Stardate 20030421.1759

(On Screen via long range sensors): Every field has its own lexicon of terms which are either totally opaque to those outside the field, or which may seem to make no sense. Such terms are needed to deal with the special problems of the field. But sometimes the underlying concepts need new terms because the ideas themselves are not in common circulation. Which is a shame because some of them are broadly important but not broadly understood.

We engineers refer to systems which exhibit "graceful failure", and tend to speak of such systems in positive terms. That may seem strange; why should any failure be considered a good thing? Wouldn't it be better to design it so it doesn't fail at all?

Well, no, actually, because it's impossible to design a system which doesn't fail. What we try to do is to design systems which are robust, and which will only fail in extreme conditions, and will then fail gracefully.

It's easiest to demonstrate that concept of graceful failure with an example. Compare two drinking glasses. One is made out of glass, the other out of thin aluminum. Strike each with a hammer with varying degrees of force.

Glass doesn't fail gracefully. Either it doesn't fail at all; you strike it and nothing happens, or it fails catastrophically (the opposite of failing gracefully) and the entire glass shatters.

Aluminum is more forgiving. When you hit it hard, you can put a dent into it or maybe make it bend, but it may still be able to be used to hold and consume liquids, even if more badly than before it was damaged.

In construction, brick houses tend to fail catastrophically and wood-frame buildings fail more gracefully, which is why the building code in California doesn't permit brick buildings unless they're specially designed. In an earthquake, brick buildings tend to collapse, but even when wood-frame buildings fail they don't tend to fail all the way. They start making noise which warns a person that something bad is happening, and may partially collapse while still retaining at least part of their original shape. A person in a wood-frame building has a much better chance of survival.

The ability of highways to carry traffic exhibits ungraceful failure. The highway is designed to have a certain capacity, and to some extent it can deal with demand beyond its capacity. But at a certain point, increasing the demand actually causes the capacity of the highway to begin to drop, and since that means that the excess demand inherently increases, it begins a process of destructive feedback which leads to traffic jams. In the extreme case it can actually cause the capacity of the highway to drop all the way to zero for short periods, which we refer to as stop-and-go traffic.

Mark Steyn writes about the latest gloom-and-doom forecast about consumption of natural resources and increases in pollution issued by the "we must use less" activists. The warnings sound extremely familiar, as Steyn demonstrates by quoting similar warnings from the past.

A lot of these warnings are based on a fundamental misunderstanding of how free markets respond to shortages. They seem to be based on a model sort of like the gas tank on a car, with a stairstep transition at the end. A car running out of gas doesn't fail gracefully; either it runs normally and consumes gas at a typical rate, or it fails entirely.

But the global economic systems isn't like that, and their projections never take that into account. The system inherently includes very graceful failure from these kinds of shortages if they pop up. It turns out to be a classic demonstration of Adam Smith's "invisible hand".

Virtually no significant natural resource is single-use. Wood is used for lumber and for paper and for chemical synthesis and for a lot of other things, and in turn paper and lumber are themselves used for many different things. Oil is used for fuel and for chemical synthesis, but as a fuel it's used for airlines and home heating and private vehicles and sea-going shipping.

Some of these are more valuable than others. If you get a shortage, then the price naturally rises, and that causes three effects. First, the system "sheds load". That's another engineering term but less opaque than many. Everyone pays the same price but not everyone actually values the product the same. As the price rises, those who value the product more will keep using it; those who don't will tend to stop.

The second effect is that a higher price will tend to encourage production. Most natural resources are available from a multitude of locations but the cost of extraction won't be the same in all cases. When the price rises, a given source which was uneconomical before may become economical, leading to an increase in supply.

The third effect is a hunt for alternatives. For many kinds of resources, there may be alternatives whose price is too high. But if the price of the primary resource rises, the alternatives may become economical and start getting used.

All of these effects mean that even if a given resource starts to run out, the system overall doesn't shatter and fail in an instant. The capitalist industrial system is extremely resilient, as long as the challenges change somewhat slowly. Where you see semi-catastrophic failures are when the challenge is big and sudden and unexpected, like the oil embargo of the 1970's.

But that was a political act, not a natural one. The oil existed, but those who controlled it refused to ship it. The natural process of running out of resources is not likely to cause that kind of failure. (And as a result of the 1970's oil embargo, the world oil market changed quite radically and the producers who used the "oil weapon" then recently refused to do so. Partly that was because their share of the market now was much lower and partly it was because they needed the sales more than their customers needed the oil. The system responded to the oil embargo by reducing its vulnerability to another one.)

This doesn't mean that we can blithely sail along and pretend there's no problem at all. What it does mean is that we're justified in being extremely skeptical when predictions of doom and gloom are used by anti-capitalist forces to push their agenda of "less is better".