Reanalysis

As part of a major replication and robustness project of articles in the American Economic Review, this fall I assigned students in my Masters Macro course at the New Economic School (Moscow) to replicate and test robustness for Macro papers published in the AER. In our sample of AER papers, 66% had full data available online, and the replicated results were exactly the same as in the paper 72% of the time.