Would it be possible to design a carbon tax that the public would enthusiastically support? That would be progressive, rather than regressive, imposing greater costs on the rich than the poor? Is it politically possible to strictly limit the total amount of carbon emitted, without rewarding past polluters with windfall emissions permits? Yes, it is. And it's fun! Politicians — Here's an opportunity to give money to your constituents, and save the world too! It works like this:

First, enact a carbon tax. Nothing fancy here, just your usual I'm-Greg-Mankiw-Wanna-Join-My-Club? "Pigouvian" carbon tax. Embedded in what drivers pay at the pump, added as yet another surcharge to heat and utility bills, would be a new Federal tax on carbon sold or used as fuel. That was easy.

Unfortunately, a carbon tax is regressive. It imposes disproportionate burden on the poor, as the higher cost of driving to work and heating a home takes a much bigger bite out of a burger-flipper's paycheck than a hedge-fund manager's "capital gain".

But, here's a trick. Just as flattish taxes are regressive, flattish subsidies are progressive. So, when we enact the carbon tax, we grant citizens the right to a refund of the tax on a fixed quantity of carbon consumed. We distribute those refunds equally among all taxpaying US citizens annually. And, we permit citizens to sell any refunds they won't need to use.

Suppose, in the beginning, we set the amount of refunds to be equal to the total expected carbon tax, given 2007 US carbon consumption. This seems dumb, right? In the aggregate, we've just created a system whereby the government collects a tax and sends it right back out again, exacting a net cost of zero from the private sector for its profligate use of carbon. All the government has done is caused transfers within the private sector. Yes. But from whom to whom? Light users of carbon end up receiving cash, from the excess permits they sell, while gas-guzzle-monsters pay up! That's likely to mean that most poorer people earn cash from their allotment, paid for by the people whose Hummers they can't see over. Moreover, note that our refunds are distributed only to taxpaying humans, not to businesses, but businesses are still subject to the tax, and can purchase refunds. That means that on net, the government will have underwritten a transfer from businesses to voters households. The vast majority of human beings will see ka-ching positive net wealth from this scheme, without any cost to the government. People who conserve more will earn more, people who conserve less will earn less, or even have to pay. Businesses will buy refunds from households, so long as the cost of the refund is less than the cost of the tax. When there are no more refunds left to buy — when aggregate carbon consumption exceeds the refund allotted — some users will have to pay the tax outright, at whatever rate the government has set.

Now of course a tax on business is indirectly a tax on households. But this is a tax businesses can minimize, by reducing their carbon footprint. That is, after all, the point, to change behavior. Plus, taxing indirectly via businesses, rather than taxing households directly, increases the progressiveness of a tax. Not all costs are passed on to consumers. Some costs take a bite out of profits, harming relatively well-off capital-owners disproportionately. (That's why we have things like corporate taxes.)

The political economy of this scheme is interesting. Since this is a tax that creates an income for most voters (earned, of course, via parsimonious use of carbon), voters might be expected to support increases in the level of the carbon tax, as this increases the value of their refunds. Increasing the tax level faster than the refund allotment makes most voters richer, and helps save the world. It also creates strong incentives for businesses to conserve carbon. As the tax level gradually grows very large, the scheme converges to a cap-and-trade, because it becomes prohibitively expensive for anyone to pay the unrefunded tax.

Would this scheme be hard to implement? Not terribly. Remember, we begin with a simple carbon tax, and we start small and build gradually, so that the refund infrastructure has time to evolve. For a while, lots of refunds would go unused. (They needn't expire quickly.) The government maintains a system of accounts, linked to taxpayer IDs, and encourages private-sector actors to implement trading systems. Carbon consumers claim refunds by submitting proof of taxes paid (bills and receipts), which the government reimburses with a deduction from the claimant's refund account. The process would be quite analogous to the value-added tax reimbursements of businesses apply for in many countries. Consumers who are too busy or disorganized to deal with the paperwork can just sell their refunds and pay the taxes (though they lose some by doing this). Initially, a small industry would spring up to ease the process of claiming refunds, in exchange for a cut. Eventually businesses would find competitive advantage in automating the process. Gas stations, for example, would have every incentive to electronically submit claims on behalf of customers, so that customers see a discount right at the pump. Fraud would be an issue, as it always is. There would be problems, scandals, and solutions.

