An Open Letter to Paul Krugman

Dear Paul:

I was reading your article How Did Economists Get It So Wrong. Who are these economists who got it so wrong? Speak for yourself kemo sabe. And since you got it wrong - why should we believe your discredited theories?

    It is a sad fact that whenever something bad happens people will claim that it means that they were right all along, and other people will listen to them. A professional prosecutor frustrated by the fact that you can't beat confessions out of suspects? Wait until September 11 and try again and this time call it the "Patriot Act." A progressive who would like to see higher taxes and more government programs? Wait until there is an economic crisis and call it a "fiscal stimulus bill." Here we are, the recession is over and we've spent 10% of the money...Not the 200% you thought we needed to end the recession.

    It is a daunting task to bring you up to date on the developments in economics in the last quarter century. I know that John Cochrane has tried to educate you about what we've learned about fiscal stimulae in that period. But perhaps a I can highlight a few other developments? You seem under the impression that economists had resolved their internal disputes before the financial crisis. So that means you haven't followed the debate about the causes of depressions between Peter Temin on one side and Timothy Kehoe and Ed Prescott on the other? You say that we think that the "central problem of depression-prevention has been solved." Has it not? Are you forecasting that this recession will turn in to a depression? But of course "More important was the profession's blindness to the very possibility of catastrophic failures in a market economy." That would be the profession that hasn't been reading what the profession has written? Perhaps you should go look at that controversial book Kehoe and Prescott [2007], Great Depressions of the 20th Century. Or you might read Sargent, Williams and Zhao [September 2008], "Conquest of Latin American Inflation". Wouldn't it be nice if people had some idea of what was being written before criticizing it? 

    Let us talk more seriously about the supposed failure of the economics profession. You say "Few economists saw our current crisis coming, but this predictive failure was the least of the field’s problems." The predictive failure is not a problem of the field - it is a problem for those who are under the impression that we should be able to predict crises. Do you number yourself in this bunch? Do physicists get it wrong because their theory says that they cannot predict where a photon shot through a sufficiently narrow slit will land? Economic models are like models of photons going through slits. Just as those models predict only the statistical distribution of photons, so our models only predict the likelihood of downturns - they do not predict when any particular downturn will occur. Saying "most economists failed to predict the downturn" is exactly like saying most physicists failed to predict the impact of the twelfth photon passing through the slit.

    More to the point: our models don't just fail to predict the timing of financial crises - they say that we cannot. Do you believe that it could be widely believed that the stock market will drop by 10% next week? If I believed that I'd sell like mad, and I expect that you would as well. Of course as we all sold and the price dropped, everyone else would ask around and when they started to believe the stock market will drop by 10% next week - why it would drop by 10% right now. This common sense is the heart of rational expectations models. So the correct conclusion is that our - and your - inability to predict the crisis confirms our theories. I feel a little like a physicist at the cocktail party being assured that everything is relative. That isn't what the theory of relativity says: it says that velocity is relative. Acceleration is most definitely not. So were you to come forward with the puzzling discovery that acceleration is not relative...

    Of course some people did predict the crisis. Some might even have been smart enough to know that if they consistently predict the opposite of a consensus point forecast, eventually they will be right when everyone else is wrong. If I say every year: there will be war; there will be an asset market crash; there will be a recession; there will be famine; we will run out of oil - eventually I'll be right. These kind of predictions are only meaningful if more people than can be attributed to random good luck got it right at the right time or if whatever method they used to reach that conclusion is replicable. Or does the ability to replicate results fall under the category of "not very interesting because that would be an elegant theory?"

    But let's turn to what you say are our deeper failures. We "turned a blind eye to the limitations of human rationality that often lead to bubbles and busts." It makes me feel physically ill that a distinguished economist could be so ignorant of his own profession. As a random example, how about my student Felipe Zurita's thesis on speculation written in 1998? There are endless papers written about bubbles and busts - some assuming rationality, some not. Some are experimental, some are theoretical, some are empirical. There are economists who have devoted their entire careers to studying bubbles. There is a fellow named Stephen Morris. He isn't what you would call a fringe member of the economics profession - he's the editor of Econometrica which, as you know, is one of the leading journals in economics. He has written extensively about bubbles. I take it you aren't familiar with his work. Perhaps you should walk down the hall and stick your head in his office and ask him about it? Each crisis - in Mexico, in South-east Asia, in Argentina - had generated hundreds of papers examining how and why the crisis took place.

