Why Pricing Carbon is Still a Good Idea

1. Introduction

My voice has been among the many who have long been arguing that a direct tax on carbon would be the best idea for regulating carbon dioxide emissions, but most of us have also conceded that pricing carbon by some other market mechanism like implementing a “cap and trade” program on emissions permits would be fine. Economic theory has taught us that the efficiencies of either would always dramatically reduce the expected cost of society’s meeting any of its politically determined and time-dependent emissions target by, instead, implementing rigid, uneven, and cumbersome “command and control” processes. Here, I explore the basis of this assertion before pondering the ultimate “Why” question circling around the entire issue – “If pricing carbon is such a good idea, why is it so difficult to get the US federal government to agree to do something in that regard?”

2. The economic theory that favors a carbon tax for macro-scale climate policies

Martin Weitzmen [1] laid down the fundamental economics of comparing these three options in the Review of Economic Studies under the title “Prices versus Quantities”; they come directly from the first principles of economic efficiency. The title should have included “under uncertainty” to convey its most important context. That would make it clear why, in the climate world, his paper can be interpreted a “tax vs cap and trade vs command and control (broadly defined)” question. Yohe [2] both expanded the title and developed a more generalizable context for its application to real policy choices.

To see how the Weitzman comparisons can be derived from first principles of economics, consider three descriptions of how each option could be designed to limit emissions of a pollutant like carbon dioxide to a specific target over a fixed period of time of a decade or three (say, from 2024 until 2035 or even from now until 2050). Assume, as is reasonable, that the target and the timeframe will have been set by a political process operating at nearly any size of government. The three options are:

1. Institute a command-and-control regime that would send strict and rigid quantity and technological regulations to every source of ambient carbon so that the sum of their emissions across all sources would be no larger than the designated target over the specified period of time;

2. Create a permit market structure wherein each source would need a license for each ton of carbon emitted. Initial allocations of the permits for each year could be distributed free of charge by the government and each source could buy more from another source or sell some at the market clearing price. The total number of licenses available for the specified time period would sum to the selected target.

3. Set a price (tax) on all carbon emissions that would likely increase over time so that the sum of expected emissions over the specified time period would match the targeted total achieved with certainty by the market scheme described in #2.

Let us make the easiest comparison by first taking explicit account of the behaviors of multiple emitters in the mix as described in Yohe [3]. Option (#1) is always the worst choice because it allows for no cost minimizing flexibility either across sources or over time. It is therefore more expensive to devise and to administer and it is also more difficult to adjust at scale over time. The other two options overcome both of these obstacles by design, and so their expected economic costs are necessarily smaller (and sometimes by quite a bit).

The choice between the other two options depends on their relative expected efficiencies emissions-decisions that sources want to make over time as market conditions for their products change from quarter to quarter or year to year. Option (#2) allows sources that want to produce more output in a particular year to augment their allocations of permits by purchasing permits from other sources who find themselves with a surplus. These sources, perhaps supplying other product markets, want to slow production and accept lower sales for whatever reason. This transaction makes both participants happy – one makes more money because it can sell more products, and the other makes more money because it can sell pieces of paper for which it would otherwise have no immediate value. Most importantly, total emissions across all sources would be fixed, and the permit market would clear every year at an equilibrium.

By contrast, the tax option always allows sources to produce whatever they want without needing a permission slip for the requisite carbon permits. All they have to do is pay a fee for every ton of carbon that they want to emit so that their dynamic output decisions internalize at least some of the external social costs of their emissions. Emissions from every source can vary over time (by more than they would facing a permit market where the price of carbon can be high or low relative to the tax). It follows that total emissions across all sources can vary from 1 year to the next, but the expected sum of cumulative emissions over the targeted time period should match the policy target (assuming that the initial year’s tax was set correctly).

The choice between these two equivalent options (on average) therefore depends on whether the flexibility in total annual emissions over time allowed by the tax saves more in reducing the expected cost of restricting total emissions to the target than the extra loss in the expected economic benefits across the targeted time period. Why this extra social cost from variable annual emissions? Depending on the curvature of their social damage functions, the expected net social benefits with variable emissions under the tax can be significantly smaller over the time period than the net social benefits of expected emissions – a level of net social benefit that can be achieved most efficiently over the selected time period by the fixed total emissions achieved by a cap and trade permit market. Both of these expected net social benefits depend directly upon the variance in annual emissions allowed by the tax.

