The Entangled Edition
1. Financial Inclusion/Household Financial Security: It seems strange that I so infrequently have items specifically on microfinance so I leap at the chance when it comes along, particularly when that chance involves one of my soapboxes. For instance: the product is what the users make of it, not what the institution wants it to be. For instance, most microcredit loans aren't investment loans, they're liquidity management tools. Which, of course, makes sense since liquidity management is a more pressing need and the structure of the basic microcredit loan is so ill-suited to business investment. But there are ways to make the standard microcredit loan structure more workable for investment purposes. For instance, borrowers from the largest MFI in China form bogus groups and then funnel all of the loans to a single member to make a larger investment. It's not a niche phenomena either: the authors estimate that 73% of groups are doing this.
Another of my soapboxes is the history of development of financial institutions that serve excluded populations, and where the modern microfinance movement fits in that history. There's a new paper from Marvin Suesse and Nikolaus Wolf on the development rural credit cooperatives in Prussia between 1852 and 1913 (I did say this was a pet interest). And here's a summary version in VoxEU. If that doesn't sound like the kind of thing you would normally click on, I beg you to reconsider. It's an interesting story about what drove the creation of a new kind of financial services institution in a setting that makes it a bit easier to disentangle causes and effects, and what effect these new institutions had on their communities. I won't spoil the ending but would encourage you to think about how their results would look if measured with an individual-focused impact evaluation.
I will spoil the beginning, though: the formation of credit cooperatives was driven by changes in the economy that increased the need for access to credit. Which brings me to a third soapbox, the Great Convergence (and there's more on that below). Here's a new report from the New York Fed on constrained access to credit in the United States, including a "Credit Insecurity Index." The premise is that access to credit is important for households to manage liquidity, manage investment and manage risk (those are my terms, theirs are "manage emergencies, take advantage of opportunities, or invest"), but that access varies geographically for lots of different reasons. The report tracks 5 tiers of credit access and changes in those tiers over time, by county. There are 11 states where more than 10% of the population lives in credit-insecure counties. It's another way to illustrate how much in common parts of the US, geographically and demographically, have in common with middle-income countries. Speaking of, I'd love to see a similar exercise done in other countries.
Finally, and keeping with the Great Convergence sub-theme, here's a new paper from Jonathan Fu looking at representative data from six "emerging economies" and five "developed economies" to look at "contextual-level" predictors of financial well-being. He finds that more sources of independent information, more competition, and specifically more competition from informal and semi-formal providers helps, and that simple access and financial literacy don't (hey, another soapbox!).
2. Digital Finance: Writing about digital finance is frequently tough because the line between what is "finance" and what is "digital finance" isn't all that clear much of the time. Thirty years ago most credit card transactions were digital (the information was passed over phone lines from modem-to-modem!) but we don't tend to think of that as "digital finance." Another of my soapboxes is that often the "digital" in "digital finance" is used as a justification to pretend the rules of finance don't apply. Here's a useful review in an unusual outlet (Computer) on the "technical potential versus practical reality" of digital finance, specifically blockchain and crypto, for low-income people. It cites some examples I was unaware of and presents the arguments for the benefits pretty clearly. But the best reason to read it is the Challenges section features a heading you almost never see from pieces that emerge from the digital side of digital finance: "Low-income groups' limited power and financial/social capital." Another thing I really like is it draws a distinction between FinTechs and TechFins, the latter being tech firms dabbling in finance.
The Economist has a piece this week on that issue specifically: "how digital financial services can prey upon the poor" with a specific focus on the potential for abuse of data gathered on poor customers who have little understanding of what is being gathered by whom or the consequences (to be fair, none of us do). To the point about the blurred line between finance and digital finance, there's not much there that hasn't been true of non-digital finance for a very long time.
The Economist piece relies heavily on CGAPs long-standing attention to these issues, and Matthew Soursourian and Ariadne Plaitakis have more to add in a look at how digital finance may require changes to competition policy in financial services, specifically as TechFins play a larger role. Oh look, they specifically call out issues of political power!
