By Roger Caiazza
According to Resources for the Future, carbon pricing is a climate policy approach that works by charging industrial sources for the tons of emissions of carbon dioxide (CO2) they emit. I have been following this for quite a while and think that despite the attractiveness of the theory there are practical reasons why it will not work as advertised.
I am prompted to prepare this summary of my concerns because a coalition has asked the Federal Energy Regulatory Commission to hold a technical conference or workshop on carbon pricing. In my opinion this coalition consists of vested interests that do not consider the effect on ratepayers.
I first became involved with pollution trading programs nearly 30 years ago and have been involved in the Regional Greenhouse Gas Initiative (RGGI) carbon pricing program since it was being developed in 2003. During that time, I analyzed effects of these programs on operations and was responsible for compliance planning and reporting. I write about the issues related to the energy and environmental interface from the viewpoint of staff people who have to deal with implementing these programs. I have followed the New York State Independent System Operator’s carbon pricing initiative since its inception and my work on that program is the primary basis for this summary.
Proponents have convinced themselves that somehow this is different than a tax but, in my experience working with affected sources, it is treated just like a tax. As a result, the over-riding problem with carbon pricing is that it is a regressive tax. In the following, I describe a number of other practical reasons that cap-and-invest carbon pricing or any variation thereof will not work as theorized: leakage, revenues over time, theory vs. reality, market signal inefficiency, control options, total costs of alternatives, and implementation logistics. In addition, The Regulatory Analysis Project (RAP) recently completed a study for Vermont, Economic Benefits and Energy Savings through Low-Cost Carbon Management, that raises additional relevant concerns about carbon pricing implementation.
Pollution leakage refers to the situation where a pollution reduction policy simply moves the pollution around geographically rather than actually reducing it. Ideally you want the carbon price to apply to all sectors across the globe so that cannot occur. I don’t think a global carbon pricing scheme is ever going to happen because of the tradeoff between the benefits which are all long term versus the costs which are mostly short term. I don’t see how anyone could ever come up with a pricing scheme that equitably addresses the gulf between the energy abundant “haves” and those who don’t have access to reliable energy such that “have nots” will be willing to pay more (as carbon taxes) just to catch up with those who have abundant energy.
For any carbon pricing scheme in a limited geographical area I think that leakage will be an insurmountable problem. I am primarily concerned about energy policy in New York and have written about New York’s electric sector carbon pricing initiative. The New York Independent System Operator (NYISO) is currently campaigning for its Carbon Pricing Initiative. Trying to force fit this global theory into just the New York electricity market is an extraordinarily difficult problem. As proposed, it will likely result in power leakage where energy and emissions are not reduced but simply shift emissions associated with power production out of the state within the inter-connected electric grid. Additionally, note that a carbon price on just the electric sector may even result in leakage if more consumers generate their own power using unpriced fossil fuel.
Revenues Over Time
A fundamental problem with all carbon pricing schemes is that funds decrease over time as carbon emissions decrease unless the carbon price is adjusted significantly upwards over time. Air pollution control costs increase exponentially as efficiency increases so it is clear that the need for stable revenues over time is acute. It has been observed that roughly 80% of the effects come from 20% of the causes and everyone knows the implications of the low hanging fruit analogy. This phenomenon has been observed with regard to New York’s observed CO2 emission reductions to date. Supporters of the RGGI point out that since its inception that New York electric sector emissions have dropped over 40% between 2006 and 2018. However, I have shown that those reductions were primarily because of retirements and fuel switching to lower emitting fuels. It can be argued that those reductions would have happened anyway because retirements and fuel switching were lower cost options without even considering CO2 emissions. New York State is now in the more difficult emissions reduction position where all future reductions will have to decrease the use of low-cost generating facilities with something that is more expensive.
This difficulty should be even more of a concern with CO2 emission reductions because at some point replacing existing fossil-fired generation not only has to consider the direct power output conversion costs but must also address dispatchability and grid support costs. When those costs are included there will be a sharp increase in total costs per CO2 reduced. Like many others, the NYISO Carbon Pricing Initiative proposes to use the social cost of carbon (SCC) as the carbon price. The SCC cost increases over time but, in my opinion, the costs over time do not increase enough to keep pace with the necessarily more expensive total costs to maintain reliable electricity to consumers.
