A Key Issue for the Upcoming Climate Conference in Santiago

In December of this year, delegates to the 25th Conference of the Parties (COP-25) of the United Nations Framework Convention on Climate Change (UNFCCC) will assemble in Santiago, Chile, for two weeks of negotiations.  The location for the Conference was changed to Santiago when the Chilean government graciously stepped in as host after the Brazilian government reneged – two months after winning the bid and one month after the election of President Jair Bolsonaro – on its previous commitment to host COP-25.

The previous year, COP-24 took place in Katowice, Poland.  As I’ve previously written at this blog (”Climate Negotiations in Poland Advanced Implementation of the Paris Agreement”, December 20, 2018), the delegates at that Conference reached consensus on a 156-page “Rulebook” that filled in important details for 28 of the 29 articles of the skeletal Paris Agreement.  Consensus was not reached on one very important part of the Agreement, Article 6, the home for international cooperation that can bring down costs, and thereby facilitate greater ambition.

This presents a major challenge for the delegates to this year’s COP as they seek to complete the Rulebook with details for Article 6; in particular, how to facilitate a robust system of international cooperation (that allows for international carbon markets) while avoiding the possibility of double counting of emissions reductions, that is, counting the same emission reduction more than once when assessing progress towards the achievement of climate mitigation targets.

This is the topic of an article that appeared very recently in Science, “Double Counting and the Paris Agreement Rulebook,” which I had the pleasure of co-authoring with an international set of colleagues – Lambert Schneider, Maosheng Duan, Kelley Kizzier, Derik Broekhoff, Frank Jotzo, Harald Winkler, Michael Lazarus, Andrew Howard, and Christina Hood.  In this blog essay, I provide a brief summary, which I hope will entice readers to check out the full version in Science (Volume 366, Issue 6462, pp. 180-183, October 11, 2019).

The Context

It is important to distinguish among three distinct yet closely related levels of actions in regard to international cooperation under Article 6 of the Paris Agreement:

First, national or regional jurisdictions can establish domestic policies, such as emissions trading systems, carbon taxes, or performance standards, for the purpose of achieving the targets specified in their respective Nationally Determined Contributions (NDCs) under the Agreement.

Second, jurisdictions can link their respective domestic policy instruments, as, for example, California and Quebec have done, allowing allowances to be traded across international borders.  Such linkage was the subject of a previous article in 2018 in Science I co-authored with Michael Mehling and Gilbert Metcalf (a more complete version of that work appeared in the periodical, Environmental Law, earlier this year).

Third, and the focus of the new Science article and this blog essay, Article 6.2 provides a potential home for accounting mechanisms (“Internationally Transferable Mitigation Outcomes” or ITMOs, and “Corresponding Adjustments”) that can properly take account of such international transfers when demonstrating achievement of national targets under the Paris Agreement.

The Risk of Double Counting

If two different jurisdictions, such as two countries with their own NDCs, were both to take credit for the same emission reductions, there would be double counting under the Paris Agreement, which would be a significant threat to the integrity of the Agreement and any carbon markets employed in its implementation.  Given that half of the Parties of the Paris Agreement have indicated their intention to participate in carbon markets, avoiding (that is, reducing the risk of) such double accounting is critical for the credibility of the Paris regime.  A robust system to account for international transfers of emission reductions is necessary.

As my co-authors and I explain in the Science article, Article 6.2 of the Agreement provides the needed accounting framework through provision for “corresponding adjustments,” which can function as a form of double-entry bookkeeping.  But despite the fact that the Paris Agreement is explicit that double counting should be avoided, some Parties to the Agreement disagree about how it should be avoided, and indeed, about what constitutes double counting.  In addition, there is some controversy related to how much international oversight is needed to ensure robust accounting

The Path Ahead

Success at COP-25 in Santiago is critical.  In our Science article, my co-authors and I propose several principles to guide the negotiations.  I will mention just two of these in this brief essay.

First, a single set of common international accounting rules should apply under the Paris Agreement, irrespective of what type of carbon market mechanism is used to generate emission reductions.

