Source: http://dilbert.com/strips/comic/2007-08-08/ (h/t John Quiggin)
The error in the Reinhart-Rogoff (2010) [RR thereafter] paper titled “Growth in a time of debt“, also published in AER. RR (2010) concluded that a public debt-to-GDP ratio above 90% drags on a country’s economic growth. More importantly, there is somewhat a nonlinear relationship between debt and economic growth. Herndon, Ash and Pollin (2013) tried to replicate the results in RR (2010) and find that when the debt-GDP ratio breaches 90%, growth slows to 2.2% and not the -0.1% that RR (2010) finds.
This error is needed to get the results they published, and it would go a long way to explaining why it has been impossible for others to replicate these results. If this error turns out to be an actual mistake Reinhart-Rogoff made, well, all I can hope is that future historians note that one of the core empirical points providing the intellectual foundation for the global move to austerity in the early 2010s was based on someone accidentally not updating a row formula in Excel. – Next New Deal.
As a researcher in this area, I’ve read that paper a while ago. Yes, it is an influential paper. Yes, a mistake is a mistake. But,
i. It is one of the many papers that have been published on this topic involving debt-GDP ratio, and implications on the economy. And, I do not really think policymakers actually based their decisions to promote austerity soley on these papers. Paul Krugman explains what I think rather well in this piece. The Altlantic provides a rather good take on this too.
ii. Debt-GDP ratios above 90% can’t be good for an economy, whether a linear or nonlinear relationship between debt-GDP ratio and economic growth exists.
Economists and researchers alike should definitely be responsible for the work they put out to public and for the work they publish. As an economist that is a perfectionist and rather pedantic, I know how easy it is to make mistakes in what you do, despite the number of times you may have checked through your work. It is scary to know that you may still make mistakes despite your best efforts, especially when you are handling lots of data. Unfortunately, in academia, there isn’t a whole lot of cross-checking going on. It is already difficult to find someone working closely in the same field, using the same methodologies.