The Convergence/Divergence Edition
Editor's Note: In the last faiV I noted that the question "How are you?" didn't seem like it could survive the pandemic. Here's an article from The Atlantic on some alternatives. But my favorite alternative so far is a different answer rather than question. Hans Dieter Seibel passed on that his colleague Marion Levy Jr. has a standard response to "How are you?" that seems especially apposite now: "Terrible. But I'm glad you asked."
–Tim Ogden
1. Existential Crisis Updates:
In case you missed it, the faiVLive on the existential crisis for global microfinance was recorded and you can watch it here. There's also a set of resources that were discussed or shared via chat during the webinar.
This all kicked off with a lament that it didn't seem like folks were worried enough about the impact of the pandemic on the microfinance sector. I can't say that's true anymore. There are a lot of people paying attention now it seems, enough that I can't keep up. I've been focused on a paper co-authored with Jonathan, Simon Quinn, Muhammad Meki, Kashif Malik and Farah Said on the specifics of the crisis in Pakistan, and the generalizable implications of those specifics (bizarrely within the last year I've had a paper published five years after submission, and now one published 5 days after submission).
Of primary concern in the crisis, I've argued, is making sure that choices and options for reacting to the pandemic are not limited by unnecessarily draining liquidity from MFIs. Here's Dan Rozas from e-MFP on that point, emphasizing the primacy of liquidity over even solvency. Here's Ira Lieberman and Paul DiLeo with their "Rescue Plan for the Microfinance Sector." Dan Rozas has a follow-up looking at how much cash MFIs have, based on MIX data. The situation, in terms of funding operations, doesn't seem to be dire if there isn't a liquidity drain. It's worth noting that the smaller MFIs in the MIX data are much worse off on this measure, and the smaller MFIs in the MIX are not the smallest MFIs in the industry.
Dan's piece is up at a special portal that CFI, e-MFP and the Social Protection Task Force have launched for discussion of the crisis. Next Billion also has a page devoted to "Coronavirus Resources for the Social Sector" and a series on "Enterprise in the Time of Coronavirus" which has a lot of specifically microfinance-related content. And here's FinDevGateway's roundup of crisis related content for the recent week.
On the FAI side, I'm excited to announce that Beth Rhyne has joined us as a visiting fellow to focus on pandemic-related issues, particularly around client protection writ large. Here's her first post for us, co-authored with David Grace, on that topic. And here's Jonathan and colleagues on whether migration patterns can help predict COVID outbreaks in South Asia.
2. Convergence/Divergence:
Last time I noted that the pandemic is highlighting the cracks in my Great Convergence thesis. The evidence is mounting on the extreme impact of lockdowns in developing countries. Ecuador seems particularly hard hit in terms of mortality. But hunger is a problem in many countries.
The impact on borrowers in the Pakistan surveys show rapid collapse in incomes, and immediate worries about food security. Julie Zollman is continuing to publish updates from follow-up to the Kenya Financial Diaries. Stuart Rutherford is doing the same with updates from the Hrishipara Diaries. Both continue to look grim. Guy Stuart and MFO have restarted the Bangladesh Garment Worker Diaries; the first round is only showing hints of the impact of the shutdown of the Bangladeshi garment industry. L-IFT is publishing data from its refugee diaries in Uganda. Here's a presentation from BRAC on their survey work in Bangladesh. BFA Global has their second wave of "dipstick" survey data from eight countries. Here's MSC's look at the impact on corner shops and farmers and traders in India.
All of that is before considering some of the knock-on effects of the pandemic and the attention it is drawing. There is still a plague of locusts in East Africa that seriously threatens food security across the region, and it's not getting the attention or resources needed to stop it. Other killers, like malaria, are likely to see a resurgence as health systems are overwhelmed.
On the other hand, the thesis wasn't that situations were identical, but that there is a portion of the United States population where the challenges have a great deal in common with middle-income countries. So what does the situation look like in the US? Concerns about hunger are growing. And food banks are running short on supplies. The government response doesn't look very effective for many workers: 71% of unemployed Americans didn't receive any unemployment benefits in March. The new education system isn't working well for low-income families exacerbating existing inequality in educational outcomes; in some cities, online participation in school is below 50%. And other programs, like small business lending, are being captured by the relatively well-off. Oh, and there are big differences in mortality rates depending on who you are.
A quote in this piece on Americans' and their limited savings and therefore limited ability to cope with the economic side of the pandemic hit me particularly hard on the convergence front: "Credit cards are the new savings account." It's not quite borrowing to save, but it is using small dollar credit to manage liquidity in a way that is quite similar to some of the patterns we see in microcredit.
