Week of August 2, 2019

The Attention is a Suckers' Game Edition

Editor's Note: Nothing particularly new to report this week, other than the faiV will be off the next two weeks. Oh and that Imbens paper on potential outcomes vs. DAGs is at least as good as expected, and there's now an NBER version.
--Tim Ogden


1. Financial Systems: I've referenced several times over the last year some work I've been doing for the CDC (the UK DFI, not the one in Atlanta) on investing in financial systems. The first public version of that work, a summary of a much longer paper that I'm still hoping to finish in the next few weeks, is now available. As a summary, it necessarily elides a lot but it does capture what I think are the essential points on the topic right now. The main one I want to highlight here is a somewhat esoteric one: the question in front of us in the sector is not whether or not financial systems matter for the poor, it's whether we know how to intervene in the development of those systems in ways that specifically benefit target populations we care about, in the timeframes and manner in which we can measure. It's an important distinction that I think is missing in too many current conversations about where we are on financial inclusion. Please do read it, and let me now what you think.
In related financial system development and development ideas, Paddy Carter from CDC pointed me to this paper from Paula Bustos, Gabriel Garber and Jacopo Ponticelli on how the financial system in Brazil channeled a productivity shock in agriculture into other sectors (which apparently is on its way to appearing in the QJE) which is exactly what one hopes a financial system accomplishes from a development perspective.
The longer paper for CDC and my research for it emphasizes the history of financial system development. A couple of 2018 books on the topic, specifically on John Lawand Walter Bagehot, are reviewed in the New Yorker by John Lanchester. Rebecca Spang has some thoughts on the continuing focus on the "great man" approach to the history of financial systems and how that misleads. Again, I hope that my work for CDC takes this into account by spotlighting what we know about informal financial systems and how to factor that into thinking about investing in financial system development.
Finally on this topic, two papers that I've had sitting in open tabs for quite some time but have never found a place for in the faiV. First, here's Anginer,Demirgüç-Kunt, and Mare on how institutions affect how much bank capital influences systemic risk (and here's the blog summary). The bottom line is that bank capital matters less when there are well functioning regulatory institutions, but higher capital requirements can substitute for quality institutions in reducing risk. Of course, those higher capital requirements limit the outreach and inclusion of those banks. Trade-offs forever. And here's Ben-David, Palvia and Stulz on how banks in the US react under distress finding that the banks generally reacted prudently rather than gambling in an attempt to revive their sick balance sheets. Which is a further argument for higher capital requirements in weak institutional settings, but creating an alternative system for financial inclusion that isn't bank-based.

2. The Corrupted Economy: My comments a few weeks ago on the "great convergence" and the "corrupted economy" in the US got more positive feedback than I was expecting. So we may now have a new regular section of the faiV.
Unequal access to a quality education is one of the areas where the US increasingly looks like middle income countries. Here's a minor, but infuriating, version of the corrupted system: wealthy parents giving up their children to "guardians" so those children can in turn apply for financial aid as if they don't have any resources. And here's a less blatantly evil version of a similar corruption: children who receive extra time on tests due to some psychological/medical diagnosis are disproportionately white and wealthy--because those are the parents who can afford the thousands of dollars required to pay a private psychologist to deliver such a diagnosis. And the issue is much broader than that because the article only briefly touches on the systemic impact on families and school districts, one I'm acutely aware of personally. I know the educational outcomes for my son, with a rare disease, are almost certainly going to be much better than many other kids in this country with the same disease, because we can afford to live in a school district that isn't so strapped for cash that they have to cut back on services, and I can be an intimidating presence in meetings with the district when necessary.
Here's a story about how the "adjustment" payments for farmers negatively affected by Trump's trade war are all going to the largest, wealthiest farmers. Here's a story about how minor criminal offenses are turned into profits and debtors prison. And here's a story about the actual labor market conditions faced by the lower half of the income distribution: a few days in the life of a meal-delivery bicyclist in NYC. Marvel at how DoorDash preys on income volatility to take tips away from riders. And how the riders' existence is pushed to margins with minimal and shrinking interaction with the customers, how they acknowledge that they are being used to generate data so they can be replaced by drones, and in the meantime how they are subject to the capricious whims of NYC police who can confiscate their bikes on a pretext at any time. And how the riders are grateful that this is a step above working directly for the restaurants. This is America.
And speaking of the Great Convergence, check out this trailer for a new Indian movie about a heroic effort to help kids break out of their corrupted economy. Then think about the long history of American movies with essentially the same plot:Stand and Deliver, Dangerous Minds, Lean on Me, etc. etc. And they are all essentially a distraction from the systemic issues.

