Hard data for hard choices

Posted on:
Nathanael Goldberg
A woman wearing a medical face mask counts rupee notes after collecting cash of financial assistance through a mobile wallet under the governmental Ehsaas Emergency Cash Programme for families in need during a government-imposed nationwide lockdown as a preventive measure against the COVID-19 coronavirus, in Islamabad on April 9, 2020.
A woman counts rupee notes after collecting emergency COVID-19 financial assistance through a mobile wallet in Islamabad, Pakistan in April 2020.
Photo: Aamir Qureshi

This essay was first published in Project Syndicate.

While high-income countries have been able to spend trillions of dollars to buttress their economies as they weather the COVID-19 crisis, many other countries do not have anywhere close to the same resources at their disposal. In the countries where the world’s poorest people live—including major middle-income economies such as India, Nigeria, and Bangladesh—the pandemic calculus is critically different.

Policymakers outside of the rich world are facing a new version of the old “trolley problem.” In that thought experiment, a subject is faced with the choice of redirecting a speeding streetcar from a track where it would kill several people onto one where it would kill just one. The choice is obvious from a strictly utilitarian perspective, but actually pulling the lever to condemn someone to certain death is not so easy.

In the case of the pandemic, the number of lives at stake means that policymakers must make decisions quickly. They are confronted with a series of apparent dilemmas: viral containment vs. the economy; young vs. old; essential workers vs. those who can shelter in place. But the choices before them in the COVID-19 crisis are not nearly as straightforward as in the trolley problem, because one cannot always tell which track will save more lives. And attribution can be hard: the “uncontained pandemic” track may lead to a clear and observable number of deaths, whereas the “economy shattered” track would lead to a diffuse and difficult-to-ascertain number of deaths from myriad downstream effects. And, eventually, if and when a vaccine is developed and becomes widely available, there may be heated debates about whether the paths not taken should have been chosen after all.

Points of departure

Data and social science can help confront the current crisis and anticipate the debates to come. For example, we already have a lot of evidence concerning what happens when low-income households suffer from a further loss of income. We also have evidence indicating that cash-transfer programs, combined with data-driven targeting methods, are highly effective in ameliorating such shocks.

Unlike in the United States, where many middle-class families have been able to shelter in place and sustain themselves temporarily with household savings, the hundreds of millions of people subsisting on less than the equivalent of $1.90 per day do not have this option. That means governments around the world are facing difficult questions. What forms of assistance are most needed to help the vulnerable survive? Who, precisely, is most urgently in need of support? How can assistance be administered as quickly and effectively as possible without increasing social contact?

As social scientists who test the impact of programs designed to help the poor, we would direct policymakers’ attention to a body of existing evidence to answer these key questions. It is now clear that the crisis will continue in some form for many months, inflicting a staggering toll on both health and the economy. Lockdowns will likely be relaxed and reinstated intermittently unless and until a vaccine arrives and is made universally available.

The expected duration of the pandemic demands that we look not only at short-term emergency measures, but also at broader opportunities to strengthen social assistance programs so that they can meet citizens’ needs as the crisis evolves. The evidence suggests that governments should make major investments in two areas: cash transfers and mobile-money infrastructure.

Fast cash

On the question of what kind of assistance people need most urgently, the best answer is also the simplest: cash. Disbursements of cash are fast, can be delivered remotely, and will give households the flexibility to buy what they need. As a means of propping up vulnerable families during times of crisis, cash transfers have a proven track record across multiple countries, continents, and contexts.

One common fear associated with cash transfers is that they will increase spending on “temptation goods” like alcohol and tobacco. But recent, careful studies do not bear out this concern. Not only does cash boost food security and overall economic wellbeing for the poor; it also yields a host of other benefits, such as reduction in domestic violence in some contexts. Moreover, cash is especially helpful in disasters, when the poor and vulnerable are even more exposed than usual.