Second, effective accounting will be greatly facilitated by all countries adopting targets (NDCs) that are economy-wide, cover all GHGs, apply to common multi-year time periods, and are expressed as GHG emissions. The Paris Agreement expressly foresees that countries will move toward such economy-wide targets over time.

If international cooperation is to combat climate change cost-effectively, the Paris Agreement needs to employ rules for international carbon markets that ensure environmental integrity and avoid double counting.  Otherwise, carbon markets may sadly undermine the Paris climate agreement.

Share

Sub-National Climate Change Policy in China

At a time when there are considerable political challenges in some countries (such as my own!) for national governments to institute meaningful climate change policies, the potential role of sub-national policies becomes more important than otherwise.  In other countries, sub-national climate policies may be a stepping stone to significant national efforts, as in China.  Partly with this in mind, the Harvard Project on Climate Agreements (HPCA) conducted a research workshop in July of this year on “Sub-National Climate Change Policy in China.”  Tsinghua University’s Institute of Energy, Environment, and Economy — directed by Professor Zhang Xiliang — hosted and co-sponsored the workshop, which was organized by my colleague at the Harvard Kennedy School, Dr. Robert StoweTwenty-seven experts from China, Europe, Canada, India, Australia, and the United States participated (see the photo below).  In addition, a group of students observed the workshop, and the Environmental Defense Fund’s China Program hosted a dinner for workshop participants.  The Harvard Global Institute provided major support for the project.  Here is a link to the full agenda (in both Chinese and English).

Background

Climate change is a global commons problem, and, as such, requires cooperation at the highest jurisdictional level — that is, international cooperation among national governments — if it is to be adequately addressed.  Participation by national governments is key, and sub-national governments can also play important roles. Provinces and municipalities around the world have undertaken initiatives — sometimes working together across national boundaries — to reduce greenhouse-gas emissions. These include jurisdictions in the largest-emitting countries — China, the United States, and India — as well as in the European Union.

The Workshop and its Analyses

Participants in the Beijing workshop examined how Chinese provinces and municipalities work with the central government to implement policy — and discussed challenges to such cooperation. They focused to a considerable degree on the implementation of China’s national carbon-pricing system, including approaches to integrating the seven pilot sub-national market-based systems into the new national scheme, scheduled to launch in 2020 (see “What Should We Make of China’s Announcement of a National CO2 Trading System?,” January 7, 2018).  Participants also addressed sub-national dimensions of other policy approaches to reducing greenhouse-gas emissions in China.

As we have done with previous HPCA research and policy workshops, participants in the Beijing event are now writing briefs on topics related to their respective presentations.  We will edit and compile these short papers in a volume to be released later this year.  In the meantime, you can view the PowerPoint presentations from the Beijing workshop:

  • China’s National Emissions Trading Program (Zhang Xiliang)
  • Ten Drivers Behind Climate Policy Making in China (Qi Ye)
  • Creating Sub-National Climate Institutions in China (Michael Davidson)
  • Multi-Dimension Post-Assessment of China’s ETS Pilots (Qi Shaozhou)
  • Political Economy Framework for Climate Change Policy in China (Christine Wong)
  • Canadian Climate Change Policy (Katie Sullivan)
  • Sub-National Carbon-Pricing Policy in the USA (Robert Stavins)
  • Integration of China’s National ETS with Provincial/Municipal Pilots (Valerie Karplus)
  • Introduction of Beijing ETS (Mei Dewen)
  • Sub-National Implementation Pathways for the National Pricing System (Goerild Heggelund)
  • Assessing Regional Implementation Pathways of National ETS In China (Wu Libo)

The Larger Context

The Beijing workshop was part of a larger initiative of the Harvard Project on Climate Agreements, supported by the Harvard Global Institute, examining and comparing sub-national climate-change policies in China and India. We will conduct a similar workshop in New Delhi next year.

