According to economists, if carbon dioxide emissions impose costs on other people you should pay a tax to cover those “externalities” so the personal cost of those emissions matches their “social cost”. And some economists have big complex “Integrated Assessment Models” that take guesstimates from climate models and combine them with conjectures from economic models, model how the models model things, and come up with a magic number called the Social Cost of Carbon, which tells governments how much tax you should pay. But since the models were first put together in the 1990s a lot of evidence has been collected about how CO2 benefits agriculture and doesn't seem to be driving temperatures up as much as people feared. So a team of number crunchers updated one of the big models and discovered that the best estimate of the Social Cost of Carbon is... zero. Or maybe slightly less. At least until 2050. So the right carbon tax is nothing, or perhaps a small subsidy.
Not that we're about to take numbers from a model at face value just because they go in a direction we find congenial. There are other Integrated Assessment Models that can be tweaked to give very high Social Cost of Carbon estimates. But what we like about this new study is that the authors used one of the Environmental Protection Agency's own models and made two simple changes, both rooted in something approaching real world data.
First, they took account of the evidence on global greening. They make the remarkable observation that, of the three big economic models governments rely on to calculate CO2 damages, two of them assume CO2 doesn't benefit plant growth at all. Right away that guarantees those models will give the wrong answer, because by leaving out an important benefit they overstate the harm. The third model allows a small benefit from CO2 fertilization, but the authors point out that lots of evidence has been collected indicating the benefits are higher than people anticipated when the model was first built.
Then they look at the empirical evidence on climate sensitivity, that is, how much a doubling of atmospheric CO2 would increase temperature. We've talked about this concept before and we've hinted that new evidence points to it being lower than was estimated to be as recently as a decade ago. Since the economic models depend on assumptions about climate sensitivity, and assume it’s higher than it now seems to be, it's long past time to update them with this new data too.
Once they did these two updates, the authors re-ran the model and found the estimated Social Cost of Carbon fell dramatically, dropping from the US$30 to 40 per tonne range to basically zero. In fact they found there's a strong probability the Social Cost is negative for at least the next forty years, meaning the benefits of extra CO2 completely offset any costs from warming, even before we tally up the benefits from using the energy in the first place.
As one of the study's coauthors, Ross McKitrick of the University of Guelph, points out on his website, "We get evidence that the marginal damages of CO2 emissions are basically zero through the mid-21st century. In other words even if you accept mainstream climate science it still doesn't justify costly policy measures."
If you doubt the “mainstream” science then of course you wouldn't support costly anti-energy policies. Now we know they can't be justified even if you do accept the mainstream.