Economists’ Advocacy for a Carbon Tax Is Misguided

Many, if not most, mainstream economists, both conservative and liberal, support a carbon tax. As economist Alan Blinder recently remarked, “It’s only a very slight exaggeration to say that all economists favor a carbon tax to mitigate climate change.” But I argue below their position seems to overlook real-world politics, bureaucratic incentives, and the “pretense of knowledge.” Advocates of a carbon tax seem to fall victim to what Ronald Coase labeled “blackboard economics”.

The benefits from such a tax may, therefore, be far less than promised by blackboard economics, and than advocates are claiming. In fact, the benefits can be negative, given what I regard as “guesstimates” (highly uncertain metrics) that policymakers rely on to represent the damage from carbon.   

Aligned with Coase’s blackboard economics, even if one can theoretically show that a carbon tax is the best way to combat climate change (which many economists would concur), it may fail in practice. A major reason is the highly speculative nature of the external costs from carbon emissions. Attempts to measure the social cost of carbon (SCC), for example, depend on predictions of economic and climatic conditions decades out in the future. Analysts refer to the damage from carbon emissions as the SCC, which is the theoretically correct metric for setting a carbon tax.

For a carbon tax, policymakers should consider the politics, the highly perplexing problem of measuring the damage from carbon emissions, and other real-world challenges, like administrative and bureaucratic behavior. 

On the blackboard, a carbon tax would offset implicit subsidies to non-clean-energy, and carbon-intensive goods and services. (The subsidies exist because their prices exclude the damage from the carbon emitted into the atmosphere.) But in the real world, the SCC is so conjectural that we have little idea of what tax would be necessary to realize this offset. 

The SCC depends on parameters that are subjective: interest groups, politicians, and bureaucrats can and do argue for a SCC that best advances their agenda. The SCC is highly sensitive to both the discount rate (which affects the future value of the benefits from reduced carbon in present value terms), and the climatic temperature change from a specific reduction in carbon, each being critical factors affecting the social-welfare implications of a carbon tax.

A major uncertainty is the linkage between the natural and economic outcomes; for example, if global temperatures increase by a certain level, the impact on Gross Domestic Product is extremely difficult to predict with any reasonable accuracy. One can describe such predictions as highly speculative, devoid of any rational policy implications. 

Overall, models that calculate the SCC are non-robust, and highly dependent on the parameters and assumptions contained in them. Interest groups, policymakers, analysts and others can easily manipulate the models to produce results to their liking. Using a discount rate of two or three percent, rather than seven percent, for example, can have a significant effect on the SCC calculation. One then can legitimately question whether predictions of the SCC actually reflect unbiased analysis. When policymakers apply an inflated SCC, restrictions on carbon emissions (whether by executing a high carbon tax or tighter controls), require society to expend excess resources on curbing emissions — that is, the additional abatement costs exceed the reduced damages from emissions, causing a net welfare loss.  

In the unlikely event that policymakers can overcome these measurement challenges, it remains doubtful that a US carbon tax by itself would have a detectable effect on climate change. The reason is that over 85 percent of carbon emissions originate in foreign countries. (Of course, other measures taken unilaterally by the US to reduce carbon emissions would have the same limitation.) 

One also cannot ignore the fact that those countries that institute a carbon tax bear the emission-reduction costs nationally, while the benefits accrue globally. The official US government’s estimate of the global social cost of carbon (the social benefit of reducing carbon emissions) is currently around $51 per ton. But the benefit to the United States would be much less, for example, around just $7 per ton, according to 2017 estimates from the US Environmental Protection Agency. If the United States or any country institutes a carbon tax alone, its citizens will likely suffer negative impacts, even if the world in total benefits. This means that setting a carbon tax at the global social cost of carbon would surely burden US citizens more than what they gain in climate-related benefits. 

Even with a carbon tax, besides arguing for the retention of existing government interventions, climate activists would likely still lobby for more inefficient mandates and other regulations, and subsidies for clean technologies. After all, the legacy of US environmental policies is heavy reliance on highly inefficient command-and-control mechanisms and subsidies for politically favored technologies. 

It may well also be true that existing government measures to reduce carbon emissions (electric-vehicle mandates, renewable-energy subsidies) have already moved emissions near or even below the optimal level. An additional measure such as a carbon tax may actually lower economic efficiency by increasing abatement costs more than the added benefits. Although this becomes an empirical question, one can conceive conditions under which this outcome becomes imaginable. This illustrates yet another example where blackboard economics has overlooked a real-world condition that could make a carbon tax less attractive. 

Some advocates of a carbon tax argue that, when found to be appropriate, it would be easier to undo a tax compared to alternative approaches (like subsidies for solar and wind energy). But based on worldwide experiences with other taxes, beneficiaries like clean-energy producers and the government will likely exert strong political opposition to the abolition of a tax — if only to avert revenue losses that would check government spending — even when warranted by new developments. 

Notwithstanding a consensus among experts on the damage from carbon, the reality is that bureaucrats and politicians would likely choose a metric aligned with their self-interest or with the interests of those who currently hold power. Look no further than the contrast between the official Biden SCC of $51 per ton (his EPA wants to increase it to $190 per ton) with the Trump Administration’s estimate in the range of $3-$5 per ton. How can government officials rely on such guesstimates when using SCC to develop policy? More than objective analysis goes into determining the level of SCC, by whomever is in power. One cannot ignore the possibility that a politically driven carbon tax, along with the high uncertainty of the SCC, will result in a sizable societal welfare loss.

In view of unrealistic assumptions, highly imperfect information, and a changing world where the optimal carbon tax would vary over time, the blackboard-economics criticism of a carbon tax seems well placed. Public choice economics would also predict the improbability of an actual carbon tax achieving a blackboard outcome: the combination of distorted incentives to appease special interests and the self-interest of government officials, the possibility of more economically rational measures than carbon reductions to attack climate change, and the lack of knowledge about how a complex and changing world works, would obviate a blackboard outcome. As in other matters, bureaucrats and politicians would identify more with their self-interest and willingly pursue it, even when the public good suffers.

Here is the crux of the problem that bears repeating: No matter how sophisticated the models and quantitative methods are in measuring the SCC, there is a great deal of uncertainty in any calculation. It seems right to ask whether one can get a reasonable estimate of SCC, since the analyst, among other things, would have to measure how society will adapt to climate change. Future technologies and other developments will determine the societal cost of climate change. Referring to Hayek’s “pretense of knowledge,” it seems that both analysts and policymakers are delusional to think that they can derive a reasonably accurate calculation for how society’s welfare would increase with less carbon emissions and a lower temperature rise. I would as far as to say that the SCC is inherently indeterminate; when one thinks seriously about the difficulty of, and challenges in, measuring a SCC, she comes to the conclusion that a reasonably accurate SCC lies beyond the capability of even the most brilliant minds. When one looks closely at how analysts arrive at their calculations, she sees how they relied on highly restrictive and dubious (e.g., devoid of economic empiricism) assumptions.

Given the obstacles and other problems (not least the massive political uncertainties) vexing conventional climate policy, such as carbon-emissions caps, carbon taxes, and targeted subsidies for clean energy, more attention should focus on measures that strengthen market signals for individuals to adapt to climate change. These measures include adaptation based on the pricing mechanism, companies satisfying consumers and investors with clean products, and governmental encouragement (or at least not discouragement) of innovations in clean-energy technologies (for instance, nuclear power, renewable energy, and hydropower) and climate engineering. 

In the end, what matters is: Would a carbon tax improve the state of affairs, and be superior to other measures to combat climate change, given the problems that I have raised? The answer is not obvious, even for a mainstream, pro-market economist like me.