In fact, many countries are already rolling out cash transfers as part of their pandemic response. In Togo, for example, the government is providing monthly cash transfers of about $17 to men and $20 to women throughout the country’s state of emergency. The policy has two overarching goals: to provide enough cash for people to put food on their tables, and to alleviate the need for people to work outside the home. Eligibility initially was determined by the occupations listed on voter registration rolls, which allowed for 80% of the program’s intended beneficiaries to sign up quickly.

The speed and size of Togo’s program is impressive. In terms of spending power, the transfers are quite large, amounting to around one-third of the minimum wage. Moreover, the government is also covering utility and water expenses, which often eat up another one-third of a poor household’s income. And with help from the nonprofit research organization Innovations for Poverty Action, Togo’s government and others are considering the use of big data to identify recipients. A combination of previously collected national surveys and cellphone data has proven to be a powerful predictor of poverty.

Another common fear about cash transfers is that, in normal times, such programs will lead to less work. That, after all, is the lesson taught in introductory economic courses: because the relationship between leisure and labor is a tradeoff, someone who already has a good income may choose to work less.

But while this assumption about the “income effect” may hold true in a hyper-stylized, overly simplified model, the data show that it does not apply in real-world cash-transfer programs for low-income households in developing countries. The irony, of course, is that now we want people to work less, in order to slow the spread of the virus. Hence, the transfer must be ample enough to “discourage work.”

Fortunately, there is good reason to think that cash transfers will work for this purpose, too. When a lockdown is in place, there is already social pressure to adhere to self-isolation and social-distancing protocols. So, the money can be disbursed in such a way as to strengthen the public-health message: this is being provided precisely so that you don’t have to go out and work.

Many governments have also opted to deliver in-kind food transfers, which can have obvious advantages in the case of a pandemic, where food supply chains have been disrupted, and where lockdowns and social-distancing guidelines prohibit market interactions. The evidence shows that children are particularly vulnerable to food insecurity, because millions are now missing out on school meals that constitute an essential safety net for disadvantaged children in normal times.

Who needs what?

Under ordinary conditions, identifying who should be eligible for a transfer can be a laborious and time-consuming process. It is also often imperfect, with transfers missing some households that need it while going to some that do not. The systems used in wealthy, information-rich countries are not viable in places where much of the population works in the informal sector and lack tax filings listing their incomes.

Needless to say, the COVID-19 crisis makes the challenge even harder, because many more people who were getting by are suddenly in a vulnerable position. Most small-business owners in cities, for example, will not have qualified for assistance before the pandemic, but soon could need a lifeline if they are forced to close down. And the same is true of many informal-sector workers.

Findings from International Care Ministries (one of our partners) in the Philippines indicate the potential scale of the eligibility crisis. ICM reports that 17% of respondents to a recent survey had gone a day without eating anything. In response, the organization is now working with local churches to identify those most in need in vulnerable communities in rural areas and rapidly distributing nutritional supplements.

Fortunately, some established methods can quickly identify households that are newly in need of assistance. When designed properly, programs inviting individuals to self-report a loss of income can be an effective and rapidly scalable screening method. Governments are also experimenting with programs that check eligibility against mobile banking balances, or that focus on certain hard-hit geographic areas (generally cities, so far). These trials are urgently needed, with programs that prove effective serving as models for others to follow.

Mobilizing mobile money

A final question is how governments should actually get money into people’s hands. Mobile money is the fastest and cheapest option, especially because access and usage have been increasing in many parts of the world. Research shows that growing reliance on mobile money has already reduced poverty rates and strengthened women’s agency under ordinary circumstances. With the current crisis, there is thus an opportunity to move even more people onto these platforms.

Here, Kenya, the country that leads the world in mobile-money usage, offers interesting lessons. Its experience has shown that governments can deliver both immediate and longer-lasting benefits by working together with the private sector to expand access to key technologies. Mobile money has been found to help households weather a downturn, and Kenya’s central bank has encouraged greater use of it since the arrival of the pandemic, issuing a directive in March to waive fees on all low-value mobile transactions. Likewise, in Togo, people enroll for benefits through their mobile phones, and then receive cash transfers in mobile money.

