What economists just don’t get about climate change

The case against cheap climate policy

Economic studies on climate policies tend to ask which policies cut the next ton of emissions most cheaply. (Illustration by The Argument using ChatGPT; photo by Jon Cherry/Getty Images)

Economists working on climate change are used to being ignored. We often blame the political right for failing to recognize the benefits of climate actions and the political left for failing to recognize the costs. And both deserve some blame.

The Trump administration has gutted federal climate policy. Meanwhile, the Biden administration’s signature climate law sidelined the market-based measures, like carbon pricing, that economists have long recommended.

But we should also point our fingers at the mirror. The trouble is that economists and climate advocates are often talking past each other. Economic studies on climate policies tend to ask which policies cut the next ton of emissions most cheaply. But most climate policy advocates are after something else: system-level changes that will put the country (or, preferably, the world) on an entirely different path, where net greenhouse gas emissions bend toward zero over the coming decades.

The case for costlier, system-level interventions stems from the physics of climate change. Carbon dioxide is a stock pollutant, meaning that a large share of it lingers in the atmosphere, causing global warming for centuries after it’s emitted. What determines how hot the planet gets is the total that accumulates over time.

The goal is not to reduce current emissions a little but instead to reduce future emissions a lot. The difference between these approaches would not matter much if cutting emissions a lot were just cutting emissions a little, many times over. But it usually isn’t. The cheapest way to reduce emissions in the near term and the cheapest way to transform energy systems over decades likely entail entirely different actions.

The problem, from the perspective of climate advocates, is that economic analyses don’t often come out in favor of systems-level interventions.

Consider the electricity system. Cheap emissions cuts can often come from switching from one fossil fuel generation source to a somewhat cleaner source — in particular from coal to natural gas. Around two-thirds of the considerable decarbonization of the U.S. power sector that occurred between 2005 and 2019 was caused by cheap gas replacing dirtier coal.

But that strategy only goes so far. These relatively easy emissions reductions are exhausted once the coal generation is gone and what’s left is gas-fired generation, and the long-lived infrastructure that remains will continue to be a major source of carbon dioxide; in other words, a policy that achieved substantial near-term decarbonization relatively cost-effectively did little to transform the power sector into a fully carbon-free system.

Deeper emissions reductions require a portfolio of carbon-free sources that can deliver power when and where it’s needed. From that perspective, the most cost-effective strategies in recent decades may have been early-stage investments in clean technologies, like solar and batteries, whose falling costs and improved performance are now making emissions reductions cheaper throughout the world.

This is old news for anyone familiar with decarbonization strategies. For decades, economists and other climate policy analysts have shown that the least-cost path to deep decarbonization often means paying for some high-cost, slow-to-move options early — like clean steel production or carbon-free jet fuel — because skipping them makes the long-run target more expensive or simply unreachable.

This insight shaped the design of measures within Joe Biden’s Inflation Reduction Act and Infrastructure Investment and Jobs Act, which focused as much on long-term transformation as on immediate emissions cuts. Yet the cost figures that dominate policy evaluations and public debates still focus on the near-term costs of reducing the next ton of emissions.

It’s time for that to change.

Cost-effective climate actions aren’t always cheap

Climate policies are commonly assessed by comparing them with a carbon price. Carbon prices encourage emissions cuts from wherever and however they can be achieved at a relatively low cost, so nearly any other policy looks expensive compared with them.

Back to the example of coal-to-gas shifting, a coal power plant emits roughly one metric ton of CO2 per megawatt-hour, about double a natural gas plant’s half-ton, so a carbon price raises the cost of running a coal plant more than it raises the cost of a natural gas plant. Even a small carbon price is often enough to flip the order in which grid operators dispatch generators to serve customers at the lowest cost, so the higher-emitting coal plants run less, and the government brings in revenue in the process.

However, carbon prices are limited tools for driving large emissions cuts cost-effectively. Vehicle emissions are a good illustration. Some can be cut cheaply from the tailpipe: Tightening fuel-economy standards wrings more miles from each gallon, and a modest carbon price nudges people to drive a bit less, delivering real but small reductions.

