Economics and Physics Envy

April 1, 2010 9:09 PM

"Take a little bad psychology, add a dash of bad philosophy and ethics, and liberal quantities of bad logic, and any economist can prove that the demand curve for a commodity is negatively inclined." MIT economist Andrew Lo and string theorist turned asset manager Mark Mueller on the "physics envy" that plagues economics, and how to stop worrying and love uncertainty.
posted by escabeche (37 comments total)

I'm studying behavioral economics right now. One of the central "theorems" of behavioral economics is that human behavior is never rational. That's what I love so much about the topic.
posted by LSK at 9:43 PM on April 1, 2010

But there's an opposite lying here - it's related to the "tell the truth to idiots and it just makes them stupider" axiom - the more we try to impose reason, the more unreasonable the cases become.
posted by sporb at 10:27 PM on April 1, 2010

Here is more information on why the coin toss isn't random. (worthy of an FPP in its own right, me thinks).

To analyze the motion of a tossed coin, Diaconis solicited help from Richard Montgomery, a math professor at the University of California-Santa Cruz. Montgomery had developed the Falling Cat Theorem -- a theory that explains how a cat dropped from any angle always manages to land on its feet. Surely this expertise in angular momentum could also apply to falling coins, Diaconis thought. Staring at a picture of tumbling felines on Montgomery's office wall a few years ago, he blurted, "You're the man for me."

A collaboration was born. Diaconis and Montgomery started meeting once a week, at Stanford or UC-Santa Cruz, to discuss their ideas. In about eight months, they arrived at a startling prediction: A flipped coin is biased to land on the same side it starts out on.
posted by water bear at 10:43 PM on April 1, 2010 [3 favorites]

btw, great article on econ. thanks!
posted by water bear at 10:45 PM on April 1, 2010

On Diaconis -- interesting that he is the person who came up with the 7 shuffles rule for shuffling cards. That was well known knowledge in the Magic: the Gathering community when I was playing, but I didn't know where it originated from.
posted by empath at 10:53 PM on April 1, 2010

I'm studying behavioral economics right now. One of the central "theorems" of behavioral economics is that human behavior is never rational. That's what I love so much about the topic.

Sounds great - because anything that gets us away from the Boston school of "rational actor" nonsense is a big advance.
posted by rodgerd at 11:19 PM on April 1, 2010 [2 favorites]

Economics as a whole would be well-served to stop pretending that it's a science.
posted by Pope Guilty at 12:29 AM on April 2, 2010 [2 favorites]

If you give me a whole lot of bad psychology, I can make anything out of it.
posted by StickyCarpet at 12:47 AM on April 2, 2010 [1 favorite]

Economics is at least as much a science as meteorology is, and probably makes predictions just as well (Not badly, as long as you don't get too far into the future or look for too much detail).
posted by empath at 12:52 AM on April 2, 2010 [2 favorites]

I just finished reading this, it's actually pretty fascinating, and not at all a shot at economics. It's really just about risk management and how to account for "unknown unknowns" to stop you from taking a bath when your model fails.

One thing that struck me while reading this, is that when you figure out the model for a physical behavior, the behavior isn't going to change because you figured it out. For example, once you figure out the law of gravity, G isn't going to change.

However, if you're modeling the stock market, and you're successful at predicting behavior -- even if you're 100% successful at predicting prices, then corporations will start pouring money into your model, which changes the behavior of the market, which invalidates the model. In fact, the better your model is, the more likely it is that it will be invalidated sooner, unless you don't apply it or you keep it secret.
posted by empath at 1:54 AM on April 2, 2010 [4 favorites]

Economics is at least as much a science as meteorology is

Meteorologists all pretty much agree on the basics of the field, and there's no prominent group of meteorologists claiming that empirical data is irrelevant to meteorology.
posted by Pope Guilty at 2:57 AM on April 2, 2010 [6 favorites]

I'm studying behavioral economics right now. One of the central "theorems" of behavioral economics is that human behavior is never rational. That's what I love so much about the topic.

The whole 'rational actor' model never made any sense. And I wonder how much "behavioral economics" is just warmed over social psychology. I mean, if you look at the "freakonomics" guys they're basically just talking about the kind of research that other social scientists have been doing for decades.

It's like they ran out of crap to spew about money and don't really have anything left to do. That's probably even more true for economists that can't do the rocket science type statistical/mathematical work.

But there's an opposite lying here - it's related to the "tell the truth to idiots and it just makes them stupider" axiom - the more we try to impose reason, the more unreasonable the cases become.

