Wednesday, August 11, 2010

Two Corollaries to Stein's Law

I see in the paper today that the US trade deficit increased.  A couple of days ago we learned that China's trade surplus "surged last month to its highest level in a year and a half".  Then, this morning, I read Michael Pettis's very thoughtful post on analogies between Chinese and Japanese growth, and the difficulty in rebalancing China's economy.  I was particularly struck by this observation:
But for a long time the problem of misallocated investment, which was whispered about in Japan but not taken too seriously, didn’t seem to matter. After all, as nearly everyone knew, Japan’s leaders were extremely smart, with a deep knowledge of the very special circumstances that made Japan different from other countries and not subject to “western” economic laws, with real control over the economy, with a strong grasp of history and penchant for long-term thinking, and most of all with a clear understanding of what was needed to fix Japan’s problems.

And look what a great job they had already done: by the early 1990s Japan had generated so much investment-driven growth that it had grown from 7% of global GDP in 1970 to 10% in 1980, and then surged to nearly 18% at its peak in the early 1990s. In about twenty years Japan’s share of global GDP was two-and-a-half times its initial share. That is an extraordinary growth story and one that can only be explained as a function of a new kind of economic thinking, right?

But less than twenty years later, after a terribly long struggle to adjust to high debt levels and massive overinvestment, Japan is about to be overtaken by China with only 8% of global GDP. Japan, in other words, has given back in less than two decades almost the entire GDP share it had taken in the two astonishing decades that preceded it (while during the same period the US has maintained its share). What’s worse, it is hard to pick up a newspaper today and read about Japanese policymakers without getting the idea that they are a totally dysfunctional, narrowly ambitious, and not especially savvy lot, much like their US and European peers. As Mortimer Snerd used to say, who woulda thunk it?
Stein's law, famously, says that "If something cannot go on forever, it will stop."  I would like to propose two corollaries that I've observed empirically in human affairs:
  • It will go on a lot longer than we think
  • It will end badly when it does stop.


Benno said...

I used to hold Corollary #1 then realized that it's actually very widely held. That may invalidate it.

A week ago an old friend of mine just lost his job at a cutting edge software company where he had worked for a decade. The office was closed and the work transferred elsewhere. It was a huge shock for him and he called me in the middle of the day deeply disturbed. But only last summer he had laid out for me in detail why it wouldn't last long. The signs were everywhere. I reminded him. His reply, "Yes, but these things always last longer than you think. Privately I really thought we could count on another 3 years."

Maybe things go longer than an "early warning" type person would think. But the end will surprise the vast majority.

Stuart Staniford said...


Yes, I agree. I think the underlying reason for these trends is that social diffusion of the acceptance of the end of the trend follows an S-shaped curve as the evidence gradually accretes (related to this post).

Eric Hacker said...

I agree with Benno that there is something not quite right with the timing corollary.

Imagine a house of cards built on a card table in the middle of the living room by your kids. You know it is not going to last; too many little bustling bodies about. Three days later you may be chagrined to see it still standing. Then while everyone is seated at the dinner table that evening, a truck rumbles by on the street and the whole thing collapses.

The problem is that the weaknesses and cascading effects are apparent, but the triggers are a black swan event. Therefore predicting the actual timing is difficult.

Does that mean it is always longer than one expects? Not necessarily, since the timing of the situation becoming apparent may affect the perceived duration of outcomes. That is, many times things may stop before we realized that Stein's law was applicable. We tend to not want to believe Stein's law is applicable to everything.

I would suggest that the timing of the end as well as the effects are chaotic. One might have expected the kids to be in tears, but perhaps they laugh out loud when house of cards collapses because they were getting tired of tiptoeing through the living room just to prove that Daddy was wrong.

Stuart Staniford said...


I guess what I really mean is that there will always be some people who realize the trend is doomed long before it's generally accepted. And I was kind of assuming that "we", and in particular, "I" will be amongst them, a perception imbued with a level of smug hubris for which I imagine the gods are already plotting my downfall...

Burk Braun said...

Very interesting.. but the thing about Japan is that they have preserved employment and living standards through this process. Their decline is only relative (to China), not absolute, and we in the US have a ways to go to attain the living standards they have in Japan (such as a working and inexpensive public health system).

adamatari said...

This is the crux of all crisis planning - what should you do if you know it's going to be a mess?

For example, it's clear that the economy of the US is not going to recover easily or soon, but at the same time it's still healthier than it was during the Great Depression. On a larger scale, energy issues will kneecap the entire world industrial economy at some point. The question is how badly and how soon, and this is essential to planning.

For example, if I think that industrial society will collapse quickly and comprehensively, I better learn to farm. However, if I think it will take 50 years, then learning to farm may be nice but I'd be better off focusing my education on something else.

If I think the economy will collapse and bring with it possible warfare, I'd better learn to shoot. If I think it's going to be a slow sag for a long time but with a still fairly functional economy, I can still shoot for using education or other means to get a higher standard of living.

With China, if it's a bubble now and pops soon, then it's in a much worse place than Japan - living standards in general are still much lower in China. They just have so many people that the top 1% or 5% is huge. I wonder if the real elephant in the room and cause of the missing Chinese consumption is not income inequality?

Greg said...

Playing Devil's advocate for a moment, from China's point of view (and looking at the big picture), all that is happening is a reversion to the normal state of affairs. China has about a fifth of world population and has traditionally had about a quarter of world GDP. There was a more-or-less permanent flow of silver into China as the rest of the world bought Chinese silk and other goods. The last couple of centuries have been an aberration.

Maybe Stein's Law applies that way round?

Robert said...

A possible example: The book Bankruptcy 1995 by Harry Figgie and Gerald Swanson

porsena said...

adamatari - I think the Chinese leadership is likely losing some sleep over income inequality. After what must have been a very difficult debate, the government has concluded that sharply increasing wages for the agricultural poor is not the right course. They've decided to accelerate urbanization of the that community instead.

High rise complexes are springing up in the middle of what used to be fields and the government will attract peasants to them by approving their relocation. What will these relocated farmers work at? Will the service sector accommodate them? Unemployment, particularly among the young, and its accompanying possibility for social unrest is the elephant in the room there.

Michael Cain said...

Re the timing thing... Keyne's version of your first corollary, which actually predates Stein's statement, is that "The market can remain irrational longer than you can remain solvent." It's amazing how long herd behavior can prop up stupid things...