Murphy on Nordhaus versus the UN
In light of Nordhaus’s calculations shown above, the apparently urgent need for “climate action” is not so urgent. It now looks more analogous to economists discovering the theoretical possibility of an “optimal tariff” but still understanding that free trade is the safest rule of thumb.
This is a key paragraph from Robert P. Murphy, “William Nordhaus versus the United Nations on Climate Change Economics,” the Econlib Feature Article for November. I think that’s a nice analogy.
A couple of other highlights follow. The first is the implications of the UN’s preferred policy for the gasoline tax, a calculation I had not seen elsewhere:
To translate this into plain English, the RFF [Resources for the Future] writers are explaining that the measures considered in the latest UN report are going to reduce particular units of emissions at a cost to the conventional economy of up to $5,500 per ton in 2030. (A carbon tax of $5,500 per ton works out to a gasoline tax of $48 per gallon.) In a standard cost-benefit approach, this would be economically efficient only if the estimated social cost of carbon were also in this range.
On how much higher than the Obama administration’s estimate the damage from carbon would have to be to justify the UN’s extreme goal:
What does the economics of climate change literature have to say about the social cost of carbon? The Obama Administration established an Interagency Working Group on the Social Cost of Carbon (SCC), in which it used leading models—including Nordhaus’s DICE, along with two others—to estimate the SCC through the middle of the 21st century. According to its last update, published in early 2017, the Obama EPA reported that the SCC in 2030, using the standard 3% discount rate, would be $50 per ton. Thus, the UN’s target of 1.5°C is implicitly treating the marginal units of greenhouse gas emissions as being anywhere from 1.6 to 100 times more damaging than the Obama Administration’s team estimated.
The whole thing is well worth reading.
Nov 6 2018 at 8:37am
So how is this an argument against a 50-100 dollar per ton carbon tax?
Nov 6 2018 at 9:09pm
Thaomas, are you familiar with the “optimal tariff” literature?
Anyway, in this particular EconLib piece, my main focus is the chasm between the UN’s calls for a 1.5C ceiling, and the work of the guy who just won the Nobel in this area.
But in the part where I make the optimal tariff analogy, my point is: In Nordhaus’ model (at least as of the 2007 calibration), the most you could possibly hope for from a carbon tax was a net benefit of $3 trillion. And yet other popular proposals–including what the UN is currently calling for–would make humanity $14 trillion worse off, relative to “doing nothing.”
So that by itself *is* an argument for not letting them get their foot in the door with a carbon tax. But for a more comprehensive critique of the idea, here’s my Cato article co-authored with two of their climate scientists.
Nov 10 2018 at 6:56am
@ Murphy Who are “them?” 🙂
Yes I have a PhD in Economics specializing in international economics from long enough ago that all the arguments for departures from free trade were well understood. I guess you mean do I understand that a carbon tax that is too high could be worse than zero as you advocate. What the optimum is obviously depend on the model used to assess the costs of CO2 accumulation.
What seems prudent both on economic and political grounds is to begin a carbon tax based on a conservative estimate of the costs, like that of Nordhaus, hence my $50-100 question. No doubt we will continue to re-estimate these costs and refine our outlook for how fast technical change affects to costs of zero carbon energy and the tax may well rise over time.
The alternative is to continue the present course: a hodgepodge of policies that may well have costs of well over that conservative carbon tax and run the risk down the line a decade or so hence that the costs will suddenly be perceived as so large that then we’ll institute all the high cost, growth destroying ideas then extant.
Nov 6 2018 at 12:56pm
That graph (Figure 1) looks wrong to me.
Consider the vertical axis. If a warmer atmosphere is bad, a really warm one will be really bad. So the bad aspects of higher temperatures should be non-linear (such as temperature squared).
Consider the horizontal axis. Many decades into the future we will be much richer and have far better technology, both of which will make it less costly (relative to today) to ameliorate the effects of global warming. (Imagine a major shift to nuclear fusion and/or fission.)
In other words, some decades out the value of keeping temperatures down will increase and the cost will decline. Therefore, I would expect all of the curves to have pronounced damped/flattened shapes.
This is not a criticism of Robert Murphy, but rather a criticism of William Nordhaus’ model. I realize it is a “crude approach that omits many crucial real-world considerations.” But it might be too crude to be useful.
