Deadweight Loss and Climate Change


Yesterday, Paul Krugman explained something that I bet could use a little more explaining for most people:

Now, a key point in all this is that the emissions tax or, equivalently, the rent on emissions permits, does not represent a net loss to society. It’s just a transfer from one set of people to another — from the emitters, and ultimately those who buy their products, to whoever collects the taxes or gets the permits, and ultimately whoever benefits from the revenue or rents thus generated. The only net loss is the Harberger triangle created by the reduction in emissions — which has to be set against the benefits of reduced pollution.

What’s a Harberger Triangle? Well, here’s a good illustration:


What’s that showing us? Well, when a market is in equilibrium you have some supply of goods being sold at a certain price. The reason people are buying the stuff in question is that the stuff is worth more to them than the money it costs. That’s the “consumer surplus.” And the reason people are selling the stuff in question is that the stuff is worth less to them than the money they can earn from selling it. That’s the “producer surplus.” If the government puts a new tax on the stuff, then the price goes up and the quantity sold goes down.

That reduces the consumer surplus and the producer surplus. But that lost surplus doesn’t just vanish, it’s basically being taken by the government and turned into tax revenue and public expenditures. But it’s not all taken by the government; some fraction of it—the triangle represented by D and E in the chart—really does just vanish. That’s the “deadweight loss” the value, to producers and consumers alike, of transactions that would have happened at the un-taxed price but didn’t happen at the taxed price. This, rather than the “cost” in taxes paid, represents the real social cost of a new tax.

But of course how much deadweight loss there is depends on the actual shape of the supply and demand curves. That chart has nice-looking straight lines. But in the real world it just depends. Different taxes have different degrees of deadweight loss. What’s more, the loss can be apportioned differently between producers and consumers. And a certain kind of tax could have a small deadweight loss at one level and a huge one at another level depending, again, on the shape of the supply and demand curves.

When you talk about a cap-and-trade system rather than a tax, you get all these same issues with consumer and producer surplus and deadweight loss. But instead of automatic revenue, you get rents associated with acquisition of carbon permits. And a key issue in policy design becomes how allocate those rents. One popular with industry proposal is to give them to industry for free. This basically compensates industry for the lost producer surplus, and in principle could actually be better for producers than the status quo ante was.

Alternatively, you could sell the permits and rebate the money to citizens—compensating consumers for the lost consumer surplus. There’s pretty good reason to believe that this would actually leave most people with more money in their pockets than they have under the status quo.

You could also auction the permits, have the government keep the money, and use the funds to reduce some other tax. If that tax has a higher deadweight loss associated with it than does the carbon cap, the overall economic cost of the carbon cap will be negative. And there’s actually pretty good reason to believe that permit auctions would be more economically efficient than many of the taxes we currently rely on, so as an abstract matter of policy design it would be relatively easy to design a pro-growth carbon cap regime. The dual problems are that the distributive consequences of going this route might be bad, and politically a tax swap is hard to pull off.

That said, as I was saying yesterday this idea is really laying out there in the street waiting for a political movement that (a) doesn’t mind redistributing wealth upwards, (b) likes to complain about the adverse economic consequences of taxes, and (c) would like to do something about climate change. We have a movement in the USA that fits (a) and (b) quite nicely, but their opposition to (c) is unfortunately so robust that they’ve somehow managed to forget everything they normally say about tax policy issues in order to concoct a theory to support the conclusion that a carbon cap would necessarily be economically ruinous.