Carbon pricing is firmly entrenched as the go-to climate policy for economists, yet many with training in other sciences and social sciences remain skeptical. This commentary surveys four recent papers by writers who are fully committed to forceful climate action but who have little if any enthusiasm for carbon pricing. Together, the papers offer important insights into disagreements about the proper role of pricing and point the way to common ground.
“Carbon pricing has been tried and the results are disappointing”
The writers in question do not reject the concept of carbon pricing altogether, but they see it as having less practical value than most economists do. As Jessica Green puts it in a tweet announcing her paper, “of the policy tools in the carbon toolbox, carbon pricing is the tiny flathead screwdriver used to fix glasses.”
Green supports her conclusion with a review of 37 ex post studies of the effects of carbon pricing, mostly from Europe and North America. Based on those studies, she finds aggregate emission reductions from carbon pricing to be “generally between 0 percent and 2 percent per year,” a degree of effectiveness that she views as wholly incommensurate with the size of the problem.
However, the carbon pricing schemes covered in Green’s paper are, for the most part, very modest in size and scope. Her own summary table is drawn in a way that makes it difficult to extract averages, but a similar study by Erik Haites provides some relevant numbers. His summary table lists carbon taxes in 18 countries that have an average level of $12 per ton of CO2, each covering an average of 46 percent of its country’s emissions, and 17 carbon trading schemes that have an average price of less than $8, each covering an average of 48 percent of emissions. The idea that placing a low price on carbon emissions from a limited range of sources does not accomplish much is unlikely to surprise many economists.
Green points to leakages as a reason that existing carbon pricing schemes have not always been effective. By leakages, she means the tendency for polluters to move production to jurisdictions where carbon prices are lower or nonexistent. Leakages are an important issue, but for three reasons, I am not sure they constitute a good argument against carbon pricing in general.
First, in an important sense, leakages are proof that carbon pricing works. Some critics claim that prices are too weak a nudge actually to change behavior, but leakages show that polluters will undertake costly measures, such as moving production operations or upending established supply chains, to avoid paying for pollution. The fact that leakages can be observed even where carbon prices are quite low underlines the sensitivity of behavior to market signals.
Second, the leakage problem is not unique to carbon pricing. Whether antipollution measures take the form of price signals or regulatory restrictions, polluters will move to avoid them, if they can. Leakages are a problem for national or regional climate policies in general, not specifically an argument against pricing.
Third, although it is a truism to say that decarbonization efforts work better if more countries participate, that effect is even more pronounced when we take leakages into account. Since opportunities for leakages diminish as new countries are added, the effect of adding each new country to the carbon pricing country is greater than the one before.
Economists and legal experts who work on the design of carbon pricing systems are well aware of the problem of leakages. Border adjustments are the most common recommendation for dealing with them. A border adjustment works by applying a tax on imports and a rebate on exports. (Such adjustments are already in wide use by countries that have value-added taxes.) Some designs include incentives for trading partners to tighten their emission control regimes. (See this analysis by Shuting Pomerleau for a brief but thorough discussion of economic and regulatory aspects of border adjustments.)
Deep decarbonization vs. efficiency
Skeptics argue that despite its theoretical elegance, carbon pricing is, in practice, incapable of achieving transformative change. Green (2021) thinks past efforts at pricing have at most produced “incremental solutions” that, “though useful on the margins, fall well short of the societal transformations identified by decarbonization scholars.” Rosenbloom et al. (2020) “question whether efficiency should be an overriding priority of climate policy.” Patt and Lilliestam (2018) argue that carbon pricing “made sense
as our primary tool against climate change when our climate policy ambitions were limited” but no
longer do so when the goal is complete transformation of the world’s energy system. Tvinnereim and
Mehling (2018) argue that pricing is “not good enough when the goal is to eliminate virtually all
emissions in the short to medium term.”
I think these objections are misguided. Deep decarbonization and efficiency are not alternatives. On the contrary, ignoring efficiency makes transformative change harder, not easier.
For an object lesson in the importance of efficiency, contrast the Israeli and Soviet approaches to agriculture in an arid climate. Efficiency has been the hallmark of the remarkable Israeli success. Its signature technology has been the widespread use of drip irrigation to make the best use of limited water supplies. Drip irrigation works together with a whole suite of complementary policies, such as reusing treated sewage for farming, finding and fixing leaks early, engineering crops to thrive in onerous conditions, discouraging gardening, and making efficient toilets mandatory. The policy mix also includes putting a price on water to discourage waste. Although individually, some of these policies only incremental contributions, together they have been truly transformative.
