Why Paul Krugman, Others Think Reinhart and Rogoff Are Wrong About Debt

BY Paul Solman  April 19, 2013 at 2:36 PM EST


Photo by Andrew Harrer/Bloomberg via Getty Images.

Instead of answering any questions Friday, I should probably address the hot topic in the world of economics this week and the subject of Paul Krugman’s column in The New York Times this morning, the Reinhart-Rogoff 90 percent controversy.

Very briefly, Carmen Reinhart is a big deal economist at Harvard University and Ken Rogoff, former International Monetary Fund economist, currently at Harvard, is an even bigger deal. They are honorable and exhaustive researchers who deserve their prominence. Their leap to internationally public prominence came after the crash of 2008, when they published an improbably best-selling book on the history of financial crises, “This Time is Different,” whose ironic title delivered the punchline: if a country borrows too much, it’s cruisin’ for a bruisin’.

As Rogoff told me in a profile of their book on PBS NewsHour, history suggested that the dire debt situation in the U.S. prior to the crash was unsustainable: “When you have a big inflow of foreign funds — and we had a massive one — you’re at risk. When you deregulate your markets rapidly, which we did in the States, that also very often happens that you have a deep crisis. The real killer is short-term debt, debt that has to be refinanced all the time. Well, that’s what the subprime [boom] was.”

From a policy point of view, the Reinhart-Rogoff prescription came a year later, with the publication of a hugely influential academic paper that found a threshold level of government debt as a percentage of the economy — 90 percent — above which economic growth falls significantly. The 90 percent threshold became a cliche in policy circles: a malign tipping point that dooms the debtor. And with U.S. government debt now around 100 percent (by the broadest measure), the sky must be primed to plummet. Better stop borrowing!


Thus, the 90 percent threshold demanded a prescription of austerity: cut back spending instead of borrowing more.

Unfortunately for austerity enthusiasts, the Reinhart-Rogoff findings turn out to be wildly overstated, if not just plain wrong. A new paper, using the original Reinhart-Rogoff data sets, asserts that it “contradicts Reinhart and Rogoff’s claim … that public debt loads greater than 90 percent of GDP consistently reduce GDP growth.”

Reinhart and Rogoff have responded, admitting their analytic errors, but insisting that their overall result is unsullied: government debt past a certain point is almost surely bad news.

So, finally, what lessons can one reliably learn? I’d highlight four:

  1. Economists these days — and especially economic historians — live for conclusions they can draw from data. If you can’t draw conclusions, you don’t have anything to publish — not a PhD thesis, or a paper that you need to get a job, or a promotion or a contribution to the field.

    And obviously, the more dramatic and policy-relevant the conclusion, the better. But the first thing I learned in graduate school remains as true as it was 47 years ago: beware the plasticity of data to “reveal” something apparently significant and satisfy your urge to find something that matters.

  2. The second thing I learned in graduate school, quite literally, is also true: Correlation does not mean causation. That is, in this case, that high government debt may be a result of slow economic growth, not its cause.

    A post by University of Massachusetts economist Arindrajit Dube makes this point brilliantly but all you have to do is look at Japan, a country with a government debt: GDP ratio of more than 200% — debt taken on almost wholly to stimulate the economy in the wake of a major recession. (A side note: that hasn’t worked either.)

  3. History teaches us plenty. As the saying goes, you can’t tell where you are unless you know where you’ve been. As Reinhart and Rogoff observe, there’s a great danger to saying “This time is different.” But, to use another well-worn truism, past performance is no guarantee of future results. To contradict Mark Twain, not only doesn’t history repeat itself, it often doesn’t even rhyme. History is people interacting with one another — millions and millions of people. You never know what they’re going to do.
  4. It follows from lessons one, two and three that we should all be careful not to take any finding in the social so-called sciences too seriously. Government debt at any given level of GDP may be ominously high. But then again, it may not be. On that statement, I’d bet the ranch.

This entry is cross-posted on the Making Sen$e page, where correspondent Paul Solman answers your economic and business questions