Skip to content

On climate change, the uncertainties multiply— literally.


Climate change faces a neglected actuarial problem. Too many conditions must be met to warrant a policy action on climate change.

The following four stipulations must each be highly probable: 

1. Global warming will accumulate at 0.12 degrees Celsius or higher per decade.

2. It is anthropogenic, due largely to carbon dioxide emissions.

3. The net effect is harmful to human well-being in the long run.

4. Preventive measures are efficient, that is, feasible at the costs not exceeding the benefits.

But even if the probability of each of these stipulations is as high as 85 percent, their compound probability is as low as 50 percent. This makes a decision to act or not to act on climate change equivalent to flipping a coin.

It is much simpler with pandemics, earthquakes, floods, tsunamis, fires, terrorism, and other natural and anthropogenic catastrophes. The probability there is only of the event itself, which is usually 100 percent in the long run in certain locations or globally, and of the efficiency of proposed measures. If the latter are efficient (for example, vaccines against pandemics, building reinforcement against earthquakes, dikes and levees against floods, surveillance and epicenter destruction against terrorism), the compound probability is sufficiently high, perhaps over 90 percent, and the preventive action is warranted.

In the case of climate change, the conditions are four. They are not random, nor are they arbitrary. To see this, one can run a thought experiment and drop or ignore any of the above foursome. At once, the entire call for action on climate change becomes pointless. If global warming is not ongoingthere is no need to stop it. If it is not anthropogenic, there is no need to curb carbon dioxide emissions. If it is not harmful, there is no need to worry. If preventive measures are inefficient, they would not help and there is no use applying them. It follows that all four conditions are necessary. If just one of them does not hold, action is unnecessary or useless.

This means they do not overlap or substitute for one another. In technical terms, they are independent events. The probability of each of them occur-ring does not affect the probability of any other occurring. Therefore, to decide whether or not a policy action is warranted, we have to evaluate their compound probability. The probabilities of all four events have to be multiplied together. This defines the calculus of choosing action or inaction. 

Alas, at the current state of scientific knowledge and global policy coordination it is hard, indeed impossible, to obtain the compound probability of the four conditions higher than 50 percent. The Intergovernmental Panel onClimate Change, a UN affiliate, expressed at various times about 90–95 per-cent confidence in the continuity, anthropogenic causes, and harm of climate change. If the broad scientific community shared this level of certainty, there would have been no controversy over climate change. The only remaining dis-pute would have been over the most efficient policy measures and their global coordination. This is not the case. Some scientists doubt that global warming will constitute a long-term trend. Others question its anthropogenic origin. Yet others maintain that the net effect may not be harmful for living conditions.

The continual fight for action on climate change is itself a proof that the scientific problems are not resolved. People do not argue whether or not to combat epidemics or terrorism, only how to do it efficiently.

To act or not to act ultimately depends on whether the estimates on the first three conditions exceed the highest bound of 85 percent and the fourth, the policy efficiency probability, is not much lower. That is, whether the compound probability is greater than 0.85 × 0.85 × 0.85 × 0.82 = 0.50. No one knows for sure whether the 85 percent bound of scientific consensus has been reached on the first three conditions. […]

Until science and policy coordination provide a higher certainty, a more modest program may suffice…

To sacrifice the economic progress of billions of people to a result on par with that obtained by flipping a coin is not a defensible proposition. In policy as in medicine the fundamental principle is: first, do no harm.

Full post