The Harvard Project has previously conducted three workshops addressing climate-change policy in — or related to — China:

  • “Bilateral Cooperation between China and the United States: Facilitating Progress on Climate-Change Policy,” June 2015.  This was hosted by China’s National Center for Climate Change Strategy and International Cooperation (NCSC).  You can read more about this workshop here, and read the full workshop report here.
  • “The Design, Implementation, and Operation of China’s National Emissions Trading System,” December 2016.  Our host was NCSC.  The participants explored technical issues related to the design of China’s emerging national system, including allowance allocation, point of regulation, and price management.
  • “Cooperation in East Asia to Address Climate Change,” September 2017.  This was hosted by the Harvard Center Shanghai, and supported by the Harvard Global Institute. You can read more about the workshop here, and read the complete volume of briefs based on the workshop here.
Share

The Future of U.S. Carbon-Pricing Policy

In 2007, I was asked by the leaders of the Brookings Institution’s Hamilton Project to write a paper describing a national emissions trading system to reduce U.S. carbon dioxide (CO2) emissions to help address the threat of global climate change.  I responded that I would prefer to write broadly about carbon-pricing instruments, including what I considered to be the symmetric instruments of a carbon tax and a carbon trading program.  But the Hamilton Project leaders said no, they would find someone else to write about carbon taxes (which turned out to be Gib Metcalf), and they wanted me to “make the strongest case possible for” what is today called a cap-and-trade system.  I did my best, and in the process I came to be identified – and to some degree may have become – an advocate for CO2 cap-and-trade.  For better or for worse, during the Obama administration transition, the design recommendations in my Hamilton Project paper became one of the starting points for efforts to structure the administration’s proposed CO2 cap-and-trade system that became part of the failed Waxman-Markey legislation, H.R. 2454, the American Clean Energy and Security Act of 2009.

More than a decade later, I have written a new paper in which I seek to approach this question as I wished to in the first place, treating both instruments in a balanced manner, examining their merits and challenges, without necessarily favoring one or the other.  On May 16, 2019, I presented this new paper at the National Bureau of Economic Research’s first annual Environmental and Energy Policy and the Economy Conference, held at the National Press Club in Washington, D.C.  My topic was, “The Future of U.S. Carbon-Pricing Policy.”  (It will be forthcoming in Environmental and Energy Policy and the Economy, volume 1, edited by Matthew Kotchen, James Stock, and Catherine Wolfram, published by the University of Chicago Press.)  In today’s blog essay, I provide a very brief summary of the paper, based upon the presentation I made at the NBER conference.  I hope you will find this of sufficient interest to download and read the complete paper.

Premises, Questions, and Conclusions

I began this research with two major premises:  first, that economists and most other policy analysts agree that carbon-pricing will likely be a necessary (although not sufficient) part of any meaningful, long term U.S. climate change policy, because of:  (1) feasibility – the necessity of affecting millions, indeed hundreds of millions, of decentralized decisions; (2) cost-effectiveness, given the tremendous heterogeneity of marginal abatement costs; and (3) the importance of providing incentives for carbon-friendly technological change.  My second premise was that there is much less agreement among economists (and other policy analysts) regarding the choice of specific carbon-pricing policy instrument – carbon tax or cap-and-trade.

This prompts two questions:  (1) how do the two major approaches to carbon pricing compare on relevant dimensions, including but not limited to efficiency, cost-effectiveness, and distributional equity?  (2) Which approach is more likely to be adopted in the future in the United States?

Having carried out an exhaustive examination, two major conclusions stand out (among others).  First, that the specific designs of carbon taxes and cap-and-trade are more consequential than the choice between the two instruments.  And second, that political feasibility affects the normative merits of the two instruments, and vice versa.

Similarities & Symmetries

Of fourteen separate issues I examine, some appear at first to be key differences (in theory), but many of these differences fade on closer inspection, and depend on specifics of design.

First of all, carbon taxes and commensurate cap-and-trade turn out to be perfectly equivalent in regard to:   (a) incentives for emission reduction (both can be upstream on the carbon content of fossil fuels); (b) aggregate abatement costs (both can be cost-effective, both provide the same incentives for technological change, and both can utilize offsets to further lower aggregate abatement costs); and (c) effects on competitiveness (both can lessen these impacts via appropriate border adjustment mechanisms).

