Pages

Thursday, September 3, 2009

Modern Economics Like Physics Without Symmetry Breaking.

There is an article by Paul Krugman that I mention for two reasons:
  1. It's Paul Krugman!  You probably cannot point to another living economist with as many honors as him.
  2. It is a perfect segway into a discussion on how symmetry breaking ruins the elegance of theoretical physics.
Theoretical physics is often described as elegant.  It is full of symmetries yielding aesthetically pleasing equations.   Learning all of this stuff can be quite fulfilling and gives you a real sense of wonder for how majestic nature is.

Then you look outside and realize all of this beauty is only approximately correct.  Many symmetries are broken to some degree.  Sure there are still some left, such as the symmetry keeping the photon massless, but most are not really exact.

From a calculation standpoint this is a real nightmare.  The symmetries make the math not only elegant but tractable.  Every symmetry that breaks creates a lot of mess that physicists have to learn to deal with.

For example, supersymmetry is so beautiful and elegant I have heard one physicist say if it isn't correct "God messed up".  However, if supersymmetry is correct, it is broken and causes so much mess and hardship that nobody knows how to properly deal with it.

Now onto economics.  One of Krugman's main points boils down to: the mathematical models in modern economics theory are very elegant, clever and inspiring.  However, they are about as accurate as theoretical physics without symmetry breaking.

To quote one example Krugman gives:
The theoretical model that finance economists developed by assuming that every investor rationally balances risk against reward — the so-called Capital Asset Pricing Model, or CAPM (pronounced cap-em) — is wonderfully elegant. And if you accept its premises it’s also extremely useful. CAPM not only tells you how to choose your portfolio — even more important from the financial industry’s point of view, it tells you how to put a price on financial derivatives, claims on claims. The elegance and apparent usefulness of the new theory led to a string of Nobel prizes for its creators, and many of the theory’s adepts also received more mundane rewards: Armed with their new models and formidable math skills — the more arcane uses of CAPM require physicist-level computations — mild-mannered business-school professors could and did become Wall Street rocket scientists, earning Wall Street paychecks.
But, as you know, this model failed to predict what recently happened.  Despite it's elegance and award winning nature, CAPM fails to take into account messes that exists in the real world.

It seems theoretical economics is about as useful as theoretical physics without symmetry breaking.  Sure there is an incredible amount of good stuff the theory gives you and its founders should be showered with praise.

However, until economists admit their models aren't perfect and that they need to work out the messy details of the real world, much like physicists have to with symmetry breaking, economic models are going to fail to predict economic predicaments such as the one we have before us.

5 comments:

  1. There is an interesting extract of a discussion between Carl Friedrich and Werner Heisenberg, from the book of the last one : “Physics and Beyond”.

    Heisenberg : “...Anyway we have understood, I did answer, that, for the person to whom the scientific and technical progress has given an important task, it is not sufficient to think only about this task. He has to observe the accomplishment of this task as a part of a great development ; obviously, he must have a positive attitude relatively to this development, since he is working on these problems. It will be easier to take some fair decisions if he considers this general context.” (See the part about the responsibility of the searcher.)
    Carl Friedrich : “This would give utterance to the fact that the searcher should as well strive to realize a contact with the public life, to exert an influence on the leading organisms of the State, since it is not sufficient for him to think correctly, but that he wants as well to act and make act the others correctly… Insofar as the scientific and technical progress becomes more important for the collectivity, the influence of the bearers of this progress on the public life should increase. Of course, that is not in question to suppose that physicians or engineers should be more capable to take important political decisions than politicians them-selves…” (See the same part)

    Heisenberg adds further, out of the conversation : “…the important thing for me was not only to obtain the greatest as possible support of the scientific research by the authorities and the public opinion ; what was for me essential as well, was to realize a penetration of the scientific way of thinking up to the inside of the governmental spheres. I thought sure enough that it had to be recalled all the time to those which will assume in our place the responsibility of the functioning of the State that, in order to govern, it was not enough to assure an equilibrium between some opposite interests ; and that there were frequently some unavoidable necessities, based on the structure of the modern world, in front of which any irrational attitude of evasion toward some sentimental and unrealistic considerations could only lead to some catastrophes.”


    In order to conclude about politics and sciences, I want to add that Heidegger has confessed in the interview he has given to “Der Spiegel” in 1966 (See Ecrits politiques, 1933-1966, par Martin Heidegger, éditions Gallimard), that philosophy (as he practiced it) should not any more have any directly useful role in the change of the world. But he has well glimpsed that the most useful role of philosophy had a future from sciences (not literary), although writing a philosophical work needs some literary talents ; on the top of that it is better to be scientific and literary, in order to have a supporting group through the works of some dead persons (what is current in philosophy) or alive, but also the support of history. Furthermore the sense of this scientific exit (which is well compatible with the political part of my work, for which I did begin an American version) is present in Heisenberg’s book “Physics and Beyond”, relating of (not literary) sciences, of moral, of philosophy, of religion, and of politics. Finally I know the USA are a bit in advance about the relation between not literary sciences and politics, if we consider Benjamin Franklin; so I hope they would help the world to go in this direction. It is time that intellectual creators, which are competent in politics, have the possibility to have more power in this field. So we searchers of the most advanced countries have to unify in order to help real intelligence to have what it deserve! We can do it, especially in a world which needs to change! But we have to take a lesson from Mr. Penn (creator of Pennsylvania), which is to not associate with the bad people.

    ReplyDelete
  2. Postscript : Before some machines and some other inventions were some slaves, I think a choice should be done.

    ReplyDelete
  3. Cartesian, thanks for the quotes.

    "It is time that intellectual creators, which are competent in politics, have the possibility to have more power in this field."

    I've always wondered what would happen if you allowed some ultra-intellectual, like Ed Witten, to be the president. :)

    ReplyDelete
  4. Well said Joe. In fact, the only ones to correctly predict this nonsense, the Austrian school of economics, are the ones whose ideas are ostracized and considered nonsensical. Ironically, the Austrian school doesn't have models for prediction. They contend that because the economy is made up of human actors there is no way to test and use models as accurate predictors.

    ReplyDelete
  5. jmb275, that's interesting. I will have to look more into the Austrian school of economics. I'm glad you pointed it out.

    ReplyDelete

To add a link to text:
<a href="URL">Text</a>