But large-scale vehicle emissions reductions are a different project that requires getting Americans to switch from internal combustion engine vehicles to electric or other alternative-fuel vehicles.

In principle, a carbon price could drive that shift, too, by raising gasoline and diesel prices so high that drivers switch to EVs. But, in addition to the political challenges of imposing high carbon prices, that approach is likely to cost more than other strategies. Alternatives could rely somewhat less on gasoline price hikes and somewhat more on policies like zero-emission vehicle standards or “feebates” that encourage the purchase of cleaner vehicles.

To see why, imagine two extremes. In the first, EVs are near-perfect substitutes for gasoline vehicles, identical to drivers in virtually every way. Then, a zero-emission vehicle standard that gradually mandates more EV sales costs almost nothing, because buyers are indifferent. In contrast, a carbon price could shift those new-car decisions only by jacking up gasoline prices for the entire existing fleet, taxing millions of drivers who own gasoline-powered cars they won’t replace for years, all to influence the far smaller group choosing a new vehicle today.

In the second extreme, there is no viable alternative to the gasoline engine at all. Now the standard would be impossible to meet, while the carbon price would at least capture the changes from nudging people to drive less and to buy more efficient vehicles.

Reality sits in between these two extremes, which is why both tools are useful. But the closer EVs come to substituting for gasoline cars, the more cost-effective standards become as a tool for achieving large emissions reductions.

A 2022 study of Canada’s vehicle market bears this out. It found the cheapest strategy capable of achieving the country’s vehicle emissions goals involved a zero-emission vehicle standard, even though the costs of this policy were estimated at hundreds of dollars per ton. Comparing that with the low costs of Canada’s carbon price would mislead, since the carbon price would not have delivered the same transformation of the fleet. (Canada got rid of its carbon price on vehicle emissions last year.)

Misleading policy verdicts

To weigh the costs of a climate policy against its benefits, you need a way to value those benefits: a dollar figure for the damage avoided by keeping the next ton of carbon dioxide out of the atmosphere. In theory, that enables a comparison of the harm caused by a ton of emissions with the cost of preventing that ton — and therefore a verdict on whether the policy in question is worth doing.

The trouble is that real-world calculations of this figure — referred to as the social cost of carbon — are not meaningful estimates of the benefits of reducing emissions. While analyses of the harms climate change will cause and how people will adapt to them are valuable, aggregated estimates of global climate damages over centuries lie far beyond the capabilities of our modeling tools.

First, the models put dollar figures on the harms economists know how to measure, like how mortality rates and energy bills rise as the planet warms, and leave out the rest, with no way of knowing whether what they have captured is one-tenth of the total damage or most of it.

Second, they use historical data as a guide for how people will adapt to a climate with no precedent in human history, even though economists have long recognized, in other settings, that the empirical relationships estimated under the conditions we have observed are a poor guide to conditions we have not.

Third, analysts make a series of moral judgments about how much to value people’s well-being decades or centuries from now, and how much weight to give to avoiding future risks and the uneven distribution of harms.

To produce a single estimate of total climate damages, the models have to settle these difficult questions one way or another. They do so in ways that keep the math manageable — not because economics can say which values are the right ones.

The result is not a fuzzy estimate of something real. It is a precise estimate of something the method cannot actually produce. It is a bit like gauging the relative size of Greenland from a flat map, except that here we have little sense of the size or the direction of the distortion.

It’s expensive to do hard things.

The upshot is that while weighing the potential costs and benefits of climate measures is essential to sound policy, empirical estimates are used in misleading ways. The near-term costs of transformational policies are compared with the costs of more incremental policies like a carbon price, or against a social cost of carbon, as if these are meaningful comparisons. They usually aren’t.

Consider the example John Thomason’s recent essay offers on renewable portfolio standards (RPS), policies many states use to require that a gradually rising share of electricity come from resources like wind and solar. He cites a prominent study finding RPS costs $86 to $211 per ton of avoided emissions and contrasts that with the much lower costs of the Regional Greenhouse Gas Initiative (RGGI), a power-sector-carbon-pricing policy that operates across northeastern states.

