31
Dec
2014
Krugman: I’m Only Keynesian to Botch 1980s Inflation Call
This one was too easy. He made the same faux pas in 2012. Krugman says it’s a myth that Keynesians were surprised by the 1980s disinflation, when there is an infamous 1982 Council of Economic Advisors memo in which he did just that.
You list three possible explanations for Krugman’s post. Isn’t there a fourth one ? Keynsians in general had a model that successfully predicted the outcome of the Volker policies, but that he specifically failed to apply that model, but (as he predicted higher inflation, not a sharper fall in inflation like the monetarists of the day) from a totally different perspective to the Lucas camp.
I would classify that under my (1) though Transformer. If Krugman remembers full well that he, Larry Summers, and other prominent non-Chicago-School guys at the time predicted high inflation and were surprised by the results, then his current tut-tutting of right-wing reflections on the period is incredibly dishonest if he carefully worded his blog post to walk through the minefield in the way you are describing.
It is certainly a bit odd that he would choose to lambast right wing economists on a subject where there appears to be documented evidence that he himself also got wrong.
I’m just not sure that one can conclude from this that Krugman is either “utterly dishonest and knows full well he is re-writing history” or forgetful (assuming we can ignore option 2), which is why I would be charitable and assume that he is talking abut Keynesians in general and excluding his own specific error (which it is possible he has simply forgotten about).
I don’t want to admit it but I’m beginning to feel sorry for Krugman after these Contra Krugman posts.
Timid guys finish last.
For all the hypocrisy and sloppiness from PK, I cannot help but think these repeated “victories” are phyrric.
For even if inflation had reaccelerated, it would not have been because of a causal process that ANY positivist model claims exists, irresspective of the political ideology of Krugman or his opponents. For let us assume that inflation did reaccelerate, but not because of any historical facts known up until the minute before PK made his infamous botched prediction in 1982, but because of some subsequent actions that took place after the prediction that prevented the expected results. PK of course would declare victory anyway, as the relationship in the model is true but only if no subsequent actions take place that would otherwise ruin the expected results. He’d point to some model somewhere in Keynesian textbooks that leave room for disinflation after the same events took place. Oh look, that is exactly what he just did.
See, the thing about the positivist mindset is as an intellectual weapon. Strictly speaking, every model is hypothetical and remains forever hypothetical. But ideological battles require certainty in the face of opponent attacks. So there is a need be certain in what they think, which is not allowed in positivism (well, superficially anyway, because the root of positivism is absolute certainty in its structure and it being the best method), but engage in that “motte and bailey fallacy” by becoming skeptics of their own ideas when pressed, and then when the opponents feel victory is at hand and scatter, the positivists go right back out into the bailey and make the same weak, flawed claims, where it is safe to do so with one’s allies. People who argue like this are most antagonistic towards having their worldviews seriously challenged, because the motte is dark, dirty and not fun to constantly retreat to every time the claims made in the bailey are questioned.
Bob, where is the definite article in that sentence this is gonna bug the crap out of me.
Nice leading with the chin metaphor,
Can you clarify what the memo has to do with the adaptive expectations Phillips Curve? I don’t get it.
I read it as Krugman pointing to future supply shocks impacting the price level. I don’t know nearly enough about commodity prices or exchange rates in the 80s to say how reasonable it was. But I do know that there’s nothing in there that’ obviously inconsistent with a expectations augmented Phillips Curve.
All Krugman is saying with Lucas I think is that Keynesians expected more of a lag than strict RE people did and that seems obviously true.
This seems to boil down to Nelson on the Simpsons yelling “haha – Krugman thought there would be more inflation” with absolutely no effort to think about the actual theoretical claims he’s referencing.
OK Daniel I’m going to write the following in 2030 and I expect you to rush to my defense when someone calls BS on me:
Left-wing economists like Paul Krugman and Brad DeLong — it’s becoming ever harder to tell the difference — have some curious beliefs about history. One of those beliefs is that the experience of low inflation after 2008 was a huge shock to Austrians, refuting everything they believed. What makes this belief curious is that it’s the exact opposite of the truth. Austrians came into the Bernanke/Yellen lowflation with a standard, indeed textbook, model of what should happen. And events matched their expectations almost precisely.
