全球智库信息集成服务系统

Joseph E. Aldy and Richard Zeckhauser, both professors at the Harvard Kennedy School, document that the world’s single-minded focus on solely reducing emissions has failed to meaningfully address climate change. Drawing on their recent research, Aldy and Zeckhauser argue that adding adaptation and amelioration strategies to existing emissions mitigation efforts is necessary if extreme losses to climate change are to be avoided. The principal amelioration strategy they consider is solar radiation management (SRM).

With this Q&A article, Aldy and Zeckhauser answer the questions about solar radiation management that attendees asked at a recent related RFF Live webinar. (Available here are the slides from their presentation.) These questions could not be addressed live, due to time constraints. Some questions have been edited for clarity and length. In the Q&A below, event attendee questions appear in bold text, and responses from Joseph E. Aldy and Richard Zeckhauser appear in ordinary text. The questions are clustered under topic headings, which are bold and in italics.

Geoengineering Methods

Event attendee question: What are the specific geoengineering solutions that you’re suggesting?

Joseph E. Aldy and Richard Zeckhauser response: The primary solar geoengineering solution that we focus on would involve the injection of reflective particles—most likely sulfur particles—into the stratosphere by high-altitude airplanes.

What about “cool communities” strategies that provide nontrivial local cooling?

These strategies could play an important role in improving resilience to and adaptation to climate change. Their potential ability to offset some climate change risks should be explored.

Is marine cloud brightening a potential solution?

Marine cloud brightening could help mitigate warming, but it is a more local-scale measure than SRM. It should be explored as a localized variant of what we discussed in our presentation.

Have you compared the benefits of geoengineering with those of social engineering (i.e., reworking the socioeconomic system so that the system operates for survival, rather than material progress)?

If the first step of reorienting our societies for survival would be to shift away from fossil fuels as the energy foundation of the modern economy, then it's clear that political leaders have revealed that the costs of doing so are high. With the high costs of climate change and the high costs of mitigating emissions, then the relatively low costs of SRM—as we document in our paper—would appear to be an appealing addition to our efforts to reduce climate change risks.

However, let us be clear about the more general lesson from our paper and presentation. Dealing appropriately with climate change will require a multi-prong strategy, which clearly includes careful reflection on the objectives of society.

Emissions Mitigation

Some might argue that past performance is not a predictor of future behavior; in other words, just because the emissions mitigation approach has not been enough in the past, doesn’t mean we have to shift course and pursue SRM.

Unfortunately, we disagree. The failure to reduce greenhouse gas emissions in the 1990s serves as a good predictor for global efforts to cut emissions in the 2000s. And the failure to reduce greenhouse gas emissions in the 2000s serves as a good predictor for global efforts in the 2010s. The underlying mechanisms—large carbon-intensive capital stock and a political economy that makes meaningful progress challenging in most countries around the world—persist and will influence near-term behavior in confronting climate change.

Adding to these challenges, greenhouse gases decay very slowly, over centuries—not years or decades. Even if none were added to current levels, our planet would be in danger. Moreover, even if current concentrations were maintained, matters will get worse, given that the oceans, which function as heat sinks, have warmed.

You talked about how emissions mitigation is not solely dependent on the industrial sector. What do you think are some other major potential sources?

Given the challenge of eventually reducing global emissions to net zero (or negative), emissions mitigation will need to occur through every aspect of the economy, such as the industrial sector, buildings, electricity, and transportation. We will also need to address emissions associated with land use change, agricultural activities, consumption patterns, and all sectors that support such activities.

If carbon dioxide concentrations drive climate change, and they will be extremely high by the time we reach net-zero emissions, why don’t you include carbon dioxide removal as a fourth prong?

Carbon dioxide removal (CDR; also known as direct air capture) is another potential technological approach to addressing climate change. Our portfolio approach would include CDR. Because CDR mitigates climate change risk by reducing the net flow of emissions into the atmosphere, we classify it with other mitigation technologies.

