Some good things came out of the COP26 conference in Glasgow earlier this month; just not enough to render as wholly inaccurate Greta Thunberg’s assessment that the discussions at COP26 amounted to nothing more than “blah, blah, blah.”
To be fair, I think there were at least three important developments to come out of COP26. First, delegates managed to finalize what had been the unfinished business of Article 6, left over from COP21 in Paris: doing so has cleared the way for building a system to price and trade carbon globally. Second, China and the U.S., the world’s two largest GHG emitting countries, announced that they had agreed to cooperate on a range of climate issues. Third, the Glasgow Financial Alliance for Net Zero (GFANZ), a group of private financial institutions representing 40% of the world’s financial assets, pledged to meet the goals set out in the Paris climate agreement.
It is easy to understand some people’s disappointment about what the parties to COP26 actually accomplished. Consider COP26’s stated goal of keeping global temperature increases below 1.5 degrees Celsius relative to pre-industrial times. As COP26 concluded, its chair Alok Sharma stated: “We have kept 1.5 alive. That was our overarching objective when we set off on this journey two years ago, taking on the role of the COP presidency-designate. But I would still say that the pulse of 1.5 is weak.”
Some argue that this weak pulse is signaling a condition that is terminal because projects already funded, when undertaken, will absorb the remaining carbon budget associated with keeping 1.5 alive.
A critical tool for maintaining a clear-eyed view of the climate change landscape is the DICE model developed by Nobel laureate William Nordhaus. DICE is an acronym for Dynamic Integrated Climate-Economy. It is an integrated assessment model, built to analyze the interactions between the global economy and the global climate.
Nordhaus usually updates DICE every few years, but the most recent version, called the 2016 model, is from the end of 2015 when the Paris agreement was negotiated at COP21. Below I describe what the 2016 model predicts about temperature change generally and about goals such as 1.5 and 2.0 specifically.
At the end of 2014, the global atmospheric temperature had risen 0.85 degrees. The 2016 DICE model forecasted that the atmospheric temperature would increase to 1.1 degrees just after 2020; and it did so a little earlier actually, in March 2020.
The 2016 DICE model forecasts that the global temperature will cross 1.5 degrees just before 2035. This is pretty sobering. Even more sobering is that the preceding statement holds, according to the model, even if we manage to institute a global system today that would price carbon at its social cost.
The last statement is controversial, and critically depends on the phrase “according to the model.” The controversy involves whether instituting a global system for pricing carbon fails to prevent, but merely delays when the 1.5 crossing point occurs.
At the heart of the controversy is the fact that according to Nordhaus’ 2016 DICE framework, the optimal price of carbon, meaning the tax rate on carbon per metric ton associated with its social cost, currently resides between $35 and $40; and will rise above $50 by 2030. In contrast, other economists, especially Sir Nicholas Stern and Nobel laureate Joseph Stiglitz, argue that carbon would need to be priced at $100 by 2030 in order to keep the global temperature from exceeding 1.5 degrees, while simultaneously achieving net zero emissions by 2050.
The 2016 DICE model forecasts that if humans continue to do business as usual, then by the end of 2035 the global temperature will have risen by 1.55 degrees. If we institute a system to price carbon at its social cost, then at the end of 2035, the global temperature will have risen by 1.52 degrees. This temperature difference, 0.03 degrees, between what happens with business as usual, and what happens with business featuring appropriate carbon pricing, is very small. I should add, however, that after 2035, the 2016 model’s forecast for temperature difference between a carbon priced economy and a non-carbon priced economy grows substantially, with the difference being significant for how the climate crisis evolves.
Most estimates of the current global price of carbon lie somewhere between $2 and $20. The 2016 model predicted that the current price of carbon would be somewhat above $2.20. According to the IMF, carbon is currently priced at $3. According to a new study from investment firm Kepos Capital, taking full account of all taxes levied, and adjusting for subsidies, carbon is currently priced globally at just under $20.
The forecasts from Nordhaus’ DICE model have historically exhibited excessive optimism, which Nordhaus has documented in reports he has published in the Proceedings of the National Academy of Sciences. However, for the period 2016-2020, actual industrial emissions of carbon dioxide, at 35.9 Gt, were a bit lower than the DICE model forecast, at 39.4 Gt. The Covid-pandemic certainly provides one possible reason for the downward bias. Notably, had carbon been priced at its social cost during this period, the forecast stipulates that industrial emissions would have been 33.1 Gt.
