It is well known that standard frequentist inference breaks down in IV regressions with weak instruments. Bayesian inference with diffuse priors suffers from the same problem. We show that the issue ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback