A recent NYT article discussed a proposal by economist Ross McKitrick to tie CO2 taxes to global temperature increases. McKitrick’s overall aim is to offer a compromise that, he argues, should satisfy those who think the government needs to take drastic action and those who think carbon emissions pose no serious long-term threat. Although McKitrick’s idea is clever, it has theoretical difficulties and (in my opinion) would certainly not work in practice.
McKitrick’s Proposal to Tie CO2 Taxes to Temperature
The NYT story does a good job summarizing the idea:
[McKitrick] suggests imposing financial penalties on carbon emissions that would be set according to the temperature in the earth’s atmosphere. The penalties could start off small enough to be politically palatable to skeptical voters.
If the skeptics are right and the earth isn’t warming, then the penalties for burning carbon would stay small or maybe even disappear. But if the climate modelers and the Intergovernmental Panel on Climate Change are correct about the atmosphere heating up, then the penalties would quickly, and automatically, rise.…
Specifically, [McKitrick] proposes tying carbon penalties to the temperature of the lowest layer of the atmosphere (called the troposphere, which extends from the surface of the earth to a height of about 10 miles). He suggests using the readings near the equator because climate models forecast pronounced warming there.…
The carbon tax might start off at a rate that would raise the cost of a gallon of gasoline by a nickel — or, if there were political will, perhaps 10 or 15 cents. Those numbers are all too low to satisfy environmentalists worried about climate change.
But if the climate models are correct, Dr. McKitrick calculates, within a decade his formula would cause the tax to at least double and possibly sextuple — with further increases on the way if the atmosphere kept heating. The prospect would give immediate pause to any investors trying to decide today what kind of cars, power plants and other long-range energy projects to finance. To estimate future profits, they would need to study climate.
“The best results will accrue to firms incorporating the most accurate climate forecasts into their decision making, precisely the kind of forward-looking behavior environmentalists want to encourage,” Dr. McKitrick writes. “Consequently, it’s not the case that we have to wait until it is ‘too late’ to respond to global warming. The market will force investors to make the best possible use of information and to press for improvements in climate forecasting in the process.”
McKitrick’s proposal is interesting, and I applaud his attempt to come up with policy suggestions that can (in principle) satisfy the concerns of most of the experts on the issue of climate change. McKitrick is trying to call the bluffs of people on both sides of the controversy, since (if calibrated correctly) his tax will stay low if temperatures do not rise in accordance with the IPCC’s favored models, while the tax will go up if those warning of large spikes turn out to be accurate. Yet despite the superficial solution to the dispute, I think McKitrick’s idea doesn’t really work.
Theoretical Problems With McKitrick’s Proposal
A crucial problem with McKitrick’s approach is the proper calibration of the carbon tax, for a given amount of measured tropospheric warming. If the IPCC consensus is correct, then carbon emissions constitute a “negative externality” (in economics jargon) whereby individual behavior imposes harms on others. To achieve a more efficient use of society’s resources (including the environment), it is optimal (at least in simple models) for the government to impose a tax on each unit of greenhouse gas emissions, to force the emitter to “internalize the externality.”
The difficulty with this, of course, is our uncertainty over the exact quantitative connection between a ton of CO2 emissions and the resulting damage to future generations. This uncertainty spills over into the proper calibration of the “optimal carbon tax.” What McKitrick’s formal academic paper suggests is that policymakers can induce an approximately optimal tax, given this uncertainty, by making the size of the tax dependent on the observed warming.
Now here’s the rub: The initial tax would need to start out quite low, in case the “skeptics” are right. In other words, if the whole point of McKitrick’s proposal is to satisfy both skeptics and fans of the IPCC, then it wouldn’t work to start out with a high carbon tax. If global temperatures stayed flat or fell over the coming decades, that means the skeptics were right, but in that case we would have suffered the imposition of a large and inappropriate tax on behavior far out of proportion to the externalities involved. So this is why the initial McKitrick tax couldn’t be too high.
But now if the initial tax starts out quite low, it would need to jump substantially in light of any rise in global temperatures, insofar as this rise gave us more confidence in the IPCC projections. This feature would be necessary to make McKitrick’s proposal appealing to those who right now are certain that human activities are pushing the world towards catastrophe. If the tax rose only modestly in proportion to modest temperature increases from this point forward, then the most alarmist climate scientists would understandably reject McKitrick’s proposal as insane. Economist Alex Tabarrok explained the point in this way after he favorably discussed McKitrick’s proposal on his blog:
As predicted most of the objections (in the comments) are from climate change proponents. In essence, they argue that the problem is so serious that we must act before the evidence is in. Aside from the obvious epistemic problems with such a position do note that a) this is a way of getting agreement where otherwise there might be none b) the tax can be non-linear so it rises (in Bayesian fashion) with the strength of the evidence, i.e. the tax need not always lag.
The problem here is that–according to everybody–the earth’s temperature bounces around a lot due to natural variability. So the more “non-linear” we design the tax–for example making it double with a further increase of 0.1C, but making it quintuple yet again with an additional increase of 0.1C–the greater the economic damage would occur if the skeptics are right and yet there happened to be a brief spike in temperatures due to natural causes. It’s true, in this scenario the high tax would wither away once temperatures naturally moved back down, but the damage would already have been done.
On the other hand, suppose the IPCC consensus is correct, and the initial McKitrick tax is far below the true “social cost of carbon.” If the underlying anthropogenic warming trend happened to be masked for an additional five years because of natural factors, then the McKitrick tax would stay far too low and allow devastating further emissions as the window of action closed.
In summary, the debate over taxing carbon emissions is so complex that it is difficult to see how we could design a McKitrick-style tax even in principle that would satisfy both sides. The natural variability of global temperatures would allow for the possibility of significant (and unnecessary) damages–either from hindered economic growth or from climate change–under any proposal that both sides could agree on. But that of course means the value of agreeing to such a proposal is greatly reduced in the present, as we contemplate various scenarios going forward.
Practical Problems With McKitrick’s Proposal
Even if we could come up with a sensible compromise that made sense theoretically, in practice McKitrick’s idea still strikes me as dubious. For example, suppose Congress enacts the McKitrick tax, and within five years tropospheric temperatures rise such that the tax draws in $100 billion annually. At that point, does anyone seriously believe that a sudden and unexpected (by the NASA scientists etc.) drop in temperatures would cause Congress to lower the tax and slash spending accordingly?
I think that once Congress got its foot in the door with a new tax such as this, it would never go away, just like the vast majority of tariffs for “infant industries” survive well into “adulthood” for the protected industries. If and when the temperature measurements called for a tax cut, we can easily imagine senators and policy wonks arguing that it would devastate the solar power industry to roll back the McKitrick tax, or that it would be shortsighted to cut the tax since scientists XYZ argue that the temperature drop is due to smog from China.
Ross McKitrick’s proposal to tie CO2 taxes to observed temperature trends is an interesting attempt to break out of the modelling log-jam, but ultimately I don’t think it will work. Even putting aside the serious theoretical difficulties in properly calibrating the tax, we could never trust the politicians to keep their word once they had enacted it.