Reinhart and Rogoff shown to be wrong -- again

By Frank Moraes 

There was a bit of economic news yesterday that was wonderfully fun and I dare say it would be a shame for you all to miss out on it just because you don't follow the economics blogs. What's more, in terms of truth, it really matters, even if, in terms of politics, it probably won't.

In 2010, Carmen Reinhart and Kenneth Rogoff, two economists at Harvard, published a hugely influential paper, "Growth in a Time of Debt" (pdf). In it, they purported to show that if the total government debt equaled more than 90% of the country's GDP, then there would be a big slowdown in economic growth. This paper was used more than any other to justify economic austerity all over the world, but especially in Europe and the United States. In fact, Paul Ryan, in his The Path to Prosperity (pdf) budget, claimed, "The study found conclusive empirical evidence that gross debt exceeding 90 percent of the economy has a significant negative effect on economic growth."

That sounds bad, right? In fact, if you look at the paper, it is even worse. Countries with a debt-to-GDP ratio of less than 30% see a growth of 4.1% whereas those with a ratio of more than 90% see a contraction of 0.1%. Of course, long ago, I had dismissed the paper for two reasons. (But the fact that I even knew about the paper should give you some idea of just how important it has been.) First, correlation is not cause. And in this case, the cause is probably the opposite of what Reinhart and Rogoff (R&R) suggest. When economies are bad, governments have to borrow money. The case is much stronger that bad economies lead to large debts rather than the other way around.

The second reason I didn't accept this paper comes from an idea that Dean Baker has been drilling into my brain the last few years: debt-to-GDP ratios are meaningless. What matters is the interest burden of the debt. A government will have no more difficulty managing a 50% debt-to-GDP ratio at 4% than it will a 100% dbt-to-GDP ratio at 2%. Or let me personalize it: would owing $2,000 on a credit card really matter if the bank only charged you $5 per year in interest?

There were other problems with the paper. Paul Krugman, as I recall, questioned the study on the fact that its entire case depended on two periods of contraction in the United States and (especially) New Zealand. But now, all the wheels seem to be falling off the R&R cart. Three researchers, Herndon, Ash, and Pollin (HAP) of the University of Massachusetts, Amherst, have published a paper that finds major problems with this earlier research. And the best part: the biggest finding of R&R -- the 90% threshold -- went away when HAP corrected a spreadsheet error R&R had made. They accidentally cut 25% of the countries in their database out of their calculations. And looking at it, they should have known. But I think when they saw what they wanted to see, they didn't probe. And that's pretty funny when you consider all the policy makers who did the same thing with their paper.

If you are interested, Mike Konczal over at The Next New Deal wrote an excellent overview of all that's going on. But the truth is, the fact that the R&R paper was build on feet of clay won't matter at all. Paul Ryan and the other conservatives who pushed austerity didn't do so because they were convinced that R&R were right. Rather, they already had their agendas, and just quoted R&R to justify it. In Europe, austerity proponents will probably stop quoting it -- but it won't change their minds. In America, they will just keep quoting it. Because, hey, facts have a well-know liberal bias.

(Cross-posted at Frankly Curious.)