I am a staffer in the Senate working on the Trillion Trees Initiative. What role do the panelists believe afforestation might play in SRM and decreasing climate risk, generally?

About 15 percent of net carbon dioxide emissions occur through land use change. Reducing this flow of emissions through afforestation can serve as a relatively low-cost way to cut emissions. It should be part of the broad portfolio of actions. And we should be taking vigorous policy steps immediately to deal with the vast deforestation that is currently taking place in various locales.

Ben Zycher, a chaired Fellow at the American Enterprise Institute, says the climate change models are not necessarily accurate. What is your assessment of the models, and what is your confidence in them?

We need to make decisions under uncertainty. Based on US National Climate Assessments, the Intergovernmental Panel on Climate Change assessment reports, and more, we have sufficient information to understand the rough level of risks we face in not taking action on climate change. As economists have known since at least 1974, uncertainty about irreversible changes in the natural environment increases the economic value of delaying such changes. Uncertainty in the science is not a credible excuse for policy inaction. Indeed, holding expected losses fixed, uncertainty about their magnitude raises the expected returns to policy action.

Risks

Can you address some of the risks of geoengineering?

We make clear in the paper that the estimated financial costs of implementing solar geoengineering interventions do not account for the potential unintended consequences. These unknowns merit significant research. They also serve as additional motivation for our “act-learn-act” approach.

How can you advocate SRM without first addressing all of its potential risks?

SRM may have unintended consequences. These risks may be serious, and they merit significant scientific investigation. Serious risks also arise due to climate change and inadequate emissions mitigation over the past three decades under the multilateral climate governance regime (the UN Framework Convention on Climate Change). Given the current understanding of climate risks—and the technological and political challenges to rapid, global decarbonization—we are concerned that failing to invest in SRM science, engineering, and policy will impose grave, catastrophic risks to people all around the world. Moreover, we recognize that some research investigations will require experiments, and those experiments will not be without risk. However, we believe that the risks of not experimenting, and therefore not learning about SRM, are more consequential.

A recent run of a major climate projection model tested the scenario of a large increase in solar-reflective aerosols. This scenario led to a reduction in the temperature difference between the tropics and the polar regions, causing weaker extratropical storm tracks, dwindling winter storms, and drought. Can we risk experimenting with this kind of geoengineering if these types of ecological impacts could result?

This is a risk-risk question. We are currently experimenting by pumping greenhouse gases into the atmosphere. We are not doing nearly enough to offset this risk through emissions mitigation, and even the most vigorous conceivable efforts would be insufficient to avoid expected losses that are monumental. Given the scope of catastrophic impacts, now is the time to seriously explore other approaches to reducing these risks, such as SRM.

Ocean acidification has significant impacts on food in many countries, with geographic incidence that probably differs from that of warming. If geoengineering reduces mitigation efforts, the rate of ocean acidification would increase. Should this effect be considered? If so, how?

First, this is a critical point that helps illustrate the importance of pushing on all three prongs—mitigation, adaptation, and amelioration. This point also highlights the importance of understanding the distribution of climate change impacts—and the distribution of SRM impacts. There will be impacts of climate change that SRM will not remedy, such as ocean acidification; thus, mitigation and adaptation must also be vigorously pursued. SRM itself is not expected to have any impact on ocean acidification and would not necessarily increase the risk of ocean acidification, as we believe that SRM may galvanize action to reduce emissions.

Second, let us make a general point. We are fiercely opposed to purposefully ignoring any critical consequences. Carefully attending to such consequences is a central implication of our two central themes: a multi-prong approach and an act-learn-act strategy.

You're evaluating the effects of geoengineering in terms of one outcome variable: temperature. But what about the effects of geoengineering on the intensity of sunlight? To weigh the impacts and risks of geoengineering, we need to understand the risks of reduced sunlight to agriculture, the food supply, and the broader global ecology.

