One of my pragmatic interests is market-based pollution control programs. In this post I am going to address the take on the Regional Greenhouse Gas Initiative (RGGI) in an influential book Making Climate Policy Work. There are also important lessons to be heeded as New York considers a Cap and Invest program.
I follow and write about the RGGI market-based CO2 pollution control program for electric generating units in the NE United States. I have extensive experience with air pollution control theory, implementation, and evaluation having worked on every cap-and-trade program affecting electric generating facilities in New York including the Acid Rain Program, RGGI, and several Nitrogen Oxide programs. 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.
Making Climate Policy Work Overview
The description of the book states:
For decades, the world’s governments have struggled to move from talk to action on climate. Many now hope that growing public concern will lead to greater policy ambition, but the most widely promoted strategy to address the climate crisis – the use of market-based programs – hasn’t been working and isn’t ready to scale.
Danny Cullenward and David Victor show how the politics of creating and maintaining market-based policies render them ineffective nearly everywhere they have been applied. Reforms can help around the margins, but markets’ problems are structural and won’t disappear with increasing demand for climate solutions. Facing that reality requires relying more heavily on smart regulation and industrial policy – government-led strategies – to catalyze the transformation that markets promise, but rarely deliver.
The authors recognize the enormity of the challenge to transform industry and energy use on the scale necessary for deep decarbonization. They write that the “requirements for profound industrial change are difficult to initiate, sustain, and run to completion.” Because this is hard, they call for “realism about solutions.” Cullenward and Victor recommend clear thinking and strategy as opposed to “Efforts spent tilting at ephemeral, magical policy solutions waste scarce resources that should instead be invested in things that work.” The goal of their book is to explain how market-oriented climate policies have fallen far short and how they might be modified so that they work.
RGGI Results
One of my first posts at this blog is still in the top ten viewed articles: Academic RGGI Economic Theory of Allowance Management. In that article I argued that economic value theory for an allowance market fails to account for the behavior of the affected sources. In particular, the owners and operators of sources treat the allowances primarily as compliance instruments and not as financial assets. 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. Compliance entities decide to buy allowances based on their expected operations in the period between auctions or, at most, the entire 3-year compliance period including a small margin for operational variations and regulatory compliance. Contrary to theory there is little attempt to make the allowances a profit center.
I have regularly evaluated RGGI performance on this blog. Last December I evaluated the 2020 RGGI Investment Proceeds report that describes the results of RGGI investments over the entire region. I found that since the beginning of the RGGI program CO2 emissions have been reduced more than 50% but that RGGI funded control programs have been responsible for only 5.6% of the observed reductions. In late December I did a similar analysis of just the New York investment proceed results and found that in New York since the beginning of the RGGI program to 2021 CO2 emissions have been reduced 39% but the reduction was 47% until the State shutdown the Indian Point nuclear station. The RGGI funded control programs have been responsible for only 16% of the observed reductions. The main reason for the reductions in RGGI and New York State has been fuel switching to natural gas unrelated to RGGI.
I also recently evaluated New York’s operating plan that guides the investment of RGGI proceeds. In the next fiscal year, the operating plan has 30 programs but only two programs claim direct CO2 reduction savings. Over the years 2013 to 2021, the total investment for those two programs is $565 million and the claimed savings are 1,684,616 MWh and 861,442 tons of CO2e with a calculated cost benefit of $656 $/ton. I classified each program relative to six categories of potential RGGI source emission reductions. The first three categories cover programs that directly, indirectly or could potentially decrease RGGI-affected source emissions. Those programs total 45% of the investments. I also included a category for programs that will add load that could potentially increase RGGI source emissions which totals 27% of the investments. Programs that do not affect emissions are funded with 21% of the proceeds and administrative costs total another 7%. In summary, even though the ostensible purpose of RGGI proceeds is to reduce emissions from RGGI-affected sources, less than half of the investments expect to do so.
Even though many RGGI proponents claim the program has been a success, my work shows that depends on how success is defined. If success is defined as significant cost-effect emission reductions from affected sources then that is not the case. If success is defined as a functional market-based system that provides proceeds then it is a success. There is no question the program components work well. The misuse of RGGI funds for affected source emission reductions is not the fault of the system but the politicians who control fund disbursement.
