Lessons Learned from Field Trials of Blockchain-Enabled Vouchers
Overview and Background
In the last decade, emergency cash programming has gained tremendous momentum across the aid sector; in 2016, the UN Secretary General called for it to be the default method of support for crisis-affected people where the situation allows. Currently, Cash Transfer Programming (CTP) including cash and voucher modalities, represents 48% of Mercy Corps’ humanitarian portfolio. In addition to mainstream CTP interventions, Mercy Corps is exploring innovative tools to supplement CTP, such as digital financial payments and blockchain based voucher programs.
Over the past few years, refugee settlements in Uganda have been growing rapidly, largely due to the influx of people from neighboring South Sudan. Since 2016, over a million refugees have been forced over the border as a result of civil war. South Sudanese refugee settlements are mainly concentrated around the nearest entry point to Uganda, a remote rural area in the West Nile sub-region of northern Uganda. Bidibidi settlement (Yumbe District) is one of the largest refugee settlements in the world, with over 280,000 South Sudanese refugees.
It is in this context that Mercy Corps Uganda, with funding from Blockchain Charity Foundation[1] (BCF), piloted the use of blockchain-enabled technology for vouchers. The donor, BCF, is the charitable arm of Binance,[2] the world’s largest cryptocurrency exchange. The pilot targeted 366 households consisting of approximately 2,200 South Sudanese refugees and Ugandan micro entrepreneur households in the Yumbe District.
Details of Pilot Design and Execution
The pilot was designed to test the theory that blockchain-enabled digital tokens could be used for value transfers under field conditions and that such usage would save costs, improve auditing, and enhance compliance. The program design did not envision delivery of a cryptocurrency (like Bitcoin) to program participants, but rather looked to create a digital token, similar to a voucher, that could only be exchanged with a pre-selected group of vendors who would offer goods and services to participants in return for the tokens. This approach was considered appropriate to mitigate several concerns. First, cryptocurrencies have an inherent risk of volatility; a risk we did not want the program participants to bear. Second, conversion of cryptocurrencies into the local (fiat) currency would present challenges because the area is remote and underserved in regards to financial services. Third, by using a token with no secondary market [value], we forestalled potential incentive for fraud.
The transfers in this trial project were made using the Humanity First Token (HFT), a unique cryptocurrency token that was created specifically for the project. The token is a layer two asset created on the Binance Chain[3]. The tokens were pegged against the value of the local currency, the Ugandan Shilling, such that 1 HFT was worth 1 Ugandan Shilling. Binance Uganda[4] spun up a trading pair on their exchange platform to enable the vendors to convert the HFT that they receive from participants into Ugandan Shillings.
The decision to use the Binance Chain for the creation of the token[5] was mainly guided by the specification from the donor, BCF. Additionally, the donor required the project to use the Trust wallet[6], a popular digital currencies wallet that was acquired by Binance in July 2018. Courtesy of the relationship between the Trust wallet and Binance, the wallet was able to provide support for the HFT token, making it usable for this trial.
According to the program design, each household would receive 72,321 HFTs (the equivalent of 72,321 UGX[7]) per month per individual in the household. Four monthly transfers (i.e., one transfer per month) were executed between September and December 2019. The HFTs were transferred directly and remotely into the personal digital wallets that are held by the program participants. The program participants would then use their digital wallets to pay the vendors directly in HFT.
The final step in the movement of value involved a cash-out option for the vendors (that is, exchanging the HFTs for Ugandan Shillings) via the Binance Uganda exchange. Exchange accounts were created as part of the process of onboarding the vendors, and were controlled by the vendors. Once HFTs were redeemed on the exchange, the vendors could either choose to keep their balances on the platform, or they could move the money out using the Yo! Payments platform[8], a popular mobile money solution in Uganda.
The value flow was mostly free of fees, so the program participants incurred zero, or near-zero costs. As the HFT is built on the Binance Chain, a nominal network transaction fee is charged to the sender each time a token is transferred. Program participants and vendors were given an allocation of digital funds sufficient to cover their transaction fees. For vendors, the Yo! Payment system also charged a transaction fee of 0.5% for funds passing through each account as vendors cashed out of the system.
All program participants had been previously qualified for participation in cash transfer programs by the Mercy Corps Uganda team and were required to have access to at least one smartphone. Prior to the transfers, Mercy Corps provided financial literacy training to program participants to build the skills needed for digital literacy in general and for managing their digital wallets in particular.
Throughout the pilot, as each token moved, its movements were automatically logged to Binance Chain. Simple transaction data was captured showing the addresses of the sender and the receiver, the amount, the fees charged, and the time and date of the transaction. The data on the Binance Chain was viewed via the Binance Block Explorer.[9] The transaction data was captured by the Mercy Corps staff and imported into Excel to analyze, identify anomalies/outliers, and reconcile records. Once in Excel, conditional formatting and formulas were used to find exceptions and outliers for further review.
Key Differentiating Attributes of Blockchain-Enabled Transfers
Generally, tokens and cryptocurrencies are comparable with other “modalities” such as cash, evouchers and mobile money, among others, yet they are characterized by key differentiating attributes, shown below. Future efforts may be focused on evaluating the relative trade-offs between these modalities.
1. “Push” instead of “pull” based payment system. Typically, if a beneficiary gives a vendor their credit card or debit card, the vendor gets access to their payment information and can “pull” (charge) money from the beneficiary’s account. A pull system creates a risk of the vendor charging more than was actually intended and using that payment information without beneficiary’s consent. With blockchain-based payments, beneficiaries can determine how much they want to spend and “push” (send) money directly to the vendor without needing to reveal payment account information (e.g. their debit card number, etc.)
