Becker-Posner on global warming & discounting

There is a useful discussion at the Becker-Posner blog site on choosing a discount rate in global warming exercises. Posner favors choosing a relatively low rate to offset the possibility of catastrophic risks while a more orthodox Becker favors discounting at the rate of return on capital.

Early studies of the costs of warming (such as Nordhaus) used a 3% discount rate (declining over time to 1%) while the recent Stern Review uses a rate of close to zero. Because many of the benefits of dealing with warming are in the distant future Stern provides much higher estimated damage costs of not dealing with warming than did Nordhaus. Nordhaus also predictably derives much less binding taxes than does Stern.

Who is right here? Well Posner has a point. He is concerned that low probability catastrophic events (such as extreme temperature rises) could have such disastrous outcomes that it makes sense to put considerable weight on them because their expected cost is so high. This point is also made by John Quiggin and argued more formally in an essay by Robert Pindyck. I tried to summarise elements of the discussion here.

If you ignore such catastrophic risks and assume that the rate of return on capital will remain at around 3% then Becker’s discussion makes sense. If the expected damages in 50 years time from climate change is $2 trillion then at 3% the value of those damages today is $500 billion so if current costs of mitigation were anything above $500 billion it would be a move bet to mitigate. For example if it cost $800 billion to mitigate today one could invest this at 3% and transfer this to those impacted on by warming in 50 years. The transfer would amount to about $3 trillion so future generations would be better off with this transfer even if they had to tolerate the costs of warming.

My intuition is to broadly back Posner’s view. The 4th IPCC Report does raise the possibility of even more dramatic possible increases in temperature than envisaged in the 3rd report. The range of possible temperatures to year 2000 has increased and the maximum possible forecast increase has risen to 6.4o C which could be catastrophic – it exceeds the range of warming experienced at the present time since the last ice age. Should society risk all to enjoy a 50% higher payoff (from $2 to $3 trillion) in 50 years? Certainly not if the risks of catastrophe are large enough and if current policies can mitigate the chances of extreme climatic outcomes.

Uncertainty over forecast future climate changes reduces the discount rate that should be applied and thereby increases the case for taking decisive action now.