I invite you, dear reader, to journey with me back in time to January 1993, to join me at the meeting of the American Economic Association held in Anaheim, California. You and I are seated in a great ballroom of the conference hotel, together with some 6oo economists and economic commentators, to listen to the wisdom of some of America’s leading macroeconomists: Laurence Summers (Harvard) in the chair, Olivier Blanchard (MIT), Robert Hall (Hoover and Stanford), and Edward Prescott (Federal Reserve Bank of Minneapolis and University of Minnesota). The topic: what caused the last recession? The topic was important, not because of the magnitude or duration of the recession of 1990-1991, but because a second dip in that recession had done in the second term prospects of President George Herbert Walker Bush.
As is usual when ‘important’ scholars are speaking, little time is left for contributions from the floor. Few individuals, other than similar demi-gods are ever recognized by the chair, ever secure an opportunity to speak. Well, on this occasion, yours truly was fortunate indeed. By waving my arms as if engaged in semaphore, I attracted the eagle eyes of chairman Summers, right at the end of the conference. He indicated that I would be allowed a brief moment to ask my question.
I rose to my feet, with 600 pairs of eyes focused on me, and addressed the panel as follows:
“Colleagues”, I advanced, with evident presumption, “I have listened very carefully to each of your three presentations. Since each of you utilizes the rational expectations approach to macroeconomics, an unanticipated shock is necessary to throw the economy off course . Each of you in turn has strived valiantly, but without effect, to find a relevant shock in this instance.
Professor Blanchard, you locate a dip in consumption, which is not surprising since the growth rate of the economy performed below trend rates; but you cannot locate the cause of that dip, the shock, if you will that triggered the dip, despite a considerable amount of empirical research.
Professor Hall, you explored empirically several possible shocks, but concluded that none of the shocks explored triggered the 1990-1991 recession. You conclude that ‘established models are unhelpful in understanding this recession, and probably most of its predecessors’.
Professor Prescott, you utilize a real business cycle model to explore the possibility of a technology shock. Unfortunately, you cannot locate such a shock, and you end your paper by talking casually about a range of other possible shocks not picked up by your model.
Well, gentlemen, I should like to ask whether you have ever contemplated the possibility that the central postulate of each of your models – rational expectations – is the problem. Should not models whose predictions are consistently falsified by the evidence be discarded in favor of alternative models? Is that not the logic of scientific discovery?”
Well, dear reader, you should have absorbed the discomfort that my simple words generated in that oxygen- depleted ballroom! The papers that were rustled, the eyes that were cast down to the ground, the covert glances projected at me by the more adventurous members of the gathering. The panel briefly consulted, clearly agitated for the first time during the session. Eventually, Robert Hall was designated to respond to my question. He rose to his feet, his body extended to its full stature, his face crimson, his eyes flashing with disapproval. ’Sir’, he politely spluttered, (probably meaning, ‘you moron’) “you have just challenged the revealed truth of the profession”.
The meeting ended without another word, the audience dispersed in an unworldly silence, and I left the room knowing exactly what it must have been like to contract leprosy before a cure became available. But, of course, I was right…
Tags: economy-wide shocks, failure of macroeconomics, falsification of theories, rational expectations, recessions
January 10, 2010 at 4:56 pm |
Then there’s the Black-Swan effect to consider. Which is sorta like, how do you what your blind spot is, if it’s your blind spot?
January 11, 2010 at 7:49 am |
I might not be much of an economist (like I have a degree but that is all) but I do see that you are right because you have followed through on the theory of JM Keynes, which these professors, especially Summers have discarded. By this I mean that Keynes recognized that his theory was not perfect and that challenges to his theory were legitimate. Your description of that conference shows that economist have become somewhat bloated in their own opinions. I could mention some Australian economist such as Ross Garnault who have a somewhat bloated opinion about their models and the like.
What has happened is that by totally rejecting Keynes they have relied upon their own models as though they are infallible, yet each seems to lack the Keynesian ability to analyze the situation at a certain point in time. This is one reason why they fail to understand these dips……
It is not hard to explain…. what happened that was important around that time? Such as legislation that changed things like welfare payments or more to the point, changes that allowed low income individuals to apply for mortgage loans without necessary security? What happened in terms of world trade? The Japanese crisis? One point here, it was Geithner who held the Japanese recovery back because of his policy prescriptions that were wrong…. did they have an impact at that time.
So many reasons that come from hindsight.
January 12, 2010 at 7:27 pm |
Your comment is excellent. I intend to respond to the issues that you raise, at least partly, in tomorrow’s column. I wrote in much greater detail about the issues in my recent book (co-authored with Nathanael Smith): Economic Contractions in the United States: A Failure of Government. It is available at http://www.amazon.com.
January 12, 2010 at 5:44 pm |
I’m hoping you’ll get into specifically why you think the theory of rational expectations is flawed.
January 12, 2010 at 7:29 pm |
I intend to deal with that issue, at least in part, in tomorrow’s column. It is a fair question.
January 13, 2010 at 6:13 pm |
There are a lot of discrete assumptions that lead to the kind of results generated by those first/second generation rational expectations, RBC models. They include:
* full information about the parameters/form of the model is available (somewhere/somehow);
* there are no information acquisition costs;
* there are no information processing costs (or calculation costs);
* agents optimize (given their expectations, information and computation costs);
* agents update beliefs and form expectations in a mathematically consistent way (i.e. Bayes Law, mathematical expecation operators).
