We've asked our friend, UMD student, and occasional Washington Post editorialist Max Epstein to contribute his thoughts on carbon pricing to the blog. Our response, by Breakthrough Generation Fellow Zach Arnold, is here.

In the wake of the failed Lieberman-Warner Climate Security Act, there has been a widespread reevaluation of whether Cap & Trade is the most effective strategy to avert catastrophic climate change. At first many promoted a carbon tax instead, but recently there has been a call to reconsider the central focus on pricing carbon itself. Following Lieberman-Warner's abrupt death in the Senate, Michael Shellenberger wrote that the new way forward should focus on making renewable energy cheap, not polluting sources expensive. In "Scrap Kyoto," Shellenberger and Nordhaus call for a massive public investment in clean technology research and deployment. Joseph Romm in Nature calls for massive subsidized deployment of existing renewable technology, relegating R&D to the "longer-term effort aimed at a new generation of technologies for the emissions reduction effort after 2040." However, such efforts would be insufficient without a price on carbon as well.

The main reason is that there is just too much market share to go around. According to the EIA International Energy Outlook 2007 reference case, total world energy consumption (including fuel for transportation) is projected to be 254.9 quadrillion BTU higher in 2030 than in 2004 (the latest date actual data is available for). To put that number in perspective, total world consumption of coal, natural gas, and nuclear power combined in 2004 was 245.4 quadrillion BTU. And even if we supplied all new energy demand by 2030 with clean energy -- a very ambitious goal considering the numbers above -- we would still not reduce emissions at all without curtailing fossil fuel use.

Substantial public investment in energy research is crucial, but waiting to begin a serious program of reductions until revolutionary technologies emerge is unwise. If we capped U.S. emissions from the transportation, industry, and electric power sectors at 2005 levels in 2012, a pace of 4.24% annual reductions would achieve an 80% cut (in those capped sectors, representing ~85% of emissions) by 2050. However, if real reductions didn't begin until 2020 -- even assuming you contained and reduced emissions growth in the meantime to 2005 levels -- the rate of reductions needed would jump by over 1.1% every year, to 5.37% over the next 30 years to hit the 2050 target. Breakthrough acknowledges the importance of achieving immediate reductions, but proposes deployment of current technology, coupled with aggressive R&D for the clean technologies of the future, rather than instituting a Cap & Trade program for CO2.

Long-term strategy, even mid-term, starts with major R&D investment today. However, besides R&D, we also need a policy that achieves two aims: 1) generate tens of billions of dollars a year to fund the sustained R&D effort, and 2) deploy available solutions as cost effectively as possible. The funding issue is not trivial; we need to discard the idea that if the government simply decides not to pay for something it's effectively free. The tens of billions needed could be funded through carbon tax or Cap & Trade revenue, or through general taxation (and again, deficit spending is just paying for it-- plus interest-- through future taxation). Here, carbon revenue funding is clearly preferable. General revenues come from taxes on labor, employment and investment, which dampen economic growth and productivity, increasing the cost of transforming our economy. Carbon pricing, on the other hand, actually promotes economic efficiency and so lowers overall costs, if you acknowledge that emissions reductions must be achieved.

Carbon pricing is also essential for the second short term aim - cost effective deployment (as well as medium and long term). First, it's important to emphasize that cost effectiveness is as important from an environmental standpoint as an economic one. If you have $100 for emissions reductions, you get twice as much CO2 reduced deploying $5 solutions than $10 ones. This is not just academic, the point holds whether the budget is $100 or $30 billion.

Cost effective and efficient deployment requires that policy encourage reductions in proportion to the amount of carbon the project reduces. A cap on carbon would be far more effective at this than any public deployment program, because the market is better at allocating resources than any central planner, even without taking politics into account. To fully satisfy the proportionality condition, public deployment contracts would have to take into account lifecycle emissions of each individual project (not just broad categories), the carbon intensity of the electric power sector in the region where new power generation would go up (and so displace existing generation), and countless other conditions. Clearly it's impossible for the government to do such an extensive audit to maximize emissions reductions with the funds available. With a price on carbon, however, these emissions get factored into costs and prices and are efficiently deployed by investors looking at the bottom line (subject to some market failures discussed later, which don't change the conclusion).

So what exactly is wrong with Cap & Trade? Most detractors first point to the European program to implement Kyoto targets, which has a $40 allowance price for carbon without achieving actual emissions reductions. Additionally, Shellenberger & Nordhaus point to the allegedly intractable "Gordian Knot" of carbon pricing: If the price on carbon is set too high, it will incur politically unacceptable (not to mention regressive) rises in energy prices. Conversely, if the price is set too low, it will fail to achieve meaningful reductions at all because most clean technologies will not be competitive.

