When “Conservative” Means “Alarmist”

Global Science Report is a feature from the Center for the Study of Science, where we highlight one or two important new items in the scientific literature or the popular media. For broader and more technical perspectives, consult our monthly “Current Wisdom.”

There is a new study out that purports to make a “conservative” estimate of the social cost of carbon and in doing so arrives at a figure nearly four times larger than the central estimate currently used by the U.S. government—the latter a figure which we and others have voluminously argued is itself several times too high. Perhaps the authors of the new report ought to look up the definition of the word “conservative.”

Recall that the social cost of carbon is supposed to represent the total value of future damages from climate change resulting from the current emission of a ton of carbon dioxide. As you may imagine, coming up with the SCC involves more imagination than actual science.

The primary “tools” used for determining the SCC are “integrated assessment models,” or IAMs, which incorporate a very simple climate model into an economics model. Writing in the journal Nature Climate Change, Jeroen van den Bergh and Wouter Botzen review elements (economic and climatic) that are poorly incorporated or missing entirely from the IAMs.

A prominent characteristic of the IAMs is that they are notoriously malleable and able produce virtually any value for the SCC that the modeler or end-user desires.

Judging from the introductory sentence of their paper

Climate change has been called “the biggest market failure the world has seen” and “the mother of all externalities.”

you can pretty much guess what kind of SCC value van den Bergh and Botzen prefer.

To support their apparent preference for a high SCC, they spend the bulk of their paper imagining bad climate outcomes—with high monetary damages—and are generally dismissive of positive climate impacts. For example:

Nevertheless, our summary of the main effects provides a clear insight, namely that unquantified negative effects of climate change tend to domi­nate unquantified positive effects. The negative effects comprise large biodiversity losses, political instability, violent conflicts, large-scale migration, extreme weather events, natural disasters and the effect on long-term economic growth. Accounting for the latter is likely to increase the SCC because large impacts of cli­mate change are expected to reduce the rate of GDP growth, partly because of negative effects on labour and capital productivity.

Unsurprisingly, when you include a lot of negative impacts along with a low discount rate, the IAMs produce very high estimates of the SCC.

In fact, van den Bergh and Botzen arrive at a “conservative” SCC value of $125. For comparison, value used by the Obama Administration for cost/benefit analyses of new regulations is $36.

Interestingly, in their “conservative” analysis, they never once mention the growing body of new and prominent scientific literature that produce updated estimates of the earth’s climate sensitivity—a measure of how much climate change we expect from carbon dioxide emissions—that are much lower and much more tightly constrained than the ones used in all of the studies reviewed by van den Bergh and Botzen.

The lower climate sensitivity estimates not only reduce the overall impacts from expected climate changes, but they do so primarily by reducing the chances of unexpected and catastrophic changes—the biggest drivers of the high SCC values in the IAMs. It has been repeatedly shown (see here, here, and here for example) that incorporating the new, lower climate sensitivity estimates reduce the IAMs’ SCC determinations by some 40 percent.

And there are lots of other things, which, if better incorporated in the IAM’s, would lead to lower SCC values.

If the positive benefits from carbon dioxide emissions on the planet’s crop production were better included in the IAM’s, the SCC value drops further.  And if arguments for the use of a higher discount rate, rather than the very low one espoused by van den Bergh and Botzen win the day, the SCC drops further still.

Add to the mix a more reasoned view of future climate extremes, and before you know it, it is an easy argument to make that the SCC value should fall significantly below the Administration’s $36 rather than some three to four times higher.

It is bad enough that van den Bergh and Botzen present a rather one-sided view of the science of climate change/climate extremes and the economics concerning the choice of discount rate, but for them to term their analysis “conservative” is really taking things too far. “Alarmist” would be a more apt description.

Our hope would have been that the reviewers for Nature Climate Change would have caught the glaring oversight of the current climate sensitivity literature (with one of the most persuasive articles appearing in the sister journal Nature Geosciences), but that didn’t happen. We’ll withhold speculation as to why that was the case.


Van den Bergh, J.C.J.M., and W.J.W. Botzen, 2014. A lower bound to the social cost of CO2 emissions. Nature Climate Change, 4, 253-258, doi:10.1038/NCLIMATE2135.