Why are natural experiments so important for Angrist's work?
Natural experiments are crucial because they provide a way to overcome the fundamental challenge of observational data: confounding factors. In a truly random experiment, like a randomized controlled trial (RCT), the treatment and control groups are, on average, identical except for the treatment itself. Natural experiments, such as those arising from policy discontinuities or geographic variations, offer situations where we can approximate this ideal, allowing us to isolate the causal effect of interest with greater confidence.
Ask Joshua Angrist the follow-up →