Contrast that with the Soviet approach to growing cotton in Central Asia. From the earliest days of the revolution, the Soviet approach to every problem was to mobilize vast resources on a clear goal, and damn the cost. In the 1960s, the Soviet planners put that strategy to work to grow cotton in the desert. Their first step was to dam the two great rivers that flowed through the region, pouring millions of tons of concrete to channel the water through a network of open canals and flooding the cotton fields. The result: Lots of water wasted, lots of low-quality cotton, for a while, followed by salinization of the soil and diminished yields. This project was transformative, too, but not in a good way. It caused the almost complete disappearance of the Aral Sea, which had once regulated the local climate, bringing hotter, dryer weather, desertification, dust storms, and the death of a once-rich fishery.
Think of carbon pricing as the drip irrigation of climate policy. Unspectacular and efficient, it works silently and out of sight, complementing other policies to produce a sustainable solution to a complex challenge.
And what of the concern that carbon pricing, however efficient, encourages only incremental changes? Patt and Lilliestam put it this way: “Carbon taxes stimulate a search for low-hanging fruit. That ceases to matter when we know we must eventually pick all of the apples on the tree. Metaphorically we need a ladder.” The ladder they have in mind is a mix of green industrial policy, regulations, and subsidies for low-carbon technologies.
My reply is that transformative technologies are all well and good, but they are no reason to disdain the low-hanging fruit. Yes, we eventually want to pick all the apples on the tree, but Patt and Lilliestam’s metaphor is incomplete. Suppose you buy a house that has an apple tree in the backyard. The first day after you move in, you notice that the apples are ripe, but you can only reach the lowest branches. If you call Home Depot and learn they can deliver a ladder tomorrow, you may as well wait until it gets there and pick all the apples at one go. But suppose, instead, that Home Depot is out of ladders and won’t have any until next year. What then? You make a deposit and reserve a ladder for delivery next fall. Then you go out and pick all the apples you can reach now. Why let them rot on the tree?
The delayed ladder scenario is a better depiction of where we are with climate change. Even the most ambitious programs for deep decarbonization do not aim to get to net zero in less than 30 years or so. (See the recent Net Zero America proposal from Princeton University for a detailed example.) Meanwhile, because of the persistent nature of CO2, even marginal decreases in emissions in the near term will mean a lower total concentration at the time emissions finally fall to zero.
All of the papers reviewed here emphasize that carbon pricing faces daunting political resistance.
Rosenbloom et al.: “Carbon pricing has also attracted political resistance among the broader public, as it is perceived to challenge long-standing practices and livelihoods, such as car-based and suburban lifestyles. Political opponents have been quick to exploit these cleavages.”
Patt and Lilliestam: “We disagree about the need for higher carbon prices, and indeed view carbon taxes and permit fees as a dangerous political distraction.”
Green: “Carbon pricing has proven a controversial policy both domestically and internationally” (citing political resistance in the United States, Canada, Australia, France, and elsewhere). “In short, it is not at all evident that limited political capital should be spent on carbon pricing when other efforts at mitigation may offer more reductions for less political controversy.”
Tvinnereim and Mehling: “Experience has shown that carbon pricing faces considerable political economy constraints … Research has therefore suggested that instruments other than carbon pricing are better able to build coalitions of support.”
They are right, of course, that carbon pricing has been a hard sell politically. As a case in point, Green notes that “Washington state had two ballot initiatives proposing a carbon tax in 2016 and 2018; both failed following heavy investments from fossil fuel industry to defeat them.” But it is disingenuous for climate activists to blame the defeat of those referendums entirely on the fossil fuel industry or suburbanites determined to protect their car-based lifestyles. The fact is that Washington’s green organizations either opposed the referendums outright or gave only lukewarm support, especially in 2016, when the chances of passage were best. (See two long analyses for Vox by David Roberts for a detailed post mortem on the 2016 and 2018 efforts.)
What is more, political pushback is not limited to price-based policies. Almost any kind of climate action gets its share, and from all sides. On the right, we saw that the Trump administration’s withdrawal from the Paris Agreement was supported by a large share of conservative Republicans. On the green side of the spectrum, environmentalists in Oregon and Idaho, seeing threats to wildlife and the beauty of the landscape, sued to stop a new transmission line intended to help utilities meet a pledge for 100 percent clean power. Still other regulatory measures draw pushback that is hard to classify as right or left. Try reading some of the hundreds of thousands of Google hits for “I hate low-flow showerheads!”