Next, the two instruments are nearly equivalent in regard to possibilities for raising revenue (cap-and-trade can utilize auctions, but given the structure of Congressional committees, revenue recycling may be easier with taxes).

And these instruments are similar in regard to:  (a) costs to regulated firms (cap-and-trade systems can freely allocate allowances, and taxes can provide inframarginal exemptions below a specified level of emissions); and (b) distributional impacts (the two instruments can be designed to be roughly equivalent in this regard).

Differences & Distinctions

Beginning with the least significant differences, there are relatively minor distinctions in terms of transaction costs (decreasing marginal transaction costs in cap-and-trade systems – such as with volume discounts on brokers’ fees – can violate the independence property, whereby the equilibrium allocation of allowances and hence aggregate costs are ordinarily independent of the initial allocation).

There are more meaningful, but still subtle differences with regard to:  (a) performance in the presence of uncertainty (for this, I urge you to read at least this section of the complete paper, because new research suggests that the implications of the classic Weitzman rule in the presence of a stock externality are moderated – if not reversed – due to the persistent effects of technology shocks, which foster positive correlation between marginal benefits and marginal costs); and (b) linkage with other jurisdictions (it is easier with cap-and-trade systems, but tax systems can also be linked).

That said, there are significant differences between the instruments in terms of:  (a) carbon-price volatility (a problem only with cap-and-trade systems, but a problem that can be mitigated with price collars and banking of allowances); (b) interactions with complementary policies (a significant issue with cap-and-trade systems, which is much less severe with carbon taxes, because the “waterbed effect” is eliminated); (c) market manipulation (there is a need for regulatory oversight in cap-and-trade systems, but tax evasion is a parallel issue in tax systems, although presumably less severe in the U.S. context); and (d) complexity and administrative requirements (cap-and-trade is certainly more complex and has greater administrative requirements, but one might ask whether a simple tax will remain “simple” as it works its way through the Congress).

Hybrid Policy Instruments and a Policy Continuum

Many of the remaining differences can diminish further with implementation.  Indeed, hybrid policies which mix features of tax and cap-and-trade blur distinctions.  For example, auctioning of allowances and the use of price collars bring cap-and-trade closer to a tax system; and quantity formula employed to adjust a tax, and the use of tax revenues to mitigate emissions bring a tax closer to cap-and-trade.  The result is that the dichotomous choice between a carbon tax and cap-and-trade can become a choice of design elements along a policy continuum, and the design of these instruments can be more consequential than the choice between the two.

Which is More Likely to be Adopted – Taxes or Trading?  Positive Political Theory

Framing this question in terms of the metaphor of a political market, it is helpful to think about political demand and political supply of policy instruments.  In terms of the demand from interest groups, first, regulated industry may oppose an ordinary tax approach, as it typically leads to greater costs than the simplest cap-and-trade (or than a performance standard, for that matter), because private industry is paying not only for compliance costs, but also for the tax on residual emissions.  Second, regulated industry may favor cap-and-trade, because it conveys scarcity rents to firms, and can provide entry barriers for potential new entrants, which can make the rents sustainable.

Environmental advocacy groups favor cap-and-trade, due to the emissions certainty it provides, but also because presumably they have a preference for policies that help obscure costs, and cap-and-trade does a better job of sweeping discussion of costs under the rug than does a tax.  However, in the era since cap-and-trade was demonized as “cap-and-tax,” this difference may be much less than it was!

Turning to the supply side (within the legislature), the revenue from either a tax or auctioning of allowances can be attractive to government.  And because of the independence property of cap-and-trade, legislators can allocate allowances to build political support without increasing the costs or reducing the effectiveness of the policy.  Of course, this important political advantage becomes an economic disadvantage if it invites particularly harmful rent-seeking behavior.  Finally, environmental policy makers tend to think in terms of pollution quantities, not prices.

