Thursday, February 20, 2014

Krugman back in 2009 on the downturn

Bob Murphy thinks I defend Krugman no matter what no matter how many times I end up disagreeing with Krugman on his blog, so he's going to think this is a predictable post. But oh well. I think it's right.

Recently Bob claimed that Krugman is "rewriting" his stimulus history. The issue at hand is the old unit roots debate between Krugman, DeLong, and Mankiw. Now DeLong actually said that he trusted the CEA forecast at the time (more on that a little later), but Krugman didn't. His post stuck pretty closely simply to what we think about the properties of different time series with respect to unit roots. It's not even like he left his view about the possibility of extended crisis unstated - he said that we can expect output to grow "if and when" slack capacity was used again. "When", sure - but "if and when"!?!?! This is not the writing of a guy that thinks there's no realistic possibility that we're not going to bounce right back!

And anyone that follows Krugman closely (and I always thought Bob was one of those guys) should know why Krugman said "if and when". From the beginning he has been bearish on the administration's forecasts.

He was talking about getting into a deflationary trap where we can't get out of slow growth two months before the unit root flap, when the initial administration forecasts came out:
"So tell me why we aren’t looking at a very large risk of getting into a deflationary trap, in which falling prices make consumers and businesses even less willing to spend."
But macro is hard and maybe a deflationary trap doesn't explicitly say, to Bob's standards, whether the administration forecasts are too optimistic and that there is a chance of a much worse crisis. Bob thinks Krugman is changing his tune and that:
"Now all of a sudden, Krugman is claiming that he was skeptical of the CEA’s early-2009 optimism about the recovery from the deep slump!"
How could anyone seriously think Krugman thought the administration was too optimistic??? Maybe from this (also two months before the unit roots debate):
"One more point: the estimate of what would happen to the economy in the absence of a stimulus plan seems kind of optimistic. The chart above has unemployment ex-stimulus peaking at 9 percent in the first quarter of 2010 and coming down through the year; the CBO estimates an average unemployment rate of 9 percent for 2010, so the Obama people are more optimistic than the CBO, and a lot more optimistic than I am."
MAYBE we got the idea that Krugman thought the administration was too optimistic because of that time early on when Krugman was saying it was too optimistic while everyone else was saying he was being too bearish. Or perhaps I'm just some hack that will defend Krugman on anything.

Look, even the readers of Newsweek got it. The readers of Newsweek. Why is Bob acting like it's absurd to say that Krugman thought the administration was wildly optimistic?:

This was Krugman's 2009 (much less fun than his 2008!): liquidity trap diagrams and being bearish relative to the powers that be.

So the unit root debate, to me, was about unit roots. Not about how bad the crisis was. Because Krugman was pretty clear on what he thought about how bad the crisis was and particularly how we could be stuck in it for a long time, and nothing in his unit root post overturned any of those arguments.

Which brings us to Brad DeLong. I also admire Brad a great deal, but he did say back in March 2009 that the safe way to bet was with the administration forecasts.

Here's the thing, guys. He readily admits that early on in the crisis he did not see all this coming. Bob often frets over Brad's rebuke to him that he should mark his beliefs to market and sit "at the feet of Paul Krguman, chanting 'om mani padme hum' until I achieve enlightenment." But here's why Brad says that - because that's where he had to drag himself when things finally set in about this crisis. It's where a lot of us on "Krugman's side" had to drag ourselves. I think because Brad is so brash (or I suppose "shrill" is the word of choice) in his blogging a lot of people miss how unbelievably humble he is. He regularly blogs about what he got wrong and how he is changing his views to mark them to market. And anyone that follows Brad knows that this issue of the severity of the crisis was a very important case where Brad used to be not-so-bearish and has changed his views.