There's been a lot of debate among the pious, which is more godly, carbon tax, or cap-and-trade? Here's a scheme that starts as a carbon tax and evolves into a cap-and-trade, that creates incentives for consumers, businesses, and politicians to reduce carbon use, that can be implemented gradually, that doesn't reward past polluters, and that leans just a little bit against inequality. What do you think?

Update: For a similar but much simpler idea, see Softening the Impact of Carbon Taxes. The paper proposes a fixed cash rebate of carbon taxes collected, rather than tradable refunds. Thanks to commenter (and author) Dan for pointing this out.

This idea was inspired by Martin Feldstein's Tradeable Gas Rights proposal, via Mark Thoma. I first suggested something like this as a comment to Mark's post.

Update History:
  • 26-May-2007, 08:35 p.m. EDT: Added update linking to Carbon Tax Center whitepaper.
Steve Randy Waldman — Friday May 25, 2007 at 10:28pm permalink
Dan (mail) (www):
I have no problem with the idea of a "flattish subsidy" you propose, but you don't need to add the expensive, complicated and fraud-prone trading market for refunds. As we propose at the Carbon Tax Center, a flat rebate accomplishes the same purpose of making the carbon tax revenue neutral and reducing its regressive effect. Or, as an alternative, we suggest using the carbon tax revenues to offset other regressive taxes. See our issue paper Softening the Impact of Carbon Taxes. As with your proposal, those who use less energy will come out ahead. In fact, they'll do even better than with your proposal since they won't have to pay all the market-makers you envision.
5.26.2007 1:30pm
Steve Randy Waldman (mail) (www):

First, thanks for stopping by and commenting.

I've just read your issue paper, which is quite nice. A few comments:

1) It's not so helpful, I think, to point out absolute dollar progressivity of a proposal. People who care about progressivity see that as a red-herring used by the tight-fisted wealthy as an excuse to pay less on a proportional basis. While the argument is true as far as it goes (and as true about income taxes as about carbon taxes), it persuades no one. People who think progressivity is important think the wealthy should pay at least in proportion to their income or wealth (and usually much more than that, in view of diminishing marginal utility). People who don't believe in progressivity aren't much interested in reassurances that the wealthy pay much more in absolute dollar terms.

2) There's a big difference between an offset to other regressive taxes that some economist somewhere calculates helps the poor and cold hard cash delivered into ones grubby hands. That is, I think your cash rebate proposal is much more compelling that a payroll tax offset (though I'd support a payroll tax decrease on its own terms). Voters will support policies that they see as the right thing to do, and clearly in their interest. Policies that seem like the right thing to do but demand sacrifice are, quite legitimately, a harder sell, and any linkage between a carbon tax and payroll tax changes would be abstract and reversible, while the carbon tax would take an immediate bite. A cash rebate, on the other hand, is a nice way to go. Promising voters that a carbon tax is not a tax increase, that all taxed funds will be returned and most taxpayers will come out well ahead (but that energy-conservers will be rewarded and gas guzzlers punished) is much better I think. It's a transparent, concrete sort of commitment. As you suggest, a flat rebate has the great advantage that it is very easy to implement and understand.

Regarding my own proposal, whether a cash rebate is better or worse than the tradable refunds is a judgment call. I think tradable rebates would be better, but that's because I'm enough of a market enthusiast to believe that the efficiency gains associated with having a market allocate costs can more than compensate for the various administrative and transaction costs associated with the scheme. But I'll concede that could turn out to be true or not, depending on how the scheme is implemented. Markets are human projects that sometimes succeed at achieving larger social goals, and sometimes collapse into destructive venues for fraud, skimming, and coercion.

That said, here are two reasons why I think tradable refunds are better than a flat rebate:

i) Tradable refunds wold permit the politically and economically sustainable level of the carbon tax to rise more quickly than under a flat refund, because intractable carbon burners would pay some of the tax at a discount. A simple carbon tax could never be raised so high that nearly no one would pay it, whereas tax-and-tradable-refund could gradually evolve into a cap-and-trade-style strict limit.

ii) Politically, I think offsetting regressive taxes is too tenuous a "payback" to ordinary voters, but a cash rebate might be too callous. Opponents would tar a cash rebate proposal as class warfare and bald subsidy, and voters can be persuaded to vote against their economic self-interest on principle. A cash rebate would seem like unearned wealth, and Americans outside of lower Manhattan are culturally averse to that. Enacting a tax but distributing waivers is not politically the same as cutting checks to people, and letting people sell waivers they don't need to use is also very natural. Also, the "market-ish"-ness of the tradable refunds might loosen some resistance on the part of the many Americans who do somewhat magically believe that markets generally do the right thing.