Experience with Carbon Pricing:  Emissions Coverage & Price in Implemented Initiatives

            There are some fifty carbon-pricing systems in operation worldwide, with equal numbers of carbon taxes and carbon cap-and-trade systems.  A quick comparison of these policies reveals two striking realities.  First, the highest carbon prices (the height of the bars in the figure below) are for carbon taxes (in norther Europe).  Second, the scope of coverage (the width of each bar in the figure) of cap-and-trade systems greatly exceeds that of carbon taxes.  Putting the two features (severity and scope) together, a reasonable measure of the relative importance of the policies is given by multiplying the carbon price (tax level or market price of allowances) by the tons of coverage, that is, the respective areas in the figure.  On this basis, it appears that political revealed preference has been weighted toward cap-and-trade (at least up until now).

Carbon Price & Emissions Coverage of Implemented Carbon-Pricing Initiatives

Which Has Worked Better – Experiences with Trading and Taxes

Based upon more than thirty years of experience with cap-and-trade systems, including but not limited to CO2 programs, lessons regarding the design and efficacy of these systems can be drawn.  In brief, there is empirical evidence for the following:  cap-and-trade has proven to be environmentally effective and economically cost-effective; downstream, sectoral programs have been common, but economy-wide upstream systems are feasible; transaction costs have been low to trivial; a robust market requires a cap below business-as-usual; banking has been exceptionally important, representing a large share of the gains from trade; price collars are very beneficial; free allocation of allowances fosters political support, with a likely transition to greater auctioning over time; competitiveness impacts can be mitigated with an output-based updating allocation; “complementary policies” are common, but in some cases can have perverse consequences, including no additional emissions reduction, an increase in aggregate costs, and suppressed allowance prices.

Turning to experiences with carbon taxes, two applications stand out.  First, there are the northern European carbon tax systems, initiated in the 1990s in Norway, Sweden, Denmark, and Finland.  Typically these were elements of broader energy and excise tax reform initiatives, and some are at the highest levels of any carbon-pricing regimes worldwide.  However, fiscal cushioning has been common for industries expressing concerns.  That said, these taxes have raised significant revenues to finance spending or to lower other tax rates, but unfortunately, there is little empirical evidence of their emissions impacts.

More striking is British Columbia’s carbon tax, initiated in 2008, which comes closest to that recommended by economists.  Currently, it is an upstream tax of $27/ton of CO2, but with important exemptions in place for key industries.  Importantly, 100% of tax revenue was originally refunded through general tax rate cuts, but over time, there has been more focus on tax cuts for specific sectors and locations.  Although there is some debate in the literature, it appears to have been effective in reducing emissions.

Empirical Evidence for Positive Assessment

Given that the normative differences between the two instruments are minimal, a key question becomes which instrument is more politically feasible, and which is more likely — in practice — to be well designed.  Based on experiences with cap-and-trade and carbon taxes, the relative masses in the figure above suggest that political revealed preference has favored the former.  Furthermore, after years of deliberation, China has chosen trading for its national program (although it appears to be a set of sectoral tradable performance standards, not a true, mass-based cap-and-trade system).  In addition, the new “Transportation and Climate Initiative” in the northeast United States was first proposed in terms of fuel taxes but is gravitating toward cap-and-trade.  Also, New Jersey is preparing to rejoin the Regional Greenhouse Gas Initiative, and Oregon is poised to enact an economy-wide CO2 cap-and-trade system this year.  On the other hand, Washington State has twice defeated a carbon tax.

But past may not be prologue.  The demonization of the Waxman-Markey trading system as “cap-and-tax” may have reduced the political advantage of cap-and-trade (that it can hide the costs).  And there is clearly increasing interest in a national carbon tax in the policy world, including several bills in Congress and the prominent Climate Leadership Council proposal.  On the other hand, the “Green New Deal” is silent about carbon-pricing of any kind.

It is worthwhile focusing on the political economy of the British Columbia carbon tax.  Its successful enactment has been attributed to “the confluence of political conditions ripe for carbon taxation”:  untapped hydroelectric potential; a strongly environmentalist electorate (as in the case of California’s move to cap-and-trade with Assembly Bill 32); a right-center government with trust from the business community (as with the George H.W. Bush administration’s SO2 allowance trading system in the Clean Air Act amendments of 1990); and a premier with institutional capacity to pursue personal policy preferences.  There has been increasing public support over time, due to the perception of emissions reductions without severe economic impacts, but political pressures have caused the evolution of the system from using revenues exclusively to cut distortionary taxes to greater use of tax cuts to favor specific sectors and regions.