    Efficient markets? Where have you been for the last quarter century? The modern theory of how financial markets incorporate information is that they do so imperfectly. The technical device is that of noise traders originating in a 1985 paper of Admati. But I think you knew of the idea earlier. In 1980 when you were a visitor at MIT, you participated in a graduate student seminar...in which I presented a paper starring noise traders...

    Do we really need some sort of behavioral model to understand why asset prices fall abruptly? If opinions about asset values change, prices must fall abruptly - it isn't irrational to run for the exits when the theater is on fire. In addition to a beautiful 1983 paper of Steve Salant there is a large literature on bank runs and contagion, not to speak of credit and collateral cycles. If there was some sort of irrationality involved in a panic, prices ought to bounce right back the next day when everyone wakes up and sheepishly realizes that they were wrong. In fact asset prices seem to be tracking news of fundamentals pretty well - gradually recovering as we get better news about fundamentals.

    Has behavioral economics offered anything that would help to solve the market failures that characterized this crisis? Was it herd behavior or animal spirits? Or was it risks that were not being priced?  Serious economists like Lasse Pedersen try to analyze how liquidity risks created systemic problems and think about how to incorporate them into our understanding of how to ameliorate future crises.  They don't shake their heads and revert to discredited static theories of the 1930's.

    Crises have been ubiquitous throughout history. While we can't forecast them we do know how to learn from them. And we certainly have a good idea what not to do in response: do what Chile did successfully - fail banks and recycle them, not do what Japan did unsuccessfully - keep the zombie banks limping feebly around. Like me you saw the bank bailout plan for what it was - not a necessary step to save the credit sector from collapse but a give-away of taxpayer money to investment bankers. But the stimulus plan? How can you be arguing for more? Since we are recovering before most of the stimulus money has entered the economy - isn't that evidence it isn't needed? How can you write as if you are proven right in supporting it?

Regards,

David

[Thanks to Tom Cooley for talking this through.]

 

 

 

Comments

[LHB58]

Professor Levine's criticism of Krugman's informative article, summing up for the general, educated public, the wrong turns that modern economic theory took during the last three or four decades calls to mind the old adage: If you laid all the economists in the world end to end, they wouldn't reach a conclusion. Professor Levine does indulge himself in a summing up toward the end of his piece, but only after a lengthy and tedious defense of complex empirical studies that likely were read by few people beyond the individuals who reviewed them favorably for publication, and which never seem to arrive at even the most basic consensus (beyond unquestioning acceptance of the basic infallibility of the inscrutable methodology of modern, mainstream macroeconomics).

A useful predicition of our current financial crisis might have taken the simple form of "continuous increases in leverage based on the assumption of constantly rising real estate prices are unsustainable" or "the allocation of real productive resources based on derivatives prices will be highly inefficient given that derivitives prices are meant to balance competing betting interests rather than the supply and demand for the underlying asset" yet few economists in 2006 were willing to risk even those sorts of predictions. In short, if there was such a thing as a free market in the theories and econometrica of mainstream economists during the last 20 years, their most well published proponents would be in the same position as Chrysyler or General Motors, which is to say: bankrupt.

[David K. Levine - Huffpost Blogger]

An economist friend's view:

We don't think that we understood and understand all the details of what happened. For example, a common view by the end of the 1990s amongst top academics was that the interbank market was so efficient there was no need for discount window lending. This was based on an idealized view of how the interbank market worked which turned out not to be true. We need to understand if the frictions that made this view fail are unavoidable or fixable by different policies. For example, the Swiss National Bank targets a term rate, 3 month Libor, did this allow for a more efficient interbank market? There are literally thousands of mundane questions like this - about how the financial sector works, about how it is connected to the real sector, and about how much of the shock was real and how much due to financial events, about how much of the bubble was due to poorly designed interest rate policies after the dotcom bust and whether it could have been avoided or mitigated. These are the questions that the serious researchers are and should focus on. There are lots of tools to study both the rational and irrational aspects of the crisis. Let's ignore the silly debate by economists who are clueless about the tools and models at our disposal about who was correct in 1978. If you want this not to happen again - move past Paul Krugman and let us do our job.

[David K. Levine - Huffpost Blogger]

More red meat: http://faculty.chicagobooth.edu/john.cochrane/research/Papers/#news

[ld]

Yes, we do expect you, as an ecomonist, can predict economic crises. What other value do you have, if not to do something that basic? No value at all.

Very lame, and your physics metaphor is especially lame.

[David K. Levine - Huffpost Blogger]

I guess biology has no value either as we can't cure cancer?

[ivo66]

I believe LD is correct when he says that your metaphor is lame.