Again, the complexity of a tradable refund scheme cuts against its sellability, and a poorly implemented market would do more harm than good. I think a tradable refund scheme could be much better than a simple cash rebate, both substantively and politically. But these are judgment calls. (The political angle could be viewed as an empirical question — carbon tax proponents could push various proposals, and see which ones come out of the sausage factory prettyest.)

Anyway, thanks again for the comment.
5.26.2007 3:03pm
mk (mail):
Sorry, am I missing something? Why do we need a market for "refund permits?" Suppose we follow your proposal. If I need to pollute a whole lot, I would buy these permits if they were cheaper than the tax rate. I would not buy them if they were more expensive than the tax rate. In other words, the permit value is the tax rate. Trading in the market doesn't do anything because the price of a permit has to be equal to the tax rate. And given this, I have no reason to buy a permit rather than pay the tax.

In other words, in terms of effects on incentives, this is exactly the same as giving out money in a flat way to all taxpayers, except there are some additional transaction costs. In other words, flat cash rebate must be the better choice.
5.31.2007 2:54am
mk (mail):
I did muck this up a bit. The permit value is going to be less than or equal to the tax rate. For example, if you gave out a ton of permits, and only a few high-polluting people were interested in such permits, there would be a lot of permit-sellers and few buyers, so the price would go down. And: not everyone would get to sell their permits (some "refund permits" actually lead to zero cash in the end).

Now suppose the opposite situation-- you don't give out many refund-permits at all, and there are many people seeking permits. The price for a refund-permit cannot go higher than the tax rate because otherwise people would just pay the tax.

I don't see what's gained by this approach. People who don't pollute much and thus have extra refund-credits will either:

1) Sell at a price equal to the tax rate;
2) Sell at a price below the tax rate;
3) Fail to sell at all because there are too many sellers and not enough buyers.

People who pollute very much either:

1) Buy at a price equal to the tax rate (kind of useless);
2) Buy at a price below the tax rate (helpful for them)
3) Don't buy permits, just pay the tax.

Essentially, your proposal is either:

1) Distributively identical to a flat cash payout (modulo extra transaction costs),
2) Distributively more favorable to high polluters, because they can bypass the tax if the refund-permit market is favorable. Less favorable to low polluters, because they may have to sell at a price below the tax rate.

I am still not sure what this has going for it.
5.31.2007 3:42am
Cassandra (mail) (www):
I have been accused of being in "the glass half-empty" camp (unfairly in my view) and so am prone to sniggering, sometimes unwarrantedly at wisdom-snippets that could be confused with overused platitudes. One such however that is germane here, I still remember, tossed at me non-chalantly by the Big Boss at my first real job. "Let not the perfect be the enemy of the good" he said nearer midnight (while simultaneous obliterating my work with yet another round of red-pen mark-ups and illegible comments). I think that the Japanese and Europeans have demonstrated that simple tax of the appropriate magnitude (whatever its form in the minor detail) is an effective behaviour modifier, and that IF one is going to err, then one should err on the side of the marginally too high, rather than the marginally too low, for humans are ingenious at denial, rationalisation, and procrastination until the threshold of economic attention has been violated. Regressivity IS an issue, but I prefer to deal with that with tax and subsidy on the product-use end - i.e. high tax on H3s rebates on hybrids; high tax on less-efficient water heaters, subsidy on solar-water heating subsidy. Less elegant than yours, but more directly tangible to human sensibilities, not to mention gets the engineers and all R&D depts thinking in the right (I mean correct) direction sympathetic with public policy and LT national interest.