For sulfur dioxide (acid rain), annual variability in emissions allowed by a tax can be very expensive when dangerous damage thresholds of acidity are crossed from time to time in lakes, across fields and forests, on buildings and cars, and so on, that is, extra damages suffered during high emissions years that exceed those thresholds can be much greater than any of the marginal damages avoided during low emissions years. For carbon, though, annual variability in emissions is essentially costless because of one well-established fact about how the climate system works on an aggregate global level: as reported in NRC [4], climate change impacts generally depend on temperature change that, in turn, depends directly on changes in atmospheric concentrations of gases like carbon dioxide. Concentrations depend fundamentally on cumulative emissions from 1850 – the marker of the beginning of the Industrial Revolution. Nearly 200 years later, even current annual emissions that are historically high make only small and nearly constant marginal contributions to historically high concentrations so that the effect of annual variability around those high levels is functionally insignificant when reflected by temperature change.

It therefore follows from economic theory that cap and trade should be preferred for limiting sulfur emissions. For carbon emissions, though, the low expected marginal cost of variability in emissions that have a zero mean over the policy period makes a tax regime the better choice.

3. Solving two important complications: Time and uncertainty

One complication arises from the obvious fact that mitigation plans are made to cover many decades or even a century or two. If atmospheric concentrations are to be stabilized at any point in the future, however, emissions must eventually fall by 80% of current levels. It follows that a carbon tax must increase and permit supplies must fall over time, but from what initial level and how fast?

Thinking about the challenge of these enormous long-term planning horizons now makes it clear that nobody should be expected to set mitigation targets in 2024 for 2100 and beyond. There is just too much uncertainty in our understanding of the climate system and in our understanding of how the global economy will evolve over centuries under any one of many possible geo-political regimes. It therefore makes sense that policy targets should be created for the long term, but policy designs should be broken into manageable chunks of time. That is to say, policymakers must be experts to make mid-course corrections based on a new understanding of what is going on.

IPCC [5] highlighted this reality when it produced a consensus agreement among the members of the United Nations Framework Convention on Climate Change (UNFCCC) that “Responding to climate change involves an iterative risk management process that includes both adaptation and mitigation and takes into account climate change damages, co-benefits, sustainability, equity, and attitudes to risk”. The two to three decadal time horizons that framed the descriptions of the options recorded above fit right into that iterative framework, so the second complication is covered.

At first blush, the first issue looks to be a little more challenging, but it was actually solved in Hotelling [6]. The “Hotelling” rule summarizes his finding: to maximize social benefits from exploiting an exhaustible resource over time, add an initial “scarcity rent” to the cost of extraction and let it increase over time at the rate of interest. That is to say, a time-dependent “scarcity tax” added to extraction cost was a control variable in his work on a naturally constrained intertemporal optimization problem just like the carbon tax introduced by Nordhaus [7] was a control variable in his seminal analysis of how to solve an unconstrained optimization problem about human contributions to global warming where the object was to maximize net discounted global welfare when carbon emissions from fossil fuel caused worldwide economic damages that would increase over time.

These two contributions to economic theory now confirm that the design of modern policies described above is appropriate even though modern emissions targets are the product of political decisions and not solutions to a welfare optimization problem. The policy goal today is to minimize the expected costs of hitting those societal targets by, in the case of cap and trade, turning cumulative carbon emissions over the design period into an artificially created exhaustible resource problem for emitters and trust that the permit market will clear each year at levels that will approximate an efficient distribution across time. We can expect, therefore, that the permit price will begin at a level that reflects Hotelling’s initial scarcity rent and trend up at the market rate of interest. By analogy, then, the equivalent carbon tax should initially be set at a scarcity rent for carbon that sets expected cumulative emissions equal to the target if it, too, increases at the rate of interest. Interestingly enough, that is exactly the trajectory that Nordhaus produced last century when he turned to minimizing the discounted cost of achieving a temperature target.

4. Evidence that the economics was taken seriously by those who speak to power

Given “the significant risks that climate change poses to human society and the environment,” members of the US Congress asked the US National Academies of Science, Engineering, and Medicine [8] early in the Obama administration to “provide a strong motivation to move ahead with substantial response efforts”; and to describe those efforts while they were at it. The result was a collection of four-panel assessments and a synthesis report. One of their most significant findings is summarized on the third page of its summary briefing document:

Notice that there is no mention of political obstacles, but there is mention of complementary policies and an entire section on “iterative risk management”. It closes with:

Getting NASEM to pick up on pricing carbon and putting it into context is one thing, but seeing actual action is quite another. Happily, there is plenty of evidence that climate change has spread into the political discourse in the United States since the Academy report. For example, the American Clean Energy and Security Act of 2009 [9] H.R. 2454 in 111th Congress named Waxman-Markey for its primary sponsors would have established a variant of an emissions trading mechanism similar to the European Union Emissions Trading System (EU ETS; [10]. It was approved by the House of Representatives by 219-212 in June of 2009, even though there was no chance of overcoming a threatened Republican filibuster in the Senate. It was more ambitious than both the Lieberman-Warner Bill and President Obama’s proposal, but it did not have universal support from the environmental community [11].

