This is another in a series of posts on the Regional Greenhouse Gas Initiative (RGGI). The program includes periodic reviews to consider program successes, impacts, and design elements. In the current program review process one issue is the potential to add a new component to the program called the Emissions Containment Reserve. Resources for the Future (RFF) and University of Virginia (UVA) had a webinar on June 14, 2017 on the results of their analyses of that component and RFF followed up with comments submitted to RGGI on July 17, 2017. This post addresses what I believe is a fundamental problem with the academic theory of RGGI allowance management.
I have been involved in the RGGI program process since its inception. Before retirement from a Non-Regulated Generating company, I was actively analyzing air quality regulations that could affect company operations and was responsible for the emissions data used for compliance. The opinions expressed in this post do not reflect the position of any of my previous employers or any other company I have been associated with, these comments are mine alone.
Dr. William Shobe (UVA) and Dr. Dallas Burtraw (RFF) described their analysis of the effect of an Emissions Containment Reserve (ECR) in the June 14, 2017 webinar. RFF used a model to project future allowance supply, demand and cost. UVA did a laboratory experiment at the UVA Economics Laboratory using students as allowance managers.
I believe they both have the same perception of the economic theory of RGGI allowance management. The UVA theory is summarized in slide 19 of the webinar slide presentation. That slide states that it is the long-run supply that counts. “In markets for storable commodities (like allowances, for example), the current price and the plan for accumulation of a stock of the commodity depend on
- The expected long-run total supply compared to
- The expected long-run total demand.”
Similarly, in the RFF comments to RGGI it is noted that the allowance banking “propagates and adjusts the value of an allowance over time in light of the opportunity cost of holding the allowance as a financial asset (versus buying one at a later point in time).”
However, I believe that RGGI allowance management is different because the affected sources do not treat allowances as a storable commodity or a financial asset in the usual sense of the term. Instead allowance management is overwhelmingly driven by regulatory requirements for the current compliance period. i.e., do I have enough allowances to cover expected emissions? Financially it is simply another cost of operating and not a potential profit center. The important difference is that the academic economic theory holds that affected sources are looking years down the road but in reality there is no such long-term time horizon for affected sources. Their decision to buy allowances is driven by their expected operations in the period between auctions and at most the entire compliance period and to include a small margin for operational variations and regulatory compliance.
I have worked in New York for a long time and I have been unable to find a single company that will admit to long-run allowance planning. In the first place, allowance purchases cost a significant amount of money. New York electric generating companies are on a tight margin with little extra money available, so the idea that money could be available to purchase allowances for needs more than three years in the future is laughable amongst my sources. It is also important to note that in New York that the non-regulated generating companies have been in a constant state of change since de-regulation began before 2000. Very few facilities are still owned and operated by the same companies that purchased when de-regulation began. As a result of that turmoil there are few incentives to purchase allowances for future needs because the expectation is that facility ownership changes will continue going forward.
In addition, the RGGI cap and auction CO2 allowance program is different than a traditional cap and trade program for SO2 or NOx. In a traditional program, allowances are allocated proportionally to affected sources based on historical operations. When the cap is established the total emissions in the State have to be lowered to that level. On the basis of the cap level, affected sources can determine if it makes sense to install control equipment or purchase allowances to comply with their compliance obligations. As long as somebody can over control and generate surplus allowances to subsidize their control investments then allowances should be available on the market for use if control costs are not cost-effective at a particular affected sources.
In the RGGI cap and auction system, everyone has to buy allowances. Even if there were an option to control CO2 emissions (and there isn’t anything available for existing sources that is cost effective) a source installing controls has to buy allowances so there wouldn’t be an offset to the control costs. As a result of all these factors the affected sources have universally adopted an allowance management strategy with a short time horizon.
So how does this affect the analyses and what could be done?
I am not sure how this affects the modeling and the lab experiment for allowance management. I note, however, that the presentation and the comments both imply decisions should be based on future expectations of allowance costs. I believe that is an outgrowth of their mistaken RGGI allowance management theory. Instead I believe that the ECR should be based on observed allowance price behavior.
I would love to see the UVA allowance tests be repeated using actual industry allowance managers and industry allowance compliance staff. It would be educational for both the academics and industry staff and would confirm or blow up my perception of the differences between academic theory and actual allowance management practice. It would also confirm whether the conclusions based on the academic theory are consistent.
One final note relative to the economic analyses done by RFF and UVA. As far as I can tell they both presume future emission cap reductions (RFF modeled a 3.5% reduction). The presentation and comments both implicitly assume that the primary motivation for the ECR is because there is a potential that future emission reductions will be greater than cap reductions leading to an over-supply of allowances and lower prices than are deemed acceptable.
I believe the more likely scenario is that emissions don’t continue at the historical decrease rate and that an allowance deficit is more likely. Consequently, I think that using the ECR to determine future emission reductions instead of arbitrarily picking a percentage reduction is less risky. The academic presumption that future reductions are “easy” also influences their recommendations relative to the ECR. At the very least they should consider a scenario where reductions less than expected.
 My rationale is that historical emissions decreases have been largely driven by fuel switching from a more expensive to cheaper and less emitting fuel. The potential for similar future reductions is largely gone. In an earlier post I showed that the experience so far of RGGI induced emission reductions is pretty low. The upper bound is an econometric model that estimates that emissions would have been 24 percent higher without the program. RGGI estimates that emissions would have been 17% higher than without a program. If you assume that all the savings in fossil fuel use only displaced natural gas use then emissions would have been only 5% higher.