faced by DSP Recipients 1. Inability to transact due to network downtime or service unreliability Many DSP programs are targeted to populations in poor and mostly remote locations, where mobile network coverage is often weak. As a result, recipients experience frequent network connectivity problems for point-of-sale (POS) devices and mobile phones. DSP recipients trying to access their periodic payments suffer acutely from such unreliability.
For example, in three digital cash and voucher programs of the World Food Programme (WFP) in Kenya and Lebanon, pervasive network failures and insufficient connectivity exposed recipients to financial loss and potential harm in a variety of ways. Interrupted transactions left payments in limbo, and when networks were down entirely, it became common for recipients to leave their card and personal identification number (PIN) with agents or merchants to complete the transaction later on, resulting in risks and reported experiences of inappropriate, and even fraudulent, agent behavior (WFP 2016;El-Huni 2014). Network or infrastructure failure was also a frequent challenge for G2P recipients in Nigeria and in the HelpAge Program in Bangladesh, which uses the bKash network to transfer funds into mobile wallets (Adewole 2015;InterMedia Africa 2015;Islam and Woodard 2014b). In the Philippines, Pantawid Pamilya recipients not only had network connectivity problems but also inconsistent and unreliable service hours of agents, even when those agents were designated specifically for the DSP disbursement (Zimmerman and Bohling 2015). Network outages and finding an agent's shop unattended added to the time recipients had to wait for cashing-out and being able to use their transfers (Islam and Woodard 2014b).
Because digital social payments (DSPs) recipients are a fast-growing, yet often overlooked, digital financial services (DFS) segment, acknowledging and addressing the most common and consequential consumer risks they face should be a priority for any program or provider seeking to unlock the potential benefits of DSPs for the poor.

Insufficient agent or ATM liquidity
DSPs are usually transferred in bulk, with most recipients typically withdrawing all of their money on a single day. This creates heavy pressure on the access point to meet liquidity demands, which is a particular challenge in remote and less secure areas. Consequently, recipients often line up and wait for several hours to collect or access their payments or are even sent back home to repeat the journey another day. This risk appears to perpetuate a vicious cycle: the shortage of liquidity erodes recipients' trust and confidence in the system, which creates an imperative to withdraw all of a payment at once and as soon as it's deposited, which thereby exacerbates liquidity issues at cash-out points.
For example, in a mystery shopping exercise carried out in WFP Kenya's Cash for Assets (CfA) Program, 21 percent of the recipients were unable to cash-out the desired amount of their transfers due to insufficient agent liquidity (WFP 2016). In another mobile cash transfers program by WFP Kenya, recipients perceived liquidity constraints as a cost factor that made the digital payments useless given the distances to and limited selection of M-Pesa agents. In the Philippines, where the payment per recipient can vary from cycle to cycle, agents often lack sufficient change to pay each recipient the exact amount of his or her withdrawal and, therefore, pay out the transfers to arbitrarily assigned groups, leaving it up to recipients to find the required change to split the payments among them, or the agent may offer them inexpensive merchandise, such as little candies, to make up the difference (Zimmerman and Bohling 2015).

Complex user interfaces and payment processes
Complex interfaces and complicated processes-which increase the likelihood of errors and losses from either incorrect transactions or recurring timeouts due to limited transaction times-create risks and a poor user experience for all types of DFS users. DSP recipients are even more likely to be negatively affected: besides being among the most vulnerable and least literate consumer segments, they are often new to and initially uncomfortable with the digital payment system, including the technology and the numerous steps required to access or use the payments (WFP 2016). Beyond the inconvenience of frequent transaction failures and repetition, these issues increase the risk of recipients being charged extra fees by agents or being victims of fraud as they have to ask others for assistance. Such issues can significantly reduce trust and make the system seem inconvenient to the recipient, making it less likely that this initial interaction with DFS will lead to broader use of formal (digital) financial services.
A CGAP study of four newly digitized social payment schemes in Haiti, Kenya, the Philippines, and Uganda found that recipients exhibited an astonishingly low understanding of how the payment schemes, and particularly the digital payment aspect of it, worked (Zimmerman, Bohling, and Rotman Parker 2014). In a mobile voucher program in Nepal, the complicated process and user interface was a major problem for recipients. Recipients had only five minutes to complete their SMS-based voucher redemption and if a timeout occurred before the transaction was finalized, the recipient had to start all over again (Murray 2013). 5 Moreover, the SMS-based system did not allow the use of local script and language, which was another challenge for recipients who were unable to read and understand the messages. 6 Two months after introducing mobile payments to WFP Kenya food-aid recipients, a survey by CGAP and WFP found that there were still recipients with inactivated SIM cards.

