- A new paper in Science combines machine learning, nightlights, high-resolution daytime satellite images, and household surveys to map poverty in Africa. Marshall Burke (one of the authors) summarizes in this blog post: “First, we use lower-resolution nightlights images to train a deep learning model to identify features in the higher-resolution daytime imagery are predictive of economic activity. The idea here … is that nightlights are a good but imperfect measure of economic activity, and they are available for everywhere on earth. So the nightlights help the model figure out what features in the daytime imagery are predictive of economic activity. Without being told what to look for, the model is able to identify a number of features in the daytime imagery that look like things we recognize and tend to think are important in economic activity (e.g roads, urban areas, farmland, and waterways…). Then in the last step of the process, we use these features in the daytime imagery to predict village-level wealth, as measured in a few household surveys that were publicly available and geo-referenced”. Over at the CGD blog, Justin Sandefur offers a nice commentary and critique.
- Also in Science, Dupas, Hoffman, Kremer and Zwane compare the relative effectiveness of prices and hassle/time costs in screening health product delivery so that only those who will use them take them. They find requiring people to show up and redeem a monthly voucher reduces the amount of chlorine given away by 60%, but with only a 1% drop in usage
- Jason Kerwin on work by Dupas, Robinson, Karlan and Ubfal on introducing savings accounts to the poor in three countries, finding very low take-up - I like his summary “Unfortunately, like many other silver bullets before it, this one has failed to kill the stalking werewolf of poverty. Indeed, it almost doesn’t leave the barrel of the gun. 60% of the treatment group in Malawi and Uganda (and 94% in Chile) never touch the bank accounts.”
- USAID has a post on my RFID technology flop, published in Development Engineering.
And finally, XKCD on linear regressions not to trust
- development impact links