Theory vs. Reality
Another problem with carbon pricing theory is that in practice affected sources may not act rationally or as theory expects. RGGI is a market-based carbon pricing program and I have written extensively on it. The academic theory for RGGI market behavior is that affected sources will treat allowances as a storable commodity and act in their own best interest on that basis. If that were true, then affected sources would be purchasing allowances for long-term needs and “playing” the market to maximize earnings. In practice, RGGI affected sources plan and operate on short time frames and have shown no signs of making allowance compliance obligations a profit center.
Carbon pricing theory claims that when the cost of using higher emitting energy increases that will provide incentives to develop alternatives and discourage continued use of existing resources. However, these incentives are indirect and again assume rational behavior in the market. While theory says that a company that currently operates a fossil-fired plant will change its business plan and develop a renewable energy facility to stay in business, there are a whole host of reasons why the company may not go that route and instead treat the carbon price as a tax, continue to operate with that constraint, and give up on a fossil-fired plant as a long term asset when they can no longer make a profit. In my opinion RGGI did not induce any New York companies to change their business plans.
A fundamental difference between any carbon cap control program and cap programs for other emissions at power plants is that there are no cost-effective add-on controls for CO2 whereas there are control technology options for SO2, NOx and most other pollutants. As a result, the affected sources have fewer options to comply with a CO2 price or cap. Ultimately, the affected source control strategy is to operate under the cap; if the cap is lower that means selling less fuel. In addition, because there are so few CO2 control options for the affected sources, this increases the likelihood that they will simply treat the costs of purchasing allowances as a tax.
A carbon price initiative for the transportation sector would try to reduce fuel use. Clearly moving to an electric vehicle is the preferred option but there is not only a large cost hurdle but a host of practicality issues as well. Paul Homewood at the Not a Lot of People Know That blog described the flaws of an article supporting a carbon tax plan that addresses this issue. He said that “The only logical reason for a carbon tax is to reduce emissions. Such a tax might help to reduce energy consumption, but only at punitive levels, because energy demand is so inelastic. Therefore, the real intention is to make fossil fuels so expensive that renewables can eventually become competitive, along with CCS, hydrogen heating etc.”
Market Signal Inefficiency
One of the underlying presumptions in any carbon price program is that the funds received will be spent effectively. I have evaluated the results of the investments made by regulatory agencies to date in RGGI measured as the cost per ton reduced. The RGGI states have been investing investments of RGGI proceeds since 2008 but their investments to date are only directly responsible for less than 5% of the total observed reductions. Furthermore, from the start of the program in 2009 through 2017, RGGI has invested $2,527,635,414 and reduced annual CO2 emissions 2,818,775 tons. The resulting cost efficiency, $897 per ton reduced, far exceeds the SCC that represents the value of reducing CO2 today to prevent damages in the future.
I looked at New York’s investments in more detail to see why those investments were so inefficient. The New York State Energy Research and Development Authority (NYSERDA) report New York’s RGGI-Funded Programs Status Report – Semiannual Report through December 31, 2018 describes how New York invested the proceeds from the RGGI auctions. That report lists the programs that are funded using RGGI proceeds in six categories: Green Jobs – Green New York, Energy Efficiency, Renewable Energy, Community Clean Energy, Innovative GHG Abatement Strategies, and Clean Energy Fund. From the titles alone it is clear that waving a pot of money in front of politically-driven bureaucracies is an incentive to build empires. I evaluated the projects within these categories and found that there were 19 programs with associated CO2 reduction benefits and another 18 programs with no claimed CO2 reductions. None of the 19 programs with CO2 reduction benefits met the $50 SCC metric for cost effective investments. Clearly the 18 programs with no claimed reductions would not be able to meet the metric either.
Theory says that the carbon price alone can incentivize lower emitting energy production and that the market choices will be more efficient than government-mandated choices. Ultimately the market signal question is whether the SCC value is sufficient to incentivize the market to invest in zero GHG emitting generation resources. There is no sign that RGGI motivated the market to act and it is not clear that the carbon pricing schemes proposed under the purview of FERC will provide enough value either. If the market signal is inadequate, then New York’s experience illustrates that government-mandated choices must be chosen carefully to ensure that the cost per ton of CO2 reduced is less than the SCC. I believe that the more targeted the investment to actually reduce energy use or CO2 emissions the more likely that SCC effectiveness criterion can be met.