Kyoto Cap and Trade: The EU Emission Trading Scheme (ETS)

The ETS certainly seems to be the embodiment of the Gordian Knot problem. With a $40 carbon allowance price and no meaningful progress on actual emissions, it appears to indicate that carbon prices would have to be astronomically high to achieve the necessary reductions. However, the ETS suffers from a number of problems that compromise the level of reductions and exacerbate costs. Over 90% of allowances in the ETS are given away for free to polluters. Reserves for specific sectors and fuel sources within the electric power sector preserve those polluting sources, while each plant retirement increases the windfall to the rest (and so keeps them around). There are even fuel-specific reserves for new power plants, which directly incentivizes the construction of new coal and other polluting sources. And since each member country makes allocation decisions according to its own National Allocation Plan (NAP), a new coal plant can cherry pick to set up where it is offered the most allocated credits (which are essentially cash windfall due to a liquid secondary market for them), and then still sell to everyone on the regional market.

These perverse incentives for continued reliance on certain polluting sources drive up the demand for pollution credits and so raise the price. The phase by phase implementation (Phase I 2005-2007, Phase II 2008-2012, Phase III 2012-2020) also drives up costs and promotes inefficiency due to regulatory uncertainty about the policies of even the near-term future. Meanwhile, emissions have continued to rise because of the use of CDM offsets with questionable verification that for the most part "do not represent real emissions reductions." These defects are not inherent to cap/trade generally and, for the most part are much improved on in even the very imperfect cap/trade bills being introduced in Congress, including Boxer-Lieberman-Warner.

The Gordian Knot

The Gordian Knot proposition that any price on carbon high enough to induce significant emissions reductions will lead to excessive price increases ignores the fact that the price increases incurred are minor relative to the revenue created. Let's run the numbers. Lieberman-Warner would have created 5,775 million allowances in 2012. An initial price of $20 per allowance would yield $115.5 billion (if all were auctioned). If you gave $200 to the bottom 80% of the country economically, it would only cost $48 billion. That's $200 per person, so $800 for a family of four. (I'm not literally suggesting this exact means of distribution, the point is just to demonstrate there's more than enough money available to offset costs.) That leaves $67.5 billion, which, if not siphoned off to buy votes from assorted regional interests, leaves plenty for advanced energy research. Compare this to the current energy R&D budget of under $2.9 billion for 2007, according to the EIA.

At $25 an allowance, total revenue is over $144 billion, which leaves over $96 billion after the $200 per person. Meanwhile, even at permit prices of $30-$80, the EIA projects average energy expenditures (though not including transportation) per household would only rise $30-$325. So any permit price in the foreseeable future would yield more than enough revenue to more than compensate the majority of consumers for the price increases incurred, with plenty left over to fund major R&D increases to ensure the pace of future reductions. The problem with Cap & Trade then (and equally for a carbon tax) is not one of incurring excessive costs, but merely of ensuring that the revenue is spent sensibly.

Letting Markets Work: Why Government Shouldn't Pick Winners

I once heard more money changes hands on global capital markets in a day than all the world's governments spend in a year. This illustrates, if not the inherent futility, at least the foolishness of trying to take on such a major and multifaceted investment exclusively with public sector financing. While public funds for deployment may spur private capital to some extent, it wouldn't do so nearly as much as further investments in research or infrastructure. Furthermore, the most effective way to motivate private capital to invest in both clean R&D and deployment is to set a price on carbon. A price on carbon literally makes pollution abatement a marketable asset. The 1990 SO2 Cap & Trade regime provides a valuable empirical example on the power of capitalist innovation in achieving emissions reduction goals. Ultimately, it shows that in achieving emissions reductions, government should set the target, but allow the market to find the means of getting there.

The 1990 Acid Rain program achieved reductions in sulfur dioxide at far lower cost than even optimistic predictions. How? In large part due to two factors. First, before the program coal was classified as 'high sulfur' or 'low sulfur;' afterwards the sulfur content was rated with much more detail, as it got factored into the price of the coal itself (inverse relationship: low sulfur became more valuable). This allowed plenty of firms to make simple, relatively inexpensive reductions just by switching to a lower sulfur coal, and only to the extent that they needed (or extra if they felt they could profit by saving allowances and selling them).

Flue gas scrubbers, the traditional means of sulfur dioxide mitigation, saw significant efficiency gains as well. First, increased competition drove down prices because they now had to compete with other emissions reductions methods, like low-sulfur coal, to stay in business. Second, scrubbers were able to be manufactured to remove fewer emissions, but at far lower cost. This led to more efficiency in reductions per dollar spent. Such a change developed specifically because every marginal reduction had a monetary value, i.e. because there was a price on pollution.

The SO2 program holds real lessons for designing a carbon reduction policy. The solutions of tomorrow will not be simple linear improvements on the technologies of today. No one predicted that the acid rain problem would be solved by a new classification system for coal, and scrubbers that actually did less scrubbing. Then as now, we had proven emissions reductions technology available that the government could have mandated or deployed. Luckily, a market based solution was chosen instead.

Picking winners, whether it's Renewable Portfolio Standards (RPS), or targeted tax credits by technology, is a bad idea. It seems now like "investing in tomorrow's solutions," but its just as likely to actually bias the market against taking tomorrow's best solutions, which we likely haven't thought of yet. The competition point from scrubbers is especially relevant. RPS standards insulate listed renewable technologies from competition with cogeneration/CHP, Demand-Side Management (DSM), and again, the things we haven't thought of yet. Shellenberger and Nordhaus write in their criticism of Cap & Trade that we didn't get the PC revolution by regulating away typewriters. Aside from the obvious distinction that typewriters did not pose serious externalities to society, its important to note that we didn't achieve the PC revolution by subsidizing mass deployment of the IBM 5150 PC either. To bring the point closer to home, you don't get a breakthrough like Concentrated Solar Power (CSP) by mass deployment of more PV panels.

