AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Problem With Randomized Controlled Trials in Development Economics
I'm not totally convinced that they've done much to reduce poverty in the world, which was the aim of them. There's a lot of problems with even internal validity and actually, you know, part of the issue is people expect too much of them and they think they reveal the truth. They give you an estimate like other statistical procedures and that estimate has a standard or a probability of having arisen entirely by chance. And that's true for all RCTs. In some ways the ideal experiment is an experiment that a physicist would do in a lab in which you control everything. But it's not the gold standard that it seems,. It's actually fairly weak alternative to what you'd really