Poor or no recourse mechanism
Recourse mechanisms, such as complaints, queries, and dispute resolution, are another particularly weak spot in DSP programs (Zimmerman, Bohling, and Rotman Parker 2014). Recipients often don't know about or are confused about recourse and support options, making it difficult for them to solve problems or get answers to questions they have about their payments. Recipients from several programs also worried that if they complained they could lose their transfers, a misperception that made them reluctant to report problems. Even where customer support or grievance hotlines do exist, recipients are either not aware of them, or have had negative experiences, such as being kept for a long time in automated waiting lines or having their call dropped before they could speak to anyone. This usually results in frustration and, in some cases, financial loss incurred from spending their airtime to make the call. In Bangladesh, for example, recipients of mobile training allowances complained that when they wanted to talk to a bKash support agent they often gave up-thinking that it was not worth waiting to express a problem that might not even be resolved (Islam and Woodard 2014a).
Recipients commonly state a preference for face-to-face problem solving-which is also rooted in cultural and personal biases-and often turn to agents or merchants for help. However, these parties cannot always adequately address problems or questions because they lack the respective training, access to a faster-to-reach support hotline, and sufficient business incentives to make time to help customers-especially to help social payment recipients who may have their own particular questions and needs (McKee, Kaffenberger, and Zimmerman 2015). In WFP Kenya's CfA Program, 69 percent of the recipients who reported a problem during their mystery shopping visit said the agent did not address the problem to their satisfaction. In 29 percent of these visits, the agent called the bank's hotline (only half reached a customer service representative), in 50 percent the shopper was given contact information so that they could follow-up on their own, and in 21 percent the recipient did not get any support at all (WFP 2016).

Fraud that targets the recipient
DSP recipients are particularly vulnerable to fraud such as unauthorized fees, price hikes at merchants, and skimming of payments (i.e., illegally retaining a portion). For example, during the introduction of WFP Kenya's Cash Lite Program, 72 percent of recipients paying at the shops with their bank cards were charged additional fees or higher prices than other customers (WFP 2016). 7 Recipients of digital G2P programs in the Philippines, Uganda, and Nigeria reported incurring extra costs-sometimes referred to as "taxes"-5 Vendors and recipients did not always enter the exact syntax needed to process voucher redemptions. If any messages in the transaction sequence were incorrectly formatted, the user would be forced to try again in a five-minute window, or the entire transaction failed. SMS logs show that 8 percent of users entered their PIN incorrectly and had to restart their transaction (Murray 2013). 6 Only 33 percent of SMS voucher users were able explain how they checked the amount charged for their purchase before entering their PIN (Murray 2013). 7 One year after implementation, another mystery shopping study revealed that 14 percent of the recipients were charged unauthorized fees by merchants.
to withdraw their payments at the agents (Zimmerman, Bohling, and Rotman Parker 2014). 8 In India, 13 percent of surveyed G2P recipients report paying a bribe to access their payments (InterMedia India 2014). 9 Recipients are often unaware of the actual charges and fees associated with their payments, nor do they know the exact amount and frequency of their benefit, leaving them vulnerable to such instances of unfair treatment and fraud. CGAP and WFP Kenya assessed recipients' awareness of transaction fees and found that 62 percent of CfA Program recipients did not know that there were transaction fees, despite paying a transaction fee of at least KSH 50 for each withdrawal (WFP 2016). 10 In the Pacific Financial Inclusion Program (PFIP) in Fiji, 68 percent of recipients were not aware of agent charges and half were not sure whether there were charges or not (Leonard 2011). And in Ghana's LEAP Program, 80 percent of recipients did not know their payment amount, and 85 percent were unaware of the frequency of their payments (Abbey, Odonkor, and Boateng 2014).
PIN protection is another challenge that is particularly prevalent in DSP programs as many recipients share their PIN with agents or third parties and do not enter their PIN into the POS device or mobile phone themselves. It appears that agents and recipients often compromise data protection procedures in this way for the sake of efficiency, especially when recipients come in large numbers on paydays and agents have to serve many clients quickly. 11 In Fiji, for example, PIN-sharing was reported as common practice, particularly by elderly people whose family members or friends picked up their benefits (Leonard 2011). In Nepal, the majority of mobile voucher recipients relied on registered "helpers" to carry out transactions. 12 These helpers entered the PIN in 89 percent of observed SMS and in 37 percent of observed smartphone transactions (Murray 2013). Research on the ARCC II Program in the Democratic Republic of the Congo documented that recipients lacked capacity and understanding to control the process of entering the purchase amount and PIN themselves (Murray and Hove 2014). In WFP Kenya's Cash Lite Program, 36 percent of recipients handed their bank card together with the PIN letter over to the merchant to carry out the POS transaction (WFP 2016). In 73 percent of the test visits in WFP Kenya's CfA Program, agents did not allow recipients to enter their PIN even though 72 percent of these recipients had memorized their PIN and knew how to enter it (WFP 2016).