Clearly, political pressures can drive up social costs with either type of carbon-pricing instrument.  On the one hand, politics may disfavor the auctioning of allowances in cap-and-trade systems, while, on the other hand, politics may disfavor cost-effective cuts of distortionary taxes in tax systems.

Does Either Carbon-Pricing Instrument Dominate in Normative or Positive Terms?

When carbon taxes and cap-and-trade are designed to be truly comparable, their characteristics and outcomes are similar, and in some cases fully equivalent (normatively), in terms of their:  emission reductions, abatement costs, revenue raising, costs to regulated firms, distributional impacts, and competitiveness effects.  But on some other dimensions, there can be real differences in performance.  The tax approach is favored by administrative requirements, interactions with complementary policies, and effects on carbon-price volatility; whereas cap-and-trade is favored by linkage with policies in other jurisdictions, and possibly by anticipated performance in the presence of uncertainty.  In the positive political economy domain, the evidence is also decidedly mixed.  Hence, there is not a strong case for the blanket superiority of either instrument.  Differences in design simply dominate differences between the instruments themselves.

Can Carbon-Pricing be Made More Politically Acceptable?

The track record of 50 carbon-pricing policies cited above should be contrasted with the 176 countries with renewable energy policies or energy efficiency standards, as well as another 110 national and sub-national jurisdictions with feed-in tariffs.  Hence, carbon pricing has not in general been the favored approach to climate change policy.  Why is this the case?  Survey and other evidence indicates that public perceptions – some of which are inaccurate – are primary factors behind aversion to carbon taxes:  “personal costs too great; policy is regressive; could damage economy; will not discourage carbon-intensive behavior; and government just want the revenues.”  So, one way to improve public acceptance could be through better information, that is, education.

But another way forward could be through judicious policy design, which may well depart from first-best design, including:  phasing in taxes/caps over time (which was effective in California and British Columbia); earmarking revenues from taxes/auctions to finance additional climate mitigation, in contrast with optimizing the system via cuts in distortionary taxes; and/or using revenues for fairness purposes, such as with lump-sum rebates or rebates targeted to low-income and other particularly burdened constituencies (a carbon tax with “carbon dividends” or a cap-and-trade system in the form of “cap-and-dividend”).

Has the Defeat of National CO2 Cap-and-Trade Initiatives Provided Openings for Carbon Tax Proposals?

Political polarization has decimated the key source of Congressional support for environmental/energy action, the political middle.  And the successful political battle against the Obama administration’s CO2 cap-and-trade legislation featured the effective demonization of that instrument as “cap-and-tax.”  Does the consequent reputational loss for cap-and-trade provide a meaningful opening for the other carbon-pricing instrument – a carbon tax?

It would seem that large budgetary deficits ought to increase the attraction of new sources of revenue, but existing carbon tax proposals have largely been revenue-neutral.  That said, it is surely true that there has been increased attention to carbon taxes from the “policy community,” with support coming not just from Democrats, but also from prominent Republican academic economists and former Republican high government officials.  But – finally – what about in the real political world of those currently holding elective office in the federal government?

It is presumably good news for carbon tax proposals that they are not “cap-and-trade.”  Perhaps that helps with the political messaging.  But if conservative opposition could tarnish cap-and-trade as “cap-and-tax,” surely it will not be difficult to label a tax as a tax!  And in addition to such opposition from the political right, it is – as of now – questionable whether the new left will want a carbon tax to be part of its “Green New Deal.”

Hence, in the short term, national carbon pricing of either type will likely continue to face an uphill battle.  Therefore, in addition to considering second-best carbon-pricing design (as I recommended above), economists can work productively to catch up with political realities by considering better designs of second-best non-pricing instruments, such as clean energy standards.

But, at some point the politics will change, and it is important to be ready, which is why – for the longer term – ongoing research on carbon-pricing is very much warranted, particularly if it can be carried out in the context of real-world politics, and focus on policies that are likely at some point to prove feasible.

Share