A lot of people like emissions trading, but I think it's a bit oversold. Implementing an emissions trading scheme is phenomenally complex, and it doesn't really reduce the regulatory apparatus required. You still have to have people figuring out what the total cap should be on a harmful substance, you still have to have people who go out and enforce to make sure nobody is exceeding their emissions limits, etc. You open yourself up to all sorts of gaming the system--people trying to claim emissions credits for things that don't actually reduce emissions. It also, as the sibling comment mentions, doesn't work for pollutants like lead, sulfur, etc, whose effects are highly localized. It doesn't really do anyone any good if polluters in L.A. buy up all the sulfur credits that result from people in the middle of Montana reducing their emissions.
The EPA generally doesn't ban things. Leaded gasoline was an unusual case because it was so harmful and also, in the grand scheme of things, quite replaceable. Usually what they'll do is specify that polluters that emit particular kinds of air pollution (sulfur, etc) must use a certain level of control technology. This creates certain market incentives for polluters to make this control technology as cheap as possible. However, this approach doesn't really reduce pollution as much as it should, since the EPA is usually quite sensitive to cost issues. But as a practical matter, some polluters should go out of business if the cost of thoroughly cleaning up their emissions outweighs the revenue from the polluting activity.
Generally, some kind of economic incentive solution works much better than outright bans. It allows the market itself to decide what instances of pollution provide enough value to justify their harm and which ones don't. The problem is, however, that while such solutions work out great for global (or at least highly distributed) externalities, as you can simply set a rate of some given number of cents per ton of CO2 released and be done with it, they don't work quite so well for problems like these because the problem is semi-localized. One gram of lead released in a suburb is not the same as one gram of lead released in the middle of the city, and one gram of lead in paint is not the same as one gram of lead in gasoline. Sure, you can tax it too, but the tax would be so ridiculously high that the effect would be the same as an outright ban.
I have yet to understand how an emissions trading system is superior to a simple tax paid on all emissions.
Trading schemes add a lot of unnecessary complexity and potential for abuse, and enshrine the idea that if you have polluted more in the past, you deserve to keep polluting more in the future. If your business model requires you to cost the society more in pollution than the value of your product, you should not be in business, regardless of how fine history your company has.
Instead just quantify the total harm caused by emissions the best you can, divide by total emissions amounts, and levy that as a tax per pound of emission. In such a system, not a single coal-fired plant would be in operation in the US.
> I have yet to understand how an emissions trading system is superior to a simple tax paid on all emissions.
They're more similar than you seem to think. Both internalize cost to the production of pollution. In a tax, the cost is fixed and the total amount of pollution is market-determined. With a cap it is the reverse. However, at the point where the prices agree, they are, in principle, equivalent.
> Trading schemes add a lot of unnecessary complexity and potential for abuse, and enshrine the idea that if you have polluted more in the past, you deserve to keep polluting more in the future.
Allocating pollution licenses to past polluters is only one possible way of distributing them. A cap-and-trade system could also to hold yearly auctions for each year's permits. The former is used many times in part because incumbents are less likely to oppose regulation. They may even be able to achieve efficiencies and sell their allocated permits at a profit because of cap-and-trade. But note that even though the incumbents are receiving the permits for free, they still have an incentive to reduce pollution when the cost of the efficiency is less than the cost of the permit.
> Instead just quantify the total harm caused by emissions the best you can, divide by total emissions amounts, and levy that as a tax per pound of emission.
Two problems is that the harm may be difficult to quantify in a price and more importantly the harm may not be linear in the amount of pollution. You could argue that the optimal solution would be for the government to estimate the harm as a function of the amount of pollution and hold some sort of complicated auction where the price depended on the quantity of pollution permits sold. However, I think this is too impractical for any real-world regulation.
"Trading schemes add a lot of unnecessary complexity and potential for abuse..."
I believe that's exactly why they are "preferred" by our political system. Whenever the political body starts to consider laws about pollution, the lobbyists for the polluters go into high-alert to influence the outcome. They don't want a straight tax, because that'll become a simple cost which is difficult to avoid. Instead they prefer a complex scheme that can be sold as if it has the same impact as the straight tax, but in reality it will provide ways to game the system to either avoid the costs or even make a profit for the polluting companies.
Emissions trading regimes can be effective for some pollutants, but not others. I'd support them for CO2 because CO2 disperses rapidly and widely, but I oppose them for mercury emissions from coal fired power plants because mercury is much heavier and much more localized; in practice, an emissions trading regime for Hg would mean that children who grew up near power plants (i.e., poor children) would have much higher Hg exposure than richer children.