But, more broadly, rapidly scaling up the use of mobile money will require some changes. For starters, registering must be made much easier. In many places, the poor lack the required forms of identification to open an account. Governments should consider loosening these regulations (if only temporarily).

One option, for example, is to leverage in-person, in-kind assistance to hold mobile-money-enrollment drives. Governments also could explore ways to remove common barriers like transaction fees, either through regulations or subsidies. And they should consider extending assistance, in coordination with the industry, to mobile-money agents—the individuals (such as storekeepers) who provide the local points of conversion from cash to mobile money.

In Jordan, for example, to reduce cash transactions, the Central Bank has made mobile money a central feature of its COVID-19 response strategy. Regulatory changes were needed to expand access rapidly, including allowing remote digital onboarding for clients and mobile money agents. This made opening accounts during the pandemic easier and simplified the know-your-customer process. As a result, merchants are able to receive payments via mobile money more easily, and mobile money account ownership more than doubled from February to May 2020.

The Jordanian government has also expanded eligibility for emergency cash transfers to informal workers and others who lost their livelihoods as a result of social distancing measures; all cash transfers are now sent to mobile wallets, unless the recipient opts for an in-person or bank transfer. As Jordan begins to reopen, the role of mobile money in delivering social-protection benefits seems to have become permanent.

Unfortunately, not every country has mobile-money infrastructure; even where they do, marginalized and vulnerable populations, including women and the elderly, may not own a mobile phone. In these cases, governments must grapple with the challenge of delivering food or other forms of in-person assistance without encouraging crowds. Here, one solution is to use coordinated and staggered distribution schedules with designated pick-up times; another is door-to-door delivery (with face masks and adherence to social-distancing protocols).

Staying ahead of the curve

All of these evidence-based approaches can ameliorate difficult ethical choices and help governments act more quickly and more effectively. Looking ahead, we can take some comfort in the fact that the countries with previous experience managing epidemics such as SARS were able to respond faster and more decisively to COVID-19. All countries, wealthy and poor, now recognize the value big data can bring to fighting pandemics and preparing for future ones. Data can help us understand the precise challenges a community is facing, and can point us to the evidence-based policy ideas that can help address those challenges. But relying on evidence and data in isolation, without carefully considering underlying conditions and theories of change, can be ineffective or even dangerous. As a common wonk saying goes, we want evidence-based policymaking, not policy-based evidence-making.

Learning from what we are all experiencing requires pausing occasionally to determine which data can be collected that could help now and in the future. For example, epidemiological models have yielded widely differing predictions of how the coronavirus will spread, depending on different lockdown policies and compliance with social-distancing recommendations. With systematic data from around the world, we should be able to learn a great deal from these disparities after the fact.

But we already can expect that not every country that adopts cash-transfer programs will succeed. Some initiatives will not provide enough support, reach enough people, last long enough time, or all of the above. All of these experiences will yield useful lessons for the future if we learn not only what is happening, but why.

For social scientists, the crisis will raise many questions. Just how bad is the problem of food insecurity today? Which anecdotes were real, and which were mere rumor, amplified by social media? Which countries’ lockdown policies deepened the tragedy, and which countries’ strategies were most effective? How did market access, population density, Internet availability, and pre-existing local support networks influence the experiences of the world’s poorest?

Differences across countries, regions, and communities will be edifying, because every government will have confronted the same basic economic and public-health challenge. Social science offers powerful methods and tools for drawing actionable conclusions from this experience. But the unprecedented nature of the crisis also demands innovation. We still have a lot to learn about improving cash-transfer programs and expanding mobile-money access.

Let us hope that the current period of lockdown and reopening causes as little harm as possible. But we must look to the future as well. We owe it to our fellow citizens, the most vulnerable and the wealthy, to learn from this tragedy so that we can be more prepared for the next one.


Copyright: Project Syndicate.

Posted by Alison Fahey, Dean Karlan, and Nathanael Goldberg, Director of Sector Programs, Innovations for Poverty Action.