First of all, physicists may not know what happens with individual electrons (i.e. velocity and position), but they have pretty accurate distribution functions that can be experimentally verified, and probabilities can be accurately calculated. Not to mention the behavior of electrons on a macroscopic level. Based on their theories, humanity enjoys the benefits of technological progress that you can see everywhere around you.

Also, while medicine/biology may not have an answer on how to cure caner, there are many illnesses that CAN be cured, and at least statistically you know what are your chances of surviving a certain surgery or procedure.

I work as an engineer, and statistical tools and distribution functions are part of a daily routine. We can design communication systems that work because have have proper models for noise or channel impairments. We can confidently say that the system will work with exactly specified level of reliability. Your cell phone, TV, radio all work based the same design procedures.

Economists, on the other hand, do not even have proper models that can be verified by empirical data. What is the closest model? Log-normal distribution? Even that should give some insight about the behaviour on the macro level, only if the model was correct.
By your own admission, you cannot predict anything, and the rest is just justification of your ideology.

[wbond]

Nice article.

I am not sure that you have relativity correct, however.

The Gaililean relativity that is a part of Newtonian mechanics - presumably not the "theory of relativity" that would be have been the conversation of "modern" cocktail parties - should cover acceleration as well as velocity. E.g. The skydiving physicist knows that whether he is moving towards the ground or the ground to him, the calculation for acceleration is identical. And no 20th century physics required.

The one-sentence synopsis of special relativity: The insight is to solve for the observation that the speed of light is constant irrespective of the movement and locations of the observers.

Cheers,

wbond

[jeffg1]

the theory of general relativity unified special relativity with newton's theory of gravity.

so, Mr. Levine, you are wrong.

[Thorstein Veblen]

Dear David Levine,

I'm just curious where you got the 10% figure from. The initial CBO estimate is that $185 billion will have been spent by the end of September, 2009, that's 24% of the total outlays. Even if it had only spent $100 billion (which is counterfactual), that is still more than 10%. In addition, the states' used knowledge that more stimulus was on the way next year when they set their budgets earlier this year -- so without the stimulus, states would have cut back more than they did.

Also, when people judge whether "the stimulus" worked, they need to look at the net stimulus. Since states and local governments have made roughly $100 billion in cuts for fiscal year 2009, the "net" government stimulus was probably less than $100 billion.

This means, of course, that what you and your comrades argued and are arguing for is for governments to suddenly cut way back on spending at the first sign of a recession. The way your write suggests you are not privy to what's going on at the state and local level.

Very respectfully submitted, Thorstein Veblen

[David K. Levine - Huffpost Blogger]

I think 10% is a month out of date. http://projects.propublica.org/tables/stimulus-spending-progress reports $98 billion spent out of $792 billion which is 12.4%. It is certainly possible that there is some additional spending in anticipation of stimulus money coming in. It's also true that a lot of stimulus money is being spent on things that would have been done anyway, which in turn might create other spending that wouldn't have taken place without the stimulus. The key point is that even $200 billion in a roughly $14 trillion economy isn't such a big deal. It makes sense for government to do things it would have done anyway earlier to take advantage of the fact that many resources are unemployed and cheap right now. But if things have turned around - and all indications are that they have - why do we want to incur another $600 billion of debt unless we are willing to pay for it? That is, why do it outside the usual budgetary process? If we want to pay for a green power grid, and it would be more cost-effective to start sooner rather than later - then say so, don't call it a "stimulus."

Some of the stimulus spending makes sense; much does not. We are saving a small number of jobs at vast public expense - $100,000s of dollars per job - and we will have to pay it back later. There are better ways to help people in need.

[Thorstein Veblen]

Not so fast, shooter.

Your link (written by folks obviously hostile to the stimulus) says they've spent $98 billion (out of about $500 billion), plus they've issued $63 billion in tax cuts. (And, btw, I spent my $400 stimulus check... I kind of doubt I'm the only one.) That's more than 20%, and that's probably not updated to the minute. You didn't see that, not because your stupid (I think you're a bright guy!), but because you've made up your mind (and announced it!) that you are against the stimulus, and so you can only see things which confirm your biases. Your incapacity to see the $63 billion in tax cuts is really just a typical behavioral anomaly common to all humans, so do not feel bad.

By the way, in the 2nd quarter, GDP was down 1%. Federal spending was up 11%. Even w/ no multiplier, why shouldn't we think this would have an effect? With interest rates pinned at zero?

