· On VoxDev, Deon Filmer, Jed Friedman, Eeshani Kandpal and Junko Onishi summarize their forthcoming ReStat paper that shows how cash transfers can have negative impacts on non-beneficiaries when markets are small and isolated. They use a RCT of a conditional cash transfer program in the Philippines, and find that program beneficiaries increased their demand for nutritious and perishable foods like eggs, which caused food prices to rise in remote markets, hurting non-beneficiaries and unfortunately increasing stunting rates among children from poor, but ineligible households in these remote areas.
· Also on VoxDev, Daniele, Mookerjee and Tommasi summarize an experiment they ran in Kolkata, India, to test the effectiveness of a food safety training program for street food vendors aimed at improving vendors’ awareness of health risks related to food vending and at reducing street food safety hazards. Like many such information related interventions, they see increases in awareness and knowledge, but no changes in actual food safety behavior.
· Andrew Gelman reviews Scott Cunningham’s Mixtape book. He makes an interesting comment that textbooks tend to focus too much on showing successful examples of applying different statistical methods, and not on discussing when these methods fail. This is something we try to do on the blog, but I agree that there is still a tendency to focus on when methods appear to have worked well.
· From July 12 to September 3, DIME Analytics will host an online course to accompany the release of Development Research in Practice: The DIME Analytics Data Handbook. The self-paced course aims to teach all users of development data how to handle data effectively, efficiently, and ethically. The content will cover the data workflow at each stage of an empirical research project, from design to analysis to publication. The estimated time commitment is of 5 hours a week.
· CSWEP has now finished its set of fireside chats with editors, finishing with different finance journals, with videos all available.