“All models are wrong but some are useful”
— British statistician George E. P. Box
In my climate change investing speaking engagements, doubters love to bring up what they think is the killer counter-argument: The mathematical models used by climate scientists and published in IPCC reports “have never been right.”
For such occasions, I have prepared a slide deck appendix that shows just how amazingly accurate and consistent peer-reviewed climate models have been over the last 40-plus years.*
Because the science of climate change has been so carefully researched and argued, there is really very little fundamental uncertainty about the most critical questions; everyone knows that future ecological conditions will range from bad to catastrophic.
Because the evidence of anthropogenic climate change is so strong and the scientific basis so clear, commercial and political interests opposing sensible mitigation policies can take only one possible tack: Attempt to minimize the potential impact of the perfectly predictable future warming.
When researching my last article, I happened across a critique of William Nordhaus’s work regarding the likely economic costs of climate change as estimated by his Dynamic Integrated Climate-Economy (DICE) model. This model, which is used by the IPCC in their assessments of the state of the global climate, was one of the pieces of work for which Nordhaus won the 2018 (fake) Nobel Prize.
The critique of Nordhaus’s DICE model came from the enormously creative, though admittedly heterodox, post-Keynesian Australian economist Professor Steve Keen. I have listened to some of Keen’s YouTube lectures about debt levels — he was one of a few maverick economists that correctly predicted the 2008-2009 mortgage crisis — but did not know that he had done any work on climate change economics.**
To be honest, I had never looked carefully at the DICE model before for one very good reason: it is clearly ridiculous.
According to DICE, the global economy will suffer an aggregate drop in GDP (i.e., all-in over the next 130 years, not annually in perpetuity) of a few percentage points even assuming a temperature increase high enough to cause agricultural output to plummet.
While Ivy League economists may not have a visceral sense of this, it is clear to your correspondent that large swathes of the workforce might be marginally less productive if they were only able to consume 500 food calories a day.
Keen is a better man than I in that, like me, he could see that the output of the DICE model was ridiculous, but still took the time and effort to crawl through the details to figure out why.
According to Keen, DICE’s egregious errors are threefold:
- The assumption that the small effects on GDP from the modest changes in global temperature to date can be extrapolated into a future of much more extreme temperature increases,
- The assumption that a complex adaptive (i.e., non-linear) system like our planet’s ecosystem would respond to extreme stimuli in a linear way,
- A dubious mischaracterization of climate scientists’ assessment of ecological “tipping points” that leads to an assumption these potentially catastrophic points will never be reached.
No admirer of the statistician Box would expect the DICE model to be anything but wrong, but — considering that the UN’s IPCC is relying upon DICE to inform the opinion of policy makers — we should at least expect it to be useful.
Reading Keen’s critique, however, I have realized that DICE acts as a literal weapon of mass destruction since it has the effect of breeding climate complacency among world leaders.
If what I say is true, why is DICE not roundly criticized and ignored?
I believe it has been adopted and its creator lauded because it represents the kind of palatable fiction that human decision makers find so comforting. Unfortunately, in investing as in life, the bug of palatable fiction sooner or later finds a windshield of harsh reality.
Or, as a more eloquent Robert Louis Stevenson wrote, “Everybody, soon or late, sits down at a banquet of consequences.” (Shout out to Kara Swisher…)
Intelligent investors take note.
* My favorite example of an early predictive climate model comes courtesy of scientists at ExxonMobil who published a paper in 1980 that forecast temperature increases caused by increasing CO2 concentrations that are amazingly accurate compared to actual temperature increases in the intervening 40 years and current state-of-the-art climate models.
Here is a quote from that report:
“The most widely accepted calculations carried on thusfar on the potential impact of a doubling of carbon dioxide on climate indicate that an increase in the global average temperature of 3+1.50C is most likely. Such changes in temperature are expected to occur with uneven geographic distribution, with greater warming occuring [sic] at the higher latitudes i.e., the polar regions. This is due to the presumed change in the reflectivity of the Earth due to melting of the ice and snow cover (See Figure 3). There have been other calculations on a more limited scale by a number of climatologists which project average temperature increases on the order of 0.25oC for a doubling of C02. These calculations are not held in high regard by the scientific community. Figure 4 summarizes the results presented in the literature on the possible temperature increase due to various changes in atmospheric C02 concentration.”
Check those forecasts with those in the most recent IPCC report. You’ll find an astounding agreement with the Exxon forecast.
Despite the outcome of the NY Attorney General’s suit against ExxonMobil in December of last year, if you are a shareholder of ExxonMobil, it might be time to start wondering if you’re on the right side of history, let alone on the right side of this trade…
** Professor Keen has a Patreon site with some excellent subscriber-only content. After reading through his eye-opening work on climate change economics, I decided to support his work with a monthly contribution.