Well, there's a difference between "telling the truth" and "Imposing reason"? What does "Imposing reason" even mean? How do you know that they're not being the ones who are reasonable and you're being unreasonable. The bankers are Goldman Sachs and other Wallstreet companies probably think their behavior is totally reasonable, and that everyone else is just acting crazy after being blinded by greed, or whatever.

---

I don't think Meteorology is a good comparison to Economics. Meteorologists are fully aware of the limits of their predictive ability, and they know it. And more importantly, Meteorology is all built on physics, while Economics tries to model mass human behavior, for which there are no hard underlying rules to build off of.

A Meteorologist can use the Naiver Strokes equations to model the weather at a given resolution and it will work. But you can't simulate a single human, much less billions interacting in the real environment.
posted by delmoi at 3:30 AM on April 2, 2010

Apropos of nothing, or very little, Roy Sorenson's wonderful article on charity and meta-charity is an example of some good philosophy about the efficient market hypothesis and the rationality of human behavior. It's long(ish) but a great read, with great counterexamples of the the "rational actor" bias.
posted by rudster at 3:32 AM on April 2, 2010

Sounds great - because anything that gets us away from the Boston school of "rational actor" nonsense is a big advance.

Chicago?
posted by DU at 4:41 AM on April 2, 2010 [1 favorite]

If you give me a whole lot of bad psychology, I can make anything out of it.

Why, I could make a hat, or a brooch, or a pterodactyl!
posted by PMdixon at 7:01 AM on April 2, 2010 [1 favorite]

I like that Sorenson article, but not for economic reasons:

"If there are no beliefs, then there are no agents and the entailment thesis `Charity implies meta-charity' is vacuously true. Thus I count all eliminativists about beliefs (such as Chuchland 1979) as degenerate co-subscribers to `Charity implies meta-charity'. But Dennett is not merely an ally on a technicality. His account of charity furnishes substantive support for the entailment thesis. After all, Dennett's theme is that the attribution of rationality is constitutive of the framework that employs beliefs and desires. His defence of charity is conceptual, not metaphysical."

:-)
posted by anotherpanacea at 7:01 AM on April 2, 2010

Human behavior is based on physics.
posted by empath at 7:20 AM on April 2, 2010

Taxonomy of Uncertainty:

level 1: complete certainty
level 2: known probabilities
level 3: discoverable or estimatable, fixed probabilities
level 4: nonstationary, time-varying probabilities
level 5: total ignorance

very insightful article with a great historical perspective!
posted by dongolier at 7:54 AM on April 2, 2010 [1 favorite]

You keep telling yourself that empath. Just keep repeating it. Pure force of will can make reductive materialism true.
posted by oddman at 7:55 AM on April 2, 2010

its a continuum not a taxonomy. the moment you put a label or tag on it, it becomes a discrete moment in time, defined by a definition, so to speak. but on a sliding scale, perhaps initially developed without metrics, one could conceivably pinpoint where a point is on the axes.
posted by infini at 8:03 AM on April 2, 2010 [1 favorite]

The "Modest Proposal" is on p.64:

most M.B.A. students at Sloan and other top business schools do not have the technical background to implement models such as Li’s (2009) Gaussian copula formula for ABSs and CDOs, nor does the standard M.B.A. curriculum include courses that cover such models in any depth. Such material—which requires advanced training in arcane subjects such as stochastic processes, stochastic calculus, and partial differential equations—is usually geared towards Ph.D. students.

... funding for finance Ph.D. students might be raised by imposing a small surcharge on certain types of derivatives contracts, e.g., those that are particularly complex or illiquid and, therefore, contribute to systemic risk. This surcharge may be viewed as a means of correcting some of the externalities associated with the impact of derivatives on systemic risk. A minuscule surcharge on, say, credit default swaps, could support enough finance Ph.D. students at every major university to have a noticeable and permanent impact on the level of financial expertise in both industry and government.

posted by anthill at 8:06 AM on April 2, 2010 [3 favorites]

Human behavior is based on physics.
posted by empath at 7:20 AM on April 2 [+] [!]

anti-eponysterical.
posted by ennui.bz at 8:19 AM on April 2, 2010 [2 favorites]

is there a MeFi meme yet on palm on forehead rather than cheek? or are we still generally saying "face" ? ;p
posted by infini at 8:20 AM on April 2, 2010

Human behavior is based on physics.

This is true, for certain human behaviors that include, for example, remaining on the ground due to gravity.
posted by Faint of Butt at 10:20 AM on April 2, 2010

Human behavior is based on physics.

