- Have we reached peak impact evaluation and peak RCT? 3ie blogs on how its database of more than 4,000 IEs shows a levelling off, which countries have most and few IEs, and more…
- Use xtreg rather than areg or reg when clustering standard errors with short panels
- IPA summary of what rapid-fire A/B testing entails and how it compares to standard RCTs.
- Chile has an innovative way of choosing a non-manipulable seed for the random number generator used for school lotteries – the magnitude of the most recent earthquake (h/t Dany Jaimovich).
- DFID blogpost and paper on how to evaluate the value of impact evaluations
- On VoxEU, Roy Van der Weide and co-authors explain how Egypt is much more unequal than its Gini would suggest, using house prices to account for the top incomes that are missing in household surveys.
Join the Conversation