2. Auditable Record-keeping. The blockchain, via the relevant block explorer, provides an auditable data trail that allows near real-time tracking of the movement of funds and eases the burden of reconciliation.
3. Near-real-time disbursement. The time required to move money from the disbursing entity to the beneficiary was estimated to be less than 1 minute. Even the slowest transactions were completed within 30 minutes.
4. Low Transaction costs. There is a relatively low transaction cost for all transactions, whether small or large.[10]
5. Reduction of 3rd-party risks and costs. With the use of digital currencies, it is not necessary to outsource payment to third parties and pay for related administrative and transfer costs. Direct disbursements help reduce privacy risks, auditability risks, potential loss risks and speed time to deliver of aid.
Challenges of This Modality
Several challenges were identified during the pilot, including those encountered during the actual program implementation, and in the audit and reconciliation process.
1. Itemized transaction records are not always possible. This trial did not record expenditures by item, in part because it was not possible to embed item information in the QR codes used.
2. Strong internet is required for blockchain enabled cash transfer programs. Accessing the Internet in the settlement remains a challenge due to limited network coverage. All transactions during the market day required the phones to be connected to 4G or Wi-Fi, which caused delays during the initial pilot, despite having set up a supplemental mobile Wi-Fi device on site.
3. Wallet applications should have broad compatibility especially with older operating systems. The Trust Wallet was not compatible with all smartphones (as the wallet requires Android 6 and up) and is a relatively heavy application to download, requiring 10MB of space. During the pilot, some households did not have phones that were compatible with the Trust Wallet. In addition to compatibility, storage capacity on the phones was also a challenge. Many households without sufficient storage space had to delete other applications or content from their phone to create space. In future interventions, a more accessible application would make the initial program implementation smoother.
4. Market days are complex and require tested systems for operation During the first distribution, program participants were prompted to scan QR codes to make purchases[11]. The combination of poor cameras on the participants’ phones and bad lighting conditions created challenges in capturing the QR codes.
5. Digital literacy and security training are essential in the onboarding phase. There are unique requirements for using token and blockchain-enabled transfers that require additional training and device specifications. Households were trained to safeguard their recovery passwords and briefed on the potential implications if their wallet was removed from their phone. Nonetheless, ten households from the pilot still removed the application from their phone and/or lost their recovery passwords when they came to the second market day.
6. Scaling up would require additional compliance and regulatory requirements. The pilot operated in a closed system of approved vendors and implemented an informal KYC for participant identification. In a more open system design, or in a jurisdiction with more stringent requirements for the use of digital currencies, there would be additional burdens on an agency like Mercy Corps to meet the appropriate levels of due diligence and KYC.
7. Protection of participant data is critical. As with all cash transfer programs, personally identifiable information (PII) must be collected. This should be done with proper guidance for program teams to ensure full protection and compliance with extant standards.
Conclusions and Recommendations
The pilot was designed to test the theory that blockchain-enabled digital tokens could be used for value transfers under field conditions, and that such usage would provide cost savings and improved auditing and compliance. The pilot project was valuable for Mercy Corps humanitarian cash program because it generated good learnings and acceptable outcomes including preliminary evidence on the potential benefits of blockchain-enabled cash transfers compared to traditional cash transfers, specifically:
1. Blockchain-enabled cash transfer distributions can help improve compliance with key standards; specifically, by reducing the workload related to administrative operations, post-distribution monitoring, and minimizing third-party risks.
2. Movement of funds are “pre-settled” so there is consistency between the transaction and the accounting. Using the blockchain explorer, the program team can calculate the funds available and remaining in the wallets at any time. The program team can also see how many transactions were made and how many actual transfers took place. This allowed Mercy Corps to proactively identify potential problem areas, as opposed to having to rely on self-reporting by participants, which then must be verified through what can be a time-consuming process.
3. The blockchain improves the program teams’ ability to track vendor receipt of funds, providing reliable validation of the transactions. This is valuable in situations involving the reimbursement of vendors.
4. Stakeholders immediately bought into blockchain-enabled technology. Blockchain is relatively unknown in the Ugandan context however, there was instant buy-in from stakeholders including program participants, vendors, local leaders, and refugee leaders. The Office of the Prime Minister (OPM) in charge of Bidibidi camp and the UNHCR field staff were enthusiastic to accept the new technology.
5. Monitoring and evaluation methods can be streamlined with data available from the blockchain. In future programs, data collection could be streamlined and consolidated with what is available from the blockchain. This would reduce the burden on implementing organizations and field teams, whilst still ensuring accurate data is captured.
The Technology for Development (T4D) team supported the Mercy Corps Uganda Team in the delivery of this pilot through generous partnership with Cisco, under a 5-year program aimed at using technology to deliver aid and development assistance faster, better, and to more people.
[1] https://www.binance.charity/
[4] https://www.binance.co.ug/
[5] A similar result could have been obtained through use of an ERC-20 token via the Ethereum blockchain.
[7] Approximately $19.80 USD at that time.
[8] https://paymentsweb.yo.co.ug/
[9] https://explorer.binance.org
[10] In the current trial, the cost per transaction was 0.000375 BNB (e.g. this converted to 21.985605 UGX on Oct 8th). It does bear noting that transaction fees vary depending on the coin or token used and can be fixed or variable; this needs to be considered in future implementations.
[11] The codes embedded the price of each item and the address of the vendor’s wallet, simplifying transactions.
