ICTworks

Many countries have launched unprecedented relief packages to cushion the economic and social impact of the COVID-19 pandemic. Information is still limited on how well the programs have functioned; in particular, there is a dearth of rapid demand-side survey evidence on the experience of beneficiaries receiving transfers and the likely magnitudes of inclusion and exclusion errors.

Nevertheless, the emerging picture provides some indications of how investments in digital systems and their deployment along the social transfer value chain have been facilitating the response.

Digital Technology in Social Assistance Transfers for COVID-19 Relief: Lessons from Selected Cases considers some initial lessons emerging from selected countries around the use of digital technology to implement these government-to-people (G2P) social transfer programs.

The review suggests that an important objective for policymakers in the post-COVID period will be to build on the capabilities developed during the crisis to strengthen sustainable social protection and payment systems that are both inclusive and effective, addressing the challenges faced especially by women who are often digitally disadvantaged.

1. Digital systems are critical in scaling payments

Digital financial services have facilitated the processing and payment of millions of grants across many countries, on a scale that would not have been remotely feasible without them. Technology has been applied to all parts of the “user journey”, from initial identification and onboarding to selection and payments.

Countries with stronger digital infrastructure, including digital ID and payment systems and social registers have generally been able to implement and disburse emergency assistance programs more rapidly than those without these assets. While most examples have been in middle-income countries, Togo offers an example of an “all-digital” emergency program in a low-income country.

2. Digital campaigns engage active beneficiaries

Multi-media campaigns inviting applications for support can mobilize people and generate awareness. Namibia, South Africa, Brazil, Pakistan and Togo all invited digital applications for emergency relief, to be made through mobiles, WhatsApp or websites. All received very large volumes of applications in a short period, sometimes overwhelming channels.

With a population of only 2.4 million,

• Namibia needed to verify almost 2.3 million applications (from a population of 2.4 million) resulting in 970,000 unique individuals eligible for further screening.
• South Africa received 13 million applications, out of which around 6 million were deemed to be valid
• The Pakistan Ehsaas program received 146 million SMS requests for assistance, with 48 million unique claimants.
• Togo’s Novissi program received 1.4 million applications and almost 600,000, mostly women, were approved.

Transferring funds to the accounts of “passive” beneficiaries saves these steps, as in the case of India’s transfers to 200 million female Jan Dhan account accounts. But active communications are still needed; especially if the accounts are dormant, some beneficiaries may not be aware that they have received transfers.

3. Digital payments are vital to social programs

Countries used a variety of approaches to deliver scaled-up payments to beneficiaries:

• financial accounts (banks and mobile money),
• digital vouchers,
• e-wallets and other mechanisms that do not offer a full range of financial services.

Some have taken advantage of the crisis to expand financial access, including by using tiered KYC to facilitate the remote opening of bank and mobile money accounts. A review of six countries with such active onboarding programs found the potential for 60 million new accounts opened since the onset of COVID-19, approximately 4% of the global unbanked population.

In some cases, remote opening had been permitted for the first time. Such regulatory streamlining and innovation can be a positive legacy of the COVID-19 digital response.

• Almost all payments in India’s scaled up response have been made through bank accounts that leverage infrastructure created for its DBT platform.
• South Africa also used bank accounts as the main mechanism for social and emergency payments/
• In Bangladesh, some important programs, like education supplements paid to mothers, are now made through mobile money.
• Togo’s Novissi program, which covered residents of three more urbanized “lockdown” areas, also used s mobile money payment platform.
• Namibia used e-wallets very successfully, with 98 percent of transfers cashed out within a short time period.

In Pakistan, social payments had previously been delivered through custom arrangements including biometric verification, though there has recently been a move to route them through multi-purpose accounts. With low financial inclusion, emergency payments to new Ehsaas beneficiaries have been delivered in cash at 17,000 dedicated payment points, also subject to biometric authentication.

