The Sumner Critique (Plus a critique of Sumner)
Now that I am starting to tangle with Krugman, I might as well go for broke, and pull out all the stops. So at the risk of sounding ridiculously over-confident, here is my grand new theorem:
The Sumner Critique (of arrogant right-wing economists and naive left-wing economists)
In any rational expectations model, government policymakers have no basis for confidence in a policy succeeding, unless it is expected to succeed (by the markets.)
Yes, the title is meant to echo the famous Lucas Critique of my former dissertation chair. And yes, I do understand that it is a bit grandiose to claim that sort of title for a theorem that on close inspection looks suspiciously like a tautology. And it’s not really original, as Robin Hanson has discussed similar ideas. But I believe it is worth thinking about, as I don’t think its implications for macroeconomics are fully understood.
For those on the right: If the market forecast of inflation suggests that monetary policy is too tight, then it is too tight. Don’t try to arrogantly assert that “right wing economists know best.” You are the guys who like ratex models. If you don’t believe the implications of your models, then change your beliefs about policy, or change your model.
For those on the left: As soon as Obama was elected, his fiscal stimulus should have started working. If it was expected to work, then spending should have begun rising on the announcement of the intent to offer fiscal stimulus. That is the lesson of Eggertsson’s 2008 AER paper. Starting in early November 2008, it was Obama’s economy. And guess what; markets didn’t expect Obama’s fiscal stimulus to work (very well.) That is why asset prices fell so sharply until March 2009, when the Fed adopted QE and China’s recovery started raising the world’s (Wicksellian) natural rate of interest.
Part 2: A critique of Sumner
I’ve been reading a Javier Marias novel called Fever and Spear, and ran across this interesting passage:
The most striking and dangerous thing about this whole business is that you, too, begin to believe yourself capable of seeing and fathoming. Boldness never rests, it waxes and wanes, it burgeons or shrivels, it slips away or subjugates, and may disappear altogether after some major setback. But boldness, if it exists, is always on the move, it is never stable and never satisfied, it is the very opposite of stationary. And its main tendency is towards limitless increase, unless kept in check or brought up brutally short, or else systematically forced to retreat. In the expansive phase, perceptions become excitable or intoxicated, and arbitrariness, for example, ceases to seem arbitrary to you, believing, as you do, that your judgements and insights, however subjective are based on solid criteria . . .
Ouch! That hits close to home. I don’t know about other bloggers, but I go through cycles. When things are going well, I start to get more confidence. People say good things about my blog. Gradually I get more cocky. I start dispensing with “maybe” and “perhaps.” I go out on a limb, play hunches. Soon I’m pontificating on things I know little about (financial regulation.) I read a few items in the newspaper, and suddenly I expert enough to offer 6 point plans to save the world financial system—right off the top of my head. The ego keeps swelling, eventually reaching titanic proportions. Then bigger than the Titanic, into James Cameron territory.
And then it all comes crashing down. A commenter who knows more than me about some area (say statsguy or 123) will put some real facts and logic into the comment section, and puncture the boldness balloon. I’m forced to temporarily scurry away like a dog with its tail between its leg. I get more cautious. But Marias is right, boldness always waxes and wanes, it’s never stable. After growing more cautious I regroup, and start building up again. In no time at all I am acting like I know more than Nobel Prize winners.
I don’t know if every blogger goes through these cycles. Tyler Cowen never seems over-confident. Others seem to have egos that are in a permanent state of inflation. Like they’re taking Cialis for the brain. I thought of this passage in Marias’ novel because whenever I think I get a clever idea, it puts me on an upswing. And I thought the application of the expectations trap to fiscal policy was kind of neat. So today is a good day. (Plus I saw the new Kore-eda film.)
Now I’m just waiting for this idea to be punctured, it’s just a matter of time before someone actually reads what I wrote carefully, and finds flaws in the argument. And then it’s back to earth. Like pulling the plug on Kore-eda’s air doll.
Tags:
4. June 2010 at 17:02
Of course your blogging reputation is an unsustainable bubble, as any Austrian Blogging Cycle Theorist could have told you. Your self-image been artificially inflated by the fiat currency of links from other websites (unlike the hard currency of, say, in-person praise over lunch), and so you’re headed for a harsh blog recession as everyone’s expectations of your posts recalibrate. Blogo-Keynesians will tell you that you can avoid this collapse in street cred by simply getting more people to link to you, but that will just make the inevitable online shunning period harsher. Maybe you should write your posts with the assumption that your nominal Gross Blogging Product will be rising steadily over time at a steady clip of around 3%, and simply post more after a Krugman-derived demand shock?