The most important thing is: Do it yesterday! There is nothing wrong with incremental policy improvement, but just make it so!
5.31.2007 10:08pm
Steve Randy Waldman (mail) (www):
mk — you are very right that the value of a refund (i'm beginning to think "waiver" would have been a better word) would vary between 0 and the price of the tax, and that this amounts to a hybrid between a progressive subsidy for the general public and a break for polluters. Both of those strike me as good things. A progressive subsidy makes the associated tax politically more palatable, and leans against inequality (which I consider to be a good thing in America today). Providing breaks for polluters is equivalent to lessening the magnitude of the tax, at an aggregate level. So, in one sense, you might (and do!) argue that it'd be better just to define a more modest tax level and provide cash refunds, since transaction costs are deadweight.

Here's a response:

1) As I suggested above, cash refunds are politically impossible in the US (crosses too many cultural redlines), and promising to offset other regressive taxes strikes me as a fool's bargain. Progressive taxation is continually gutted by the well-organized wealthy, and any reduction in payroll taxes linked to a carbon tax proposal will just get undone in when "entitlements crisis" becomes the problem-du-jour again in Washington. If we have a regressive tax, I want to compensate by putting something in citizens' grubby hands that politicians will find hard to take away. Tax waivers, refunds, whatever, are in an economic sense subsidies, but in the political sense, they are something else, something more legitimate, simply a right to be left alone up to a certain point. My claim is that Americans can (politically) accept a subsidy in the form of an exchangeable right not to be taxed, but they cannot (again politically) accept a cash subsidy.

2) Having two degrees of freedom makes possible more political incrementalism than a single degree of freedom allows. If we set a clear carbon tax too low, we will have a serial massive political struggles to raise it, and it over time, it really will have to go quite high. Increasing the tax just a little and diminishing citizens' allowances just a little may be equivalent to a large net tax increase.

3) The subsidy itself creates large a constituency for carbon tax increases, since greens can sell those increases as good for the world, and but not harmful to ordinary people (helpful, even, as the value of waivers rises towards the tax-defined ceiling). A cash subsidy would do this too, but again, I don't think a cash subsidy is on the table and even if it were, politicians would have a harder time, rather than easier time, increasing the taxes. Naked self-interest is a loser in politics. Self-interest must be worn like a sexy dress, as artful in what it conceals as in what it reveals.
6.1.2007 7:56pm
Steve Randy Waldman (mail) (www):
Cassandra — I don't disagree with anything you say. Although I am concerned about inequality in America, I'd support a simple carbon tax today, no problem. I'd support a cap-and-trade scheme, although less enthusiastically. As you say, do it yesterday!

I don't think in my initial post I did a very good job of emphasizing that this doesn't need to be complicated. If a carbon tax is the medicine, the waivers are intended as just a spoonful of sugar. By all means enact a carbon tax. If no one ever uses a waiver, great! We'll deal with snowballing inequality some other way, eventually. But maybe promising citizens that they'll get waivers, "to reduce the burden on middle-class America", might make a carbon tax more palatable to those with high sunk costs in SUVs. The government could send out old-fashioned coupons, and let people do what they do -- use them, sell them to more middleman who promise quick cash, manually collect fuel and utility receipts and apply for refunds, whatever. The idea is not to let the best be the enemy of the good, it is just to help something happen, both initially, and to create a dynamic where incremental improvements (that is increases in the effective tax) are politically likely.

My biggest concern with this proposal is deadweight costs of bureaucratic work (mail-in rebates are, after all a terrible waste of human energy) and excessive skimming by marketmakers. But, the proposed "security" e.g. an offset to 1-ton-C02 equivalent carbon tax) is so homogenous, its valuation so well constrained, I think that very liquid and efficient markets would in fact develop. Probably it would end up making economic sense for most consumers to simply sell, and for the most-efficient redeemers (industrial users) to deal with the bureaucracy. But that's great! It's just like giving people cash for accepting a carbon tax, but it's politically plausible. (And, as the tax grows large, it becomes equivalent to a cap-and-trade where, rather than auctioning permits to firms, permits are given to the general public to sell.)

The idea is to make a carbon tax sellable in America, not by appealing to the concerns of mass polluters, but by making sure the change is a benefit to the general public, especially in proportion to the degree households and consumers conserve.
6.1.2007 8:18pm
Note: You don't have to register to comment! Just use the pull-down menu to comment as "Guest".

Post as: [Register] [Log In]

Remember info?