Another consideration in effectiveness is timing. New York has a legislative target to generate zero GHG emissions from electricity production by 2040. Even if investors do come forward, are they going to be able to develop alternative generating resources in the time frames necessary to meet the ever more aggressive goals set by states competing to be the most ambitious?
Cost Shifting Total Costs
As noted previously at some point replacing existing fossil-fired generation not only has to consider the direct power output conversion costs but must also address dispatchability and grid support costs. When the carbon pricing proposal simply increases the cost of the energy generated, I think that approach will lead to cost shifting where the total costs of fossil fuel alternatives are not addressed.
Consider an electric system carbon price. In that approach any generator that emits CO2 will have to include a carbon price in their bid which serves to provide the non-emitting generators with more revenue. However, solar and wind generators are not paying the full cost to get the power from the generator to consumers when and where it is needed. Because solar and wind are intermittent, as renewables become a larger share of electric production energy storage now provided by traditional generating sources will be needed but there is no carbon price revenue stream for that resource. Because solar and wind are diffuse transmission resources are needed but solar and wind do not directly provide grid services like traditional electric generating stations. Energy storage systems could provide that support but they are not subsidized by the increased cost to emitting generators.
There are ways to address this. The carbon price could be modified to direct revenues to energy storage systems. However, when you do that the direct cost will go up and those least able to afford energy price increases will be hit with a regressive tax. The simplest solution would be to require all electric power sold to the grid to be dispatchable. In other words, require wind and solar to only sell power through their own dedicated energy storage systems. That won’t be popular for those resources because it effectively doubles their cost but the fact is that someone, somewhere will have to pay for those services so why not them.
I also believe that there are significant logistical issues associated with carbon pricing. In order to set a carbon price, you have to know what the carbon emissions are for every source providing energy to the market. For a global all-sector pricing scheme, you could set the price as the fuel is produced so that everyone pays the cost all the way through its end use. On the other hand, in the NYISO proposal they have to set the carbon price as electric energy as it is sold. Tracking emissions on that real-time basis is a non-trivial problem. In New York, NYISO knows which generators are running and has a pretty good idea of their emission rates. However, the final emission numbers are not available real-time because the emission values reported to prove compliance are not finalized until quality assurance post processing is complete and that can be months after the fact. The more significant problem is that NYISO has no way to calculate imported electricity carbon emissions on a real-time basis so cannot assign a carbon price value that accurately reflects how imported electricity is being generated. These issues have been glossed over to date.
The sources affected by RGGI had a long history working with cap and trade programs such as the Acid Rain Program before RGGI was implemented. On the other hand, if carbon tax schemes are implemented for other sectors the affected entities may not have experience with this kind of regulatory program. I believe that this increases the likelihood that affected sources will simply treat this as another tax.
Vermont Regulatory Analysis Project Carbon Management Study
There are not many critiques of carbon pricing schemes but there is one that deserves recognition. The Regulatory Analysis Project (RAP) recently completed a relevant study: Economic Benefits and Energy Savings through Low-Cost Carbon Management for Vermont that raises relevant concerns. The introduction describes the genesis of the analysis:
In the 2018 legislative session, the Vermont Legislature called for a study to examine the possible methods, costs, and benefits of using carbon pricing to address the problem of carbon pollution in the state. Resources for the Future (RFF) was commissioned by the legislature’s Joint Fiscal Office to conduct that study, using the economic models and approaches available to RFF.
The Regulatory Assistance Project (RAP) has been asked to assess the RFF study and its conclusions, and to offer suggestions for action based on its results and our expertise in energy and climate policy. RAP has, over the past 25 years, examined these issues not only in Vermont but across the globe. Our observations and recommendations are based on that broad base of experience.
For the purposes of this report, in the short time available, we commissioned two expert studies. The first, on low-carbon transportation, was completed by M.J. Bradley & Associates (MJBA), which has conducted several studies on this topic across our region and beyond. The second, on opportunities for energy savings in housing and public buildings, was completed by the Energy Futures Group (EFG), an expert consulting firm based in Hinesburg, Vermont. We are grateful to these two firms for lending their expertise to Vermont and offering leading insights to this review.