Here is a particularly thoughtful and detailed admission of wrongness from Brad in 2011:
"Although I worked for three years in the Clinton Treasury Department, and am a card-carrying member of the economist guild, I predicted none of this. Like most of my peers, I was wrong. Yet the most interesting thing is that I could have -- should have -- been right. I had read economist John Hicks; I just didn’t quite believe him. 
Hicks, one of the clever young Brits dotting i’s and crossing t’s in the writings of John Maynard Keynes in the 1930s, was responsible for the workhorse formulation of Keynesian economics -- the IS-LM model -- that has been the bane of many an intermediate macroeconomics student. It was his version of the IS-LM model that formalized and elevated a key insight: that interest rates paid by creditworthy governments would remain low after a financial crisis. This formulation holds even in the face of enormous budget deficits that greatly expand the supply of government bonds. 
A financial crisis initiates a sudden flight to safety among bondholders -- widening interest-rate spreads, diminishing the private sector’s desire to sell bonds to raise capital and encouraging individuals to save more and consume less as they, too, hunker down. Thus bond prices rise, and interest rates drop. As rates fall, firms see that they can get capital on attractive terms and so issue more bonds; households see the low interest rate earned on their savings and lose some of their desire to save. The market heads toward equilibrium... 
I had read Hicks. I even knew Hicks. But I thought that his era, the Great Depression, had passed. Sitting in my first graduate economics class in 1980, I listened to Marty Feldstein and Olivier Blanchard -- two of the smartest humans I am ever likely to see -- assure me that Hicks’s liquidity trap was a very special case, into which the economy was unlikely to wedge itself again. Yet it did. 
On my shelf is a slim, turn-of-the-millennium volume by Paul Krugman titled “The Return of Depression Economics.” In it he argued that we mainstream economists had been too quick to ditch the insights of Hicks -- and of economists Walter Bagehot and Hyman Minsky. Krugman warned that their analysis was still relevant, and that if we dismissed it we would be sorry. 
I am sorry." 

 So please - let's drop this.

You might be able to squeeze a blog post or an Econ Journal Watch article out of this, but nobody that follows Paul and Brad is buying it.


  1. Krugman wrote about how conservatives read until they come across something they dislike and stop, or assume if they have a position their opponent must take the opposite because if they didn't there would be no disagreement. Perhaps they just don't read him at all but excerpts of him from their compatriots. That always works to understand someone /sarcasm

  2. Daniel, what percentage of Krugman's readers do you think read his post on DeLong/Mankiw and came away thinking, "Paul thinks that DeLong/CEA are totally wrong in their forecast and Mankiw is right"?

    If you agree that his post was unbelievably misleading, and was part of "Go Team Keynes," then I will agree that he is not now re-writing history regarding his *other* statements about the CEA at the time.

    1. I think the unit root post did not explicitly weigh in on the depth of the recession. I believe there were definitely people that probably came away thinking that from that post. But I think those people had not paid very close attention to everything else he wrote in - say - the first six months of 2009 about the Obama administration's take on the crisis and the depth of the crisis.

      I also think they probably disregarded the "wonkish" label, which means he is talking at least somewhat technical economics that is likely not accessible to some of his audience.

      This was March. A lot of his criticism of the administration came first thing in January. If in April you took a poll of Krugman readers and asked if Krugman thought the administration was too optimistic or too pessimistic about how serious the crisis is, the very large majority of them would say that Krugman thought the administration was too optimistic.

      I don't know if this satisfies your trade, but I don't really think a trade is necessary. If you think it, say it. Don't wait on me to say something. And if you don't think it, don't say it even if I say something. What you think should not be conditional on what I think.

    2. Bob, let me repeat what I said on your blog:
      "No, I don’t think readers would be able to gleam Krugman’s skepticism of the Romer estimate from that post, but that’s just because his intention in the post was not to assess the plausibility of the CEA estimate, but rather to criticize what he perceived to be bad reasoning on the part of Mankiw. And that’s why he seconded Delong’s sigh, because he was concurring with Delong’s assessment that Mankiw’s analysis was “deliberately obtuse”."

    3. I don't think the post was misleading, but I do think it was "Go Team Keynes" in the sense that the message was "Look how deliberately obtuse the people on the other team are being in their arguments."

  3. Daniel, I agree that it is not correct to say that Krugman was particularly bullish in the early days of the recession. However, there are good reasons to emphasize this episode and to deny its relevance might involve a bit of...deliberate obtuseness.

    1. This isn't some nobody who challenged Krugman to a bet; this is Greg Mankiw. He was blogging more frequently then and had one of the most popular blogs. Surely Krugman heard about Mankiw's response. Why didn't he clarify his position? Now we're stuck here five years later and no matter what happened, Krugman can claim to be right.

    2. The context of these administration growth projections is the estimation of future deficits. The logic is that if the future is bright, we should borrow money today and finance it with future taxes. If Krugman believed the administration's outlook was too rosy at the time, he should have also been emphasizing the underestimation of deficits that we are experiencing today.

  4. It's funny, I was in my undergraduate intermediate macroeconomics class at the time and we did some very basic calculations on the board, where we took the shortfall in demand as measured by change in gdp for the previous two quarters, and then calculated what the stimulus size would need to be in order to counteract the shortfall in demand. At the end we came up with the number ($700 B per year). My teacher than said, so it appears that the stimulus offered by the administration was exactly in line with what would be needed given a shortfall.