Your point is a valuable one. We focus on temperature in part because it is the focus of international agreements (e.g., the Paris Agreement goal to limit warming to well below 2°C) and in part to maintain simplicity in communicating our key points. We support a broad portfolio of policy actions that will mitigate climate change risk, and we recognize that climate change risk addresses many dimensions beyond temperature, such as ocean acidification, changing precipitation patterns, and subsequent changes to natural and human systems. Consistent with what we say above about considering the broad array of consequences, assessments of SRM impacts should look beyond temperature and consider its impacts on these other dimensions.

What prevents a rogue actor from trying out geoengineering and injecting particulates in the stratosphere?

The governance literature has raised questions about the potential of "rogue actors." Continued research to improve our understanding of the climatic impacts of various solar geoengineering interventions would help policymakers assess the potential risks of rogue actors, and hence the need for the management and deterrence of rogue actors. Such research would also help inform the development and implementation of a coordinated management of SRM.

We should also be explicit about a danger inherent in your question. It takes time to work out international governance regimes, as experience with international agreements on mitigation makes clear. It would be exceedingly imprudent to delay substantial research efforts on SRM until an agreement on governance was in place, or even until a clear path to an agreement was identifiable.

It seems like a major barrier to any kind of change is a general inability to buy into and implement technology and methods toward any of the three prongs that change the economic status quo. How can advocates be supported?

We recognize that a variety of ways exist to support advocates. As researchers, we believe the most effective way the two of us can support advocates is by providing rigorous analysis and evidence that will enable advocates to make the case for better, more effective climate change policy. This will certainly not be sufficient, but it will very likely be necessary.

Implementation

I like your advocacy of an "act-learn-act" approach, but how do you see such an approach working for SRM, in which the "learning" could involve planetary-scale impacts? How can SRM be best managed to maximize the net gains from learning?

The development of SRM will occur in stages, and we will need to plan for and implement rigorous evaluation at each stage. Any experimental elements of policy will begin with very small steps. In designing the policy apparatus governing SRM—from the pilot stage to full-scale implementation—we should integrate data collection, analysis, and publication protocols. The transparency of the data and analysis will be critical, both to enable third-party verification and to enhance the legitimacy of this approach. The process by which such learning is incorporated into future policymaking will also need to be developed to further support the credibility of this prong.

Regarding implementation: suppose Pareto optimality is reached at a level of SRM well below what would be individually optimal for most countries. What happens then?

Just as in the emissions mitigation context—in which countries will differ in both their benefits and costs from reducing greenhouse gas emissions—there will likely be differences in impacts across the planet from SRM interventions. The distribution of the impacts of climate change and the distribution of the impacts of SRM to offset climate change will each influence the political economy and international relations dimensions of SRM policy design and implementation. It will be important for future research to build on a nascent literature characterizing the distribution of SRM impacts.

Finally, let us make a technical economics point. In the absence of massive side payments, there will be a broad range of Pareto optimal levels for climate and SRM. In light of that prediction, first achieving Pareto optimality is hardly a “Eureka” moment.

Isn't there a budget constraint? If society has a limited willingness to pay for addressing the climate emergency, won’t investing in this third prong mean less resources for the other prongs?

We posit in our paper that SRM could galvanize the public to support more ambitious emissions mitigation policies. Suppose that the typical person believes climate change is serious, but also believes that other important issues need to be addressed now. (This is generally consistent with recent polling.) If the government announces that it has to begin injecting sulfur aerosols into the upper atmosphere to cool the planet, some people—including the typical person—may be shocked that the situation has become so dire that the government must take such an awful action. This "awful-action alert" could make the need to reduce emissions more salient and immediate and drive greater ambition in future climate change policy. It's also important to recognize that the apparent costs of SRM are orders of magnitude less than for emissions mitigation and adaptation. If climate change risk mitigation were a zero-sum game, then funding SRM would have only a negligible impact on mitigation and adaptation spending.

We should also remark that concern for budgets is far from a compelling argument against SRM. Let us observe what analysts widely agree on and that we point out in our paper: The expected costs of even the most ambitious SRM efforts would be minimal, relative to the expected losses otherwise from climate change, or even relative to the expected costs of optimal mitigation and adaptation policies.