Indeed, even at their highest levels in the history of the policy, recent RGGI carbon prices around $35 per ton are a bargain for CO2 reductions (and the costs of the policy are smaller once the productive use of revenue is considered). That’s by design: RGGI’s rules are meant to keep prices at levels that cause only small increases in electricity rates. But it also means the carbon prices cannot rise high enough to encourage much additional carbon-free generation.

A $35-per-ton carbon price adds only 1 or 2 cents per kilowatt-hour to the costs of burning natural gas, far short of the incentive needed to encourage clean electricity sources to grow at scale in the region.

So, especially in states with no remaining coal to shift away from, RGGI is an instrument that can efficiently reduce emissions a little, but it cannot reduce emissions a lot. One reason RPS policies are more expensive is that they are trying to do the harder thing — build carbon-free generation — rather than the easier changes that RGGI can provoke.

Now, imagine you are a policymaker in a New England state where RGGI, an RPS, and a host of other clean energy measures operate. You want your state to be a climate leader, but you do not currently have a policy portfolio that is aligned with your ambitious long-term emissions goals, and high electricity prices leave little room to add climate policies without noticeable economic trade-offs. You badly need guidance on how to build a cost-effective strategy from here.

In the power sector, that means understanding the trade-offs among measures that could bring on more solar, wind, batteries, and nuclear — or transmission and pipelines to access low-cost energy from neighboring regions. A backward-looking comparison of RGGI and the RPS answers a different question from the one you are asking.

An economics of humility

None of this excuses the all-too-common practice of ignoring the costs and trade-offs of climate actions. More careful thinking about costs and benefits could help avoid policy blunders, like the reflex to block fossil fuel production wherever possible, even when doing so is expensive and barely affects global emissions, because the production simply moves elsewhere.

This instinct is driving some of the opposition to the Bipartisan American Affordability and Jobs Act, recently introduced in Congress. That permitting reform bill would speed the build-out of clean energy that deep decarbonization requires, but because it would also make fossil fuel infrastructure easier to build, parts of the climate movement have lined up against it. Fighting a bill that would help the climate because it also helps fossil fuels, rather than working to improve it, is just the kind of mistake that weighing costs against benefits helps to prevent.

But if the goal is to get climate advocates to weigh costs and benefits, scolding them for ignoring economic analysis may be counterproductive. Economists should press for more cost-benefit thinking from a place of humility about what we can and cannot deliver.

While economists will and should continue to analyze the near-term costs of incremental changes, they should also build tools that better capture the factors that large-scale emissions reductions depend on. On this front, many economists are already at work.

For example, one recent study in a top economics journal shows that the cost-effectiveness of electric vehicle subsidies improves dramatically once the dynamic effects of accelerated innovation are taken into account (though EV subsidies still cost far more than subsidies for wind), and that the cost-effectiveness of energy efficiency programs depends heavily on how clean the regional electricity grid is.

Perhaps more importantly, a humbler approach to encouraging more cost-benefit thinking would mean showing up to policy debates as contributors rather than as arbiters handing down verdicts. For reasonable policies with the potential to make genuine progress on the parts of the problem that actually require transformation — like getting oil out of vehicles or cutting natural gas use in electricity and heating — economists should skip the ritual comparisons to a carbon price or the social cost of carbon, which they may have previously used to brand policies as inefficient.

For a proposed clean vehicle standard, for instance, acting as a contributor instead of an arbiter means helping to improve the design with flexibility measures that reduce compliance costs, analyzing how it will affect vehicle and fuel prices, and flagging unintended consequences, like distortions that could result from counting different fuels as clean or dirty.

Economic tools can produce genuinely useful information to help policymakers weigh the costs and benefits of cutting emissions. What they cannot produce are comprehensive estimates, or anything close, of the costs and benefits that policymakers actually care about. We should be far more transparent about those limitations.

Counterintuitively, scaling back the ambition of our analysis may be what finally makes it useful.

Recommended reading: