If you spend much time around economists and the conversation turns to global warming, the odds are high that at least some of them will advocate a tax on carbon to reduce carbon usage and thereby slow or halt global warming. Even some economists who are skeptical that global warming will do much harm often think that carbon taxes are a good idea. When I taught an energy economics class at the Naval Postgraduate School, I, too-- someone who thinks global warming will be far from catastrophic--made a case that if the government is going to “do something” about global warming, carbon taxes are the least bad measure.

I now think I was wrong. Necessarily, therefore, I think the many economists who advocate carbon taxes are wrong. My argument does not rest on the idea that global warming will not be very harmful. Rather, my argument is that taxing carbon makes sense only if reducing carbon is the most efficient way to forestall global warming. It might be the most efficient way, but there’s a good chance that it is not.

Let’s first dispose of an issue that has probably confused many non-economists and even some economists: the idea that taxing carbon is the same thing as “pricing carbon.” Carbon is already priced. Natural gas, oil, and coal all have prices and their prices are somewhat related to the amount of carbon they contain. To be sure, adding a tax to carbon would raise the prices of all those fuels, just as adding a tax to alcohol would make your tipple more expensive. But just as setting a tax on alcohol does not “price alcohol,” setting a tax on carbon does not “price carbon.” In my more cynical moments, I wonder if advocates of a carbon tax sometimes call such a tax a price to mislead people into thinking that a carbon tax is a market solution rather than a tax solution.  

Economists are naturally drawn to the idea of taxing carbon. Their thinking rests on the ideas, from early last century, of a British economist named Arthur Pigou. Pigou addressed situations where the actions of some people—call them Group A—damaged other people—call them Group B—but people in Group A had little or no incentive to take account of that damage in their decisions. He, like the British economist Alfred Marshall before him, called the effect of Group A’s actions on Group B “negative externalities.” Pigou’s thinking in these matters dominated economics through the rest of the 20th century and is still vaunted today. Indeed, Harvard economics professor N. Greg Mankiw has formed what he calls the “Pigou Club,” which he describes as “the elite group of pundits and policy wonks with the good sense to advocate higher Pigovian taxes.” Among its members are the Hoover Institution’s John Cochrane and George Shultz.

Here, briefly, is the argument for their position. Burning carbon creates carbon dioxide; increasing the concentration of carbon dioxide in the atmosphere leads to more global warming; more global warming leads to bad consequences in the long run; the majority of people who use carbon don’t take these bad consequences into account when they decide how much carbon to use; therefore they will overuse carbon unless some level of government taxes us for using it.

Advocates of a carbon tax argue, correctly, that a tax is a much better way to reduce carbon usage than any system of regulations could be. The reason is that every use of carbon that creates carbon dioxide imposes damage, and a carbon tax based on the amount of carbon dioxide created will cause everyone who creates carbon dioxide to, indirectly, take that damage into account. The carbon tax beautifully scales the payment to the damage: those who create more damage pay more in carbon taxes.

By contrast, a system of regulations or subsidies that favors one energy source over another, or one energy use over another, depends on central planners having information that they cannot possibly have. So, for example, if government planners require that a certain percentage of energy usage come from solar, they do not take account of the cost of solar and they foreclose ways of producing energy, such as natural gas, that produce less carbon dioxide than other ways, such as coal. A carbon tax, by contrast, automatically pushes electricity production away from coal and towards natural gas. As a bonus, if, under a carbon tax regime, solar energy turns out to be less expensive than natural gas, the market will push energy production more in the direction of solar.

That reasoning is fine—as far as it goes. But it doesn’t go far enough. The reason is that economists who advocate Pigovian taxes take as given that the most-efficient way to forestall global warming is to reduce the amount of carbon used. But what if their assumption is incorrect?

There are at least three important reasons to conclude that the assumption is wrong. First, cow farts. That’s right: cow farts. A far more potent greenhouse gas than carbon dioxide is methane. Methane, which is present in cow farts, warms the planet much more quickly than carbon dioxide before decaying to carbon dioxide. The UN’s Intergovernmental Panel on Climate Change (IPCC), which is generally thought of as the scorekeeper on global warming, estimates that over the approximate decade before it decays, methane warms the earth by a whopping 86 times as much as CO2. To be sure, the CO2 lasts much, much longer than methane, but the fact of methane’s huge short-run potency surely suggests that a tax on carbon may not be the cheapest way to forestall global warming.