Making Climate Policy Work and RGGI
I wondered if this book talked about RGGI and how they rated its results relative to my analyses. I went through the document searching for and documenting every reference to RGGI to see whether I agreed with their description and evaluation of the program.
The first chapter describes the vision and the reality of carbon reduction market-based policies. Three example policies are described, including RGGI. The RGGI description states:
RGGI’s vision is the most realistic and generally applicable precisely because it is the most pragmatic about what is able to be achieved. The program encompasses states with varied political interests around climate change, ranging from the highly ambitious to the cautiously engaged. It covers only the electricity sector – where the technologies for cutting emissions are most mature – with transparent and predictable program rules. Even in the power sector, however, RGGI is not the only or even main show in decarbonizing its participating states’ electric grids. Other policy programs are having a bigger impact, including state renewable portfolio standards; subsidies that keep nuclear power plants, which are prodigious suppliers of zero-carbon power, from shutting down; and other government-managed regulatory and procurement efforts all aimed at making the RGGI states’ power infrastructure less carbon-intensive. In many respects, the RGGI system represents the high-water mark for what subnational markets can do: RGGI supports the broader goal of deep decarbonization, generates discretionary revenue streams for participating governments, and increases the static economic efficiency of a policy portfolio – all in a single sector. Its benefits are clear and relatively modest. Among purists, RGGI is often mocked because its prices are low (about $5–6 per metric ton of CO2 emissions in 2019) and coverage is limited to just one sector. We see the experience through a completely different lens: RGGI works because its architects knew what they were doing and designed a system that is politically feasible and durable.
I have slightly different takes on some of these points but overall I agree with their characterization.
The next two chapters and Chapter 5 only mention RGGI in passing. Chapter 2: Ambition makes the case that the theory of flexible and economically efficient carbon markets should make them ideal for maximizing the effort to control carbon pollution. This chapter explains why carbon markets have failed to live up to the expectations. The only reference to RGGI discussed the political process that underpins participation. The RGGI framework is flexible enough so that the addition and deletion of participating states when political regimes change does not affect the viability of the overall program. It concludes: “Firms and governments participating in RGGI know that states may come or go, with the consequences managed through an informal political process rather than a legal one.” Chapter 3 on coverage and allocation notes that RGGI is limited to the electric sector. Chapter 5 on offsets notes that even though offsets are allowed they have not been a factor in RGGI. I agree with their characterizations.
Chapter 4: Revenue and Spending delves into the disbursement of funds collected in the market. The total RGGI cumulative auction proceeds at the time of this writing is $5,895,274,757.14 since the first auction in September 2008 so RGGI has successfully generated revenues. With regards to spending the chapter notes that “How societies spend the money raised through these sales is vital to understanding the politics of emissions trading.”
The chapter discussion on RGGI points out that each state controls its revenue spending. There is a graph from the 2017 RGGI proceeds investment report that describes revenue uses in three categories: general funds; revenue recycling (earmarking revenues for spending that benefits citizens); and green spending (energy efficiency, clean energy, and climate mitigation). Given the difficulties I have had trying to interpret the RGGI proceeds reports, it is not surprising that there isn’t more detail.
The authors did pick up on some of the revenue problems in RGGI:
The RGGI program also reveals some of the political dynamics that can emerge when political leaders decide to re-purpose funds. The Governors of New York and New Jersey have both diverted RGGI revenues to the state’s general fund at points in the program’s history, raising concern from environmental NGOs and others who have supported a green spending agenda.
In a section within this chapter titled “Why green spending becomes green pork” the authors explain that there is not much scrutiny how the money is spent. They define pork as an expenditure that is designed to disproportionately benefit a special interest rather than the broader public good. They claim that “the organizations that spend RGGI funds are better designed to provide more discipline and accountability on how those funds are spent” than the other example programs discussed. While that may be true with respect to RGGI as a whole, it is not the case for New York. For example, the authors did not manage to tease out the fact from various unclear reports that New York uses RGGI funds to cover costs that were covered by general funds, i.e., a hidden diversion of revenue to the general fund. I am sure that had the authors looked into New York’s operating plan for RGGI auction proceed expenditures they would have agreed with my conclusion that green pork is a prominent part of New York’s expenditures.
