Tuesday, May 19, 2015

A plug for the Economics Anti-Textbook

It's only been a few weeks and yet I'm really bummed by economics - Mankiw's textbook especially, but the patent idiocy of the field more generally.

One problem I have is that economics seems to consider itself "scientific", and yet all I have seen in refereed papers is steady-state formulas.

Now, in electronics, a steady-state formula (like P=VI) is only a ciggypack first approximation. In fact, nothing in electronics happens instantaneously, because an infinite first derivative violates the laws of physics. E.g., you can't have an instantaneous change in torque, or else the machine will rip itself off its foundation and explode in all directions in a nuclear fission reaction. Everything happens over time.

So we were taught (and remember this is in community college, the place that University professors consider beneath them) that every true model needs to take time into account. Even when all you have is a DC power supply, an on-off switch and a resistor, you have to model that circuit with the understanding that VR does not go up instantaneously: the wires have an impedance, hey maybe even the switch has a capacitance of some sort, and thus you get an exponential curve from 0V to the on-voltage. It maybe only take a nanosecond, but in electronics we work with lots of stuff that operates in the nanosecond range (CPUs, microwave xmitters, etc).

It seems there is no such acceptance of the existence of time in economics. So, for example, when a demand curve shifts, it is modelled as a t=0 and t=1 problem with absolutely no time in between, so you get no explanation of how the equilibrium price shifts from E0 to E1.

You can not call yourself a science if all you can present are steady-state formulas with no prediction of what happens between t=0 and t=1. In fact, a theoretical model that takes time into account will actually make your theory at least marginally testable without a control group - just tell me what the data should look like from one day to the next, and as long as you have sufficient resources to follow the data and can sufficiently control for noise in your signal, you can prove the model just by showing that it successfully predicted the movement of a variable over time.

(Of course then you will see the problem, obvious to those of us who've studied control systems, that in fact no variable can be left out of your equation: in a feedback system, it's possible for any input variable to rule the output. But let's leave Laplace transforms and feedback math out of this, we don't want to make the poor economists' heads assplode.)

Even worse, the "comparative advantage and gains from trade" topic in every first-year text doesn't even seem to include the idea that the tradeoff of trade specialization involves not just foregone revenue in the market where you have comparative disadvantage, but also foregone growth and revenue into the future. It seems Krugman recognized this and used it to invent his idea of "dynamic comparative advantage", though I haven't seen the paper yet: but how pathetic is that when it takes a Bank of Sweden prize winner to insert the letter t into an equation? He's done this other times too: apparently stickiness at the ZLB is only a solvable problem by including present and future expectations. Did economists only discover time recently?

I am thus dumbfounded by the apparent utter ignorance of the entire economics profession. Plus I am enraged by how much first-year textbooks serve only as indoctrination texts for the Republican party. Remember, a lot of students take first-year econ: engineers, finance students, business students. They're learning stuff that's utterly childish and wrong. It's even worse: this propagandizing provides the support for the loony right-wing to pursue policies that actually destroy the economy.

Thus, I'm happy that I managed to get myself a copy of The Economics Anti-Textbook. The authors (a pair of econ profs from UNB) are also fed up with the pathetic state of intro econ textbooks, and so they wrote an entire book explaining how first-year econ subject matter is completely disproven by more advanced econ.

It's not commie pinko disinfo either: Robert Prasch and Joe Stiglitz (two people referred to in this book whose journal articles I've started reading) aren't marginal economists, they're serious people.

So thankfully I've found that, despite the clownish idiocy of Mankiw's Intro Macro propaganda pamphlet, there are some few economists who are actually trying to study economics seriously. So maybe I can continue on with this degree.

Though I wonder if it's possible to sue a university for teaching patently false subject matter. I mean, don't econ profs have a duty of care to ensure that they don't teach stuff that's disproven in upper years? I mean, it's as if a physics department's first-year class taught that the sun revolves around the earth. May I please get my first-year tuition back?

Apparently the authors of the Economics Anti-Textbook are discussing doing a new edition for 2015. I think what they should do instead is write a new intro econ textbook, following the formula of the existing books, but going beyond them by incorporating real theory and illustrating the actual disagreement and the faultiness of simplistic models.

Anyway, it's a great book, well recommended.

No comments:

Post a Comment