[Then I insert a screen shot of Rothbard talking about supply and demand in the market for cash balances.]
So were Austrian economists feeling amazed and dismayed by the events after 2008? On the contrary, they were feeling pretty smug: lowflation had played out exactly the way the models in their textbooks said it should.
Indeed, it was the other side of the macro divide that was left scrambling for answers. The models the Keynesians were promoting said that prices should have fallen with accelerating speed as long as the economy was stuck below the full employment level. But a few years into the Great Recession it was obvious that those models were unsustainable in the face of the data.
But anyway, what you need to know is that post-2008 were actually a decade of Austrian analysis triumphant.”
And remember Daniel, I’m not even going to mention that I lost a CPI bet when I write the above. So get ready to defend me when people say I’m rewriting history.
Nice thoughts, Bob.
It would be really embarrassing if two years ago today I wrote a post about how the low inflation instead of high inflation after 2008 is totally consistent with ABCT and if two years and two days ago I made the same point and said that Austrians shouldn’t be so down on ABCT over the inflation thing.
It would also be really embarrassing if Krugman had said on multiple occasions that the lack of deflation demonstrated an obvious failure of Keynesian theory.
I hope – to spare you the embarrassment – neither of those things is true.
You are saying that Krugman’s note of Sept 9, 1982 and his warning of rising inflation from 1983 onwards was not based on Keynesian analysis?
Thus when Krugman says “what you need to know is that the 80s were actually a decade of Keynesian analysis triumphant” he also implies something like “wish I had known how to do Keynesian analysis back in the 1980’s.”
Tel – I thought I was pretty clear. I’m saying that Krugman’s 1982 memo was about commodity shock and exchange rate effects on inflation. He may be wrong – I honestly don’t have the foggiest of what was going on with all that. It’s not my area. For the sake of argument let’s say he was dead wrong about it. But I definitely don’t think his views on commodity prices and inflation don’t contradict expectations augmentation of the Phillips Curve.
If you disagree with me about the implications for expectations augmentation could you explain what I’ve got wrong?
What you are doing wrong is just introducing new factors any time the model needs some retrospective cleanup. Since there’s no limit to the number of new concepts you can keep throwing into your model, this approach will guarantee to fit all historical eventualities, but you can’t predict anything with it because you still don’t know what the next special case is going to be, until after it happens.
To put it mathematically, you have as many degrees of freedom as you have data points, and it doesn’t disturb you in the least that you are adding more degrees of freedom whenever new data points come along.
The expectations adjusted curve is a classic example… how much do we adjust for expectations? Just as much as we have to, so that the curve fits.
Augmented Philipps curve theory is Keynesian theory?
Bob’s analogy is spot on. Bob’s reaction to Krugman’s post and memo is a lot like people who make big dismissive statements about Austrian economics because we didn’t have a lot of inflation after 2008. He hit the nail on the head.
Now he and I agree such a view of Austrian economics is highly problematic, so I don’t know why he’s having so much trouble with the Keynesian side of things. It’s his analogy after all!
If I’m being too cryptic here, my point is that you don’t have to wait until 2030.
I still believe, as I did two years ago today, that when you think there will be 10% inflation and there’s way, way, not 10% inflation it’s probably worth revisiting some of the things you thought. Doesn’t necessarily mean you need to rethink ABCT. As I’m already on record saying (and as you’ve written above) lots of Austrians predicted low inflation.
More broadly, my views on Austrian economics are 100% clear and on record. I very painstakingly laid out what I buy and don’t buy about it in my Critical Review article. Criticism over inflation predictions has not been featured.
So what makes you worry that I might not defend you on these grounds? Surely nothing in my record.
