Suboptimal, certainly. But not really complicated. The way Washington really bleeds (my mother actually worked at DOE in renewables) is that huge quantities of money get sucked out invisibly through schemes that are too intricate for anyone to understand. It’s hard for USG to just give away money to lobbyists. It has to come up with something clever.

Pigouvian taxes are extremely ripe for all sorts of chicanery. And they are dwarfed by carbon credits, which are already a swamp of pure Third World sleaze. Note that just as Washington prefers taxes to a simple quota scheme, it prefers carbon credits to taxes. Negative selection wins again.

Of course, it relinquishes control over the (Pigou-equivalent) tax rate. Because there is no way to sample the demand curve except where it intersects, any policy must lose control over one of these numbers.

Aside from this, I think the main attraction of “oilpartheid” is that it has no obvious way to grow hair. Taxation schemes develop loopholes and turn into complexity furballs. Buying oil is just buying oil. It is relatively hard for Beltway bandits to horn in on.

And if there is no premium on domestic oil, obviously either (a) the quota is unnecessarily high, or (b) you’re done.

What you’re proposing is just about exactly equivalent to a tax. If you know how many barrels per year you want to country to consume, you can simply set the tax to the difference between the market price and the price your auction would provide. Of course, it would take a few years to adjust this right, but presumably your version would be introduced over a few years too.

One difference is that the Pigous ask that Uncle Sam reduce the income tax rate to keep his revenue constant. (You could do the same.)

A bigger one is about where this goes… in your scheme, the consumption of oil will be forever the size of the auction lot. But in a P scheme, if somebody manages to make solar/nuclear/geothermal/biofuel just a little cheaper than oil+tax, then the consumption of oil will fall off dramatically. Which is presumably good.

As for things like aerosols and black carbon, I sincerely hope that all parties to the debate would be moved by the law of unintended consequences. I suppose the law applies everywhere, but it seems to apply in spades to trying to engage in another polluting activity to fix global warming.

The most common problem with prediction markets, at least of the imaginary sort peddled by Professor Hanson and his cronies, is that they bet on events in the emperor’s-nose category. For example, imagine a prediction market for earthquakes. Since no one can predict earthquakes (at least for certain values of the word “predict”), the market will be pure noise and completely useless. There is no way for your market to breed a population of good earthquake predictors.
You said they pick poor topics and then gave an example of a topic they didn’t actually pick. And aren’t geologists or seismographers or some other specialization supposed to know something about the occurrence of earthquakes? This indicates it’s plenty imprecise and sketchy but better than the unknown nose example.

You might be interested in A Manipulator Can Aid Market Accuracy. Hal Finney made the same point about asymmetric information (though he referred instead to no-trade theorems) in the first link of this comment.

Like “climatology,” the entire field of economics is a quack science.
You have a lot of respect for Austrians despite that though.

Its practitioners should remember that it’s never too late to build a real career in a real field.
I hear the job-market in the private sector for economists is quite good. How do you determine what’s a real career vs a fake one?

In the productive sector, oddly enough, there is not so much emphasis on making all decisions, even important decisions, as the result of objective scientific processes.
The first prediction market was among the employees of the Xanadu project. The organization best associated with prediction markets now is Koch Industries, which is supposed to be the most successful private company in the country. Yahoo, Microsoft and Google are also significantly into prediction markets. They don’t use them as much as they could, but you yourself are fond of pointing out that modern corporate governance is not optimal. I don’t know of any governmental agency that actually implemented one yet.

You may be willing to argue racism with Will Wilkinson, but you’re too chicken to actually go toe-to-toe with any prediction-market proponents, even when you knew the author of Predictocracy was simultaneously arguing his case at two different blogs.

Regarding the government entering the oil business, I think it’s an awful idea simply on public choice grounds. Why trust them with anything like that? It sounds like carbon credits. A Pigou tax is simpler, though I don’t trust them to use the revenue for its intended purpose.

How will they control the inevitable black (no racial overtones intended) market?

