Week of October 11, 2019

The Regressive Edition

1. Microfinance: October 2nd was the 10th anniversary of what I consider to be an underappreciated but critical moment in the history of the microfinance movement--David Roodman's piece on how Kiva actually worked. David had already been working on a book about microfinance that was going to be very influential--his open book blog as a whole is a remarkable contribution to the public good, one I wish many more people had decided to replicate--but the Kiva post (based on it being one of the most read blog posts in CGD history according to Justin Sandefur) brought a huge amount of attention to questions about how not only Kiva, but microfinance as a whole, actually worked. I re-read it this week and it's as good as I remember it and definitely makes me pine for the brief glorious time where the development blogosphere was a thing.
There's another important anniversary this week for global microfinance though with a less arbitrarily neat number--Muhammad Yunus's Peace Prize was 13 years ago. Today many were surprised that Greta Thunberg didn't win. The explanation seeming to be both timing and the fact that there is not a direct link between climate change and conflict. There may be a narrowing of the scope of the Peace Prize given that there is certainly no connection between microcredit and reduced conflict. In case you didn't know the winner was Abiy Ahmed, the Ethiopian Prime Minister, who has done some pretty impressive things directly related to peace, like ending the conflict between Eritrea and Ethiopia and freeing thousands of political prisoners. For what it's worth the Economics Nobel announcement is Monday so expect to see more about that in next week's faiV. Some favorites with particular applicability to the faiV include some combination of Donald Rubin, Josh Angrist, John List and Guido Imbens for kicking off "the credibility revolution" and Michael Kremer, Abhijit Bannerjee, Esther Duflo and/or John List for kicking off the experimental revolution. Of course, I'm hoping for the latter because it would likely give a pretty significant boost to my book sales.
But back to microfinance. Banerjee, Emily Breza, Townsend and Vera-Cossio have a new paper (presented at NEUDC) that uses the Townsend Thai village data and the expansion of a credit program to further bolster what should be the clear consensus on the effect of microcredit: on average not much, but very high returns for some. In this case, they find that there are very large gains for high productivity households who get access to credit (1.5 baht increase in profits for every 1 baht increase in credit) and even higher for those outside agriculture. This is broadly similar to earlier work, now in an NBER paper form, by Banerjee, Breza, Duflo and Cynthia Kinnan on Indian microfinance. Keep in mind, as we continue to see these results, that there is another side of the coin: is there a business model that can reach the high productivity borrowers more exclusively?

2. Inequality: If you think about within-country inequality, you think about taxes. Since the United States has had a huge explosion of income and wealth inequality in the last few decades, and there is a presidential election (hopefully) just over a year away there is a lot of discussion about the US tax system and how it has contributed to the growth of inequality and how it might be used to reduce it. This week there has been a lot of focus particularly on whether the US tax system is progressive or regressive, which seems intuitively like it should be a pretty straightforward question to answer. But the US tax system is so complicated, including not only collecting but distributing cash, it's a controversial question. Emmanuel Saez and Gabriel Zucman make the case that since the 1950s the US tax system has shifted dramatically toward being regressive. Here's David Leonhardt's shorter version of their argument with cool animated graphics. But not everyone agrees and those differences can't be traced just to ideology. Here's a thread from Jason Furman, former chair of the Council of Economic Advisors under Obama debating Zucman on methodology and interpretation. Here's David Splinter with a more in-depth analysis illustrating why Saez and Zucman get such different numbers than the traditional approaches to analyzing progressivity.
Meanwhile, there is an entirely different question about whether taxes can be used to effectively address inequality (Saez and Zucman's book is all about how the wealthy evade taxes). There's a new NBER paper on the response of rich taxpayers to an increase in the California tax rate. It finds that just under 1% of those subject to the higher taxes moved out of state, and those who stayed found ways to avoid the tax, so that total income from the tax was about half of what it would have been otherwise. Here's Lyman Stone's Twitter summary.
It's not clear how to think about that 50% cut in additional revenue: on the one hand, there is a big increase in tax collection, on the other hand you have to expect that over time people are going to get even better at evading the tax. Here's Lily Batchelder and David Kamin with a comprehensive review of wealth taxation in implementation with hope that wealth taxes can work.