When people object to rational expectations they are often objecting to the final two points which correspond broadly to “rational” and “expectations” respectively. However to the best of my knowledge there is no reason to believe that any difficulties in explaining the data cannot be solved via questioning just the first three points (without giving up on either rationality or mathematical expectations). Agents are uncertain of the true model, they don’t know its parameters, without exerting significant effort they get only very crude signal’s of the economy’s state, agents find computation expensive (in terms of time, or the costs of paying someone to do it for them). Models have been developed investigating the implications of all these departures from the idealised first generation RBC world.
It is not particularly productive however to give up either rationality (conditional on information and computation costs) or mathematical expectations (under imperfect information), since departures from either of these cannot fail to be ad hoc. Giving up rationality in economics is much like invoking god in science: it explains everything and nothing. I am prepared to admit that in 100 or a thousand years time we may understand the implications of relaxing the first 3 assumptions above well enough in order to know we have to give up on the final two. We are nowhere near that point yet however.
January 13, 2010 at 6:19 pm |
>>’Sir’, he politely spluttered, (probably meaning, ‘you moron’) “you have just challenged the revealed truth of the profession”.>>
Hmmm … in physics, chemistry, and medicine, that’s what they give Nobel prizes for.
January 16, 2010 at 7:31 pm |
They gave Daniel Kahneman the Nobel prize in 2002 for challenging the notion that people are rational: http://nobelprize.org/nobel_prizes/economics/laureates/2002/kahnemann-lecture.pdf
This is accepted (and repeatedly demonstrated) wisdom in psychological research, by the way…
January 13, 2010 at 9:34 pm |
How Markets Fail: The Logic of Economic Calamities by John Cassidy.
January 14, 2010 at 12:44 pm |
[...] http://charlesrowley.wordpress.com/2010/… [...]
January 14, 2010 at 4:03 pm |
What about the Gulf War, the rise in oil prices, and a very vulnerable banking sector stretched by RE loans (i.e. S&L crisis)?
January 14, 2010 at 4:41 pm |
Those are excellent observations. Surely a well-calibrated macroeconomic model would have picked them up if they were indeed significant shocks. None of the three models presented at the AEA meeting identified them. My own theory about the 1990-1991 recession is that the Fed failed to respond to a tightening in M2 and that adaptive expectations, not rational expectations, drove the outcome. I was not allocated time to mention that alternative hypothesis at the meeting.
January 15, 2010 at 12:57 am |
I humbly suggest you read the late Jude Wanniski’s work at polyconomics.com.
January 15, 2010 at 10:09 am |
All of this advice would be sensible if we were trying to explain a few data points or if there was some simple universal law that we were trying to uncover.
Please dont pretend that what is to be explained in macroeconomics is a small set of data points.
There are billions of different savings decisions, billions of lending decisions all at different interest rates. People face different prices for goods across all markets. There are billions of individual relationships that are not easily substitutable, etc. We also have data on all these things. So we abstract, general equilibriumize to make sense of this mountain of data etc. We use computers rather than worry about an analytically tractable straightjacket etc. We seek explanations rather than definitive answers. We put some pride of place on explanations that are model consistent – including the dreaded rational expectations hypothesis – big deal. by this measure I think there are many macro economists doing much of what you advise.
Please do ask them to make the models more simple and add more stuff – that is logically impossible. However, general checks of robustness is useful and we should be open to new kinds of robustness checks.
I think we should not get too worried when one of the big dogs barks – bob hall definitely has bad hair days when it comes to public discussions and is known to flip flop. I worry that you think he has this affect on the profession.
January 17, 2010 at 1:05 pm |
Bagehot wrote about this in 1876. Of course what can he possibly have to say to us today, in our so much more sophisticated era! Bagehot just didn’t know enough maths to be a real economist.
“effectual political economy … is a very difficult thing; something altogether more abstruse and difficult, as well as more conclusive, than that which many of those who rush in upon it have a notion of. It is an abstract science which labors under a special hardship.
Those who are conversant with its abstractions are usually without a true contact with its facts; those who are in contact with its facts usually have little sympathy with and little cognizance of its abstractions. Literary men who write about it are constantly using what a great teacher calls ‘unreal words’ – that is, they are using expressions with which they have no complete vivid picture to correspond.
They are like physiologists who have never dissected; like astronomers who have never seen the stars; and in consequence, just when they seem to be reasoning at their best, their knowledge of the facts falls short. Their primitive picture fails them, and their deduction altogether misses the mark – sometimes, indeed, goes so far astray, that those who live among the facts, boldly say that they cannot comprehend ‘how anyone can talk such nonsense’.”
http://www.moneyhoneyblog.com/where-is-bagehot-now-that-we-need-him/
February 2, 2010 at 8:52 pm |
Hi. Very interesting Blog. Not really what i have searched over Google, but thanks for the information. Can you email me back, please. Thank you.
February 8, 2010 at 7:18 pm |
[...] Just look at these links, many of them absolutely excellent and the last two are my favourite (*, *, [...]
June 15, 2010 at 2:00 am |
Thank you for this post. I will be sure to add this to my favorites. Please look at my musical concert August 21st at UC Berkeley!
January 2, 2011 at 12:44 pm |
[...] How Macroeconomics Lost Its Way 1: Theory Ignores Evidence January 2010 18 comments 4 [...]