One of the major intellectual underpinnings of the austerity movement may have had its legs swept out from under it. On Tuesday, one of the most cited studies by the pro-austerity movement, a paper by Harvard University economists Carmen Reinhart and Ken Rogoff that claims that when the debt to GDP ration of a country’s economy rises above 90% growth slows dramatically, was found to have serious computational and transcription errors.
Three economists from the University of Massachusetts, Thomas Herndon, Michael Ash, and Robert Pollin, conducted the critique of Reinhard and Rogoff study, which can be found here. Over the course of their study they identity errors that range from selective exclusion of data to actual errors in the spreadsheet used to calculate the results. The results cast much doubt on the widely cited conventional wisdom that an increasing debt load hurts an economy.
The first error of Reinhard and Rogoff is that they excluded data from their dataset with no explanation given. The years excluded were times in three countries: Australia (1946-1950), New Zealand (1946-1949), and Canada (1946-1950), which were the countries with high debt and solid growth. For example, the data for New Zealand in the paper's high debt data set changes dramatically, from a dreadful -7.6% to a respectable 2.6%. This is a 10-point error.
This in turn is important, because the effects of high debt are examined in only seven countries that cross into Reinhard and Rogoff's high debt area. Each country’s data is weighted the same. Therefore, correcting for the missing data actually eliminates most of the falloff in growth that Reinhard and Rogoff "proved" was the case by adding 1.5% to the average growth rate of high-debt countries.
The final error is something that should be found in an undergraduate paper, not something published by two Harvard economists. A coding error in the data spreadsheet completely excludes five countries: Australia, Austria, Belgium, Canada, and Denmark. The error can be seen below,
In the last row where it says AVERAGE(L:30:L44) it should say AVERAGE(L30:L49). This is an embarrassing mistake that should have immediately been caught by peer review and had the paper sent back to Reinhard and Rogoff.
So what are the actual results when all the errors are corrected and the model is run again? To quote Herndon, Ash, and Pollin, "the average real GDP growth rate for countries carrying a public debt-to-GDP ratio of over 90 percent is actually 2.2 percent, not -0.1 percent as [Reinhart-Rogoff claim]."
The 90% debt-to-GPD ratio was sighted by many as proof of the necessity of austerity in fiscal matters. Paul Ryan's 2012 "Path to Prosperity" budget plan cited the paper and specifically cited the 90% figure. European Union Economic and Monetary Affairs Commissioner Olli Rehn cited the study when urging EU member nations to cut their budgets, saying, "Serious empirical research shows that public debt above 90% of GDP acts as a permanent drag on growth." Dean Baker of the Center for Economic Policy and Research said of the errors, "How much unemployment was caused by Reinhart and Rogoff's arithmetic mistake?"
Reinhart and Rogoff issued a statement acknowledging the errors. They also reiterate that they were not attempting to push an austerity agenda. But in a piece written for Bloomberg in 2011, they came out against fiscal stimulus and in support of austerity.
No word if Paul Ryan is going to change his budget plan now that one its pillars is crumbling.