Sean Casten’s  criticism of Cap-and-dividend[1] seems to indicate that he had a really bad day. Implying that former  former CEO Peter Barnes, and former software corporate executive Senator Maria Cantwell are Marxists is simply not a propitious way to begin a critique or proposal.  The substance does not seem any better.

One of his criticisms of the Cantwell-Collins bill: “The closer an incentive/penalty to the behavior, the more efficient the incentive/penalty. In the Cantwell-Collins bill, CO2 is not taxed at the point of CO2 release, but rather at the point of fossil fuel extraction/import. The theory says that these prices will all ripple perfectly through the system making this identical to a tax on CO2. By that logic, we ought to stop criminalizing gun violence and put a tax on bullets.” That analogy is self-evidently flawed: the difference between a tax on bullets and criminal penalties for violence is that one is upstream and the other downstream? Really?

But in addition, I’ve answered this point before WITH[2EMPIRICAL[3]evidence, specifically in response to Sean. Economic theory on  tax incidence  does NOT say that markets are perfect and distribute 100% of costs to consumers. What tax incidence theory says is that how a consumer tax is distributed between consumers and producers depends on elasticity, on the ability of consumers to push back against price increases. Whether the taxes are levied upstream or down in these cases makes no difference. If consumer demand drops in response to price increases then producers will drop their price and absorb some of the cost, regardless of whether the tax is on consumers or producers. If there is no price elasticity, and consumer demand won’t drop in response to price increases then such taxes will be passed 100% to consumers regardless of whether they are levied upstream or downstream. That is not only standard economic theory. It is backed up by empirical studies (which you will find citations for in the previous two links). As far as anyone can tell in this case people actually behave as economic theory suggests. I’m not blaming Sean for not remembering my rebuttal of his point. I don’t expect anybody to have a perfect memory. I know mine fails on occasion.

But I do expect Sean to either admit he is wrong or provide some good empirical evidence that he is right. Repeating discredited talking points after they are disproven is something unprincipled hacks do. Because Sean is not an unprincipled hack, I’m sure he will either prove his point or back down from it.

Let’s move to the alternative proposal suggested. The proposal takes the commonplace insight that price alone is not a sufficient driver for emissions reduction. The obvious conclusion from this is to support public investment and rule based (as opposed to price based) regulation as additional means of driving the move to clean energy. Instead Casten takes the flaws in pure price-driven proposals as reasons to suggest pollution fees should be directed to those who clean up pollution. Sounds good right? When the proposed means of doing this is to issue 100% free permits to pollute, the creation and distribution of those permits to be based upon various outputs, it stops sounding so sweet.

The proposal is described as a “carrot and stick” proposal. A formula based on your activities automatically allocates you free pollution permits. However this proposal does not have even a nominal cap. Instead of a fixed number of permits being issued, permits are created by exceeding an efficiency standard,  meaning that permits are issued in ratio to total economic output. Create less pollution than you are allowed under the standard, and permits are automatically generated you can sell. Increase production and maintain that same ratio and you have more permits to sell. If  you can’t maintain the right ratio, even if your overall pollution is small then you have to buy permits.Under these standards very large polluters could end up selling permits to people who pollute very little. Further the attempt to find an automatic way to generate permits and avoid conventional regulations ends developing completely infeasible efficiency standards.

Let’s start with manufacturing (excluding electricity). If we want to lower manufacturing pollution via efficiency, then ultimately we want to reduce emissions per dollar of manufacturing output. There are many problems with setting a direct standard on that basis in practice, which is why the proposal does not attempt it. But Sean’s proposal  sets  a standard for heat delivered to work process compared to carbon emissions. Deliver enough  BTUs to a your furnace or kiln or whatever per unit of pollution and you have permits to sell. Deliver too few and you need to buy permits. But  heat delivered to work process is not a reasonable proxy for value of what is produced. The number of BTUs delivered to produce a dollar’s worth of wood, a dollar’s worth of steel, a dollar’s worth of plastic and so forth bear absolutely no relationship to one another. And, as we would expect with that fundamental a problem, it simply fails to provide any incentive for many means of lowering industrial emissions. Reducing the use of paper, metal, plastic, glass and other raw materials via longer lifespan for consumer goods, increased recycling, decreased packaging, substituting on-line reading for paper printouts and so on would  not show up in this system. Amory Lovins, the Wuppertal institute and many a great many other experts in manufacturing have documented indirect savings through these and other indirect means that dwarf direct savings.