Much later, the Inflation Reduction Act of 2023 [12] funded 10 years of climate-friendly investment in infrastructure, alternative energy, electric vehicles, and additional charging infrastructure to promote the demand side of the vehicle market. Most notably, for present purposes, it did not include pricing carbon despite some encouragement to do so. Perhaps most importantly, it was complemented during President Biden’s first term by an increasing number of programs that look a lot like creative applications of command and control mechanisms:

• The Security and Exchange Commission [13], after years of drafting a review, finally implemented new reporting rules for material risk from climate change and climate policy. They rely on climate savvy investors in financial markets to motivate corporations to become better climate citizens because it is in their own best interest;

• The Federal Reserve System [14] continued the process of characterizing severe global recession scenarios accompanied by a period of heightened stress in both commercial and residential real estate markets (caused by sudden enormous contributions to global sea level rise from accelerated melting and disintegration of Antarctic Ice Sheets). They rely on word of successful strategies to prevent emerging climate risks from spreading quickly across the entire banking system (see [15]);

• The Environmental Protection Agency [16] continued to strengthen a variety of environmental exposure regulations that improve human welfare one threat at a time, with a recent focus on fossil fuel-generating power systems; and

• The Department of Transportation [17] has recently announced a series of higher fleet mileage requirements that will move supply and demand sides of transportation markets toward increased fuel economy and corresponding lower emissions of hazardous pollutants including greenhouse gases, and so on.

The list can be extended, but the point is clear – climate mitigations of many types have been successfully implemented across the United States for more than a decade, notwithstanding persistent intransigence in the legislative and executive branches of the federal government.

5. Some concluding thoughts

Intransigence in Washington does not mean that pricing carbon is a bad idea for federal-level policy. It is not. After all, some nations, like the entire European Union, have been quite successful in pricing carbon. They were the first to take that approach. They are now joined by the United Kingdom, Australia, New Zealand, South Korea, and China, which are on the list of countries whose daily spot prices for carbon emission allowances are listed by CarbonCredits [18]. On May 11, 2024, the EU ETS price was just over $77 per ton of carbon. Even the northeast region of the United States implemented its own limited multistate Regional Greenhouse Gas Initiative (RGGI) in 2009; RGGI [19] reported that the market clearing auction price in March of that year was $16.00 per ton of carbon.

As reported in Macrotrends [20], United States carbon emissions have been on a short downward trend since around 2020. That trend, even after COVID-19 impacts have passed, shows that things have been getting done across the United States to hasten its mitigation efforts. It has not been enough, though, for any number of reasons that do not really matter for this paper. More has to be done, and pricing carbon is still the cheapest way to do that.

Americans have been allergic to adding new visible taxes since the Boston Tea Party in revolutionary days, but that seems to have been changing recently across much of the population with regard to carbon emissions. As reported from public surveys from late in 2023 by Yale Climate Communication [21], 72% of the US population understands that global warming is real, and 68% think that fossil fuel companies should pay a tax on the carbon that they deliver into global energy markets. Indeed, as shown in Figure 1 for 2023, a majority of people in nearly every county in every state feel that way (with some states topping off above 80%).

Debates about the structure of a carbon tax or cap-and-trade regime continue, of course, but they are mostly about what to do with the revenue to ameliorate the regressive character of their incidence. Prasad [22], for example, reviews their content. Of particular note is the well-taken recognition of the danger in using carbon revenue to reduce other tax sources of government funds – the expectation that those proceeds would diminish over time if the policy is successful in reducing emissions.

What can be done? The United States is supposed to be a representative democracy, but the climate views of people who represent the Americans whose views are reflected in the Yale surveys are certainly not seeing positions being captured in either climate actions or even carbon policy design. Their contrarian representatives in the House and the Senate have been standing in the way for years, but that obstinance has been exaggerated by the new administration in the White House.

Scholars cannot give up pushing for what would work better in even the status quo because they face a political impediment born in the last century that still persists. They cannot stop pushing to price carbon nationally by putting climate change and climate action front of mind in every race being contested in every election, starting from the next one. And they can be buoyed by new evidence that policies designed to promote emissions reductions work if somebody pays for doing otherwise [Stechemesser et al. [23]]. {SCIE} They must push at the ballots in every state where races for the President, Senators, and House Representatives can be found. They must continue to get the word out so that contrasts across candidates can be honestly assessed. They can celebrate every victory and mourn every loss for one day but then start all over again when they wake up the next day.