Taxes al Carbon
| Steven Postrel |
Let’s suppose you’ve been swept up in the recent frenzy and decided that it actually makes sense to apply coercive regulations to reduce human carbon dioxide emissions. Let’s further suppose that you’ve caught up to the 21st century and know that imposing specific technology standards on particular sources of emissions is a sign of policy incompetence: You know that market-ish mechanisms can do a much better job than technology standards of allocating clean-up tasks to the lowest-cost producers; you know that market-ish mechanisms provide incentives for private innovation in emissions control while technology standards stifle better ideas.
Congratulations! You are now about where the public policy debate has fallen these days — naive about the quality of the natural science involved but possessing a sound insight about the smartest way to do a foolish thing.
Still, you have an important decision to make: Should your market-ish mechanism work by taxing CO2 emissions? Or should you instead issue tradeable permits that cap the total amount of human CO2 released? Oddly, there’s been little public discussion about this choice, and certainly very little that applies economic reasoning. Business groups are plumping for permits. Economists are mostly talking about taxes. There are three important criteria to consider in choosing the correct instrument:
1) Which policy is superior from a political economy/public choice point of view? It would be better not to create new incentives for costly rent-seeking behavior. It would be better not to exacerbate existing biases toward growth of the public sector.
2) Which policy is superior from an institutional/governance point of view? It would be better not to adopt systems that are hard to enforce. It would be better not to impose huge transactions costs on firms, consumers, and regulators.
3) Which policy is superior from an information/efficiency point of view? It would be better not to set up incentives that turn out to be wildly off the mark, giving us way too much or way too little emission control.
1) Permits vs. taxes is hard to evaluate on the political-economy criterion, but it leans in the direction of permits. Obviously, giving the politicians a new thing to tax isn’t good for reducing the growth rate of the public sector; new taxes increase the problem of collective action by raising the amount of wealth that can be fought over politically. Worse, a new tax on CO2 could easily drift above its optimal level as politicians see it as another way to raise money to bribe the electorate. Even if we could get other taxes, such as the payroll tax, iniially reduced in return, they would likely creep up again over time. Positive tax rates have no Thomas Schelling-style “bright line,” as you can always add a few tenths of a percentage point without raising a big fuss, whereas zero tax rates on a given activity are easier to defend.
Permits don’t create a continuing revenue source for politicians, although if auctioned off they would generate a one-time windfall. That is a strong point in their favor. On the other hand, there are all kinds of ways that firms can engage in lobbying and rent-seeking in how a permit system is set up. For example, if firms are issued permits proportionately to their current emissions, then current low-emitters get hurt and high-emitters get a windfall when they implement reductions (as they can then sell their surplus permits). Or if firms get permits that reward them for being efficient now, then they can immediately sell them off and make a profit. Probably an auction would be the best way to avoid such arbitrary redistribution (and raising of rivals’ costs), but the chances of that happening are almost nil, as industry would go nuts due to the uncertainty of auction prices and the wealth transfer from their shareholders to the government.
2) On the institutional/governance side, taxes have the advantage. First, they can be applied quite easily to millions of transportation users by taxing at the pump, whereas permits would get pretty complex if imposed at a retail level. Second, the infrastructure for taxation and auditing is already in place, so there would be little additional paperwork involved, whereas for many industries keeping track of permits and emissions would be complex. Third, the government would have strong incentives to enforce taxes because they like getting revenue, while enforcing permit restrictions would be much less motivating. This might be especially true in countries where the public sector is more corrupt or less competent. Fourth, there will often be uncertainty about the future cost of permits, which would make investment planning more difficult than with a tax, which would be more predictable.
3) Here is where I would have expected to hear much more from Mr. Pigou and his friends. The blackboard economics of regulating by price (taxes) or quantity (permits) was first laid out in 1974 in a famous paper by Marty Weitzmann.
The basic idea is that it’s usually very uncertain ahead of time how costly it’s going to be to reduce emissions. If you then have a situation where the marginal benefit of emission reductions falls very sharply at some threshold — e.g., you have a chemical where if it ever gets above 10 parts per million everybody dies, but if it’s less than that there’s no effect — then you want to use a permit system and pin down the quantity at 10 ppm minus your safety factor. The primary welfare issue is getting the quantity right, not accommodating the cost of emissions control. If you put in a tax, the uncertainty in marginal emission control costs means that you could easily overshoot or undershoot, by a lot, the well-defined optimal quantity of emission control. You just aren’t sure what tax to put on to give firms the incentive to hit the quantity target.
On the other hand, if you have a situation where the marginal benefit of emission control falls slowly over a wide range of emissions — e.g., where every part per million removed makes things a little bit better without any thresholds — then the optimal level of emissions control is largely determined by what marginal costs turn out to be. If it turns out that it’s not that expensive to take out emissions, you want to end up with lots of emissions control, and if it turns out to be very expensive, then you want to end up with only a little bit of emissions control. Under these conditions, a quantity ceiling is likely to be way off, either too lax or too stringent. An emissions tax, on the other hand (which is incentive-equivalent to offering the firm a positive price for every unit of emissions control), assures that marginal control costs will be equal to the tax.
Back in 1997, William Pizer, writing for Resources for the Future, analyzed this tradeoff for CO2 control. He looked at a hypothetical international control regime. His conclusion was that the marginal benefit curve for CO2 control was quite flat according to the climate models available at the time (and I have read nothing that suggests that that assessment has changed in the meantime). As a result, taxes were far more desirable than permits, yielding $337 billion in global welfare gains versus only $69 billion for permits. (A hybrid system, with taxes levied on each unit of emissions above the permit limit, could do even better.)
While I would not hang my hat on the absolute numbers generated here, the relative magnitudes are likely to be in the ballpark given the logic of the argument and the flatness of the marginal benefit schedule for CO2 emissions control.
So the final score is: Permits get a moderate edge on political economy/public choice issues; taxes have a big advantage on institutional/governance issues; and taxes deliver a big can of whipass on traditional economic efficiency concerns. So conditional on accepting the weak case for CO2 emissions control, the Pigou people have a strong case against the cap-and-trade brigade. Maybe they should start making it.