But even in direct savings, a BTU standard misses many types of saving. For example, Direct Reduction is more efficient than Basic Oxygen Furnaces for processing iron ore into steel. Some of the efficiency is due to fewer emissions per BTU delivered to work process. But another large part of those savings are delivered because Direct Reduction produces more steel per BTU than Basic Oxygen Furnaces. In short BTUs delivered to work product is not just a poor standard in the abstract. It misses most of the potential savings in industry in practice.

OK, but surely manufacturing is a unique case. We have decent proxies in transport (emissions per ton-mile and per passenger mile) and decent proxies in buildings (emissions per square feet or per person). The problem here is that single largest source of emissions in transport are light vehicles, and those are mostly driven by consumers. The largest source in buildings is residences. And even in commercial buildings most office space is leased rather than owned. You are not going to find many consumers who are anxious to cut their driving in order to generate permits. You are not going to find many consumers who are anxious to insulate their homes to sell permits. And quite frankly you are not going to find a lot of commercial tenants anxious to cut energy use in order to generate permits for sale. Now I know what supporters of this “carrot” idea will reply. Businesses will come along and offer to insulate homes or businesses for free to gain permits for to sell or to use. But just as most homeowners require paybacks of less than two years before they will invest in insulation, I suspect you will find that most business that offer free insulation will require similar paybacks to compensate for risks, and to ensure profits commensurate with this risk.

Also I know David Roberts likes to sneer at the idea that transaction costs in carbon trading are a significant downside. And maybe when transactions are large, there is an argument to be made. But real world experience with “white certificate” within the EU s
hows the standard dead weight from various transaction costs for efficiency improvement tradable certificates is 15%. Note this is not management or administrative costs of actually saving energy. That is the cost of financialization, of turning energy savings into financial instruments. When it comes to the transportation and building sector where transaction sizes are smaller, it seems unlikely that financialization transaction costs will be lower than this.

The only economic sector where this “carrot and stick” idea is not obviously infeasible is the electricity sector, because you have two factors coinciding. KWhs are a reasonable proxy for economic output of the electric generation industry (not perfect, but good enough). And almost all electricity is generated by commercial entities (even if sometimes non-profit) rather than consumers. So permit sellers and permit buyers are overwhelmingly businesses, and don’t have the major problem of split incentives that you find in buildings . So, when it comes to electricity, this proposal is no worse than any other form of cap-and-trade with 100% giveaways of permits. (Please note that this is not a compliment.) For example giving away permits, contra Coase, increases volatility vs. auctioning[4].  The main takeaway is that the “carrot-and-stick approach” is infeasible compared to normal cap-and-trade in every sector but electricity, and has all the worst problems of  cap-and-trade with 100% giveaways in the electricity sector.


[1]Sean Casten, Grist, Cap-and-Dividend is the worst possible way to regulate greenhouse gas emissions, https://grist.org/article/cap-dividend-the-worst-possible-way-to-regulate-ghg-emissions/

[2] Gar Lipow,  Grist, https://grist.org/article/Levy-carbon-taxes-upstream

[3] Gar Lipow, Grist, https://grist.org/article/Levy-taxes-or-permit-fees-upstream-II

[4] Gar Lipow, Grist, https://grist.org/article/cap-and-trade-permit-giveaway-hurt-waxman-markey-effectiveness/