October 10, 2013
In a blog post at Econbrowser, Princeton Economist Angus Deaton complains about how his work on the impact of inequality on health outcomes was challenged by Michael Ash, an economics professor at the University of Massachusetts. He compares this challenge to the more recent challenge posed by Ash, along with Thomas Herndon and Robert Pollin to the work of Harvard professors Carmen Reinhart and Ken Rogoff. People may recall in this latter work, Herndon, Ash, and Pollin showed that the results in Reinhart and Rogoff’s original 2010 paper on the relationship between national debt and growth were due to both an Excel spreadsheet error and a peculiar method of aggregation.
Deaton joins the criticisms of the two papers:
“In our case, as in Reinhart and Rogoff, neither the coding error (in our case there was none) nor the choice of weights has any effect on the main results. … With Reinhart and Rogoff, they referred only to an early paper, ignoring updated results. But the effect is the same, to magnify a tiny or non-existent problem and claim that it threatens the whole enterprise whereas, in fact, nothing of the sort is true.”
Deaton then complained that the criticisms of Reinhart and Rogoff did not even take place in refereed journals, but rather through blog posts (yes, I’m one of those guilty) and the media.
Perhaps Deaton is unaware of the impact of Reinhart and Rogoff’s work on the debate over stimulus and deficits. Otherwise it is difficult to see how he can trivialize the importance of the criticisms from Herndon, Ash, and Pollin (HAP) to the public debate on this issue.
Contrary to what Deaton implies, Reinhart and Rogoff highlighted the idea of a cliff at a debt to GDP ratio of 90 percent from their first paper. Above this level, their earlier work showed a sharp falloff in growth. This paper had enormous impact on policy debates in Europe and the United States. In fact, it was the explicit basis for the debt targets set out by Erskine Bowles and Alan Simpson, the co-chairs of President Obama’s deficit commission.
While Reinhart and Rogoff’s later work did not provide evidence of such a cliff, this conclusion from their original paper continued to guide public debate uncorrected by Reinhart and Rogoff. The spreadsheet error uncovered by HAP managed to bring this issue into the public spotlight. When the error was corrected, any basis for claiming a large growth penalty for debt to GDP ratios in excess of 90 percent disappeared. There was in general a negative correlation between growth and debt levels, but the sharpest tradeoffs were at relatively low levels of debt (under 30 percent of GDP), not the 90 plus level highlighted by Reinhart and Rogoff.
HAP also inspired a series of papers that examined the direction of causality. All of these papers showed evidence that the causation overwhelmingly went from slow growth to debt rather than in the other direction. In other words, countries that were growing slowly tended to accumulate lots of debt, rather than the other way around.
The result is that policymakers and the general public now have a much clearer view of the potential limits posed by debt. The fact that this exchange of analysis among economists occurred outside of professional journals would seem to be a criticism of professional journals and their limited relevance for policy debates. If a side effect was that two prominent Harvard economists were publicly embarrassed, that seems a small price to pay.
Addendum:
Reinhart and Rogoff might have saved themselves and the rest of the world much grief if they had made their data publicly available in January of 2010 when their work first began to have a major impact on policy debates.
Comments