Nov 6 2018 at 4:26pm
Charley Hooper wrote: “…I would expect all of the curves to have pronounced damped/flattened shapes.”
They do. The “Base” is just the default with no controls, so your reasoning doesn’t apply to that line.
The 2.5C and Stern lines (which basically overlap) clearly fit your prediction.
The “Opt” line does too, but it’s harder to see. In the early years it barely deviates from the baseline, but over time it begins to turn. I think if Nordhaus extended the graph to the year 2200, it would be more obvious that the optimal temperature flattens out, just as you are expecting it too.
Nov 6 2018 at 4:59pm
My reasoning doesn’t apply to the Base line only if the Base line assumes that all interventions originate with the government (“…if the governments of the world took no major action to arrest greenhouse gas emissions.”). In other words, there will be no government interventions and no private interventions.
I once bought an ultra-low emission Honda not due to government taxes, but because I simply preferred to pollute less (and it was a good car). Honda received kudos for manufacturing a ULEV vehicle and I received kudos for buying it. We both understood that pollution was bad and we both voluntarily scaled back our emissions. The same thing is already happening with carbon emissions, as people try to reduce their “carbon footprint.”
Nov 6 2018 at 5:31pm
Hi Mr. Hooper,
Well, I guess I’d repeat my answer then: At a big enough timescale, I think you *would* see that pattern. E.g. at some point it becomes virtually costless to have zero emissions, and so I think the “no controls baseline” would show the temperature increase from human activity ultimately flatlining.
In the DICE model Nordhaus has assumptions about a “zero emissions backstop technology,” but I don’t remember the details off the top of my head.
Nov 6 2018 at 4:46pm
This is an excellent and important article. It’s amazing that there can be so many different assessments by different parties (Nordhaus, IPCC, Obama Administration, the media) and yet we are told that there’s a clear consensus, when there is no such thing.
Nov 6 2018 at 9:12pm
Thanks for the kind words. For another mindblower, check out this LA Times reaction to Nordhaus’ award.
Nov 7 2018 at 11:48am
Yikes! What a jumble of half-truths, misunderstandings, and accusations.
Nov 7 2018 at 12:46pm
One thing about all these analyses is that assumption of CO2 emissions under the “No controls – baseline” case is ridiculously high, and therefore the temperature trend for that case is probably also ridiculously high.
From Table 1 of Bob Murphy’s article, “William Nordhaus vs IPCC” there is an option, “Limit CO2 to 560 ppm.” Assuming that 560 ppm is actual CO2 concentration (rather than CO2 equivalents), that should be the “No controls – baseline” case. Therefore, the Present Discount Value (PDV) of abatement costs should be zero for that scenario, and the total cost of the scenario should be $15.97 trillion, rather than $19.92 trillion.
Using a reasonable “baseline – no controls” value of 560 ppm for the CO2 concentration in 2100 will also change the PDV of abatement costs and PDV of total costs for the other scenarios. If I unexpectedly have lots of free time in the near future, I’ll try to re-create Table 1 using the more realistic “baseline – no controls” value of 560 ppm.
P.S. How do I know that the 560 ppm value is a much more realistic “baseline – no controls” value? It’s very simple. Way back in 2006 I predicted that global CO2 emissions over the course of the 21st century would be 712 gigatonnes of carbon, resulting in an atmospheric CO2 concentration of 558 ppm. The world is very much on a course to produce approximately 700 GtC emissions over the course of the 21st century. Current global emissions are approximately 10 GtC per year. If global CO2 emissions plateau at this value for another 10 years, and gradually fall to approximately zero by 2100, total emissions over the century will be approximately 600 GtC. No one who knew anything about the subject ever thought that global CO2 emissions would continue to increase throughout the 21st century.
Nov 9 2018 at 2:01pm
“In a standard cost-benefit approach, this would be economically efficient only if the estimated social cost of carbon were also in this range.”
Let’s assume the carbon tax (of whatever amount) offsets (replaces) other existing taxes. Doesn’t this change the equation? Are all other taxes less distortive and more efficient? If the carbon tax replaces other taxes, is the gross amount of the carbon tax really a dollar-for-dollar “cost to the economy”?
Comments are closed.