by David Wojick
The Social Cost of Carbon is absurd but entrenched.
The monster of all climate modeling efforts is the so-called Social Cost of Carbon (SCC). This involves what are called Integrated Assessment Models or IAMs. These models reflect the fact that, from a policy standpoint, the issue of climate change is simply an enormous environmental impact assessment.
An IAM begins with a CO2 emission model, then adds a climate change model, plus a physical damage-from-climate change model, followed by a global economic damage (from the physical damage) model. This complex combination of models is the ‘integration’ part.
The projected global economic damage is then discounted to present value. Finally, in a true tour de force, part of this far future damage is allocated to present day emissions on a per ton basis, even though a lot of future emissions are required for the damage projection. The result is a specific dollar amount of discounted future damages attributed by ton to today’s emissions. The SCC is based on averaging three IAMs, all of which project great future damages.
The SCC is much more than an exercise; it is a dangerous policy cancer spreading throughout the US Government. Federal agencies are increasingly using it and so are the Federal courts. But the computer modeling behind SCC is quite literally absurd.
The principal absurdity is that the modeling is over a 300 year future timeframe. This includes modeling of emissions, climate change, economic development and economic damage, all for the next 300 years. Apparently it is necessary to go this far into the future because the purported climate change damages come on very slowly, especially sea level rise. Given enough time and the necessary assumptions we can sink many major cities.
In normal impact modeling the discount rate serves to limit the timeframe to a few decades, but in SCC this does not happen. A combination of economic growth, population growth and damage growth largely offsets the discount rate. This opens the door to fantastic multi-century forecasts.
The thing is that we cannot possibly forecast this far into the future. Consider especially the issue of forecasting technological progress, because given the right technologies we can adapt to climate change, should it actually occur.
Three hundred years ago George Washington was not even born yet. The people living in the early 1700’s could not possibly have imagined today’s world. In the same way, there is no reason to believe that we are in a position today to look 300 years into the future. Yet this is precisely what the SCC computer models claim to do.
In fact it is commonly said that the pace of change is quickening. If so then the changes over the next 300 years are likely to be much greater that those over the last 300. We may actually be in a worse position for forecasting than the people were in 1717. We have more science, but nothing that lets us predict how the world will be in 300 years.
Under these circumstances it is obviously impossible for a computer model to provide meaningful 300 year projections for policy purposes. Yet this is precisely what the SCC models claim to do, or rather what the Federal agencies claim they do. It is not clear that the owners of these computer models endorse these claims, but neither do they publicly oppose them. The agencies may even be funding them.
Clearly this computer modeling is pure speculation, not science. There are no testable hypotheses here. Nor is there settled scientific support for the alarmist assumptions built into these computer models. Programming alarmism into a computer does not make it scientific. Simply put, it is absurd to base public policy on these 300 year projections.
Moreover, the specific Federal SCC results are suspicious, for several reasons. To begin with there are the big adjustments. The SCC modeling numbers were first announced in 2010. Three years later they were adjusted upward by a whopping 50% or so.
Then too the SCC per ton damages are remarkably similar to the per ton carbon tax numbers that have been proposed in Congress. The SCC numbers come in a range, based on the discount rate used. The carbon tax proposals also come in a range and the two ranges are roughly identical. This suggests that the models have been tuned to match the tax proposals.
Clearly the Trump Administration should try to rectify this situation, but it may not be easy. If they simply change the damage estimates, say by using a higher discount rate, then they are endorsing the absurdity. If they try to ban the use of the SCC, say via an Executive Order, they will run into the Court orders. Unlike Court orders, Executive Orders have no force of law; they are merely administrative.
It may be that only Congress can ban the use of the absurd SCC. It certainly should.
JC note: Here are several recent articles on this topic that I’ve flagged:
- Article on SCC by David Roberts at Grist
- WaPo article on the coming battle surrounding SCC
- Blog post by Willis Eschenbach on the SCC, with a fascinating discussion with Richard Tol in the comments
Moderation note: as with all guest posts, please keep your comments civil and relevant.