“I read it as Krugman pointing to future supply shocks impacting the price level.” – Daniel Kuehn
Hmmm…let’s see…
“Our very rough guess is that correction of these distorted relative prices will add five percentage points to future increases in consumer prices and about two percent to the GNP deflator. This estimate is conservative in that it assumes stable oil prices.” – 1982 Krugman memo
We’ve got Daniel Kuehn and Krugman on the record claiming that cutting out-of-control spending and funny money growth causes a TEMPORARY increase in unemployment which eventually comes back down more or less to its original level. Thus, because the market does not fail, there is no point to the spending and funny money growth in the first place which will only cause a catastrophe later.
Not the cleanest dynamics in the world, but the basic point should be clear: cutting inflation would require a temporary surge in unemployment. EVENTUALLY, HOWEVER, UNEMPLOYMENT COULD COME BACK DOWN TO MORE OR LESS ITS ORIGINAL LEVEL; this temporary surge in unemployment would deliver a permanent reduction in the inflation rate, because it would change expectations [actually, inflation will come down due to the change the rate of money growth].
But anyway, what you need to know is that the 80s (and the teens) were actually decades of Austrian analysis triumphant.
As I go into some detail below, the author of Krugman’s textbook Rudi Dornbusch was saying much the same thing: that changes to the monetary system won’t control the long-run equilibrium, but it can influence the path we take reaching that equilibrium.
Dornbusch seems to have advocated stability for the most part… that is to say, take an orderly path to the equilibrium point, and avoid over-reaction to changes. You will find a few cases where Hayek advocated the same thing.
Bob Murphy,
Brilliant reply to Daniel Kuehn.
Looking forward to his response?
I hope you liked it Adrian,.
I wish I could make it more interesting for you, but I no more shill for Keynesians than I thoughtlessly dismiss Austrians.
Let’s get back to the topic at hand: The memo raised concerns about commodity prices and the valuation of the dollar. The model is about the augmentation of the Phillips Curve trade-off by expectations of inflation. I guess Krugman was wrong about commodity prices and the dollar – to be honest I haven’t the foggiest what was going on with either in the 80s – it’s not my area. Let’s say for the sake of argument he was completely wrong. How in the world does that say anything about what he thought of inflation expectations?
I said it in my original post and I’ll say it again – you guys are like Nelson from the Simpsons just laughing at very superficial facets of this discussion and you seem to have no idea what the actual issues at stake are here (not Bob, I know he knows).
My bigger HET question here is not about Keynesians – the Volcker disinflation and the inflation run up to the Volcker disinflation was clearly considered a win for the expectations augmented Phillips Curve. No serious question there. what I’m curious about is the claim about RE and Lucas. Does anyone know how they reacted to the late 70s/early 80s? Honest question – I don’t know.
Volcker himself said the Phillips curve was outright bunk.
http://sigmundholmes.com/the-cantankerous-volcker/
Even if you do want to believe in some expectations-adjustment, since you can’t measure expectations in any repeatable way, then the theory fits anything you throw at it (only retrospectively) but becomes completely useless for prediction. Volcker clearly didn’t use it in the 1980’s and no one else has been able to use it since.
I don’t want to sound like an ass so if I do keep in mind this is an honest question: did you not read the question he is responding to or do you not understand it?
I was not attempting to answer your question.
When you said, “clearly considered a win for the expectations augmented Phillips Curve”, that’s bunk, and I was pointing out why.
But it’s very obviously not bunk and he isn’t saying it is, and that’s precisely why I ask Tel. It’s better to think of him as criticizing a set of a assumptions or, if you like, a particular version of the Phillips Curve rather than the Phillips Curve itself. And here’s the thing: *everyone agrees with Volcker on that*.
I don’t know you and I don’t know how much of our convo here is ignorance of the subject vs bad argument
Or it could be a third option: our friend Mr. Kuehn here has a number of – shall we say – “odd” personal views on the importance and validity of the Phillips Curve.
The Volcker interview makes it quite clear that he’s also rejecting the notion that an expectations-adjusted Phillips Curve has much of any policymaking value. (See: “And if it is expected, then everyone adjusts, and it doesn’t do you any good.”) But if you hold odd personal views about the Phillips Curve as Kuehn does, you’re liable to start doing other odd things such as accusing the person stating a perfectly reasonable and conventional view of the Phillips Curve of “ignorance” for declining to confirm your original oddities.