While the COVID-19 response has triggered innovations in G2P payments systems which have succeeded in getting funds rapidly to people, only some country responses provide an impetus towards a sustainable digital payments system for the longer term.

4. Digital payment limitations in COVID-19 relief

The main limitation of digital payments for COVID relief is the same as before the pandemic – a high propensity to cash out benefits immediately. With increased payment volumes, this has resulted in crowding at many pay-points, increasing the risk of contagion.

Relative to the option of cash handouts from a limited number of public pay-points the flexibility of cash-out arrangements for financial instruments probably results in less crowding and congestion as well as more space to manage liquidity.

In Namibia, for example, the transfer of e-wallets was broken into batches to reduce crowding, so that beneficiaries would not all receive their notification at the same time. This also gave banks the ability to top up their ATMs several times each day.

Other countries such as Peru and India allowed phased cash withdrawal based on ID and bank account numbers, thereby reducing overcrowding at payment points reportedly with limited success.

A second limitation, in some cases, has been the capacity of digital systems to handle large, sudden, increases in the number and value of payments. This was reported as a problem in India, for example, with a sharp surge in transfers resulting in high rates of failures of cash-out attempts through business correspondents using the Aadhaar-Enabled Payments System (AEPS).

It was not clear whether the these reflected failures for individual attempts as opposed to actual payment failures (which could be lower) and whether they were being driven by authentication failures, network limitations or inadequate incentives for banks and mobile business correspondents to process transactions.

Recent large-scale survey evidence provides some useful insights. On the one hand, cash-out failure rates were found to be moderate, around 11-12 percent, and not too different across banks, ATMs and correspondents.

On the other, many households had been slow to cash-out, citing concerns over health and crowded pay-points, while some who had cashed-out found difficulties in purchasing essential goods because of economic disruption. These impediments appear to have been more problematic than the cash-out failures themselves.

Reflecting the principle that no system is perfect, households in India have been fortunate in being able to access both cash transfers and in-kind support through the ration system. The survey found that only 1 percent of poor urban households and 4 percent of poor urban households were not covered by one or other of these approaches.

5. Payment integration raises privacy concerns

With highly integrated social benefit systems linkable to other databases through the national ID, countries like Pakistan, South Africa and Brazil have been able to implement a coordinated response to reach a large number of people, screening for unique beneficiaries and to ensure that beneficiaries are not receiving support from multiple programs.

India offers a notable example of a highly digitized social protection system but one that is far less integrated. As noted previously, India’s extensive and overlapping system of benefits offers high coverage across its poor population. And, with the Aadhaar entrenched as the ID system serving all social benefit programs, it can ensure that the beneficiaries for any one program are unique.

But, because of legal restrictions on sharing databases, it cannot check, for example, that a woman receiving an emergency COVID-19 payment into a Jan Dhan bank account is not also receiving other additional benefits or even that she is actually poor. This has led to criticism that, even with a very large number of payment beneficiaries, many poor women might not be covered by this emergency program.

Another limitation revealed by India’s disconnected COVID-19 response is lack of portability. The residence-based nature of its state-implemented programs has opened up major gaps in support, particularly for internal migrants.

In Bihar state of northern India, the government announced a grant for returning migrants provided that they produced a local bank account as proof of state residency. As a result, those who had opened bank accounts outside the state became ineligible for the transfer, even though they might have fulfilled other core criteria of the program.

These examples point to a tension between the short-run imperatives of the COVID-19 response and longer-run concerns around privacy and data protection.

From the short-run perspective, the ideal is a highly integrated national system with existing and new programs able to seamlessly onboard applicants for assistance and to allocate resources efficiently among them.

While the issues raised by dynamic screening go beyond the COVID-19 response, it will be important to ensure that such emergency responses do not undermine measures to preserve data privacy in the longer run.

The tension will only increase over time as an expanding digital footprint for citizens increases the ability of governments to draw on a wide range of data, potentially extending to real-time monitoring of activities such as mobile phone use patterns.