carmen reinhart

Carmen M. Reinhart

Larry Summers has decided to rush into the fray to defend his “colleagues and friends” Carmen Reinhart and Ken Rogoff. Unfortunately, the only way for Summers to defend his friends is by trying to rewrite history. Summer tried to pretend the problem was simply a matter of making some errors. From Summers:

First, this experience should accelerate the evolution of mores with respect to economic research. Rogoff and Reinhart are rightly regarded as careful, honest scholars. Anyone close to the process of economic research will recognize that data errors like the ones they made are distressingly common. Indeed, the JP Morgan risk models in use when the London “whale” trade was placed appear to have had errors similar to those made by Reinhart and Rogoff. Going forward, authors, journals and commentators need to devote more effort to replicating significant results before broadcasting them widely.

The issue was not that they made mistakes, it’s that they actively hid their mistakes. If Rogoff and Reinhart had released their data earlier, like many economists asked them to, their obvious “errors” would have been found out right away. As a result, their incredibly inaccurate claims would not have had a big policy impact.  Instead they refused to release their data even as many claimed they were unable to reproduce it.

Rogoff and Reinhart behaved like anything but careful or honest scholars. In any real scientific field, which apparently doesn’t include economics according to Summers’ interpretation of how acceptable scholarship is performed, research is never taken seriously until methodology and data is published. Science by its very nature needs to be reproducible.

In science anyone making bold claims, but refusing to release their research data, should always be treated as a charlatan until proven otherwise. This Rogoff-Reinhart incident proves how valuable this basic piece of wisdom is to research. That should be the real lesson going forward.

Photo courtesy of Wikimedia Commons