Chapter 6: Market Links discusses the “institutional challenges of managing cross-border market governance”. With regards to RGGI I agree with their characterization:
Critically, what holds this system together is not law and the creation of robust, tradeable property rights, but rather a shared vision of parallel efforts at low levels of ambition. Design decisions are made according to the evolving political views of current and prospective participants. And because RGGI features so many parties – none of which hegemonically dominates the group’s overall agenda – the program must be transparent and predictable. The largely egalitarian cooperation of RGGI states works because it is anchored in stability-oriented market design features that make market behavior more predictable and risk management more tractable.
Chapter 7: Getting the Most Out of Markets explains how to increase program ambition, for example, attracting more jurisdictions or setting more ambitious targets. The RGGI discussion does a good job explaining how the program addressed an oversupply condition:
The northeastern United States’ RGGI program takes a similar approach through a pair of one-time cap adjustments, as well as a dynamic intervention that resembles the Market Stability Reserve. Like the EU ETS, RGGI experienced market oversupply conditions and very low prices in the 2010s. The situation with RGGI was more extreme, however, because this cap-and-trade program only applies to the electricity sector and the United States’ electricity sector began a profound transformation alongside (but not because of) RGGI. Not only did many of its participating states implement aggressive renewable energy and energy efficiency regulations, but also the rise of cheap natural gas from fracking dramatically accelerated the replacement of high-emitting coal-fired electricity with relatively clean natural gas and zero-carbon renewables. Emissions have been falling steadily, despite – not because of – anemic RGGI prices. As emissions fell owing to exogenous forces, the market became oversupplied. In response, RGGI’s two cap adjustments removed almost 140 million allowances – about two years’ worth of total emissions – from the supply of allowance budgets through program year 2020.[1]
In addition to these one-time adjustments, RGGI also developed a dynamic mechanism to alter the supply of allowances.[2] This additional market feature is triggered by observed market prices, rather than the EU ETS Market Stability Reserve’s measurement of excess allowance supplies. Like the EU ETS Reserve, RGGI’s approach is two-fold: RGGI features a Cost Containment Reserve that releases 10% of the program-wide allowance budget into the market if prices reach $13 per allowance in 2021; and if prices fall below $6 per allowance in 2021, an Emissions Containment Reserve will absorb 10% of the program’s annual allowance budget and remove these allowances from circulation. When the market remains in between the two triggering prices, allowances supplies are fixed – just as in the EU ETS, where supplies are fixed so long as the total number of surplus allowances stays within a specified range. (Both triggering prices increase at 7% per year to increase ambition over time, but not even the high-end prices are significant when compared to the policy incentives supporting renewable or nuclear energy in participating RGGI states.)[3]
The final chapter is entitled “Rightsizing markets and industrial policy”. One of the problems identified in the book is that the level of expenditures needed to implement the net-zero transition vastly exceeds the “funds that can be readily appropriated from market mechanisms”. The chapter describes RGGI as the “the cap-and-trade system whose design is most purely oriented around generating and spending revenue”. The authors note that the October 2019 report “The Investment of RGGI Proceeds in 2017” indicates that New York has mobilized just $100 million per year for green spending. My review of the latest plan to invest New York RGGI auction proceeds indicates that the design plan is supposed to “support the pursuit of the State’s greenhouse gas emissions reduction goals”. Of the five goals listed, only one addresses emission reductions. The others are vague cover language to justify the use of RGGI auction proceeds as a slush fund for hiding administrative expenses and costs related to Climate Act implementation at the expense of programs that affect CO2 emissions from RGGI affected sources.
Making Climate Policy Work and New York Cap and Invest
Governor Hochul recently announced a plan to use a market-based Cap and Invest program to raise funds for the Climate Leadership & Community Protection Act. I submitted comments on the Draft Scoping Plan that made opposed the recommendation for such a program. My initial impression of the Cap and Invest program is that it is more style than substance. If I had read this book before drafting the comments or my initial impression article, I would have highlighted the findings in this book as part of my arguments against this approach.
The program public relations summary claims that “A Cap-and-Invest Program is the most feasible, efficient, and affordable method to attain a more sustainable future.” I have been surprised by the amount of support for the plan. At the February 14, 2023 New York Senate Environmental and Ways and Mean legislative public hearing on the 2023 executive budget the majority of the speakers supported the proposal. I don’t think that any of the comments that support the program realize the many flaws in that proposal that are described in this book.