This is true, for certain human behaviors that include, for example, remaining on the ground due to gravity.

Look, you have two options:

1) Humans have a metaphysical "soul" of some sort that provides for behavior. I'm only aware of the theologically-proposed soul, which most materialists reject out of hand for lack of evidence. I know of no study that has provided positive, direct evidence for the existence of a metaphysical soul of any sort.

2) Humans' do not have a metaphysical soul, and so our personality and behavior is seated in the brain. Given that the brain is a physical object, operating on physical rules, then our personality and behavior must be based on physical rules.

Now, I'll grant you that in Scenario 2, there is a great deal of unexplained and poorly-understood complexity in the system. But, high complexity does not lead to the conclusion that humans' behavior is seated in a metaphysical construct any more than the complexity of atomic theory leads to angles on pin heads.

Please don't confuse your romantic notions of humanity's primacy with analysis of our cognition.
posted by Netzapper at 11:10 AM on April 2, 2010 [1 favorite]

The whole 'rational actor' model never made any sense. And I wonder how much "behavioral economics" is just warmed over social psychology.

Economists who want to acknowledge they have a whole lot more in common with social sciences are a win, don't mock 'em for it.

Chicago

Stupid American city names.

Pure force of will can make reductive materialism true

Let me know when you can think yourself into flying, and I'll be take your snide quips about "materialism" seriously.
posted by rodgerd at 11:26 AM on April 2, 2010 [1 favorite]

I just came in to thank the OP for quoting from Samuelson's Foundations. Far too neglected, that book is. (Yes, I see the irony in praising the book that ushered in the great wave of physics-envy, but it really was an amazing work, and in many places outlined the limits of the methods).

Haven't read the full FPP article yet, as I am at work (on a quick lunch break - hi bosses!), but as far as "physics envy" goes as a criticism of much modern economics, see much of the work of Phil Mirowski, including his delightful More Heat than Light. [Double warning: pdf link to mises.org]
posted by dilettanti at 12:07 PM on April 2, 2010

Why so much hate for econ on MeFi?
posted by treeshar at 2:06 PM on April 2, 2010 [1 favorite]

Economics as a whole would be well-served to stop pretending that it's a science.

Anti-intellectualism is just as ugly when it comes from the left.
posted by ripley_ at 3:33 PM on April 2, 2010

It's not anti-intellectualism. Any hard science with as much of a contempt for empiricism and as much dissent about the very basics of the field would... well, it wouldn't be a science, would it? That you are systematic about something does not make you a scientist; that you come up with ideas about the world does not make you a scientists. The current state of Econ is as if Physicists couldn't agree on what Force times Acceleration equals, or if chemists couldn't agree on how many valence electrons Potassium has. Economists can either start acting like scientists or start having a better predictive power than coin flipping.
posted by Pope Guilty at 4:07 PM on April 2, 2010

And their Nobel prize is a licensed knock-off.
posted by StickyCarpet at 5:54 PM on April 2, 2010 [2 favorites]

The current state of Econ is as if Physicists couldn't agree on what Force times Acceleration equals, or if chemists couldn't agree on how many valence electrons Potassium has.

What exactly are you referring to? I don't see as much dissent about the basics of economics as you do. There isn't much controversial content in any introductory micro textbook, macro perhaps slightly more.

I think that what you're seeing is greatly magnified by what Alan Blinder refers to as Murphy's law of economic policy: "Economists have the least influence on policy where they know the most and are most agreed; they have the most influence on policy where they know the least and disagree most vehemently."

Rent control? Trade policy? Minimally distortionary tax policy (i.e. consumption taxes)? There isn't much dissent on those topics within the economics profession, but modern nations that attempt to follow economic consensus on any of those issues are few and far between. I see this as more of a political problem than anything else.
posted by ripley_ at 5:57 PM on April 2, 2010

Economics is pretty spectacularly successful at prediction of behavior within limited domains and markets. It's just very, very difficult to make accurate predictions as you add more variables and move out further in time.
posted by empath at 6:51 PM on April 2, 2010

If the author of this article and Brian Greene had a child it would be called The Unified Theory of Everything in the Universe: We Don't Know Everything, So Let's Take Into Account, I Guess?
posted by water bear at 1:04 AM on April 3, 2010

he would undoubtedly be a faithful agnostic.
posted by water bear at 1:06 AM on April 3, 2010

i like acrostics
posted by infini at 2:20 AM on April 3, 2010

there was a nice interview with lo here :P

cheers!
posted by kliuless at 8:32 AM on April 4, 2010