Three Basics to Mitigate Risks and Open up the Financial Inclusion Gateway of DSPs
While more needs to be done to understand the nature, incidence, and consequences of consumer risks for recipients, three basic principles emerge that would build a more solid foundation for effectively mitigating these risks and, thereby, enabling financial inclusion in DSPs for the poor: reliability, communication, and monitoring. These are critical for consumer risk mitigation in most DFS deployments, and are certainly not the only critical elements of well-functioning DSP schemes. Yet, getting these three basics right will not only help improve the overall functioning of these schemes, but will also help open a potential pathway toward financial inclusion. The following are important points about these basics and examples of existing promising solutions that programs and providers have implemented: 1. Ensure reliability of the payments experience first and foremost. DSPs reach recipients through complex systems that can include agents, merchants, mobile phones, POS devices, and ATMs. Both the value chain of actors and the payment systems rely on strong network connectivity to successfully process real-time digital payments. Up-front risk assessment and contingency planning can help to address some of the aforementioned risks at the design stage, such as (i) making the customer interface more user friendly; (ii) ensuring agent service quality, training, and float and liquidity management; and (iii) clearly defining roles and responsibilities for risk mitigation among actors. For example, MTN Uganda added 15 new network towers-five specifically for its SAGE Program-in previously poorly connected catchment areas (Zimmerman and Bohling 2014). 2. Improve communication channels between recipients and providers. Even more than for typical DFS consumers, social payment recipients require relevant, and often substantial, training and communication first when a new digital payment system is introduced and then continuously after rollout. For more self-control and confidence, recipients need to know their payment amount and frequency. Equally as important (and at times overlooked), they need to know how the system and payment mechanism are supposed to work and where to go if they face problems. This means that responsibilities for specific problems should be clearly defined among those involved in the social payment value chain. A well-organized grievance and complaints mechanism for recipients can also be very useful for improving the program design based on recipient feedback, as identified by USAID in the case of the LEAP Program in Ghana (Abbey, Odonkor, and Boateng 2014). More programs are investing in toll-free service hotlines and communication channels specifically for recipients. For example, Digicel in Haiti doubled its call center staff on paydays for the Ti Maman Cheri Program, and Save the Children and WFP in Malawi gathered representatives from each nongovernment organization, the bank, and the mobile network operator during disbursements to support recipients who had problems or questions (Zimmerman and Bohling 2014;Almazan 2013). Banco Davivienda in Colombia trains and employs former G2P recipients to work for the support hotline to encourage reporting of complaints and improve resolution (CGAP 2014). 3. Institutionalize monitoring and prepare to adjust as needed. Introducing digital payment mechanisms influences the behaviors and incentives of recipients and payment providers alike. Effective monitoring and evaluation of experiences, preferences, and behaviors is therefore necessary for successful program delivery. 13 As important, however, is that the DSP systems and partnership structures are flexible enough to make required adjustments, whether they are small tweaks or larger process changes, to quickly rectify problems. For example, to prevent fraud, WFP in Lebanon collaborated with a partner bank to monitor participating merchants' transactions in nearly real time and freeze a merchant's POS device if transactions exceeded certain threshold (El-Huni 2014). After discovering that agents and merchants frequently charge unauthorized fees and treat social payment recipients unfairly, WFP Kenya designed posters with pictures depicting payment procedures and selfprotection rules that will be displayed prominently in agents' and merchants' shops (WFP 2016). When Westpac in Fiji found that its agents passed their POS transaction fees onto PFIP recipients or requested minimum purchases, it modified its fee structure to reduce the costs agents incur when serving recipients (Leonard 2011).
These emerging examples of solutions show how a variety of DSP programs and providers are working to get the basics right for successful delivery of DSPs. They address observed shortcomings on the supply side, on the one hand, and demonstrate the importance of strengthening the demand side-the recipients-to self-protect and become vigilant and empowered customers, on the other. Going beyond these basics to achieve meaningful financial inclusion outcomes will require solutions that do more to build trust and confidence in digital payment services, and ultimately add value to recipients' lives and create interest in other financial services. These issues will go unresolved if winning a fee-for-payment government contract is the only motivating factor for providers, or if the program values the lowest fee service proposal over customer-centric systems that can add the most value and best service for recipients. Social payment programs and providers are responsible for ensuring the reliability, convenience, and safety of DSPs. Doing so will imply tradeoffs and require investment of both time and resources, but may be the key to unlocking the elusive win-win-win for all stakeholders.