What have we found? Based on the plain facts of Vermont’s physical and economic conditions, we conclude that an attempt to reduce Vermont’s carbon emissions based on carbon pricing alone will cost more, and deliver less, than a program of carbon reductions that is based on practical public policies—policies that attack the main sources of carbon pollution through tailored, cost-effective programs geared to Vermont’s families, businesses, and physical conditions.
Although the focus of the RAP study was on transportation and energy efficiency the over-arching conclusions are also applicable to all carbon pricing proposals. The report raises the important policy question: What does a climate policy cost consumers per ton of carbon avoided? Their answer is relevant:
Many advocates of carbon pricing begin with the proposition that the main point is to charge for carbon emissions “appropriately” and that carbon reductions will surely follow in the most efficient manner. While carbon pricing is a useful tool in the fight against climate change, there is now substantial experience to suggest that wise use of the resulting carbon revenues is equally important, or even more important, if the goal is to actually reduce emissions at the lowest reasonable cost. One of the principal conclusions of the RFF study is that, even if carbon charges were set as high as $100/ton, the reduction in carbon emissions achieved statewide would be only about 10 percent below the expected business-as-usual case.
This seems to present us with an insoluble problem. On the one hand carbon pricing is said by many to be the “best” and “most efficient” way to drive down emissions in line with global targets and Vermont’s statutory goals. But on the other hand, as common sense and studies—including even RFF’s analysis—conclude, carbon pricing alone will be a weak tool to deal with the realities of consumer behavior, our historic buildings infrastructure, rural settlement patterns, and the many barriers that working families and businesses face in choosing to invest in energy efficiency or other low-carbon options.
I believe that the RAP analysis supports my concern about carbon market pricing signal investment efficiency. Even though they still claim that “energy pricing can be married to public policies”, the high hurdles of leakage, reduced revenues over time and the disconnect between the theory and reality are unaddressed.
Carbon Pricing Cautionary Summary
Proponents claim that “An increasing number of organizations recognize this unique, market-based solution as a viable, scalable option for helping to reduce carbon emissions market-based solution”. I frankly don’t think most of those organizations have had actual experience with a carbon pricing initiative logistics and have not evaluated whether the carbon prices proposed will provide the market signals necessary to spur the necessary renewable development needed to meet any CO2 emission reduction goals as a viable, scalable option for helping to reduce carbon emissions. The success of any carbon pricing scheme boils down to the question whether the carbon price set will provide enough of an incentive for projects that produce emission reductions that displace today’s generators and eventually covers the costs to provide the dispatchability and grid support functions provided by today’s generation mix.
In my opinion, carbon price support is based on parochial interests. In the case of the NYISO carbon pricing initiative they appear to believe it will simplify the cost accounting for New York’s renewable implementation efforts. I think they have under-estimated the difficulty implementing the infrastructure necessary to accurately track the price of carbon and have ignored the potential that the complex scheme needed to reduce leakage will lead to unintended consequences. Other support appears to be based on the potential to make money and it is not clear that is in the best interest of the New York’s desire to reduce CO2 emissions as cost-effectively as possible.
To summarize, carbon pricing will always be a regressive tax. I also think that there are a number of practical reasons that carbon pricing will not work as theorized. Because a global program is impractical, leakage is always going to be a problem. All carbon pricing proposals need to address the problem that as carbon emissions go down revenues go down relative to the fact that reductions get more difficult and expensive as control efficiency increases. The academics who support carbon pricing seem to be blissfully unaware of the realities of the energy market that are at odds to their theories. Based on observed results I think that indirect market signals are going to lead to less cost-effective reductions in the time frame necessary for the aggressive reduction rules. To date, carbon pricing for the electric sector only considers generation costs which leads to cost shifting the additional costs to supply electricity when and where it is needed to be covered outside the carbon pricing framework. Finally, supporters under-estimate the very real problems of implementation logistics. My concerns about carbon pricing are supported by the RAP study.
Roger Caiazza blogs on New York energy and environmental issues at Pragmatic Environmentalist of New York. This represents his opinion and not the opinion of any of his previous employers or any other company he has been associated with.