You are an economics student (Thomas Herndon) who has been asked by your professor to take one well-known paper in economics and replicate its results as homework.
Detail of a desk after studying. (Photo credit: Wikipedia)
You decide to take one of the most famous papers in economics in recent years (“Growth in a Time of Debt”) by two eminent Harvard professors (Reinhart and Rogoff). This paper has been widely cited, not just in academic circles, but also in policy circles because it apparently showed that a country’s economic growth tended to slow dramatically when its government debt exceeded 90% of GDP. It has been widely cited by finance ministers and policy makers around the world as a justification for austerity measures and cutting back on public spending.
You try and try, but try as you might you do not succeed in replicating the paper’s results. So If you are Thomas, do you:
(a) Take another paper and see if you can replicate its results instead.
(b) Decide economics is too difficult for you and switch to something that does not involve numbers.
(c) Kick the computer and anything else handy.
Thomas did not do any of that. At first Thomas and his professors were sure he was the one making the mistakes, but when Thomas eventually got the original data from Reinhart and Rogoff, it eventually dawned on all of them that the problem lay in the original use of the data. The most glaring problem was Reinhart and Rogoff had omitted key data in their Excel analysis, but there were other problems such as a dubious method of weighting data.
Thomas and his professors reworked the Reinhart and Rogoff analysis and came to the conclusion that yes, Reinhart and Rogoff’s data did show that economic growth tended to be slower when government debt exceeded 90% of GDP, but the effect was usually relatively mild and nothing like as dramatic as claimed by Reinhart and Rogoff.
Thomas published his analysis with his professors Michael Ash and Robert Pollin in a draft working paper on the 15th April and the reaction was dramatic, immediate and world-wide. The arguments were complex and (apart from the spreadsheet error) not always clear cut, but the on balance most informed commentators found more merit in the Herndon/Ash/Pollin analysis than in Reinhart/Rogoff's.
The controversy is just a few days old and the dust is still to settle on all this, but since the Reinhart and Rogoff paper had been used by many as one justification for current austerity measures, it is more than just an academic debate. In fairness to Reinhart and Rogoff, all they claimed to show in their paper was an association between low growth and high debt, it was some policy analysts and policy makers who went even further and saw high public debt as causing low growth, rather than possibly the other way round. But Thomas’ persistence has still refueled a major debate on the economics of austerity.
However, there is one thing that I have not seen in all the already extensive public debate on this.
What grade did Thomas get for his homework?
Or do they have to invent a new grade for him?
(You can keep up with the ongoing debate here by pasting Herndon Reinhart Rogoff into Google News.)