In their case it's the political power that the market power of TechFins brings, but it's not just the political power of corporations that becomes worrisome in digital finance. The political power of governments is even more concerning to the extent that it enables even more channels for surveillance, oppression and exclusion. Here's a story about Kenya's digital ID initiative that is excluding many marginalized groups from getting the IDs that will soon be necessary for many aspects of life including access to the financial system. But even those people who are included may end up excluded because the government lacks the tools and expertise to protect the very sensitive data that goes into the biometric IDs.
3. Our Algorithmic Overlords: The line between Digital Finance and Our Algorithmic Overlords is often as blurry as the line between digital finance and finance. Security looms large in both. Here's a story about how the UN's network was hacked last year and databases containing sensitive data on employees and humanitarian and human rights organizations was compromised. Oh, and the UN didn't warn anyone whose data was leaked.
Substantially shifting gears to another part of the Overlords domain, robots are coming for someone's jobs, but whose? Well, "the new warehouses will be built around A.I. robots and not humans." Here's a paper from last year from Acemoglu and Restrepo arguing that the quotes that abound in that article are indicative of a larger problem--too much effort being put into the "wrong kind of AI" which puts "insufficient focus on creating new tasks where labor can be productively employed."
Those warehouse jobs certainly aren't great (real-life crossover: my wife is currently starring in a production of The Glass Menagerie, which is about Tennessee Williams' dreams of escaping his warehouse job) and one of the reasons cited for automation is that's it's hard to find humans who actually want the jobs. The automation of such jobs is a problem for low-skilled workers in wealthy countries, but it's potentially even more of a problem for low-skilled workers in developing countries, with profound consequences for the global economy. From there I could pivot to either the domestic or global perspective (more entanglement!), so I flipped a coin and...
4. Corrupted Economy: A major feature of the Corrupted Economy is the pressure on low-skill jobs, and low-skill wages. Which in turn is part of the story of "Deaths of Despair." I noted last week that there are a number of competing claims about the size and extent and specificity of Deaths of Despair, which I've had a hard time parsing. Anne Case wrote me last week with some compelling data from her forthcoming book with Angus Deaton that pretty thoroughly debunks the claim made in Kevin Drum's piece, linked last week, that the problem is limited to white women in the South.
Doing some additional reading, I found this report (and news article) about mid-life mortality rates from 1959 to 2017 from JAMA, which shows those deaths increasing, and life expectancy decreasing from 2010 to 2017. Increases were concentrated in areas struggling economically. That trend stopped in 2018, and overall life expectancy rose slightly for the first time in four years, according to the National Center for Health Statistics, with deaths from drug overdoses falling for the first time in 28 years. Now to map those data against the credit-insecure data linked above.
As best I can tell, much of the debate over deaths of despair depends on how finely you cut the data, how you adjust for overall population trends, what time series you look at, and what your point of comparison is. What doesn't seem to be in any doubt is that if you benchmark the US against other countries, the situation looks grim. A growing gap in life expectancy has opened between the US and other rich countries. I'm sure I'll have more to link to on this as the Case & Deaton book approaches publication.
But back to those drug deaths, and a Corrupted Economy. If you're wondering about causes of the opioid epidemic, well, what if I told you that an opioid manufacturer paid a software company to tweak it's algorithm to encourage doctor's to ask about patient's pain levels and then recommended prescribing opioids? And it did that more than 230 million times in three years.
5. Global Development: So now we're back to global perspectives. What are the major insights from the last decade of development economics, according to the World Bank's Development Research Group? There are 13, covering program design, implementation, evaluation and even some on-the-ground successes.
Success is often in the eye of the beholder, and there's no better illustration of that than interpretation of results from a policy experiment in Liberia--privatizing government schools. Here is a Twitter thread from Justin Sandefur on the results of the impact evaluation (and the working paper, with Mauricio Romero). Here's a point-by-point Twitter thread challenge to Justin's interpretation from Kevin Starr of the Mulago Foundation, who helped fund the experiment. And a blog post that goes into more detail. And here's a blog post from Wayne Sandholz (and his job market paper) on how voters reacted to the experiment and it's results. All of those links are highly recommended.
Graphic of the Week
From Justin Fox, here's a look at how James Hansen's 1988 prediction of where global temperatures were headed as a result of anthropogenic global warming. Pretty damn accurate. I would snark about social science, but this is really just too scary to joke about. Source: Bloomberg.