Within this third prong of geoengineering, what is the earliest development timescale you could foresee? Would concerns over governance and public opinion further shape this time range?

The critical factors in the time required to develop SRM will likely be various political, behavioral, and ethical considerations, rather than engineering development timelines. It is difficult to speculate how long it would take to address these non-engineering factors—such as the status quo bias, the political reservations associated with acts of commission versus acts of omission, and government coordination—in order to be ready to implement SRM. If engineering were the sole constraint, SRM should be available for implementation within a decade.

Due to pandemic-related delays, the National Academies report that a research program and governance recommendations for solar geoengineering will now most likely be delayed, perhaps until December.

Good point, and something to keep in mind. Given the urgency of the climate-change problem, we are eager to see the report.

How do you respond to the criticism that SRM measures must be continued indefinitely to maintain the effect and that stopping certain SRM measures would lead to a dramatic atmospheric “rebound”?

We emphasize the importance of pursuing each of our three prongs—mitigation, adaptation, and amelioration. Solar geoengineering buys the world some time—and helps avoid the worst of the catastrophic impacts of climate change—so that future mitigation technologies can eliminate net global greenhouse gas emissions. In the long run, negative emissions technologies (e.g., bioenergy with carbon capture and storage, direct air capture) have the potential to reduce the atmospheric stock of greenhouse gases. If and when those technologies prove feasible, we may see a reduced need for eternal SRM interventions, thus precluding the risk of an abrupt and unexpected stoppage of SRM. We view SRM as a necessary instrument to help the planet get through the next few decades. Our crystal ball on future policies goes cloudy beyond that time frame.

Are any countries already applying the three-pronged approach?

A few national governments, including the US government, have made modest research efforts and commissioned studies. Australia is experimenting with marine cloud brightening, a more localized and lower-altitude variant of SRM. Otherwise, no countries are currently pursuing all three prongs beyond the discussion level. However, we should observe that discussions of the three prongs in many nations are now much more vigorous than they were merely five years ago.

How will SRM be governed?

The question of governance over the development and deployment of SRM technologies is an important emerging question in both academic and policy circles. Our analysis abstracts from the governance issues. We do recognize the evolution of governance in the multilateral UN climate change negotiations, from the 1992 UN Framework Convention on Climate Change (UNFCCC) through the 1997 Kyoto Protocol, the 2009 Copenhagen Accord, and most recently the 2015 Paris Agreement. These governance efforts initially focused narrowly on mitigation, primarily from Organisation for Economic Co-operation and Development (OECD) and former Soviet bloc nations. Over time, multilateral governance has expanded to emissions mitigation among all countries. We could imagine that, as solar geoengineering becomes more prominent in the public debate, national governments for major economies will get involved, and multilateral negotiations will get under way. As we suggest above, we expect international governance arrangements to experience a slow evolution.

What would you see as next steps for considering the technical and governance options, risks, and uncertainties of SRM?

We will need ambitious research on SRM that attends to a broad array of issues and calls on many disciplines. Scientific and engineering efforts have been initiated in many quarters but are still at very modest levels. We expect related research to burgeon. Initial efforts to think through governance issues will likewise benefit from continued research by social scientists as well as individuals with professional experience and expertise with international governance issues.

Relevant to our presentation, specifically, and growing attention from RFF and similar research organizations more generally, we believe that a more ambitious research program drawing from economics and decision sciences will be critical. The outputs could move us from our preliminary understanding of the potential for and risks associated with SRM. We are continuing our research, applying tools from these disciplines, to better understand SRM through the Harvard Solar Geoengineering Program and the Resources for the Future Solar Radiation Management Dialogue. We expect vigorous parallel efforts from readers of this article and from many dozens of others who feel, as we do, that SRM is a critical policy and an intellectually intriguing issue.

Available here are the slides from Aldy and Zeckhauser's recent RFF Live presentation, "Three Prongs for Prudent Climate Policy."