UNCERTAINTY STRENGTHENS THESE CONCLUSIONS
The conclusions above were stated without repeated hedges about the vast uncertainty surrounding the effects of climate change. Climate science and climate change policy, however, are plagued by uncertainty. We do not know how much temperatures will go up with emissions. The range is large. We know even less about the physical effects of temperature increases. Major issues such as the extent and speed of the melting of Greenland’s and the Antarctic ice sheets defy understanding. We also do not know the economic and social harms that these effects will cause. And we do not know the costs of reducing emissions. It is not wrong to say that the core problem of climate change is one of making choices in the face of deep, irreducible uncertainty.
To develop firm conclusions in the face of uncertainty, I tried to consider the best and the worse cases given what we know and see what conclusions follow. If, for example, even under the best case (i.e., low climate sensitivity) temperatures continue to increase with emissions and that at some point the resulting harms will be intolerable, we know that emissions must eventually be reduced to zero.
It is important, however, to understand whether and how our uncertainty about the effects of climate change might affect these conclusions. Perhaps if we really do not know what is going to happen, we should be cautious about spending the substantial resources needed to transform our currently well-functioning energy system. We have plenty of other problems to solve in the meantime. One hundred years is a long time. Maybe we should stay the course and hope for inexpensive clean energy to come along. One hundred years ago, who could have imagined today’s world?
The uncertainties in climate change, however, are not symmetric. The possible (but unknown) downsides are almost unbounded. If we do nothing and get unlucky— temperatures are highly sensitive to carbon dioxide and the harms from temperature increases turn out to be worse than expected—the costs may be terrible. The possible downsides from reducing emissions too much, however, are bounded. We can estimate the cost of replacing our fossil infrastructure with clean energy assuming the worst case, highest possible replacement costs, so we know, within some bounds, what the worst case is for doing too much.17 That is, if we do too little and we get unlucky (i.e., climate change is far worse than expected) the effects will be very bad. If we do too much and climate change turns out not to be much of a problem, the effects are limited. Moreover, the fossil fuels we will have kept in the ground are still there to be used.
Uncertainty, therefore, should make us more cautious than otherwise. The core uncertainty is that climate change may be far worse than we expect, so we should be willing to spend more to avoid these very bad cases. That is, the conclusions above, which did not consider uncertain but very bad outcomes, were (intentionally) conservative. If anything, we should be doing more than was suggested.
There is also uncertainty about the costs of clean energy technology, as already mentioned. We might invent much less expensive energy sources or, perhaps, ways of capturing atmospheric CO2 and storing it in a safe place. If we spend resources today to reduce emissions, those expenditures will have been wasted. Perhaps we should not sink costs into reducing emissions when waiting might reduce those costs or reveal more information about the best way to meet our goals.
This uncertainty, however, is counterbalanced by uncertainty about harms. Carbon emissions are effectively permanent, so they are effectively sunk as well. We have the choice of two sunk costs: the costs of clean energy infrastructure or the costs of atmospheric CO2. Waiting might reveal that either one is cheaper or more expensive than we expect. Whether these offsetting uncertainties make us want to speed up or slow down reductions is hard to say.
The problem of deep uncertainty is challenging and deserves far more attention than I can give it here. I do not, however, think a resolution of the problem would alter the conclusions except possibly to strengthen them, to argue for faster reductions in emissions than I have suggested.