Maybe that wasn’t as funny as I thought it was. For what it’s worth, I really like your blog and learn a lot from it, especially when you debate Krugman/DeLong (or maybe “get snapped at by” is the proper term).
4. June 2010 at 17:07
“If the market forecast of inflation suggests that monetary policy is too tight, then it is too tight.”
Here you are conflating tightness in monetary policy and CPI inflation expectations. Looseness in monetary policy and CPI aren’t perfectly related. Supply side issues have strong effects on CPI, but are separate from monetary policy.
Its very possible to have very loose monetary policy as measured against other currencies or financial assets while still having a very low CPI increase. This was what happened in the 90’s. You are still making the mistake of arguing from prices (which is what CPI is).
4. June 2010 at 17:39
Embrace Your Boldness. You definitely deserve it. Thanks for what you are doing with this blog.
Just try not to let your tone or style veer over into the zone of being permanently indistinguishable from an adolescent arrogant jerk bully who treats everyone that disagrees with him at all as being either completely stupid or completely evil and generally unworthy of being allowed an existence.
Not that I’m pointing fingers at anyone or any two…
4. June 2010 at 18:50
“As soon as Obama was elected, his fiscal stimulus should have started working. If it was expected to work, then spending should have begun rising on the announcement of the intent to offer fiscal stimulus. That is the lesson of Eggertsson’s 2008 AER paper. Starting in early November 2008, it was Obama’s economy. And guess what; markets didn’t expect Obama’s fiscal stimulus to work (very well.) That is why asset prices fell so sharply until March 2009, when the Fed adopted QE and China’s recovery started raising the world’s (Wicksellian) natural rate of interest.”
Except things were worse than the models the administration were using predicted without any fiscal stimulus. In other words, it’s not that the stimulus had no effect. It did, and that can be seen clearly in figures for job losses. What happened was that the economy was worse than predicted.
And none of this is a point against people like Krugman. He was consistently saying that ARRA was too small. And Romer’s own figures suggested it should’ve been more than $1.2 trillion, but they shied away from it because of “sticker shock”.
And I think these expectations of yours are moving at faster than light speed. Firstly, Obama didn’t announce a stimulus until late November. Secondly, why should that announcement affect the economy so quickly? Businesses can’t turn around their business model that quickly. And surely successful legislation is far more important in expectation terms than political announcements. An announcement still has huge uncertainties around it, with the market not knowing what it will mean when it goes into action if it gets through Congress.
But for what it’s worth, the market did quite like Obama proposing a fiscal stimulus:
http://uk.reuters.com/article/idUKN0831597420081208
So I submit that you now have three options available to you:
1. Rational expectations do not hold in the real world.
2. The market expected the stimulus to be successful, and so by the Sumner Critique, confidence in the stimulus is absolutely fine.
3. The Sumner Critique is wrong.
4. June 2010 at 22:02
@Alex:
‘1. Rational expectations do not hold in the real world.
2. The market expected the stimulus to be successful, and so by the Sumner Critique, confidence in the stimulus is absolutely fine.
3. The Sumner Critique is wrong.’
You are assuming the question. The efficacy of fiscal stimulus is very much open for debate, and most non-keynesians consider it damaging, not just ineffective.
4. Fiscal stimulus has a negative effect, and the market expectation of this is precisely why the economy crashed so hard. You have to remember, this is the largest fiscal stimulus made in the history of the west. We have had far worse financial crashes in the past that didn’t have nearly as bad economic repercussions, and one possibility is that the reason for this was precisely because this time we tried to fiscally stimulate our way out.
5. June 2010 at 02:15
4 is not an option available, since I showed that the market expected the stimulus to be a good thing. Read the Reuters link.
And I would like evidence provided for this statement:
“We have had far worse financial crashes in the past that didn’t have nearly as bad economic repercussions”
5. June 2010 at 02:32
And I’m not sure it’s fair to say that “this is the largest fiscal stimulus made in the history of the west”. Larger than both the New Deal and the Marshall Plan? Are you sure?
5. June 2010 at 07:13
@Alex:
“We have had far worse financial crashes in the past that didn’t have nearly as bad economic repercussions”
The S&L crash was worse for example at first, but it didn’t drag on as long nor have as long an economic impact.
