FAQ: Reinhart, Rogoff, and the Excel Error That Changed History

Photograph by Gregor Schuster

Harvard University economists Carmen Reinhart and Kenneth Rogoff have acknowledged making a spreadsheet calculation mistake in a 2010 research paper, “Growth in a Time of Debt” (PDF), which has been widely cited to justify budget-cutting. But the authors stand by their conclusion that higher government debt is associated with slower economic growth. Here’s what you need to know:

How big is this mistake?
Reinhart and Rogoff wrote in their 2010 paper that average annual growth was negative 0.1 percent in countries with episodes of gross government debt equal to 90 percent or more of GDP between 1945 and 2009. Liberal economists have been critical of their work for years (just economists being their usual cranky selves), but now three economists at UMass say Reinhart and Rogoff made several mistakes and omissions. According to the UMass scholars, the “corrected” number is positive 2.2 percent—which means GDP still grows, even when debt levels are very high.

Do Reinhart and Rogoff admit they got it wrong?
They admit they accidentally excluded five rows from an average in their Microsoft Excel spreadsheet, but not the other charges. Fixing the spreadsheet error would lift growth in those high-debt countries to about 0.2 percent annually (still not that good). Adding country data that wasn’t available when they did the 2010 paper would boost growth further, to perhaps 0.5 percent (ditto). The UMass economists get all the way up to 2.2 percent by using a counting method that gives more weight to a few countries that experienced long periods of high debt. Reinhart and Rogoff insist that their method is better.

Yes, and Reinhart and Rogoff argue that it’s beside the point anyway. They put more weight on other data covering longer time periods, which finds that growth is about 1 percentage point lower in episodes of high debt compared to when debt is below 90 percent of GDP. (U.S. government debt is over 100 percent of GDP when you include debt owed by one part of government to another, such as the Social Security trust fund.) They say that even the UMass researchers found that high debt and low growth go hand in hand.

You mean that high debt causes low growth?
Not necessarily. Reinhart and Rogoff are careful in their academic work not to claim causation. It’s possible that slow growth leads to high debt rather than high debt causing slow growth. (Or maybe the causality runs in both directions.) One complaint of their critics is that in Op-Eds, interviews, and other non-academic work, Reinhart and Rogoff have sometimes flatly asserted that high debt leads to slow growth when other explanations are possible.

Why does this academic dispute matter?
In politics, academic studies are used as weapons. The Reinhart-Rogoff work has been widely cited by deficit hawks, including House Budget Committee Chairman Paul Ryan, who said, “Economists who have studied sovereign debt tell us that letting total debt rise above 90 percent of GDP creates a drag on economic growth and intensifies the risk of a debt-fueled economic crisis.”

Reinhart and Rogoff are deficit hawks, then?
Yes and no. They are concerned about the long-term U.S. budget outlook, but they also say a serious deficit-reduction program should not begin until the economy is healthier.

What happens next?
Probably not much. Deficit hawks will argue that the kerfuffle is meaningless. “We can debate an exact dollar amount,” Douglas Holtz-Eakin, a former director of the Congressional Budget Office, says in an e-mail, “but the simple fact that debt ultimately hinders growth is unchanged.” On the other hand, economists and policymakers who never liked the Reinhart-Rogoff conclusion in the first place are having a field day. Paul Krugman wrote four posts about it in two days on his New York Times blog, one titled “Holy Coding Error, Batman.” As I wrote in a magazine piece this week, “It’s too soon to judge the political fallout, but this can’t possibly be good for Team Austerity.”

    Before it's here, it's on the Bloomberg Terminal.