The government should stick to targets and let the market figure out how to achieve them. Government action should be targeted to address specific market failures. Carbon pricing is the obvious and necessary first step due to its externality cost for society. Research and development is another role for government based on a similar dynamic - since research inevitably yields benefits that accrue to other firms beyond what the researcher can capture for profit, it provides an externality benefit for society. Thus, research would be undersupplied if left to individual profit-seeking firms. Ditto for large scale demonstration projects, which feel out technical, regulatory and supply stream issues, the resolutions of which benefit all firms that come afterwards.

Deployment generally does not fit in this category, though occasionally it will. Subsidizing more efficient appliances may be justified because consumers choose many appliances without regard to energy consumption, and so are not purchasing "rationally" in the economic sense. It's unreasonable to expect every shopper to crunch potential kwh savings for a certain efficient computer or dishwasher against additional cost, correctly forecasting future retail electricity prices, and all the while carefully accounting for the varying time value of money. So where the government has compelling evidence that economic efficiency gains could be realized by making more efficient products more cost effective, or even mandating standards, it is justified in taking such steps for deployment. Another example might be small scale distributed power generation, which can face discriminatory access to local transmission, impeding its ability to market its product.

We should not confuse the roles of public and private capital investment in efficiently steering our economy towards a more sustainable path. Not every expenditure that we hope to see private companies make is a good choice for government action. With limited exceptions, deployment of clean technology is a perfect example of this distinction. Internalizing the externality cost of carbon will lead the market to deploy carbon-reduction technologies more efficiently than the government would.

The Immediate Adverse Consequences of the Investment-Only Approach

It's no secret that just the fear of carbon prices in the near future has stalled plans for construction of new coal plants. Clean generation is not picking up the slack either, because in general the investment doesn't offer sufficient returns at current electricity prices. The result is major projected gaps between electric generation and demand in many electricity markets across the country. If the US were to initiate the investment centered approach of "making cheap energy cheap instead of dirty energy expensive," the immediate consequence would be to usher forth many new coal plants. In the absence of a price on carbon (or risk of one in the future), they are often a profitable investment for needed supply. And once these coal plants go online they will be very hard to get rid of, even with major breakthroughs in clean technology.

A power plant doesn't get built unless it's investors predict they can sell the power at a high enough price to cover: 1) their capital "sunk" costs, 2) the operating costs of the plant, and 3) a good return on their investment (profit). However, once a plant is built, it will sell power at the highest price it can get, though be willing to sell as low as just its operating costs if need be (for example, if new supply hits the market and bids down the price). If a coal plant was built for $1000 and has operating ("variable") costs of $0.03/KWh, it would like to sell above $0.03/KWh to pay off its capital investment and turn a profit. It's "levelized cost" of electricity might be around $0.08/KWh. However, if new supply hits the market, that coal plant will continue to operate so long as it can profitably sell above that $0.03 operating costs. By selling below its "levelized cost," it falls behind on paying off its capital costs, but those capital costs are "sunk" - those costs (unlike operating costs such as labor or fuel) will not be mitigated by closing down the plant.

Since most of the cost of a coal plant is in its capital expenditures, it is very hard to push it out of the market because its operating costs are so low. This is why even where a natural gas plant has a lower levelized cost than a coal plant nearby, the coal plant is the baseload source that runs all the time and the natural gas plant is the peaking plant. Coal's operating expenditures are cheaper than gas's, because the fuel is so much cheaper, and dispatch happens on the basis of operating costs, not "total" cost.

This has major implications for any policy that might usher in a new wave of coal plant construction in the short term. Any coal plant built today will continue to burn coal until there is enough power to serve the entire market selling power below just the operating costs of the coal. This may not seem like such a big deal - after all renewables have the lowest operating costs of all, their fuel is free. However, new clean generation plants won't be built unless they anticipate selling power at a high enough price to recoup all capital costs and earn a good return on investment. This means even if renewables become cheaper than coal (on a levelized cost basis), their deployment would still have to be subsidized to build enough to actually displace the coal plants, if no price on carbon is adopted. Otherwise the market would stop building them when existing clean power bid down the price of power to around operating costs - where existing plants continue to operate (including coal) but new ones will not be built to displace existing ones because of lack of profit opportunity.

This is on top of the major market share issues introduced in the beginning, because it is certain there would be more than enough market share for all the clean power we could build and still keep on building coal plants (not to mention keeping existing ones) as well if carbon pollution was left unpriced.

The reorganization necessary to fundamentally revolutionize our economy would bankrupt us if not pursued efficiently. And to do that, we must cap carbon and let the market work to find lowest cost, scalable solutions. If we don't price the externality cost of carbon, we won't need breakthroughs, we'll need miracles.