Knowledge, or how to clobber an economy

Everything we do requires that we can predict the future, at least a little bit. If we try to take one step forward, we need the ground to be where we think it’ll be. If we’re wrong, we trip. If we’re really wrong, we die.

We don’t know the future. We’ve never seen it. We don’t even know the present. At best, we know the recent past, because it takes time for signals to get to us and more time for our brains to realize what’s going on – or more precisely, what was going on just a moment ago. If we think of our life as a journey, then it’s a trip we take facing backwards. We don’t know what the road looks like ahead, we can only try to predict it based on what it looks like behind.

We’re not very good at predicting the future. We’re best at predicting the future in the short term and on a small scale. Drive down the road with your eyes open and you can predict the future well enough not to crash – most of the time. But shut your eyes for a second, two seconds, three seconds, and you will start to panic, because it gets hard to predict the future even three seconds ahead. And even if you keep your eyes wide open, you have no idea what is happening in the next street over.

And yet we do work that will not bear final fruit until some remote time in the future, and often somewhere far away, to benefit someone we’ve never met and may not care about. We manage to do that even though we’re bad at predicting the future and know so little about the world around us.

The knowledge that we do not possess nevertheless must be there in some form. As Milton Friedman said, “Nobody knows how to make a pencil.” But the pencil is made, so in some way, the knowledge is there. Bits and pieces of the knowledge (e.g. how to chop down trees, how to mine graphite) are in people’s heads, but those bits and pieces don’t constitute the knowledge of how to make a pencil just by sitting around in different people’s heads. Those bits of knowledge are stitched together by the economic relationships between people.

But how did those relationships appear? They evolved from slightly more primitive relationships and so on back to a time when people were drawing with a piece of charcoal which they personally dug out of a dead fire.

The evolution of long chains of production like that which makes a pencil doesn’t require very much from the people involved. It does not require that we have much awareness of the world or of the future. All it requires is that we can learn the ropes of some job or other and that, on occasion, somebody does something new which constitutes an incremental improvement on what he was doing before. It can even be a chance discovery.

When a government does something that impacts the economy, it disrupts those economic relationships, and in doing so, disrupts the knowledge they contain. It induces partial amnesia, as hitting someone in the head with a baseball bat might do.

As mentioned earlier, we’re not very good at predicting the future beyond the short term and the small scale. How can legislators possibly know in advance what impact their legislation will have? They can’t. They can at best have a vague notion, and even that is probably wrong. Economic legislation is a heavy weapon swung wildly in the dark. It is born from ignorance, destroys knowledge, and turns back the clock on progress.

Share this

There is no clear separation between polity and economy.

What about when an everyday individual "does something that impacts the economy"? Say, stabbing someone...that can sure "disrupt those economic relationships" too. A heavy weapon swung wildly in the dark is also "a heavy weapon swung wildly in the dark", if you catch my drift. All legislation is "economic" in part, and some of it may even foster business performance by deterring violent or otherwise counterproductive action.

It seems possible that having some particular set of de jure rules could serve to foster roundabout production processes; if that's true, and if we're saying that we really like roundabout production processes, then the desirability of de jure rules cannot ruled out of bounds a priori.

I liked this piece better

I liked this piece better when Hayek wrote it.

No, seriously, I've never read Hayek. You argue uncertainty is bad. That's groovy, but there's no reason--at least not from what you've said--to limit the culprit to the government. The market can be plenty unpredictible too--that's why I've got all these John from Cincinnati mousepads in my closet ($5.99 per pad, see me after class). And, by the same token, legislators can be plenty predictible.

Empirical evidence

Empirically, we can see tha the market has evolved systems for producing pencils, cars, computers, and so on. Empirically, we can see that government regulation has caused more harm than good.

Theoretically, we at least have a mechanism by which the market can accumulate knowledge by small steps. It is learning by trial and error on a small scale, with preservation and spread of successful behavior, a similar mechanism to that used by biological evolution. But legislation is too vast in scope. Any trial and error that goes on there is extremely slow at best. Maybe in twenty thousand years there will be significant legislative evolution by the mechanism of trial and error. Of course, there are other factors preventing legislators from trending toward good law.

There's also market-based disruption, as I mentioned in my other comment (e.g. what people are calling "disruptive technologies"), but the disruption is accompanied by, and occurs because of, some improvement. People switch from ice boxes to refrigeration, people switch from horses to automobiles. The thing disrupted ought to be disrupted - the knowledge forgotten (e.g. how to make a horse-driven carriage) ought to be forgotten, because something better has come along. So yes, the market itself induces amnesia (forgetting of obsolete technologies and practices), but it's to make room for better knowledge.

It's not the unpredictability of legislators that's the problem. Kleptocratic and hard core socialist governments are extremely predictable too, and the results of their predictable behavior is the almost complete and continual destruction of their local economy, which never lets it get off the ground.


For the sake of readability I figured I would give briefer answers.

[Jason] the desirability of de jure rules cannot ruled out of bounds a priori

I was not ruling out de jure rules, I was talking about legislation. They're not the same thing.

[Scott] legislators can be plenty predictible

The problem I pointed out wasn't their own unpredictability, it was that the consequences of new legislation are hard to predict, and additionally that it damages the accumulated knowledge of the market.

[Scott] The market can be plenty unpredictible too--that's why I've got all these John from Cincinnati mousepads in my closet

And they should stay in your closet. Thank God for the market. This is part of the accumulation of knowledge. If you passed a law forcing people to take those mousepads off your hands (to preserve predictability), then that would conceal the knowledge that nobody cares anymore.

I think the problem is that your one-sentence summary of my point ("uncertainty is bad") is just not what I said. In the case of the mouse pads, legislated predictability would be bad, and would be bad for the reasons I outlined.

I, Pencil

As Milton Friedman said, “Nobody knows how to make a pencil.”

For the record, that Friedman video was based on an essay by Leonard Read.