Top 4 Things I learned at NEUDC (North East Universities Development Conference)

I just came back from NEUDC (North East Universities Development Conference), which is arguably the best academic development economics conference in the US. It was held at Northwestern University on October 5–6. I thought I would share four interesting things I learned. This is not at all a comprehensive list — for a full recap of the research presented, read Center for Global Development’s round-up. I attended so many interesting presentations and there were even more great ones I missed. But these four stood out because they are relevant to my work or work I’m interested in doing.

1. Susanna Berkower presented her job market paper (joint with Joshua Dean), which showed, surprisingly (for me!), positive results from a clean cookstove Randomized Controlled Trial (RCT) in Kenya. The kicker is that cookstoves had an estimated 14x return on investment, coming from savings on charcoal. Also, use of the stoves (and the monetary savings) persists for at least 18 months. This is in contrast to a paper by Hanna, Duflo, and Greenstone showing no effects (due to decreasing usage) of improved cookstoves in India. Maybe this new stove (the Jikokoa by Burn) in Kenya is just more attractive to its users than previously-studied ones?
2. Erich Battasin presented a paper (joint with Michele de Nadai, and Nandini Krishnan) showing that using diaries for accurately measuring household consumption is no better (and at times worse) than recall-based measures. I had always thought that in theory diaries would be better at accurately illustrating spending habits — they were just very hard to implement well. But they show that diaries really provide no benefits to data quality, but have much higher costs. That’s great for all of us trying to get accurate measures of consumption.
3. I, unfortunately, missed this in person, but Jared Gars presented results from an RCT (joint with Ram Fishman, Avinash Kishore, Yoav Rothler, and Patrick Ward) assessing the impacts of India’s government’s massive soil health card scheme. Unfortunately, they found that this scheme (as currently implemented) has no effects on farmer behaviour. This might not be so surprising for the IDinsight teams working on this program, as their qualitative work has shown that farmers have a lot of trouble understanding the cards. We are currently working with the Ministry of Agriculture to try to improve this program by testing redesigns of the cards to make them more accessible. Will this be enough to move farmers’ behaviour? We’ll need more research to figure that one out.
4. John Loeser presented a paper (joint with Florence Kondylis, Maria Jones, and Jeremy Magruder) on the impacts of a large-scale irrigation scheme in Rwanda, which uses a cool gravity-induced discontinuity design. On the positive side, they find large impacts on revenue for adopters in the dry season, driven by a switch from bananas to high-value vegetables. On the other hand, they find large increases in household labor for adopters, no revenue effects in the rainy season, and modest adoption of irrigation overall. Depending on how you value household labor (which is super-tricky with no good answer!), profits could be very large or near zero. Incomplete adoption of irrigation has led the government of Rwanda to design and test the impacts of complementary interventions, such as labor-saving technologies and output market linkages, to optimize the returns to these large investments (Evidence-> Policy!).

Thanks to all the organizers and presenters for a great conference!