Arnold Kling  

How an Irrational Fad Became Entrenched

PRINT
Austerity Within Reason... My Essay on the Foreclosure Sc...

Scott Sumner asks,


How was it that just a few years later it was almost impossible to get anything published without assuming rational expectations, efficient markets, etc.

Let me tell you. For a period of about 5 to 10 years, you could not get hired as a macroeconomist at a major university without a Ph.D disseration that used Euler equations. Everybody was in love with the technique. It didn't matter whether the young graduate students using the technique had any economic intuition or not. Those of us who didn't jump onto the fad simply could not get placed out of graduate school. At MIT, Dornbusch and Fischer were the ones with the power to place graduate students in academic jobs, and so those jobs went to their dittoheads. Who then proceeded to submit and referee articles for journals and--guess what?--just about every published article conformed to what Olivier Blanchard aptly called the "haiku" of the technique.

The rational expectations revolution was carried out by scarcely more than a handful of economists dispersed among key graduate departments, whose genes were the only ones that stayed in the pool. In a different world, Clower and Leijonhuvfud would have been ancestors of a significant part of modern macroeconomics. Had that transpired, I believe that the field would be in much better shape now.


Comments and Sharing





COMMENTS (15 to date)
Luis H Arroyo writes:

Clower & Leijonhuvfud!!! Mmmmm my God!! What a great difference! I remember well...So, they were (or are) so good as I used to think!

Will Ambrosini writes:

To be clear: having an Euler equation means the economic agents in your model optimize... it says nothing about your models assumptions regarding rational expectations, efficient markets, etc.

Quit beating up on Euler equations! :-)

ThomasL writes:

This bears a striking similarity to many of the critiques of the current state of climate science.

You can't get "in" and get published unless you please a fairly small number of very influential people--people that all think very much alike.

Voilà, out pops a "scientific consensus."

Joey Donuts writes:

Something from Kenneth Boulding

CANTO XIL
A matrix is a set of roles
Arranged like banks of pigeon-holes.
It is a box of empty boxes,
Filled with nothing but paradoxes.
Their mathematical operations
Involve elaborate permutations
In which the x’s and the i’s and j’s
Dance algebraical ballets,
And in the course of their gyrations
Solve simultaneous equations.

CANTO XIII
It gives a piece a certain unction
When each relation’s called a
Function,
And if prestige is what we seek,
We write the functions out in Greek.
These intellectual acrobatics
Are fun, but are not mathematics,
For there are subtle traps immense
Where all that’s symbol is not sense.

kurt writes:

Now if we could only do away with Modern Portfolio Theory. Fat tails ftw.

Rebecca Burlingame writes:

@Joey Donuts,
You made my day. I'm going to have my husband make up a copy of Canto XIL and XIII with his beautiful calligraphy.

Roger Sweeny writes:

XIL? That doesn't conform to the rules of Roman Numerals. Perhaps a mistype for XII?

Google is your friend. That is indeed Canto XII.

http://quod.lib.umich.edu/cgi/t/text/pageviewer-idx?c=mqrarchive;cc=mqrarchive;rgn=full text;idno=act2080.0

Bob Murphy writes:

Arnold, this is great stuff. May we have another, sir?

Jeff writes:

I've commented on this over at my new blog Acme Quality Assurance.

ajb writes:

Why just complain about Macro?

IO was destroyed by all the structural modeling as even Joskow and others at top schools have noted.

Many top papers in political economy publishable in AER or JPE up to the early 1980s would be unpublishable today. [Just look at the current J of Law and Econ because it is the last major journal that publishes low tech empirical work of the sort common to JPE in the 1970s.]

Or look at how the identification police have taken over applied metrics. Yes IV and random discontinuity studies are helpful. But that does NOT mean that correlations without IV or diff and diff are useless in the right context.

Or look at the endless theory papers in the top journals that are basically philosophical exercises whose point could be made with a simpler, less formally rigorous model. It's a wonder that stuff like Leeson's occasionally gets into a good journal. For the most part, to get published without high tech, you have to be a) very lucky or b) very well connected. Usually both.

The easiest way for a schmoe in a low ranked school to get into the top journals is a technical piece that successfully tweaks the cause du jour of the professiariat.

Most outsiders don't understand. The ideology of mainstream econ isn't socialism or liberalism.

IT's MINDLESS TECHNIQUE.

Chris R writes:

I'm with Will here about halfway. The problem is that it's the same three Euler equations showing up everywhere, usually based on some sticky price model which has little empirical support. This model is nothing if not simple. Interesting macro, meanwhile, has gotten pushed toward the sidelines.

Bob Gordon has called for a return to 1978 era macro; I think that 1998 era macro is more appropriate. By 1998 people were telling coherent interesting stories about search-based unemployment, labor hoarding, financial frictions, news-driven business cycles, limited participation, increasing returns, and the like. Something happened between 1998 and 2008 to push these things toward the sidelines.

Mo writes:

Arnold's decision to not go with the profession a few years before I was born has greatly increased my utility the last few years via this blog.

I have a feeling he wouldn't be thinking and saying the things he does today had he ended up climbing the academic ladder.

Thank you!

Rebecca Burlingame writes:

@Mo,
I agree. And it's an exciting time to be thinking about things economic...will only be more so in the years ahead.

joeedh writes:

Has no one seriously thought of doing real, field experiements?

Come on people! We have the internet! It's not hard to do macroeconomic experients! Just look at all the barter sites, World of Warcraft, and Second Life!

If economists do not learn this soon, they may be tasting a little "creative destruction" of their own. :) For example. . .anyone ever thought of a website to test different monetary systems (e.g. Keyens's bancor or more recent ideas) and see which ones best promote sustainable, balanced trade?

I'm thinking of something like that myself. There is demand for alternative monetary systems (e.g. the return of the local currency movement) now that our recent (incredibly mild) balance of trade crisis has wrecked the dollar.

Dan in Euroland writes:

ajb,

Did you miss the whole IV movement? pretty straight forward stuff right there.

joeedh,

just about every economist in the world would love to run experiments. but there is some difficulty with actually doing it. The whole N=1 earth thing.

While there is research on artificial communities even that is tough to do. All those gold farmers will not be too pleased if WOW starts debasing. they will then switch games.

There are many legitimate criticisms of modern econ. None are appearing in this thread.

Comments for this entry have been closed
Return to top