Monday, December 28, 2009

Newton's Laws Do Not Ignore Friction

Usually, if this was made by some obscure individual on some obscure website, I wouldn't have given it a second thought. After all, you can find a lot of such garbage all over the 'net. But this is from a supposedly-respected financial publication, and made by someone from New York University. Therefore, it deserved to be addressed and corrected. It is also another example on why the general public often have faulty knowledge of physics, even basic physics that we teach kids in high school.

This article comes from, and talks about the future of the Dow Jones. Inevitably, people who don't know enough physics will want to make either a comparison, or an analogy, to some aspect of physics.

"Economics used to ignore liquidity risk, like Newton's laws ignore friction in physics," says Lasse Pedersen of New York University "However, now people are realizing similar frictions are central to what is going on in the economy."

This, of course, is totally wrong. Any physics student taking intro physics in college would have known this. Example: body on an inclined plane problem with friction involved. The whole point of drawing the free-body diagram in such a case is so that one can find the NET force acting on the body and applying F=ma (which is ONE of Newton's laws) to find the dynamics of the body. So Newton's Laws DO NOT ignore friction. The frictional force is like any other forces and there's nothing special about it.

This is another black eye for those in the economic/financial sector. It seems that, based on the past few posts in this blog, that I'm taking aim at this field. I'm not! It is the people who represent such a field and making stupid statements that are giving this field a bad name. If they just stick to what they know and keep their mouth shut on things that they don't, we would get along very well. But nooooo.... they somehow want to show off their ignorance of physics and hoping that those who read what they say don't know any better. And the sad thing is, they'll probably get away with these things too.



Peter said...

The article is behind a wall that I prefer not to go through, so I can only respond to your post. As far as I can tell, liquidity risk is not closely analogous to friction, but perhaps the physics side of the analogy could be put correctly as "Physicists often ignore frictional forces when applying Newton's laws"?

The important proviso is that in such cases an estimate should be made of when this simplification will lead to significant differences between the model and experiment. As I understand it, there nothing comparable to Newton's laws in economics, although it's arguable whether Newton's laws have any content without some sort of force law being given (such as Newton's inverse square law for gravitational force). That is, Newton's three laws of motion establish a set of conventions that allow force laws to be formulated. Economists frequently construct differential equations to model the variation of observed variables over time, but they appear not to be constrained, for example by anything comparable to symplectic structure, and there are no concise rules that adequately decide what would be the most worthwhile observables and frames of reference to use.

Having written the above, I note that constraining models to have symplectic structure and a conserved Hamiltonian is applicable only when forces are conservative --- that is, there are no non-conservative (frictional) forces. Some physicists would consider a model that includes frictional forces "not fundamental". On such a view, a model that is fundamental would account for the frictional forces using a conservative statistical mechanics.

Having teased this out as well as I can in a quick blog comment (though being as careful as I can to avoid the wrath of Zz), I wonder whether liquidity risk might be analogous to friction in that it is time-asymmetric? That would be a fundamental mathematical property of the role of liquidity risk in economics models that it would be good to know. It would be (barely) reasonable to make an analogy between liquidity risk and friction if both are time-asymmetric, although from the Wikipedia entry on liquidity risk it's not clear to me that there is any analogy with statistical mechanics, in which case the analogy is very weak.
Happy New Year, Zz.

bbbl67 said...

I think the economist was just being a little casual with his terminology, but he's right. In Newton's Laws all forces are considered in isolation, so gravity is isolated from friction and calculated separately. That's as opposed to Aristotle's Laws, which were the prevailing theory of physics prior to Newton, which mixed up frictional forces within other forces.

ZapperZ said...

But that makes even LESS sense. ALL forces, not just friction, can be "isolated". So if that's the reason, why not say that in a mass-spring system, Newton's laws do not consider, say, the restoring force, since that isn't part of gravity.

There is no excuse or justification for someone to make that kind of a silly statement such as in that article.


Michael F. Martin said...

This is a bit lieralist for my taste. One must ignore dissipative forces in the lagrangian before reproducing Newtonian equations of motion, no?

Actually, the parallels between neoclassical theory and netownian mechanics are rather profound. The economists' assumption of homotheticity, for example, is similar to an assumption of global symmetry.