Tuesday, January 27, 2015

Michael Pettis on (Some of) What’s Wrong With Economics

This is in the context of dueling predictions of future Chinese growth, where the bulls are bullish, Pettis says growth has to slow way down because of excess debt accumulation, and then the bulls claim victory because China hasn’t collapsed.  Pettis didn’t say it would but has a theory for why people influenced by economics would think the choice is between robust growth indefinitely and collapse.
....most orthodox economists find it very counterintuitive that the performance of an economy can be systematically affected by the structure of the balance sheet. They tend to think of economic activity as a linear process that tends towards equilibrium, and this path towards equilibrium is perturbed only by external shocks. Equilibrium is assumed to be determined by fundamental factors, and so this is why most economists focus primarily on the fundamentals — i.e. the asset side — and how these fundamentals are managed. External shocks are of course unpredictable, and so they simply incorporate them into their analyses as they occur.
If you think about the economy as a dynamic “system”, however, then it is easier to understand how institutions, including the balance sheet, can create systematic biases and how imbalances evolve and must ultimately reverse themselves. When you read Minsky, for example, he spends a lot of time insisting that the economy is a system, and that it has its own internal dynamic, which is one of the reasons why he said stability is unattainable. It is also why it takes a huge “event” that only Minsky can explain to bring him into the mainstream debate. In ten years he will no longer be part of the debate.
....The framework most orthodox economists have does not permit systematic biases and unsustainable growth. When a finance guy says that the growth in the debt burden is implicit and necessary to maintain growth, and also that it cannot continue, it doesn’t matter how many times I say that I do not expect China to collapse, I don’t think they are able to combine the two statements. They genuinely believe that there is no difference between saying that growth is unsustainable and that China will collapse in a few months. To them, either China is following a stable equilibrium, or it is on the verge of collapse — nothing else is possible in the orthodox framework.


Bruce Wilder said...

"a linear process that tends toward equilibrium"
If the economy is a dynamic system, driven by non-linear processes, then the linear model we have for it will work well until it doesn't. And, when it doesn't, we call that a collapse. Or, crisis of capitalism. Or, adjustment.
It is not necessarily the economy that collapses -- it is the model, though sometimes the collapse of the linear model can take the economy with it.
Saying that the performance of an economy can be affected by the structure of the balance sheet is saying that economy as a system and our models of the economy are not neatly separable, ontologically different things. That's the closely related illusion: that there's a concrete real economy and an illusionary, money economy, and the latter is an optional part of any analysis.
Seeing the economy as a linear system stabilized by the attractor of a (unique?) long-run (or very long-run) equilibrium is also part of seeing the economy as an ergodic system, in which history doesn't much matter and there's sufficient information to decide what has to be decided. It's a deck of cards continually re-shuffled, its history disappearing rapidly into the dust left behind even as the future emerges from the fog ahead. It can be in short-term disequilibrium, but it always working its way back to the path of a steadily moving long-run equilibrium.
Just the opposite might be true: the economy as a system might be no where near its (very) long-run equilibrium, but reasonably close to a short-run equilibrium, close enough that linear models are good enough to get on with the business at hand without demanding that anyone know more than anyone has yet learned.

rosserjb@jmu.edu said...

This is the first time I have read anything by Michael Pettis, who seems to be an MBA who is very good at promoting himself and bloviating widely about the Chinese economy. Anybody who puts "system" in quotes is making a fool of himself and seems to think that his audience are completely ignorant morons, which maybe is true of most who pay close attention to him.

His remarks on Minsky are particular silly and just plain wrong. Minsky most definitely never said that it takes some really big shock to collapse a system, although maybe I misread this. He says Minsky will be absorbed by the mainstream? Really?

I tend to agree with Pettis that it is not obvious that China faces a choice between permanent super growth and collapse, but I am also not aware of all these people he seems to be debating who supposedly make such claims. Does he name any names who argue this or just call unidentified people names?

Of course, there certainly are plenty of conventional economists using linear DSGE models who make silly arguments and failed predictions, but they do that about pretty much all economies, not being especially wrong about the Chinese one more so than others.

Peter Dorman said...

My guess, Barkley, is that you snorted at the same thing I did, which was the misuse of "linear". I put it aside (lots of people misuse it after all), but this is your bread and butter, so I'm not surprised it set off all sorts of alarms.

But take a closer look. He absolutely does not say that Minsky said it takes a big shock etc. He did say that it takes a big shock for most economists to appreciate Minsky.

More generally, Pettis, IMO, is no bloviator. I think dumping on him for not having a PhD is also ill-advised. If you go back and read the source article, you'll see he has good things to say, quite in line with Keynes and Minsky, about the sometimes devious forms that speculation takes in modern economies. He is also quite honest in acknowledging the limits to his knowledge.

If there's a second flaw with this quote, it's that it leaps quickly over many intermediate steps in his argument. In particular, simply invoking the concept of a "system" doesn't explain how you could have time-inconsistent behavior be the norm among the actors he examines. But the argument is fairly obvious when you think about it, and in fact is quite JEBO-ish (in the good old sense). So you two are really both on the same side.

I read the guy regularly, I think I have pretty good BS radar, and I don't usually have a problem with him.

rosserjb@jmu.edu said...

Maybe I was unfair, Peter, particularly in being arrogant. I grant that poking at somebody for lacking academic credentials is not a proper way to argue, although I do it with Robert J. Samuelson all the time, the faux economics reporter of the WaPo. I should be better behaved...

john c. halasz said...

This is Pettis' book that made his reputation:


He seems to be working from a Wynne Godley framework and is an astute analyst of global CA imbalances.