Week of July 23, 2018

The In Bloem Edition

Editor's Note: My writing hiatus from the faiV continues. This week Jeff Bloem, a PhD student in the Applied Economics Department at the University of Minnesota, and faithful reader of the faiV, takes over. You can follow Jeff on Twitter and, yes, via his own blog. Be sure to check out his review of what sounds like a fascinating, must-read book.--Tim Ogden

1. Food Fights and Methods: First, over on the Economics That Really Matters blog, Paul Christian and Chris Barrett summarize their paper on US food aid and conflict. They call into question the results of an influential paper finding a causal link between US food aid and conflict. The authors follow up with a methodological note on the use of instrumental variables with panel data.
Next, the most recent issue of the American Journal of Agricultural Economics (AJAE) has a nice article, comment, and response. In the article Ore Koren finds that it is food abundance, rather than food scarcity, that causes conflict across Africa. Marshall Burke writes in a comment that the effect sizes are implausibly large and are at odds with previous research. Koren responds to these comments by offering three explanations for the "implausibly" large effect sizes.     

2. Randomistas are our new Algorithmic Overlords: At the development economics section of the NBER Summer Institute, Esther Duflo delivered a lecture entitled, "Machinistas meet Randomistas: Some useful ML tools for RCT researchers". Slides from the lecture are available here, and Dina Pomeranz was live Tweeting the lecture. The paper it was based on is here. On the surface it may seem like machine learning and RCTs are interested in different parts of empirical research--the former focused on prediction and the latter focused on causal identification. Duflo highlights a couple areas where using machine learning when analyzing an RCT can be beneficial.

3. Informal Insurance: In a recent article on VoxDev, Kaivan Munshi and Mark Rosenzweig summarize some of the insights from their 2016 paper on the impact of rural informal insurance networks on rural-urban migration in India. The authors first point out that the rural-urban migration rate is relatively low in India compared to other similar countries. The explanation for this is the presence of well-functioning rural informal insurance markets. In order for these informal markets to function well, however, mechanisms must exist to prevent households from reneging on their obligations to their network. A key way this plays out is in restrictions on mobility. This raises a question: What would happen if formal insurance were introduced? Munshi and Rosenzweig run policy simulations and find formal insurance arrangements may increase rural-urban migration. Relatedly, in a new AJAE paper, Kazushi Takahashi, Chris Barrett, and Munenobu Ikegami study how the introduction of formal index insurance affects informal risk-sharing arrangements in rural Ethiopia. They find little evidence of a crowding out of informal insurance from formal insurance products.

4. The Power of Hope: This has already been shared quite a bit, but if you haven't read Seema Jayachandran's summary in the New York Times of the literature on hope and aspirations… you should. The article briefly discusses three recent studies. First, a study in Kampala, Uganda examined the effects of students watching the movie "Queen of Katwe" on educational performance. Second, a study in Oaxaca, Mexico showed an inspirational documentary to women eligible for small business loans. Finally, a study in Kolkata, India designed a psychological training program for sex workers. In each of these studies, these treatments lead to measurable changes in internal characteristics--for example: self-esteem, aspirations, self-efficacy, optimism--and also concrete changes in educational attainment, enterprise revenues and profits, and personal savings.

5. Sanitation: NPR's Planet Money podcast has a nice story about work in Dakar, Senegal helping improve the market for septic tank cleaning. The podcast frames the work as breaking up a "poop cartel." This is a catchy way frame the story, but the truth is urban sanitation is a very important issue in many developing countries. In many places with rapid urbanization and sprawling city footprints, organizing and funding the provision of environmentally-friendly sanitation services is a vexing challenge. This work offers some important insight in how cities might boost sustainable improvements of their sanitation services. More information about this research is available here.


From VoxEu, some data on something we all knew, but perhaps not this specifically and depressingly. Via  VoxEU .

From VoxEu, some data on something we all knew, but perhaps not this specifically and depressingly. Via VoxEU.

Return to the Weekly faiV