In my opinion, a fundamental flaw in the Scoping Plan is that it does not include feasibility analyses to determine whether the laundry list of control strategies will be feasible. The Plan does not demonstrate that the proposed strategies will be maintain current standards of reliability and safety or can keep energy costs affordable. This lack of analysis extends to the Cap and Invest proposal. Proponents claim that it is the most feasible option but that is relative to a short list of options and does not necessarily mean that it will work as proposed. The preface of the book notes the importance of feasibility:
In telling the story of how market-based climate policy works in the real world, we adopt the premise that idealized markets would be desirable if they were feasible. We hope this choice allows us to reach readers who identify strongly with the power of market forces, since we hope to change their minds. We want them to understand how political forces constrain what market-based policies can do, especially at the early stages of deep decarbonization, because wishing those forces away isn’t practical and hasn’t worked.
The Cap and Invest fact sheet notes that this program will be similar to RGGI that “has helped reduce greenhouse gases from power plants by more than half and raised nearly $6 billion to support cleaner energy solutions”. As noted previously my analyses show that RGGI was only a minor cause of the observed emission reductions. Chapter 1 this book also argues that RGGI is not the primary cause: “Other policy programs are having a bigger impact, including state renewable portfolio standards; subsidies that keep nuclear power plants, which are prodigious suppliers of zero-carbon power, from shutting down; and other government-managed regulatory and procurement efforts all aimed at making the RGGI states’ power infrastructure less carbon-intensive.” Based on my work I believe fuel switching has been the primary cause of New York observed reductions but there are two aspects to consider. The reductions were because natural gas was a cheaper alternative than coal and oil. However, the subsidies for nuclear power plants kept emissions from rising. That is until the State made the irrational decision to shut down 2,000 MW of nuclear power at Indian Point. Since 2019, when the staged closure began, New York electric utility CO2 emissions have increased 5.8 million tons or 23%.
The Scoping Plan recommendation for an economy-wide strategy to address the financing and emission limitations is based on a naïve understanding of market-based programs. Cullenward and Victor explain the reality:
Market-based policies on a planetary scale, the theory goes, would empower firms and governments with the flexibility to focus investment on the least expensive options for controlling emissions. Flexibility would reduce costs, allowing more environmental protection with fewer resources; in turn, frugality would make it easier to mobilize business and voter support for ever-deeper climate pollution reductions.
They go on to explain that this vision has completely failed:
Many pollution markets exist, but nearly all are smokescreens that create the impression that market forces are cutting emissions when, in fact, other policies are doing most of the real work of decarbonization. Almost everywhere that market systems are in place they operate at prices that are so low as to have little impact on key decisions such as whether to invest in or deploy new technologies.
The Cap and Invest solution is being marketed as both a compliance and financing tool. The belief is that the cap will establish compliance limits and the auction will provide the funding to make the reductions. There are issues with these tools.
The use of the cap as a binding compliance mechanism is unprecedented. Consider, for example, the EPA Cross State Air Pollution Rule (CSAPR). This cap-and-trade program is in place to limit nitrogen oxide (NOx) emissions in the eastern United States for ozone compliance. There have been multiple iterations of this rule that have progressively reduced the cap. The distinction between CSAPR and a binding cap is that EPA evaluated emissions, existing control technology, and potential improvements or additions for all the sources in the CSAPR-affected states. The cap was determined using this control technology evaluation to set a feasible limit. A binding cap is one chosen arbitrarily without any such feasibility evaluation. In 2030 New York GHG emissions must be 40% lower than the 1990 baseline but this is an arbitrary target mandated by the Climate Act.
There is another aspect of any GHG emissions reduction program. There are no cost-effective add-on control technologies available for existing sources. The only options available for an affected source are to change the fuel to something with lower GHG emissions, make the system more efficient, to reduce operations, or shut down. As noted previously, New York reduced its electric system emissions significantly because of fuel switching but that strategy is tapped out for any future significant reductions. In order to get more reductions from the electric generating system, zero-emissions resources must be deployed to displace the fossil resources. This is particularly difficult because the loss of Indian Point’s zero emissions generation has increased recent emissions. The control strategies are similar for all other sectors.