Second, one important technological development over the last decade has been “geo-engineering.” The idea here is to change other things in the atmosphere that are easier to change than the amount of carbon used. Consider what we learned from the June 1991 eruption of Mount Pinatubo, in the Philippines. That eruption poured 20 million tons of sulfur dioxide into the stratosphere. Over the next two years, the effect of that one eruption was to reduce the earth’s temperature by about 1 degree Fahrenheit. That might not sound like much but it’s actually over half the 1.4 degree warming that has happened over the last century. What if every year we could put sulfur dioxide in the atmosphere so as to permanently prevent the earth from warming? In their 2009 book Superfreakonomics, University of Chicago economist Steven D. Levitt and writer Stephen J. Dubner point out that if we could get just 34 gallons per minute of sulfur dioxide into the stratosphere (which translates to about 100,000 tons per year), that would reverse warming in the high Arctic and reduce it in much of the Northern Hemisphere. Why focus on such high latitude areas? Because, note Levitt and Dubner, “high-latitude areas are four times more sensitive to climate change than the equator.” How would you get the SO2 into the atmosphere. Levitt and Dubner cite the thinking of Nathan Myhrvold, at one time the chief technology officer for Microsoft. Mrhrvold argues that if we had a big enough hose, we could do it.

Is such a technology feasible right now? Maybe not. But if it were, it would be incredibly cheap. Myhrvold’s organization, Intellectual Ventures, estimated that it could be set up in two years for $20 million and an annual operating cost of about $10 million.

Those are not large numbers, especially compared to the cost of a carbon tax. In December 2018, the Congressional Budget Office estimated that a carbon tax of $25 per metric ton for the 10-year period from 2019 to 2028 would generate cumulative revenue for the U.S. government of $1.1 trillion. The cost of building and operating the SO2 hose would be a rounding error on that number.

But isn’t geo-engineering risky? Couldn’t it have negative unintended consequences? Yes and yes. But everything we do is risky, including imposing more taxes on carbon. Moreover, taxing carbon is an indirect form of geo-engineering.

The third low-cost way to rein in global warming is by planting trees. Trees absorb and store CO2 emissions. You could call the tree-planting strategy geo-engineering, but it would count as such in a very low-tech form. According to a July 4, 2019 article in The Guardian, planting one trillion trees would be much cheaper than a carbon tax and much more effective. At an estimated cost of 30 cents per additional tree, the overall cost would be $300 billion. That’s large, but it’s a one-time cost. Moreover, writes The Guardian’s environment editor Damian Carrington, such a tree-planting program “could remove two-thirds of all the emissions that have been pumped into the atmosphere by human activities, a figure the scientists describe as ‘mind-blowing’.” A carbon tax, by contrast, would simply slow the rate of emissions into the atmosphere.

It’s true that there would still be the problem of how to pay for that extra trillion trees. Thirty cents per tree is small but one trillion is large. Yet my point is the one I started out with: there are potentially much cheaper ways to deal with global warming than a carbon tax.

Advocates of such a tax might argue that it could be set up to give money to people per tree planted. It could, but notice that the key is not that it’s a carbon tax but that it’s a per-tree subsidy. The carbon tax would still be much more expensive, for a given amount of global warming prevented, than such a subsidy.

Perhaps you find too uncertain the idea that geo-engineering, either high-tech or of the arboreal variety, is at the stage where it would slow or reverse global warming. That may be. But here’s what we know. First, imposing a tax on carbon is very expensive. Second, taxing carbon gives little incentive to find other solutions that are less expensive. Wouldn’t the incentive to find other solutions be high because doing so would help major energy companies argue against a carbon tax. They would certainly have the incentive to argue, but the incentive to discover lower-cost solutions would be small. To the extent a major energy company found such solutions, and to the extent it persuaded governments to back off the carbon tax, it would be helping its competitors and billions of consumers, not just itself. These other companies and consumers would be free riders on that one company’s effort.

It’s possible that somewhere down the road, we will conclude that, for reasons we don’t yet know that make geo-engineering infeasible, a carbon tax is a good idea. But we’re not close to that conclusion yet. A carbon tax is not a slam dunk.

overlay image