3. Evidence-Based Policy: Let's talk about the inverse relationship between farm plot size and productivity in developing countries. That's admittedly a strange place to start a discussion about evidence-based policy, but I view it in the same bucket as whether the US has a progressive or regressive tax system. These are questions that seem easy to answer--and very important to answer for policy purposes--but are remarkably complicated to actually answer.
For background, what is known as the IFSP hypothesis started way back with Amartya Sen making the observation (well before the credibility revolution) in the early 1960s about Indian agriculture. Lots of papers over the years have tried to chase this observation down and document it in India and around the developing world. Here's a helpful summary that you really should read even if this is not even close to your area of interest. That was written in 2018, and still there are papers coming fast and furious that related to this basic question. Here's Gollin and Udry looking at measurement error as a big factor in measures of farm productivity, explaining about 90% of the differences between farms (e.g. they aren't really differences). And from NEUDC here's Kibrom Abay, Leah Bevis and Christopher Barrett with a different take on measurement error, where it comes from and how it affects the related policy questions (which are big! Like how much should you redistribute land!). And here's Milu Muyanga and TS Jayne using data from Kenya to document that there at least the U-shaped relationship holds no matter how you measure productivity.
So what is the evidence-based policymaker to think when such basic inputs to policymaking as progressivity of the tax system and whether there is a IFSP are subjects of debate for decades? If you are waiting for "settled science," well the economics publishing industry isn't going to help much. Here's new data from the Journal of Political Economy which, in the fine print, says that the average time from submission to publication--EXCLUDING AUTHOR REVISIONS--is more than 450 days.

4. SMEs and Human Capital: A couple quick hits on two favorite topics of mine these days--especially as they come together. Girum Abebe, Fafchamps, Koelle and Quinn had a paper at NEUDC about how to get management experience to young people in Ethiopia. The paper is, unfortunately from my perspective, most framed around the methods and the algorithms to use to produce the best results, rather than what I find most interesting: is it possible to match young managers with more experienced managers in order to spread the human capital of management? They find that matching high-ability trainees with high-management score firms yields meaningful increases in self-employment (in other words that these trainees go on to start and manage their own firms). What I really want to know though is how much better everyone got at managing and what the effect within firms of better management was.
There's a new paper that starts to answer this question with larger firms in Mozambique--by offering finance training to managers of "medium and large enterprises." Claudia Custodio, Diogo Mendes and Daniel Metzger find that a short executive education style course for these managers has meaningful effects on working capital management and small but still there effects on long-term investment. Perhaps there is hope for the management trap!

5. More from NEUDC: From guest faiVer Jonathan Morduch who actually made it to NEUDC:
Most people know the NEUDC conference by its acronym. Every year, it’s the biggest global gathering of researchers on development economics, drawing participants from around the world. But it started in 1967 as the Northeast Universities Development Consortium and has stuck to the northeast USA. This year, though, the NEUDC landed in Chicago with a successful event at Northwestern. You wouldn’t have been alone if the 4 parallel sessions (most with 4 papers each), had left you longing for a nap. It turns out that naps are powerful. In an NEUDC paper based on a field experiment in Chennai, urban poor individuals averaged 5.6 hours of sleep a night. For those who could get an afternoon nap, the researchers saw “improved cognition, psychological well-being, and productivity. Naps also reduced inattention to incentives and increased patience, as measured by a real-effort task and financial savings.” The big, lurking question is whether behavioral biases are often a function of sleeplessness?
A different NEUDC paper finds that microcredit flexibility may be over-rated. An RCT introduced flexibility into microcredit contracts in Bangladesh. The setting is one with highly seasonal income, and the flexible contract allowed borrowers to pay less during the lean season. I was surprised (and the authors were too) that the intervention made little difference to financial and economic outcomes.
Before I rethink all the lessons on volatility from financial diaries, here’s a different finding that steadiness can in fact be very helpful. The big news in cash transfers from the summer was that Mexico’s long-celebrated conditional cash transfer (CCT) program was shutting down. It was the CCT program that launched dozens of other CCTs -- and also gave early credibility to RCTs (and RCTs of CCTs). Prospera’s demise came despite the NEUDC paper showing that Prospera recipients were much better protected against anticipated income shocks than otherwise similar households. However, households receiving transfers were still sensitive to unanticipated shocks, in line with the Permanent Income Hypothesis.

1950 tax.png
The "money" charts on how the total tax rate has changed in the US since the 1950s that I mentioned in the first item. Source:  NYT

The "money" charts on how the total tax rate has changed in the US since the 1950s that I mentioned in the first item. Source: NYT

Return to the Weekly faiV