The concentration of carbon in the atmosphere is a critical metric for global temperature. In 1988, climate scientist James Hansen testified before the U.S. Congress, stating that 350 parts per million (ppm) would be a high, but acceptable concentration level. At the end of 2015, carbon concentration was about 400 ppm, a level not reached during the last 3.2 million years, until now of course. The DICE model forecast was that concentration would grow to 418 by the end of 2020. At the end of 2020, actual concentration was 415 ppm; and it currently stands at 417 ppm.
The 2016 DICE model forecasts that even with carbon priced at its social cost, annual emissions will peak around the year 2050, at which time atmospheric carbon concentration will be about 517 ppm, and the global temperature will have risen by 2 degrees. It has been more than 16 million years since atmospheric carbon concentration was more than 500 ppm.
It is important to understand that according to the DICE model featuring optimal carbon pricing, annual emissions are forecast to peak in 2050; that is, they do not fall to net zero in 2050, but in 2105! Therefore, the model predicts that both atmospheric carbon concentration and global temperature will continue to rise after 2050. By the end of the 21st century, the DICE model forecast for carbon concentration is 628 ppm, with the global temperature rising by almost 3.5 degrees. It has been 56 million years since atmospheric carbon concentration last crossed 600 ppm.
If the scenario just described is not already bad enough, it becomes much worse if there is a delay in pricing carbon at its social cost.
For contrast, consider how the long term emissions forecast developed by the consulting firm Wood Mackenzie compares to the one from Nordhaus’ DICE model. Wood Mackenzie refers to their best guess forecast as their base case, and it features emissions peaking around 2027, and then gradually declining thereafter. The Wood Mackenzie analysts point out that achieving either the 1.5 goal, or the 2.0 goal, would require emissions to decline much more sharply than they do in their base case, beginning around 2022 to achieve the 1.5 goal and around 2027 to achieve the 2.0 goal. Notably, the Wood Mackenzie forecast for the global temperature at the end of the decade lies in the range 2.5 to 2.7 degrees, well above 2.0.
There is an important reason why the estimate which Nordhaus provides for the trajectory of the social cost of carbon is lower than the estimate provided by other economists. Nordhaus built his model assuming that people would evaluate tradeoffs associated with climate change in much the same way as they evaluate other tradeoffs. By this I mean having a present-bias focused perspective when weighing the costs of consuming less today in order to provide the investment that will generate benefits in the form of even more consumption in the future.
The assumptions used by other economists are a little different, in that they assume less present-bias than does Nordhaus, a perspective consistent with a longer term view of both human life on the planet as well as broader issues such as biodiversity. My sense of this is that Nordhaus seeks to model human nature as realistically as possible; and keep in mind that to date, actual decisions about climate change have been inferior to the decisions which various versions of DICE have been recommending for decades!
Relatedly, let me mention that Nordhaus’ model treats climate-related damages as significant, but not catastrophic. This statement holds even for global temperature increases that are as high as 4 degrees, which the DICE model forecasts will occur during the next century. In this regard, the model predicts that the atmospheric temperature will peak at around 4 degrees in the year 2145, assuming carbon is priced at its social cost, beginning now.
The “beginning now” qualification is important, because a delay in pricing carbon at its social cost leads to a forecast for temperature change which is higher than 4 degrees. The model’s 2145 temperature forecast for a non-carbon priced economy is 5.5 degrees, on its way to 6 degrees and above by the year 2165, assuming that human life will be possible in such an environment. I say assuming, although the science summarized by Nathaniel Rich in his book Losing Earth suggests that even at 4 degrees, human existence will be in question.
Models are almost always imperfect representations of reality. It is possible that the DICE model seriously underestimates future damage from climate change. It is also possible that the DICE model underestimates the climate benefits from new technologies that will emerge from investment.
These two types of imperfections are incredibly important. Their presence offers reasons to be both fearful, because of potential catastrophic damage, and hopeful, because of the major benefits future technology might bring.
In describing his general framework, Nordhaus speaks of two policies which work together, one for pricing carbon and one for supporting low carbon technologies. That said, there is good reason to believe that currently we are not doing enough when it comes both to pricing carbon sensibly and to investing appropriately in low carbon technologies.
Yes, there was progress in Glasgow on Article 6, on cooperation and coordination between China and the U.S., and on the backing of the financial sector for the Paris agreement. And yes, United Nations diplomats know that the rhetoric at most meetings far surpasses the accomplishments. Still, the agreement negotiated at COP26, while it points in the right direction, appears totally inadequate relative to the great challenge before us.
Perhaps the accurate characterization of what emerged from COP26 is blah, blah or to be more charitable, just blah.