    Given my readings of Krugman (I had just started following him at the time), I knew the sentence my teacher had just spoken was wrong, wrong, wrong. The "per year" part was the key. The stimulus was scheduled to release over a span of roughly 3 years, so the stimulus was approximately 1/3 the size it needed to be given the shortfall in demand from a Keynesian standpoint. At the time I was not a fervent Krugman supporter, but over the intervening three years I found that what Krugman had described in January of that year 2009, was more or less exactly on target with what occurred (yes, except for deflation which was not the key point in Krugman's analysis). The point I'm trying to make is that, even as an undergraduate, it was clear to me at the time from my passing reading of Krugman, that Krugman was screaming at the top of his lungs about the inadequacy of the stimulus. I don't know how he could be read in January any differently.

  5. Daniel D, I am being dead serious, you have made my point exactly. (Not sure if you clicked through the links or are just relying on Daniel Kuehn's summary.) Of course Krugman from Day #1 was saying the stimulus was too small. But my point is, if you go and read his arguments, he wasn't saying, "I think their baseline trend is way too optimistic, so we need a much bigger stimulus." Rather, he was saying, "Even on their own terms, unemployment is still going to be high in 18 months, and the Republicans are going to have a field day."

    So your statement here backs up what I am saying.

    (Granted, it gets muddier because--as Kuehn rightly points out--Krugman *also* thought the Romer/Bernstein baseline forecast was too optimistic. But that wasn't the foundation of his now-famous warnings that the stimulus was too small and would lead to Republican gloating.)

    1. @Bob,

      Wait, what? So because he said both but one more loudly than the other he doesn't get credit for both? I read his posts as saying, for sake of argument let's assume that we take your overly optimistic viewpoint as the truth (not that he was agreeing with their forecast, just that he was conceding that point because forecasting is really hard and even if he was somehow wrong and cea was right he could make an argument for stimulus anyway), than your stimulus is still too small. Maybe he didn't say this in every post where he was making an argument for stimulus, but read in the context of all the other posts he had made at the time I was taking away from it what Daniel Kuehn was taking away from it. 1. That their forecasts were overly optimistic, and 2. even if they're not the stimulus is still too small. Sometimes to win an uncertain argument you have to make multiple points given that one of your assumptions could be wrong in order to make your case stronger.

    2. @Daniel D - don't expect a fair reading of Krugman from Bob. He is blinded by his hatred of Krugman.

  6. A very good defence of Professor Krugman, Daniel Kuehn. Unfortunately, it seems that he and other like-minded economists were proven correct.

  7. Are you guys for real? I am going to post an update on my blog about all this, because Daniel raised some good points; it wasn't right of me to say "Krugman is now rewriting history."

    But we have to say SOMEthing about his handling of Mankiw vs. CEA and DeLong. If Krugman actually thought the CEA and DeLong were crazy for predicting a recovery in output in a few years, and that Mankiw's analysis was totally right except for labeling it "unit root hypothesis," then Krugman's post at the time truly should win an award for the art of misleading.

    1. "Are you guys for real?"

      I don't know if you are including me in the "guys". "Absalon" is, of course, an alias but behind it there is a very real person who means every word. There is something fundamentally wrong with your approach to Krugman's writings.

  8. There is a typo at the top of the blog - no closing quote mark on the keynes quote
    Also, this should not be a quotation, but attributed to; ifaik, there is no published source for this quote

  9. Daniel,

    FWIW, I read Mankiw, DeLong, and Krugman on this issue back in 09 and thought at the time that Krugman was agreeing with DeLong and disagreeing with Mankiw on the underlying question of whether the administration's forecast was overly optimistic. Mankiw presumably had the same misimpression, because he offered to bet Krugman on the path of the economy.

    Reading Krugman's other pieces, I think it's pretty clear that you're right that Krugman probably did agree with the second premise of Mankiw's article and with the conclusion, or would have agreed if he had read it more closely. My wild guess is that Krugman read DeLong's piece, and either didn't read Mankiw's or didn't read past the first few paragraphs, so he assumed that Mankiw's argument rested solely on unit root. It would have been nice for clarity if he had responded to Mankiw's bet offer by clarifiying what he had meant, but maybe he didn't hear about it.


All anonymous comments will be deleted. Consistent pseudonyms are fine.