California is paying people with IOUs. Pennsylvania didn't pay its workers in July. Florida raised taxes on retirement homes. New York City raised fares for disabled people. Sacramento laid off teachers. Public employees everywhere are taking a haircut. State shortfalls alone (not counting local, which are also substantial), are $170 billion for FY2010. Yes, this means that the net fiscal stimulus is quite small. But it also means you are pushing for far more drastic spending cuts and tax increases in a recession.

That's nuts.

[ILibertine]

Those suggesting the recession is over must have a glossy view of the world in general. It's tantamount to suggesting high fructose corn syrup is the cure for all that ails you. Nothing has changed insofar as the essential well-being of the American consumer through stability in their occupation or real income; nor, apparently, in the integrity of the financial system or federal monetary policy. Any temporarily perceived sense of stability will likely be rapidly altered through energy prices if consumers might begin spending appreciably.

The maladies inherent in the downturn that is an on-going recession to America in general, I suggest, have not been addressed and it appears congress and business merely wish to play kick-the-can and cross their fingers for an upturn while doing nothing other than aiding the rich through a period of inconvenience.

One decrying a particular perspective might be well advised to avoid a similarly flawed viewpoint that seems to me to be based upon indicators rather than realities for rank and file America, i.e. recession over; although I would be happy to be wrong on my concerns.

[dynwit]

So David Levine's best response to Krugman is: "We never said we knew what we were talking about." Hmmmm..... that's not exactly an impassioned plea for the continuing existence of economics as we knew it. Which was exactly Krugman's point, I believe.

[David K. Levine - Huffpost Blogger]

My response is: we know what to do in response to recessions. We don't yet know how to predict them and we will never be able to predict financial crises. I understand that in troubled times people want to turn to charlatans who promise a quick and easy fix. That doesn't make it a good idea. The medical profession once thought that drawing blood was the cure for all ills. Just as they learned it is wrong, so we have learned that fiscal stimulus does more harm than good.

[DiogenesOfAlaska]

Now I am confused:

here you write that you don't think that we will ever be able to predict financial crises. But below you wrote that the financial crisis was inevitable given the reckless risk taking of the financial sector with implicit government guarantee (with which I agree).

So there are circumstances in which financial crises are predictable. For example those that were ignored by Greenspan under the pretense of the philosophy of efficient markets - which basically rules out the financial sector in the form in which it was the subject of Greenspan's object of oversight.

There's a problem with this philosophy. No matter how you look at it. It may be desirable as a fiction, but that's also true for a classless society. It may be a useful benchmark for certain kinds of arguments, but it is at best an approximation. And finance is where it fails.

The only surprising thing about this is that it is so totally unsurprising. Like an empress totally surprised that people are watching her taking a bath when she's stark naked.

[RandVictims]

You're right, we need more Reaganomics, not jobs.

[David K. Levine - Huffpost Blogger]

For those are interested there is an impassioned reply to my post here: http://delong.typepad.com/sdj/2009/09/the-intellectual-bankruptcy-of-the-chicago-school-infests-st-louis--or--the-huffington-post-needs-a-quality-control-filter.html. The gist of the post is "A well-functioning labor market--one not deep in market failure--matches workers looking for jobs with jobs looking for workers. A well-working labor market will keep both the number of unemployed and the number of vacancies together in balance, and at relatively low values. That is not what is happening now. The number of vacancies--jobs for which employers are actively searching for workers but have not found any to hire yet--is very small. The number of unemployed--of people who want to hold jobs and are qualified to hold jobs--is very high."

We are moving from a situation where - among other things - a lot of people were building houses financed by borrowing money from the Chinese. The Chinese are not so enthusiastic about lending to us right now, so there is going to be a lot of dislocation while people move to different occupations. Does Brad think that the government - those guys that brought you Iraq, Afghanistan, and the evacuation of New Orleans, the financial crisis, etc. - is the right agency to figure out what sort of industries will rise in the wake of the recession?

[David K. Levine - Huffpost Blogger]

If you want to see Brad DeLong debunked, look here: http://economistsview.typepad.com/economistsview/2009/03/debate-at-uc-davis-between-brad-delong-and-michele-boldrin.html - my favorite line: if a guy has a broken arm, you don't put a band-aid on his butt.

[dynwit]

You have just confirmed what the average Joe/Jane has already been thinking: that economists don't know WTF they're talking about.

[Doofus]

'Like me you saw the bank bailout plan for what it was - not a necessary step to save the credit sector from collapse but a give-away of taxpayer money to investment bankers. But the stimulus plan? How can you be arguing for more? Since we are recovering before most of the stimulus money has entered the economy - isn't that evidence it isn't needed? How can you write as if you are proven right in supporting it?'