3. FinTech and Social Investment: The systemic issues are something I really struggle with, and it came up this week as I was asked to review some applications for a FinTech incubator. I'm not going to name either the incubator/investor or the applicants, but it was impossible to miss the disjuncture between the systemic issues that were the motivation for the program and the proposed solutions. Those solutions ultimately boil down to a theory of change that rests on individuals being primarily responsible for their financial distress--and therefore apps that get their attention or "gamify" savings are somehow "solving" the problem. Now, I think there are some people that are going to be helped by an app that draws their attention to not missing payments and harming their credit, or who aren't saving not because their wages are volatile and well below what they need to afford housing and healthcare, but because it's not fun enough. But I have a hard time caring much about those people. Especially when the business models of many of the FinTech apps I see seem to be built on gaining trust of users and then profiting from referral fees paid by other financial services. I have to wonder: what kinds of financial services firms are going to be interested in paying for access to these kinds of customers? I doubt it's going to be ones that are offering high-quality, low cost services that are good for people--for no other reason than those products aren't going to be profitable enough to pay referral fees.
Many of these apps also raise an issue I've been concerned about in the application of behavioral science since I wrote a review of Scarcity: if everyone recognizes limited attention and behavioral barriers and tries to address those, where do we end up? I think it's likely that attention-focused interventions are going to be revealed as a sucker's game: you constantly have to do more and spend more to compete with all the other people trying to grab attention. Case-in-point: a large scale intervention to grab students attention and redirect it to studying shows no effect. But if you look back at the article about delivery riders, you'll notice that those apps are doing a great job of using behavioral tricks to take advantage of riders. The takeaway from the "studying" study is that you should shift your priors toward high-touch financial coaching and away from FinTech.
One more quick related rant: the whole process reminds me that there is a long way to go in thinking rigoroulsy about social investment and social capital, and I feel better that this chapter Jonathan and I wrote on that topic is worthwhile. (By the way, I took advantage of a plane ride earlier this week to read through most of the rest of the chapters and the whole book is worthwhile.)

4. SMEs: While filling out my evaluations for the FinTech incubator I couldn't help linking to the now published (and open access for a limited time) paper from David McKenzie on how hard it is to pick winners in business competitions, and that experts and machine learning are both bad at it. Inspired by David's earlier work on management, here's a piece on how management consulting could be the best form of foreign aid (factor into your research theories of change how long it's been since the research cited in that article was done).
And here are two recent articles from Next Billion on SMEs--on the difficulties of scaling up local manufacturing in Uganda, and from TechnoServe on building links between SMEs and foreign firms.

5. Global Development Miscellany: I'll confess this is self-indulgent, but it's on topic: the NYT covers the rising tensions in northern Colombia as Wayuu people cross the border fleeing from Venezuela. The village in the story is a couple of hours from where I grew up and only a few miles from where my mother was born--and for the record the descriptions ring very true to me.
Here's a story that I hope gets attention in proportion to it's past history. Prospera, that staple of CCT discussions, and of evidence-based policy, is being abolished. The story seems to be a political economy one combined with fairly significant mistargeting. I hope to read a lot more about this.
There were concerns that DFID was on the verge of a major demotion under Boris Johnson, but that didn't come to pass, yet. Large concerns remain about the future of the British development agency, though not as large, of course, as concerns about the future of British anything.
And finally, you can consider this a global development story if you squint hard, but it's fascinating never-the-less: there is a form of communicable canine cancer that has spread all over the world, and biologists have mapped that process of globalization with remarkable similarities to economic development.

Historical context really does change perspective, as Matt Yglesias suggests that too many of today's policymakers "spent their formative years in a period of anomalously high interest rates." Though I would argue that the long term perspective also means that short-term rates really are concerningly anomalous. Via  Matt Yglesias .

Historical context really does change perspective, as Matt Yglesias suggests that too many of today's policymakers "spent their formative years in a period of anomalously high interest rates." Though I would argue that the long term perspective also means that short-term rates really are concerningly anomalous. Via Matt Yglesias.

Return to the Weekly faiV