3. Policy Convergence/Divergence:
A specific version of the convergence/divergence question comes down to policy. The IMF has a list of countries and their policy responses so far. On the finance front, AFI has a spreadsheet of central bank responses.
As uncertain as the past six weeks have been, the next six weeks are likely to bring even more uncertainty as policy paths diverge (cause or effect of the pandemic patterns?). There's clear consensus from economists in the United States that opening up too soon will be worse than continuing lockdowns. But that doesn't really hold for developing countries. Given the collapse in incomes in many countries it seems unlikely that lockdowns will be able to continue. South Africa, for instance, has announced plans to begin opening up. Here's Ray, Subramanian and Vandewalle on options for India. Vietnam is an interesting case study, little noted so far. Sweden has gotten more attention but is still interesting.
Zooming back out for comparison's sake, Brookings has a pretty comprehensive overview of labor market impact in the OECD 20. Here's a Twitter thread with the major findings. The convergence/divergence comes down ultimately to how much of a difference lockdowns make, and how exactly they do it. Here's a thread from Lyman Stone, based on a new NBER paper among other things, arguing that "shelter-in-place" orders don't have much effect on people's behavior—it's the declaration of an emergency and publicizing the pandemic that makes people behave differently. Scrolling through Lyman's Twitter feed is definitely a useful source for different perspectives—here's another thread trying to distinguish the effects of social distancing vs. masks. Another key input on policy is understanding not just COVID-caused deaths, but excess deaths. Here's the first thing I've seen on that, though obviously with very limited data, and only for the US. I expect there are years and years of papers to come on exactly how excess deaths are measured around the world.
4. Methodology (and Financial Diaries):
It's not pandemic all the time around here. I am trying to pay attention to other things too. How about some methodology for a quick break from existential crises? Duflo, Banerjee, Finkelstein, Katz, Olken and Sautmann have a new working paper suggesting the need for moderation in the use of pre-analysis plans for RCTs. In a different time I might note some historical parallels on concern that a too doctrinaire approach to PAPs "may mean that some [studies] do not take place, or that interesting observations and new theories are not explored and reported." Instead, I'll recommend that you also take a look at this thread from Cyrus Samii reacting to the paper and laying out a different rationale for PAPs.
This is far from new, but it came to mind this week in the context of all the financial diaries-style work being done trying to get a pulse of what is happening on the ground right now. I worry a lot about how to interpret data coming from phone surveys during a crisis, but the only evidence I know of suggests that high frequency phone interviews don't yield markedly different data than high-frequency in-person visits. That being said, there is a lot of high-frequency data being gathered this way suddenly and I think there is a lot to learn through collaboration on that front. Another project we've launched at FAI is an attempt to bring together folks doing high-frequency data collection on low-income households and SMEs during the pandemic. The intent is to work together to improve methodology and data interpretation and ultimately to amplify the impact of the work. If you happen to be among those gathering high-frequency, diaries-style data please be in touch and join us.
More broadly, IPA, J-PAL, CGD, IGC, Y-RISE and NGPRL (that may not be their preferred name, but sometimes you just have to fit in to the acronym parade) have jointly launched a portal called RECOVR (see acronyms!) for "surveys, evaluations, and datasets documenting the impact of COVID-19 in low- and middle-income countries."
5. Digital Finance:
Working on the Pakistan paper gave me an opportunity to think more about the implications of the pandemic for digital finance. Last time I highlighted a lot of questions I had (and still have) about whether the pandemic will boost digital finance growth or slow it down. Since then I've been thinking more about the implications for the traditional microfinance sector.
MFIs obviously haven't been leaders in deploying digital finance—it's expensive in both financial and human capital and MFIs are short of both. Now they are going to be even shorter as they attempt to rebuild their cash flows and recapitalize. Making the transition will also be more expensive for existing players because they cannot simply switch to a new system; they have to maintain dual systems for as long as their customer base does. The question is whether funders will be willing to continue to invest in traditional MFIs who are not making much progress toward digital, or if the investors and funders will gravitate toward new entrants who are less handicapped by the crisis and by existing systems. If so, then traditional MFIs could see their path to recovery get even harder as these new entrants draw off the best and most profitable (and most digital) customers, leaving the MFIs to serve the most excluded, poorest and least digital customers. As I mentioned you can read more about this in the Pakistan paper.
If you're interested in facts on the ground more than my speculations, MSC has two posts very much worth your time. First, with Caribou Digital they have a report on the role of DFS agents in the pandemic. They find that transactions are falling and agent incomes are suffering along with everyone else's. Just as bad, the agents aren't getting good guidance or help in protecting themselves. Here is more on the situation for Cash In/Cash Out agents, with a suggested policy framework for supporting them. I'm hopeful that the next faiVLive, on these topics, will be in the next few weeks.