Be happy to offer an argument, though I’m not sure whether or where I’ve seen it, and it may not be sufficiently “extended”:

The central point has to do with the cascade of uncertainties that continue to surround the issue:
- is it actually occurring?
- if so, are the models associating it with human-forcing correct?
- if so, do its costs outweigh its benefits?
- if so, do we know the costs and the likelihood of success of any attempted amelioration, and how they would compare with the net cost of doing nothing?
- and even if we know or think we know both those costs and probabilities now, do we know how they will be affected by another 50 to 100 years of scientific, technological, and economic advance (which is the sort of time scale involved here)?

Given these uncertainties, particularly the last, I’d say it’s clear that the best policy at this time is to do nothing apart from ongoing research on technological fixes and/or adaptation. I’d be cautious even about the size of those “huge” prizes.

Give it to Treasury – they conduct auctions every week. Or the Fed. It’s practically the same code they’re using for the TAF.

And storing oil is also a core competence of USG. Think of the SPR. It’s actually one of the few problems that Uncle Sam would be the honest-to-god low bidder on.

It’s not like I’m proposing that the Federal government should open a fast-food chain, or something that was actually difficult. I’m talking about running an oil futures market. If that’s beyond Uncle’s powers, it really is time to take to the hills with your gold, women and ammo.

Actually I thought the David Leonhardt article in the Times today, about the guy who kicks Intrade’s ass, was quite good. Prediction markets are not magic. They work, when they work, because a market over time goes through the natural Darwinian process of selecting for people like Mr. Ravitch, who have a clue, and against people like the ‘tards who bid Ron Paul stock up to 10%. Taking money from suckers is the oldest game in the book. I believe the article explained this, quite well as I recall.

The most common problem with prediction markets, at least of the imaginary sort peddled by Professor Hanson and his cronies, is that they bet on events in the emperor’s-nose category. For example, imagine a prediction market for earthquakes. Since no one can predict earthquakes (at least for certain values of the word “predict”), the market will be pure noise and completely useless. There is no way for your market to breed a population of good earthquake predictors.

More cracks appear when you try to go from prediction markets to decision markets. Decision markets are not useful for most problems. If they were, someone would use them. If you have to beg the Federal government to roll out something that should, according to all your theories, be useful to the private sector, the odds are pretty high that you’re a charlatan, a quack or a shill. Or even all three.

There are two main problems with decision markets.

One is asymmetrical information. By definition, you should not trade in markets where you think other players have asymmetrical information. The fact that DoD actually tried to roll out a prediction market for terrorist attacks, which as criminal conspiracies are pretty much the ultimate in asymmetrical information, is perhaps a good indication of the level of mineralization in the Orange Line brain. I’m starting to suspect that all that hard Potomac water causes some kind of heavy buildup in the ventricles.

Two is that a decision can be worth quite a bit to any group of people, and this may well be worth the cost of “burying the corpse” while manipulating the prediction market. Any market can be manipulated by anyone who is prepared to build up an arbitrarily large position.

So, for example, if Intrade numbers fully took the place of polls as indicators of momentum in the US presidential election, as they already to some extent have, I can easily imagine how the campaigns might participate in the market with ulterior motives. It’s all fun and games until someone is stopped or margined out.

But since the whole idea of a decision market is to make good decisions which cannot be influenced by sinister hidden conspiracies, the entire design is simply an engineering disaster. You might as well just auction the decision. It’s like something out of the President Camacho school of government.

In the productive sector, oddly enough, there is not so much emphasis on making all decisions, even important decisions, as the result of objective scientific processes. When we have a decision that has to be made, we tend to just give the responsibility to someone smart who has a clue, and let them make the call. Sometimes when the question is really important and totally up in the air, we convene a board or a committee, and they take a show of hands.

It’s backward, I know. It would never fly on the Hill, at Harvard or in Foggy Bottom. But it’s actually kind of fun. I believe there’s some kind of German name for the principle – I can’t think of it right now.

I do kind of like the dutch auction idea. (Arguendo, of course.)