Cullenward and Victor make the point that it is easier to make reductions with existing technology:
In a few places, carbon prices from market-based policies have been powerful enough to induce some changes in emission patterns – such as when firms decide whether to produce electricity from high-emission coal plants or lower-emission rivals. Those impacts, however, have nearly always involved commercially mature technologies competing in stable environments and under other highly restrictive conditions.
In order to meet the 2030 GHG emissions target technology that has not been proven commercially viable at the necessary scale is needed. This challenge is a problem with the Climate Act deep de-carbonization targets that the Scoping Plan recommendations ignore:
On another front, what markets do best – creating transparent, marginal price signals that encourage firms and households to optimize their choices – is misaligned with the industrial challenges facing deep decarbonization today. In most sectors the world is not far along with deep decarbonization: key technologies, demonstration projects, and the emergence of new firms to back low-carbon technologies are fledgling at best (see Figure 1.2).9 Industrial firms and consumers aren’t waiting for a faint, marginal signal from markets to nudge their behavior. Instead, they need active programs to mobilize and apply resources to new technologies that, with time and effort, will launch the global process of deep decarbonization and displace incumbent industries. Well-designed market signals, at best, are good at encouraging optimization when technologies are commercially mature and strategic choices are clear – such as when the UK electricity market had a signal to select mature renewable energy technologies and gas instead of coal. The hardest challenges of deep decarbonization involve redirecting investment toward technologies and businesses that are the opposite: beset with risk and danger for first movers. Creating those new industries requires a policy strategy – industrial policy, in effect – that is focused on the problem at hand, rather than inducing marginal changes in behavior with known technologies and production methods.
The authors address three issues related to the fact that the existing systems have failed to live up to expectations. The first issue is related to the technology issues noted above:
We explain why idealized, “first-best” designs for pollution markets envision systems that produce high carbon prices as a powerful incentive for change. In the real world, the outcome has been the opposite: prices are low and often volatile, which undercuts the incentive to invest in ambitious new technologies and to make changes in production methods beyond those that are straightforward with few risks. First-best visions for pollution markets also imagine that markets should cover many sectors simultaneously, allow extensive interconnection with markets overseas, raise large amounts of revenue, and spend those revenues efficiently to offset distortions in the economy. On every front the real world has produced outcomes that are the opposite from theory: markets are fragmented, links are few, sectoral coverage mostly is narrow, and revenues raised are small.
Details for the proposed Cap and Invest program are sketchy but my impression from what I have heard is that it will also be the opposite of this theory.
As an alternative, the author describe how to make market-based programs more effective. Their second issue is necessary market reforms:
Some reforms are needed to make market signals more reliable – an outcome that requires shifting away from cap-and-trade systems, where market structures create volatile prices, and toward systems where prices are managed within narrow bands. In effect, cap-and-trade systems can be made more effective when they are designed to behave more like taxes; it is no accident that the few jurisdictions with the highest prices and the greatest level of effort use taxes, not cap-and-trade. More stable prices will make it easier for firms to invest in anticipation of market signals and to build political coalitions that are supportive of that investment. Systems that are designed like taxes also perform better in the real world where market policies are implemented alongside other regulatory programs. In that setting, cap-and-trade schemes merely trade the residual and get little work done in cutting emissions – they are Potemkin markets. Tax approaches, by contrast, create a clear incentive for change (the specified tax level), which persists even as other policy instruments have big impacts on behavior as well.
This approach is basically RGGI without a binding cap. Unfortunately, Climate Act proponents are convinced that the transition schedule is possible despite the lack of any evidence supporting evidence and that the climate crisis necessitates the aggressive schedule of the Climate Act. Even though New York GHG emissions are less than one half of one percent of global emissions and global emissions have been increasing by more than one half of one percent per year this rationale for the Climate Act schedule is a major obstacle against this common sense approach.
In addition to the compliance mechanism the proposed Cap and Invest program is intended to provide revenues for the transition. I have no doubts that the program will generate revenues and suspect that the Hochul Administration will decide the revenue targets based on just how much they think they can get away with rather than basing them on the results of their RGGI auction proceeds. Cullenward and Victor address this aspect:
Our playbook for market reform offers some insights into why so many of the visions for market-oriented climate policy won’t happen under real-world political conditions. For example, many advocates for market-based policies imagine that the adoption of market schemes will occur alongside massive policy reforms that roll back regulation. We explain why, politically and administratively, those regulatory and industrial policies are not easily rolled back. Moreover, we explain why pushing for that outcome would be a bad idea – since those other regulatory policies, in fact, are doing most of the serious work in cutting emissions.