In my view, the lesson to be drawn here is not that we should rule out all price-based carbon policies or all regulatory measures. Rather, the lesson is that we are still in the early days when it comes to designing policies that are both economically effective and politically feasible. Conceptually, we should start with the policies that generate the least political pushback per ton of carbon saved, but we don’t know exactly what those are. We are not starting from zero, though. Lots of people have looked at pieces of the problem and offered promising ideas.
Consider, for example, the large literature on CAFE standards vs. carbon pricing as alternative ways of reducing automobile emissions. (See here for an overview.) CAFE standards force changes in vehicle design that push up the price of a new car, while carbon pricing raises the price of gasoline. The supposed political advantage of CAFE standards is lower salience – people buy a new car only every few years, whereas they buy gas every few days. Yet, dollar-for-dollar, higher fuel prices save more emissions. The reason stems from what economists call the “rebound effect.” Once you buy a fuel-efficient car, you have an incentive to drive it even more miles than you drove your old gas hog. The extra miles cancel out part of your new car’s fuel efficiency. Higher gas prices, in contrast, incentivize both the purchase of thrifty cars and less driving. On balance, salience or no, it is not obvious that CAFE standards generate less political resistance per unit of emission reduction than do higher fuel prices.
Another widely discussed idea for mitigating political resistance is to use some or all of the revenue from pricing schemes to compensate key constituencies. The Citizens’ Climate Lobby, for example, proposes to rebate the entire proceeds of a carbon tax equally to everyone. Other plans limit payouts more narrowly to low-income households, which tend to spend a higher share of their budgets on home heating, gasoline, and other types of energy. Targeting rebates on low-income households would leave more revenue available for other policies that would help build political support, such as job-creating infrastructure projects or research to help industries transition more smoothly to a zero-emissions future. (See here for a full discussion of the compensation issue, with charts and sources.)
Proper policy sequencing is another idea for reducing political resistance to carbon pricing. Jonas Meckling, Thomas Sterner, and Gernot Wagner (2017) make a case for beginning with green industrial policies such as support for research and development, subsidies, loan guarantees, and direct mandates. Such policies, even on a modest scale, would build support for adding carbon pricing to the mix later. One reason is that they would build a constituency among producers of low-carbon goods and services, who would then back carbon prices to boost demand for their products. Another reason is that early investments in research would help low-carbon technologies move “up the learning curve and down the cost curve,” so that the incentive effects of carbon prices, when they were introduced, would produce greater gains at a lower cost. Early investment in infrastructure would also help. For example, a price incentive for wind and solar power would become more effective once a transmission network was in place to move it from where it was most cheaply generated to where it was most in demand.
Danny Cullenward and David Victor, authors of Making Climate Policy Work, suggest a sectoral variant of policy sequencing. Recognizing both the theoretical appeal of carbon pricing and the reality of political resistance, they suggest that pricing should be introduced first in sectors like electric power, where political resistance is relatively low. Other sectors (transportation, for example), “where the public is inordinately sensitive to sticker prices,” should, in their view, be handled with nonprice strategies, at least initially.
The case for a mixed strategy
Theoretical discussions are sometimes taken to suggest that climate policy should consist solely of setting the right price for carbon (possibly a very high one) and then letting the market do the rest. In practice, however, there is a strong case for a mixed strategy that combines green industrial policy with carbon pricing. That case is especially strong when political realities are considered and when the goal is deep decarbonization at the earliest feasible date.
Of the four papers that have been the focus of this commentary, Tvinnereim and Mehling come down most strongly in favor of a mixed strategy. As they put it,
This article does not argue against using carbon pricing as a mitigation instrument, but rather to caution against placing too much faith in a theoretically compelling policy instrument. …
Over time, as targeted policies increase tolerance and support for higher carbon prices, it may eventually unleash more of its potential as a cost-eﬀective tool of climate mitigation. Finally, the presence of a price on carbon emissions can send an omnibus signal that policy makers are serious about tackling global warming, and that long-term investments need to be made with expectations of future carbon constraints in mind (pp. 187-188).
