When the Stern Commission came out with its famous report on global warming last year, suggesting drastic and immediate action to reduce emissions, Sir Partha Dasgupta issued one of the two most notable critiques. (The other came from William Nordhaus of Yale).

Now, however, Sir Partha is critiquing Bjorn Lomborg's new book. I regret that I cannot read the entire paper, which is behind a pay barrier, but Mark Thoma has a substantial excerpt:

Unfortunately, Lomborg's thesis is built on a deep misconception of Earth's system and of economics when applied to that system. The concentration of CO2 in the atmosphere is now 380 p.p.m., a figure ... in excess of the maximum reached during the past 600,000 years. If there is one truth about Earth we all should know, it's that the system is driven by interlocking, nonlinear processes running at different speeds. The transition to Lomborg's recommended concentration of 560 p.p.m. would involve crossing an unknown number of tipping points (or separatrices) in the global climate system. We have no data on the consequences if Earth were to cross those tipping points. They could be good, or they could be disastrous. Even if we did have data, they would probably be of little value because nature's processes are irreversible. One implication of the Earth system's deep nonlinearities is that estimates of climatic parameters based on observations from the recent past are unreliable for making forecasts about the state of the world at CO2 concentrations of 560 p.p.m. or higher. ...

These truths seem to escape Lomborg. His cost–benefit analysis involves only point estimates of variables..., implying that he believes we shouldn't buy insurance against potentially enormous losses resulting from climate change. ...

The integrated assessment models of Earth's system on which Lomborg builds his case are arbitrarily bounded on either side of his point estimates. It can be shown that if those bounds are removed (as they ought to be), even a small amount of uncertainty — when allied to only a moderate aversion to uncertainty — would imply that humanity should spend substantial amounts on insurance, even more than the 1– 2% of world output that has been advocated. If the uncertainties are not small, standard cost–benefit analysis as applied to the economics of climate change becomes incoherent, even if those uncertainties are judged to be thin-tailed (gaussian, for example); this is because the analysis would say that no matter how much humanity chooses to invest in protecting Earth from passing through those later tipping points, we should invest still more.

The catastrophic unknowns are, to my mind, the biggest economic challenge of global warming. As Sir Partha points out, we literally don't know what we don't know. And we have no very good economic framework for dealing with catastrophic risks when we cannot evaluate either the extent or the probability of the disaster. We may misunderestimate the probability of thousand year storms, but at least we've got a pretty good idea what a hurricane looks like.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.