One of the most important contributions of markets is among the least appreciated today: well-designed market schemes can raise revenue. A politically savvy strategy for market reforms requires paying closer attention to how program revenues are spent – and specifically to allocating funds to activities that will build experience with new technologies and thus also catalyze new interest groups that are supportive of accelerating deep decarbonization.
Because of the enormity of the challenge another issue is discussed. In particular, what else is needed:
The key is to channel resources into the sectors that are critical for deep decarbonization. Rather than link all sectors together into a common market system, each must be treated independently because each has its own political economy and state of technology. In sectors where technologies are immature, industrial policy should focus on research, development, and demonstration (RD&D) in a diverse array of options – an approach that yields knowledge and also builds political coalitions around new low-carbon industries.
The New York Climate Act covers all sectors. It may be possible to breakout the sectors based on such a recommendation. However, the looming problem is that a binding cap will limit emissions even if the zero-emissions resources are not available to displace the existing emissions. Carbon dioxide emissions are directly tied to fossil-fuel combustion and energy production. If for any number of reasons, the zero-emissions are not deployed fast enough in all the sectors there won’t be enough credits available to cover the emissions necessary to provide the energy needs. In the worst case, an electric generating unit needed to keep the lights on will refuse to operate because they have insufficient allowances.
The obvious solution to this concern is a feasibility analysis of the schedule for technological innovations necessary to maintain affordability and reliability. The authors suggest “Doing better requires recognizing the structural limits to what is achievable with market-based approaches – limits that are rooted in how the politics and technological opportunities are organized in each sector.”
Conclusion
The Hochul Administration proposes a Cap and Invest program that will provide revenues and establish a compliance mechanism. I agree with the authors that the results of RGGI and other programs suggest that the Cap and Invest proposal will generate revenues. However, we also agree that the amount of money needed for decarbonization is likely more than any such market can bear. The problem confronting the Administration is that in order to make the emission reductions needed they have to invest between $15.5 and $46.4 billion per year. I don’t think that range is politically palatable.
The use of Cap and Invest as a compliance mechanism is more of a problem. The Hochul Administration has not acknowledged or figured out that the emission reduction ambition of their Climate Act targets is inconsistent with technology reality. Because GHG emissions are equivalent to energy use, limiting GHG emissions before there are technological solutions that provide zero-emissions energy means that compliance will only be possible by restricting energy use. Unless a miracle occurs in 2030 when there are insufficient allowances someone has to choose who gets to operate.
This is a good book and I recommend it to anyone interested in energy and climate policy and emissions trading programs.
[1] The Regional Greenhouse Gas Initiative, “Elements of RGGI,” https://www.rggi.org/program-overview-and-design/elements; see also The Regional Greenhouse Gas Initiative, “RGGI Program Review: Summary of Proposed Changes to RGGI Regional CO2 Allowance Budget” (Nov. 21, 2013); The Regional Greenhouse Gas Initiative, “Second Control Period Interim Adjustment for Banked Allowances Announcement” (March 17, 2014).
[2] The Regional Greenhouse Gas Initiative (2014), supra note 11.
[3] New York and Illinois (the latter of which is not in RGGI) created the first zero-emission credit (ZEC) subsidy programs for nuclear energy in the United States. See Nuclear Energy Institute, “Zero-Emission Credits” (Apr. 2018). These policies were challenged in court and ultimately upheld in two parallel cases. Coalition for Competitive Electricity v. Zibelman, 906 F.3d 41 (2nd Cir. 2018) (New York); Electric Power Supply Association v. Star, 904 F.3d 518 (7th Cir. 2018) (Illinois). Following these favorable outcomes, New Jersey (once again part of RGGI) adopted a similar program. Robert Walton, “New Jersey moves ahead on nuke subsidies, approving ZEC application process,” Utility Dive (Nov. 21, 2018). For an overview of state renewable energy policies, see Galen L. Barbose, “US Renewables Portfolio Standards: 2019 Annual Status Update,” Lawrence Berkeley National Laboratory (2019), https://emp.lbl.gov/projects/renewables-portfolio.
Pragmatic Environmentalist of New York 