A pretty interesting piece, in that a 'freshwater' professor is at least standing up to the all-
powerful 'saltwater' Krugman. Clearly, both are 'distinguished', although only one has a
Nobel. But about that 'recovery' that's happening. When does employment kick in with that?
Or do we need to move elsewhere for jobs?

[ShadowatDusk]

Professor Levine, I have your book Against Intellectual Monopoly , I found it very persuasive (it is one of my favorite books on economics). I respect both your and Dr. Krugman's opinions. However, neither his article nor your (or Dr. Cochrane's for that matter) seems to contain enough solid evidence (for me at least) to distinguish which side is more correct. It resembles an old argument brought to the surface but with the premises, key for those unaware of details involved with the argument, missing; assumed away due to a natural wish to avoid redundancy. It is quite frustrating, not being able to choose a side in this debate

[David K. Levine - Huffpost Blogger]

The main purpose of my post was this: Paul Krugman has a Nobel Prize in Economics. It isn't unreasonable to think he might know what he is talking about when it comes to economic crises. He doesn't, never having studied them, nor having followed the literature. His views are what he was taught in graduate school in the late 1970s. I have a pretty good idea what that was, having arrived at MIT just as he was leaving. Nothing of the "neoclassical synthesis" taught at that time has survived the test of time and evidence.

There are many serious macro-economists, most of them young and more focused on figuring out what is going on than engaging in public debate. People bandy around statistics as if they mean something - actually making sense out of them is a difficult enterprise that has taken decades, and it isn't as if we are finished. If I want to know what is going on about things I'm not expert on - for example with job vacancies and unemployment - I wouldn't ask Brad, I'd ask someone who knows - for example Rob Shimer http://robert.shimer.googlepages.com/.

This isn't going to be a popular comment: but on the scale of important human events, the current crisis doesn't rank. Help people in need, yes, but let's not do something so crazy it will rank on the scale of important events.

[JanusDaniels]

What Brad DeLong said:
http://delong.typepad.com/sdj/2009/09/the-intellectual-bankruptcy-of-the-chicago-school-infests-st-louis--or--the-huffington-post-needs-a-quality-control-filter.html
... A well-working labor market... will keep both the number of unemployed and the number of vacancies... at relatively low values...
That is not what is happening now...."
Shorter Brad:
If "recovering" meant "we haven't bottomed out yet" then David K. Levine makes sense.

[macrumpton]

The recession is over? What planet are you from? We have double digit unemployment and bankruptcies are at levels not seen since the depression and several of our major industries are only holding on because we have pumped a Trillion dollars into banks, insurance companies and the auto industry.

[DiogenesOfAlaska]

The problem itself is certainly not a semantic one, but your response is still based on or at least explained by the terminological meaning of the word 'recession'. Clearly, this matters, because it's no use to argue that things get better based on the misunderstanding of words. I recommend the link given in JanusDaniels comment above.

Apart from the proper interpretation of what 'recession' means, there is also another important point left out in the optimist reading that the stimulus wasn't needed: as you point out, the recovery rests strongly on the anticipation of banks and corporates. Absent the stimulus, their anticipation of the future would be a different one. And the same holds for investors, of course. At least in case they're rational.

But it's not so simple, because the same also holds for investors in US government debt...

There's really nothing simple at all about this, and all I can recommend is to 'bless the mess'.

[marinara]

what happens when an speculative bubble covers up an important drain on the American economy? The problem festers.

[Duncan Cameron]

I was surprised to read this:
"it isn't irrational to run for the exits when the theater is on fire."
The rational response would be to file orderly from the cinema, so that at least some could escape. The likely response would be to run for the exits, and risk having everyone block each other from getting to safety.
I think this makes Krugmans point: macroeconomists don't handle rationalit­y/non-rati­onality very well.

[DiogenesOfAlaska]

I think it's even better: the example shows that a narrow understanding of the theory of rational expectations isn't going to provide an answer without smuggling in a lot of further ad-hoc assumptions.

But I also think that D.K. Levine is among the first to point out such limitations of narrow interpretations of the theory.

[Eric D Dixon]

Actually, you're getting rationality wrong here.

The rational response would be to recognize that groups of individuals responding to the threat of imminent danger have not had a chance to coordinate a group strategy for maximum escape success. Any individual actor in this instance will be responding to a new expected (or suddenly manifest) equilibrium that results from the uncoordinated actions of a group of people in panic mode. Given that the likely response by that group would be to make a mad rush for the exits, it's the most rational strategy for self-preservation by any individual actor would probably involve an attempt to reach the exits first.

[David K. Levine - Huffpost Blogger]

What Eric D Dixon said.