Rosenbloom et al. are more skeptical about carbon pricing, but they, too, leave room for a mix-and-match approach:
In summary, the dominant logic of contemporary climate policy, in which carbon pricing is the central policy response, is deeply flawed. Given the aforementioned shortcomings, carbon pricing should not be the primary policy strategy to combat climate change. Instead, carbon pricing should be used as part of a policy mix that promotes innovation and decline, accounts for political dynamics, varies between sectors and over time, and aims at profound system change (p. 8668).
An intriguing new working paper by Emil G. Dimanchev and Christopher R. Knittel of MIT clarifies the case for a mixed policy. Those authors, like many before them, find that a pure carbon pricing policy is more cost-effective than one based on standards alone. However, using both theory and modeling, they add an important twist. They find that beginning from complete reliance on standards, the benefit of adding a carbon price to the mix is large at first, and then gradually diminishes. That, they say, demonstrates that even a modest carbon price can have large efficiency benefits. Whereas Green, citing Cullenward and Victor, argues that regulations, not pricing, “do the majority of the work” in mixed policy regimes, Dimanchev and Knittel stand that view on its head. Their work suggests that the modest pricing component is exactly what allows the regulatory part of existing mixed regimes to work effectively.
In view of these arguments in favor of a mixed strategy, I think other authors go too far in their skepticism. I do not agree with Green’s view that carbon pricing is only a “tiny screwdriver.” In my view, she inadequately accounts for the synergistic interactions between carbon pricing, green industrial policies, and performance standards. I also strongly disagree with Patt and Lilliestam’s characterization of carbon pricing as nothing but a “dangerous political distraction.” On the contrary, it seems to me that the need for wholesale transition is a reason for enthusiastically supporting the use of every weapon in the decarbonization arsenal.
I do agree with Patt and Lilliestam in one particular, however. Without going so far as to say that carbon pricing is already “outdated,” as they do, I agree that one day we will be able to look at it as such. That point will be reached when zero-carbon technologies, not just in the energy sector but everywhere, are so cheap that they can outcompete carbon-intensive methods even if the latter are not required to pay for the harm done by emissions. At that point, the revenue from a carbon tax would fall to zero, as would the market price of emission permits, if the pricing strategy were based on cap-and-trade.
In that sense, I would like to see carbon pricing implemented so effectively, with such a strong suite of supporting polices, that it becomes outdated as soon as possible.
 The focal papers are purposely chosen from academic rather than popular publications. They include Jessica F. Green, “Does carbon pricing reduce emissions? A review of ex-post analyses,” Environmental Research Letters (2021) (accepted manuscript); Anthony Patt and Johan Lilliestam, “The case against carbon prices,” Joule 2, no. 12 (2018); Daniel Rosenbloom, “Why carbon pricing is not sufficient to mitigate climate change—and how ‘sustainability transition policy’ can help,” PNAS 117, no. 16 (2020); Endre Tvinnereim and Michael Mehling, “Carbon pricing and decarbonization,” Energy Policy 121 (2018).
 Some of the studies that Green examines report the effects of carbon prices as declines relative to a rising business-as-usual trajectory, while others report absolute decreases from the level of emissions when the price was implemented. It is not clear from the paper which type of reduction the 0 to 2 percent range refers to. In private correspondence, Green writes that “The 0-2 figure captures the range of both types of reductions.” The “generally” seems intended to allow for the fact that a few of the studies she considers report considerably larger decreases relative to business as usual or synthetic baselines.
 The incremental effects of carbon pricing may not be individually transformative, but they are not trivial, either. For example, the International Energy Agency estimates that from 2008 to 2018, fuel-switching from coal to gas alone, just in the United States, saved 250 million tons of carbon emissions. That is equivalent to putting 100 million zero-carbon electric vehicles on the road. Those gains were achieved without carbon pricing, but as Lilliestam, Patt, and Bersalli find in a 2020 paper, carbon pricing does increase the rate of fuel-switching even in cases where it does not stimulate transformative technological change. Presumably, then, the rate of fuel switching in the United States would have been even faster under a pricing regime. Multiply the IEA findings by additional countries, and add in other price-induced behaviors like installing rooftop solar panels, improving home insulation, or switching unnecessary business trips to Zoom, and the savings begin to add up.
 This commentary has focused on papers published in the scientific literature. In popular writings, carbon pricing skeptics have expressed themselves even more forcefully. For example, writing for Jacobin, a magazine offering socialist perspectives on politics and economics, Green says that carbon pricing has done more harm than good and contributed to the polarization of climate policy. In a long piece for the Boston Review, Matto Mildenberger and Leah C. Stokes call carbon pricing a “political disaster.”