It’s all really a matter of perspective. Plus a touch of self-awareness, of which he has very little.
RussJ – Well there are lots of Phillips Curves with expectations in them, some of which have good news for policy some of which don’t. That’s another discussion. Clearly Volcker: (1.) supports including expectations adjustments in the PC, same as Krugman, and (2.) is skeptical about the trade-off, which most people are most of the time with some special exceptions.
You are correct, it was too easy. Further investigation reveals layers within the trifle.
From Krugman:
The textbook Krugman pulled out to show “Keynesian analysis triumphant” was written by Rudi Dornbusch and Stanley Fischer, both of whom studied at the University of Chicago during the 1970’s.
It gets better. From Wikipedia’s biography of Robert Lucas:
… then from Wikipedia’s biography of Stanley Fischer:
Note the wording there, Stanley Fischer was never fully Keynesian, but he became a central figure because his work reconciled some of the Keynesian ideas (particularly sticky prices) with classical ideas. Robert Lucas wanted to reconstruct macroeconomics as based on a proper foundation of microeconomic principles, plus collective aggregation; Stanley Fischer was fully onside with this and never argued against it, he merely introduced additional factors (such as long-term work contracts) that had the effect of changing the dynamics of how the economy adapts to monetary shocks.
Speaking of dynamics, let’s look at the other author of that textbook Rudi Dornbusch, who was also at University of Chicago around 1975 at the same time as Robert Lucas. Dornbusch was bringing the ideas of dynamics into macroeconomics and wrote a famous article on overshoot that was amazing to economists, but pretty ordinary stuff for engineers at the time. People like Harry Nyquist had done this work 50 years earlier outside economics, but still classical economics maintained a focus on equilibrium and nothing but equilibrium. Dornbusch was no dope, I’m not running him down, and it’s a great thing he helped economists catch up with the concepts of dynamics and why they are useful; just that economics in general is very resistant to taking on ideas from outside. Indeed I’d argue that a lot of Dornbusch’s work on dynamics has been lost or abandoned by modern economists, so that when Steve Keen started using it, he thought this was some amazing discovery he’d just stumbled onto.
http://lpsa.swarthmore.edu/Transient/TransInputs/TransStep.html
There’s what engineers are learning (skip the math and just look at the pictures) you see something very similar to the textbook that Krugman is holding up. Note the use of Laplace transforms which were invented in the 19th Century. In Krugman’s picture the red line “gradualist” would be called “input shaping” by engineers and it is a standard technique to reduce overshoot and vibration. There’s evidence that real central bankers use this technique, and I’m in favour, especially when it works.
Yes, I’m happy to see the “New Keynesians” have incorporated a useful mix of ideas from eclectic sources, but only a tiny bit of that actually came from Keynes himself. This is really the convolution of many sources, including a very large component from the Chicago school of the 1970’s. Krugman must be well aware of this, he just expects (as usual) his readers won’t do the legwork to call him out. The typical leftist urge to rewrite history (like Bob caught Piketty doing).
Finally, I’ll point out that there’s nothing in any of Dornbusch & Fischer’s work that disproves, nor even disagrees with the long-run neutrality of money. Dynamics are not about the long run, they are about how the system reacts in the short term. It might be very sensible for a central bank to impose some short term stabilizers to the economy, but this will not fix structural problems, it can’t. The best government can do is soften the blow of change, slow the rate of change, but at some stage get back out of the way and let the inevitable change happen.
That’s the thing that Krugman (and his ilk) just won’t accept. In their opinion now is NEVER the time for small government, and NEVER the time to refrain from interference. He does not properly respect the work of the economists who have come before him, Krugman is a political operative and an entertainer.
Bagging out Krugman for dodgy analysis is a competitive market.
http://www.cato.org/blog/another-oops-moment-paul-krugman
Nothing shocking there then.