Larger than both new deal and marshal plan? Yes in terms of % of US GDP. Once you add the stimulus package and the bailouts, you get about ~1.4T (I lump them together because they were in the same time frame).
Our GDP is about 14T. So about 10% of GDP.
The marshal plan was about ~5% of US GDP.
The New Deal wasn’t a single stimulus but a combination stimulus plan and structural change in the economy. There was also several New Deal plans, specifically referring to the “first 100 days” plan, yes I believe this stimulus was larger.
I guess now thinking of it, WW2 was larger than this stimulus, but Barro’s work has shown that WW2 specifically had a negative stimulating effect.
5. June 2010 at 07:17
Re: Alex
“4 is not an option available, since I showed that the market expected the stimulus to be a good thing. Read the Reuters link.”
You are making the mistake of assuming that the converse of a statement has the opposite truth value as that statement. The Sumner Critique doesn’t mean that if the markets think something will be a good thing that it will, but rather if they don’t think it will be a good thing then it won’t.
‘In any rational expectations model, government policymakers have no basis for confidence in a policy succeeding, unless it is expected to succeed (by the markets.)’
This means any policy expected to not succeed, won’t succeed; but it doesn’t mean that ones that are expected to succeed will succeed.
5. June 2010 at 07:35
Aaron, I like that.
Doc Merlin, I agree, but I meant tightness in terms of what others see as the Fed’s objective. Right-wingers think it should be inflation. OK, then pay attention to inflation forecast if you are a Ratex economist.
Thanks Indy, And I hope commenters will set me straight when I step over the line. Otherwise the head swells too much.
Alex, When did Obama say he was going to do a stimulus package, it not when the details were announced. You asked:
“And I think these expectations of yours are moving at faster than light speed. Firstly, Obama didn’t announce a stimulus until late November. Secondly, why should that announcement affect the economy so quickly?”
It’s not me that makes that claim, its the economist (Eggertsson) that Krugman and Thoma like to cite. Eggertsson would say that current AD is heavily affected by future expected AD. If a future stimulus is expected to lead to recovery, then this should give the public confidence and the recovery should start now. Eggertsson made that argument about FDR.
If I win the Nobel Prize, and am told I’ll get $1,000,000 in 6 months, I’m going out to the most expensive restaurant in Boston tonight–I won’t wait 6 months.
I agree about Krugman, but he also said something like “how can the Republicans blame Obama, he’s only been if office X months.” I don’t recall the exact quote, but it was from 2009.
Regarding the market reaction to Obama’s program, I also recall that it was positive. My point was that it was far too small to have any significant effect on events. The overall market performance during this period was very weak. We didn’t need a 3% bump in stock prices, we needed a 60% bump, or a 80% bump. The sort of bump FDR’s dollar depreciation gave the stock market. I used the “very well” qualifier in my post, as I knew a shrewd commenter would bring up this point.
So markets didn’t expect the stimulus to be successful, they just expected the recession to be a tiny bit less severe.
Doc Merlin, And example is the 1983 stimulus, which was half as big (as a share of GDP) and produced twice as fast a recovery.
Alex, I see the market response to the stimulus as a massive vote of no confidence in monetary policy. They saw the fiscal stimulus as the only game in town. But they didn’t think it would “work” in the sense that I mean work–prevent a severe recession. I claim that monetary policy could have “worked,” if it had been tried. No politically plausible amount of fiscal stimulus could have made much of a dent.
5. June 2010 at 08:30
“Tyler Cowen never seems over-confident.” Shouldn’t that be: “Tyler Cowen never seems *especially* over-confident”? It seems to me that Tyler maintains an even keel of overconfidence.
5. June 2010 at 16:38
Thanks Scott.
5. June 2010 at 17:30
Alex: the trouble with your analysis is that it is unfalsifiable: you can always claim “ah, but it would have been worse if they hadn’t”. After all, the result has been the US economy has performed worse than the Administration predicted without the stimulus. If the largest peacetime stimulus in US history has not stopped this being the largest post-war downturn then your evidentiary ground is very weak.
Particularly as cross-country comparisons within the OECD are not exactly encouraging.
6. June 2010 at 06:07
Philo, I agree that he is stable. I suppose we are all overconfident, but he at least seems aware of our biases.
Doc, You are welcome.
Lorenzo, I think the burden of proof should be on those who want to spend a trillion dollars. And I agree with you that the evidence is weak.