Reinhart and Rogoff get pasted

One of the better summaries of the entire mess comes from Josh Barro at Bloomberg. This really highlights the need for better analysis and critiquing before papers with such dramatic predictions get out. Two errors: one, in missing out one country, RR get to show a large cliff effect of high debt/GDP ratios, while adding that country back produces results that are not as obvious; two, RR’s original paper seemed to imply a causality between high debt and slowing growth. In the paper they do not make such a claim but their silence when so many others read that causality could have been avoided. Incidentally the small sample size is an issue (RR already acknowledge this).

The fact that the EU has been citing this paper (maybe indirectly) to support fiscal consolidation / austerity during this crisis shows the possibility that bad political decisions may have been made based on poor statistical analysis.

See also Andrew Gelman at this post. There are many good reports out there.

Leave a Reply

Your email address will not be published. Required fields are marked *

Connect with Facebook

This site uses Akismet to reduce spam. Learn how your comment data is processed.