Arnold Kling  

A Difficult Book

Another Reason To Doubt Infant... Villains, Victims, and Economi...

In the foreword to Imperfect Knowledge Economics, by Roman Frydman and Michael D. Goldberg, 2007 Nobel laureate Edmund S. Phelps writes,

The authors argue that if we aspire to build models that apply to modern economies--economies whose central functioning is the manufacture of change through their innovative activity and their adoption and mastery of the innovations made available--it is contradictory to adopt the rational expectations postulate that whatever change takes place in the future is already knowable and known in the present...

Keynes saw the rate of return as quite unknown and the demand for investment funds as driven by enterpreneurs' "animal spirits." Hayek saw that every participant has little or no knowledge of how the economy works as a whole, contrary to rational expectations; that a participant is apt to ahve only some highly specialized knowledge...such knowledge may permit creative person to conceive some new business strategy or new business product that is not in the air, not already known by all...Keynes and Hayek were right but did not carry the day.

Unfortunately, once I moved from Phelps' foreword to the actual book, I could not understand it, because of all the fancy equations that are not accompanied by sufficient intuitive explanations (for me). So I cannot really write an informed review. The book jacket has endorsements from Jean-Paul Fitoussi, Ken Rogoff, Kenneth Arrow (another Nobel laureate), and Alan Blinder, any one of whom could understand the book if they were motivated to do so.

I have a hard time believing Blinder was motivated to really delve into the guts of the book. Rogoff probably was, because the authors focus a lot on the empirical performance of exchange rate models, where Rogoff did some of the most important research.

Exchange rate markets differ from most markets in that currencies are traded in central locations, so that anyone with any sort of opinion or exposure can participate on pretty much equal terms. So the uncertainty is going to consist of different models of the world. I thought that this sort of situation had been explored by Mordecai Kurz, but I do not see Kurz listed in the list of citations.

Because participants in financial markets cannot be certain which model of the world prevails, I have always thought that the sorts of variance bounds tests pioneered by Shiller are silly. Those are tests that compare the observed variation of fundamentals (dividends in the case of stocks) to the variation in market prices. It seems to me that if you make a small change to your dividend growth forecast, you can get a big change in a stock price. Back in the day, people pooh-poohed my criticism. They said I was just worrying about nonstationarity. I did not understand what they were saying, and I was never convinced that they understood what I was trying to say.

In any event, my view of the world now is that financial intermediaries, including ordinary corporations that raise funds in the capital markets, are not perfectly transparent (contra Modigliani-Miller). So investors have to guess what sorts of risks they are taking. Risk premiums move around as innovations occur and as investors revise their ways of making guesses about the intermediaries. The subprime mortgage kerfluffle is an example of this sort of movement.

So I agree broadly that it is a shame that Keynes' and Hayek's notions of uncertainty and imperfect knowledge got trampled under the steamroller of rational expectations. However, the "new" mathematical approach of Imperfect Knowledge Economics does not speak to me.

Comments and Sharing

COMMENTS (6 to date)
Greg Ransom writes:

The un-scientific mistake is to make the _philosophical_ demand of the "mathematical" economists that this individual knowledge and lack of knowledge be put into mathematical notation. This is not only unscientific, it is bad math and bad statistics. In other words, the economic profession has been captured by bad _philosophers_, folks who are incompetent as scientists because that are bad philosophers first and foremost.

spencer writes:

I agree with you completely.

Often the greatest thing an entrepreneur has going for them is their absolute confidence -- often based on ignorance -- that they can not fail.

But this mean that all the models that business investment is based on a rational calculation that the expected return will exceed the cost of capital is really meaningless. I will stick with my old bromide that any corporate vice president worth his salt will never propose an investment that does not promise a 33% to 50%
rate of return. They just build into their forecast that sales when the project comes on line will be strong enough to generate those returns. I'll never forget that as a stock analyst in 1981 going to corporate presentation after corporate presentation and having CEO after CEO explain that the 15% to 20% bond yields were having no impact on their investment budget.

spencer writes:

sorry the sytem was showing that my comment had not been accepted.

Andreas writes:

"Bad math"- in my opinion- is, if somebody doesn't follow the mostly very strict rules of algebra, calculus and so on.
"Bad statistics"- in my opinion- is either, if someone doesn't use statistical tools (like variations and such) the way they are supposedly used, or if someone interprets statistics in a way, which are out of their scope.

Of course it would be unsientific, if either of those things were true for any production of knowledge.

Now if someone builds a model, which follows the rules of maths and statistics, but over- interprets it (which all of us know is happening all the time and isn't at all limited to the economic profession), it's in my oppinion "just" bad- science, but not "unscientific".
If someone tries to express things mathematically, which can be only put into equations in a very reduced or unuseful way, than it's "nothing more" than over- interpretation.

So many of those people may be bad in terms of interpreting their own models and results, but I don't think that this makes them "incompetent as scientists" in general.

Anyway, I like maths because it makes things easier to understand.
But probably I'll change my mind, after trying to read Imperfect Knowledge Economics .

ps.: I think it always looks much more scientific, if you put something into an equation than expressing it in full length sentences. ...and isn't that all that counts ;)

Jeff Hallman writes:

Rational expectations does not require that future change be knowable and already known, unless you're talking about change in the model itself. RE is a kind of consistency requirement: you can't say here is how the world works, but people don't know that's how it works, so their expectations are different from the predictions of my model.

A somewhat weaker form of RE was given by Abe Lincoln: “You may fool all the people some of the time, you can even fool some of the people all of the time, but you cannot fool all of the people all the time.” Lincoln's formulation is really all you need to get a vertical long run Phillips Curve.

Matt writes:

"entrepreneur has going for them is their absolute confidence"

This is what needs to be quantified. Why does an entrepeneur have a different comfort zone for acceptable risk?

It would seem to me of enormous evolutionary advantage if we all had the same comfort zone for acceptable risk. Imagine the great cooperative avantages if small groups all operated on the same expectation of success in planning.

The prediction, then, is that when the number of transactions are very small, participants cannot find arangements that puts everyone in the proper comfort zone. Or, on the other "thick tailed" extreme, the wleath is so near the comfort zone that any quanta of wealth applied will nearly double the relative wealth of the individual.

In other words, quantum effects, because all of our brains work best with a constant uncertainty in important everyday affairs, and efficiency falls at greatly than non-linear rates when confort levels are too safe or too risky.

The quantum effect implies there is a fixed, constant quanta of wealth that the economy recognizes, and the usual explanations that this effect can be ignored when transactions are large holds. The quant would be an absolute fraction, say, like 1/20, meaning we all expect about 1 out of 20 things to screw up today, all of us, regardless of wealth. Outside of this range, we will readjust our lives and investments to bring the uncertainty back into comfort zone.

Comments for this entry have been closed
Return to top