It’s Time to Change Focus From Reinhart-Rogoff Witch Hunts to Krugman’s Contradictions

You may have seen the new skirmishes this weekend in the very public debate about Carmen Reinhart’s and Kenneth Rogoff’s government debt research. Reinhart and Rogoff (I’ll call them RR) posted a letter to Paul Krugman on Reinhart’s website, in response to a piece that Krugman wrote for The New York Review of Books.

Krugman, of course, is one of the pundits who last month published “incomplete, exaggerated, erroneous and misleading” reports about RR’s research, as I explained at the time. Unlike some of the others involved, he kept the smear campaign alive and all but guaranteed RR’s latest response. I’ll recap the weekend’s exchange in a moment, but not before offering my take on Krugman’s MO.

Observations on Krugman’s campaign for more deficit spending

I’ll start with an excerpt from my April 27 article, “More Reasons to Call Off the Reinhart-Rogoff Witch Hunt”:

[The] question of how much debt is too much is an extremely important question. And RR’s many critics make no attempt whatsoever to answer it (if I’m wrong about this, please send me the research), even as they bash two researchers who’ve been leading the way to possible answers since well before their 2010 paper.

It seems to me that you can say two things about the many pundits on this issue:

  1. Commentators in the U.S. who’ve both looked under the surface of our fiscal challenges (see the problems with official numbers in articles such as this and this) and seriously considered the question “How much is too much?” have come away alarmed.
  2. Of the commentators on the other side – those now crowing over the misinformation that’s spread all the way from Rortybomb to the Colbert Report – I don’t know of any who’ve attempted to do the same due diligence and answer the all-important question.

Well, I still haven’t found an RR critic who’s made a genuine effort to estimate how much debt is too much. But I did force myself to search a little further, and my search coincidently focused on Krugman. I read End This Depression Now, his 2012 book, from cover to cover.

This wasn’t my idea of a good read, but I try to make my reading list as balanced as I can stand. In this case, I was looking for counter-evidence to my assertions above. I wanted to see if there was anything more to Krugman’s positions than über-Keynesianism and boasts that his adversaries were proven wrong. He writes incessantly about inaccurate inflation and interest rate forecasts made by so-called “austerians” – those who’ve dared to express concerns about America’s soaring government debt – while arguing that these erroneous forecasts somehow prove his policy advice correct.

But his logic is full of holes. Near-term inflation and interest rate predictions have little to do with one’s beliefs about the mostly long-term risks of excessive debt. I don’t doubt that some forecasters have made faulty predictions. I know for a fact, though, that there are many who’ve opposed Krugman’s deficit spending recommendations while at the same time either accurately forecasting inflation and interest rates or not making any predictions at all. By my estimates, the second group is large and the first not so much.

How Reinhart and Rogoff Spoiled My Holiday

So, what does all of this have to do with RR’s letter to Krugman?

Well, just when I’d gotten through End This Depression Now and thought I could go back to reading stuff that makes sense to me, I learned of the RR post and felt compelled to read through Krugman’s latest attacks. And then I felt compelled to write – not just to offer my two cents on this weekend’s scuffle, but also to draw some parallels between the current debate and Krugman’s book. So much for my planned holiday reading of either The Bankers’ New Clothes or more of The Great Deformation. (I know, I’ll get over it.)

In any case, here’s my list of takeaways from RR versus PK, Memorial Day weekend edition:

  • Using Internet archives (the “WayBackMachine”), RR demonstrated that their data was publicly available as far back as 2010, contrary to PK’s accusations that this data was withheld. (As I said in earlier posts, I’ve downloaded their spreadsheets in the past and knew they were available.)
  • RR showed that they’ve recommended policies intended to reduce austerity, contrary to their portrayal by PK as tireless austerity advocates.
  • RR counterpunched on the question of debt-to-growth causality. (And yes, everyone recognizes this isn’t determined wholly by correlation.) They responded to the accusation that they put too much weight on the causal effects of high debt on growth with an argument that PK puts too much weight on the causal effects of low growth on debt.
  • Krugman eventually retreated to a core position that RR should have prevented impressionable policymakers from over-interpreting their 90% debt-to-GDP threshold. (I offered my opinion on these accusations here.)

For longer summaries, I recommend reading Econbrowser (James Hamilton) and economicprinciples (David Warsh).

Here’s the discussion we should be having

Now I’ll jump from RR versus PK to F.F. Wiley versus PK. (And no, I don’t expect a response, but I’ll say my piece anyway.)

Not surprisingly, End This Depression Now reads like a longer version of Krugman’s blog. If you’re one of his followers, you’ll be even more convinced that poor inflation and interest rate forecasts by the “austerian” crowd provide the evidence needed to justify massive deficit spending. Never mind those of us who’ve forecast inflation and interest rates correctly, or not at all, and yet still believe his fiscal policy views are too extreme.

As I said above, I couldn’t find a genuine effort to estimate how much debt is too much. But Krugman’s book does include a section titled “What about the Burden of Debt?” This is two pages long, running from the top of page 141 to the top of page 143. Here’s an excerpt:

The key thing to bear in mind is that the $5 trillion or so in debt America has run up since the crisis began, and the trillions more we’ll surely run up before this economic siege is over, won’t have to be paid off quickly, or indeed at all. In fact, it won’t be a tragedy if the debt actually continues to grow, as long as it grows more slowly than the sum of inflation and economic growth.

To illustrate this point, consider what happened to the $241 billion in debt the U.S. government owed at the end of World War II … this amounted to about 120 percent of GDP (compared with a combined federal, state, and local debt of 93.5 percent of GDP at the end of 2010). How was that debt paid off? The answer is that it wasn’t.

Krugman goes on to discuss post-World War II debt developments, while providing some incredibly misleading debt service calculations, but I’ll leave those for another day. (Update: I subsequently wrote about the debt service calculations here.)  I’ve shared the excerpt because it seems to tie into the RR debate. Consider the following:

  • By using the post-World War II experience to dismiss our current debt problems, Krugman essentially suggests a debt threshold of 120% of GDP. He says nothing about debt above 120%, but he’s suggesting that we shouldn’t worry as long as we’re below 120%. It’s a threshold, all the same.
  • For all the bellyaching about RR’s data choices, they’ve built the world’s most extensive government debt database (to my knowledge). Their most recent paper on debt and growth, published in 2012 and largely ignored by their critics, examined 26 high debt episodes in 22 countries. Krugman’s threshold, on the other hand, is based on one high debt episode in one country.
  • There are many reasons to believe that our World War II debt was far less troubling than today’s challenges. One of these is that we were running gigantic non-defense budget surpluses, which meant that we balanced the budget after the war by merely bringing our soldiers home and returning factories to civilian use. As I discussed here and here, balanced budgets and financial repression were the critical ingredients in our success at whittling away war debt, and were more important than the “mild inflation and substantial economic growth” cited by Krugman.
  • Outside of the natural slippage that occurs in recessions, our post-war leaders truly detested budget deficits. Presidents Harry Truman and Dwight Eisenhower restored surpluses within two years of every 1940s and 1950s recession, while mostly dismissing Keynesian theories that were being shaped in academia at that time. Truman even raised taxes to help pay for the Korean War (what a novel idea!), while Eisenhower went so far as to claim that “continued deficit spending is immoral.”

In other words, there are several layers of contradictions in Krugman’s reliance on the post-World War II period to support calls for more stimulus. Perhaps most glaringly, the reduction to a 60% debt-to-GDP ratio by 1957 (about half of the 1946 figure) was achieved partly by avoiding the types of Keynesian stimulus measures that have since helped push debt back above 100% of GDP. If Krugman’s ideas about deficits existed at all in the 1950s, they certainly had no influence in policymaking circles, where old-fashioned principles of fiscal discipline still held sway. And without Truman’s and Eisenhower’s un-Krugman-like beliefs, the success story that he cites as a reason to be unconcerned wouldn’t have unfolded the way that it did.

Krugman’s use of this decidedly non-Keynesian episode of debt reduction to justify his Keynesian beliefs reminds me of the many times (too many) that I’ve encountered “circular reference” errors in Excel. His logic is no less flawed or “circular.” But since there’s no spreadsheet involved, I’ll call it an error of induction.

Getting back to RR versus PK, Krugman refuses to walk away from a smear campaign that’s based on overblown accusations of questionable thresholds, selective data use and a spreadsheet error.

I think it’s time to change focus and consider the questionable thresholds, selective data use and induction error in Krugman’s work.

Bookmark and Share
This entry was posted in Uncategorized and tagged , , , , , , , , . Bookmark the permalink.

5 Responses to It’s Time to Change Focus From Reinhart-Rogoff Witch Hunts to Krugman’s Contradictions

  1. lemmy caution says:

    The issue with the 90% threshold was that going over it under RR’s old model was a big problem. Things went to crap so quickly once debt-to-GDP ratio got above 90% that 90% was effectively a hard cap for reasonable policy makers. That hard cap threshold goes away once the data is interpreted correctly.

    You are just making up the part about krugman having a threshold of 120%. There is nothing magical about 90% there is nothing magical about 120%. Lower debt-to-GDP ratio is better than higher debt-to-GDP ratio but there is no discontinutity. That is something RR just made up so asking where krugman’s discontinuty point is is kind of silly.

    You are right that the cut spending during good times part of Keynesian is the hard part, but at least then it will be the good times.

  2. ffwiley says:


    Thanks for reading and commenting.

    I didn’t address the whole concept of thresholds in this article because I did this in my earlier posts, but I’ll take another stab at it here.

    First, I would say that most laypeople and virtually all academics (one would hope) understand that economics is one of a handful of fields in which there’s no such thing as magic numbers. When academics talk about thresholds in these fields, they don’t mean magic numbers. Nor do they necessarily mean that when you chart two variables – say, debt and growth – you’ll see a sudden discontinuity.

    Some examples: Doctors don’t expect to see a discontinuity in heart disease risk at a cholesterol reading of 200 mg/dL , even though the American Heart Association advises people to keep their cholesterol below this threshold. Public safety experts don’t expect to see a discontinuity in automobile accidents at a blood alcohol level of 0.08%, even though this is the legal limit for driving while impaired. And economists shouldn’t have expected to see a discontinuity at 90% debt/GDP. As I said in an earlier article, the scatter plot included in the famous critique of RR’s 2010 paper is exactly the picture that a competent economist would have expected to see based on the results that RR reported.

    So why do we have these markers if they don’t pop out at you on a scatter plot?

    I think the answer is that people need markers to help them manage complex risks, and they look to professionals to provide them. Whether you’re talking about cholesterol or blood alcohol or government debt, it doesn’t do much good to say that these things are bad and stop there. In fact, they’re all acceptable in moderation. The key question is: How much is too much? If a researcher in any of these areas isn’t trying to answer that question, then their research probably isn’t of much use.

    And that’s my long way of saying that there was nothing out of the ordinary in RR’s approach. There’s a clear correlation between low growth and high debt, and it’s natural to estimate some threshold beyond which that correlation becomes meaningful. This always involves subjectivity and interpretation, but that’s the nature of the beast. That’s why economists often welcome competing estimates and report them alongside their own, exactly as RR did in their papers. And RR’s 90% happened to be in close proximity to other researchers’ estimates. So not only is it quite normal to flag a threshold when you’re publishing research on risk, but there were other researchers applying the same approach to the debt-growth relationship and coming up with almost identical recommendations.

    As long as you accept that the term “threshold” isn’t a synonym for “discontinuity,” then Krugman has certainly offered one. As I said in the article, when he claims there’s nothing to worry about because we’ve been at 120% before and did okay (and you’ll see this in his blog posts as well as his book), that’s no different than RR suggesting that we should start worrying at about 90%. And I’m not faulting the use of a threshold in either case. I’m just suggesting that 1) Krugman’s criticisms of RR’s approach are hypocritical, and 2) his 120% threshold is highly flawed, so much so that I wrote an article about it two months ago (“7 Fallacies About the Lengths of Things”).

    Finally, I doubt that I’ve said anything that Krugman doesn’t realize. The smear campaign that he’s helped lead is a clear attempt to encourage policymakers to go deeper into debt by discrediting RR (and the authors of the original RR critique said it about as clearly as that). That’s why it’s important to separate spin (and pure fabrication in some cases) from substance.

  3. Labropotes says:

    The most troubling aspect of this is how hard it is to drive misinformation from the public dialog once introduced. PK has propagated nonsense at a far lower cost than what is needed to round it up again.

  4. Followed from DK says:

    “I know for a fact, though, that there are many who’ve opposed Krugman’s deficit spending recommendations while at the same time either accurately forecasting inflation and interest rates or not making any predictions at all. By my estimates, the second group is large and the first not so much.”

    Why in the world would you link those two groups in the same sentence? It’s nonsensical.

    “I know for a fact that there are many who opposed the Yankees by playing against them in Red Sox uniforms or by watching on televion. By my estimates, the second group is large and the first not so much.”

    • ffwiley says:

      Thanks DK, let me clarify. The second group consists of those who’ve made accurate forecasts or none at all. The first group consists of those whom PK calls out for bad forecasts (from the sentence before the one that you included in your comment). You read the first group differently and I can see that my wording can be confusing – if I see a quick edit that works I’ll make it.

Comments are closed.