Paul Krugman has a blog post about one of my favorite economists, Marty Weitzman. He has the central point right, which is that “on any sort of expected-welfare calculation, the small probability of catastrophe dominates the expected loss.”
But Krugman’s general lack of understanding of global warming — and his willingness to believe anything Bj¸rn Lomborg says — undermines his entire analysis:
Bjorn Lomborg … says that climate change will reduce world GDP by less than 0.5%, so it’s not worth spending a lot on mitigation.
Weitzman’s point is, first, that we don’t actually know that: a small loss may be the most likely outcome given what we know now, but there’s some chance that things will be much worse. (Marty surveys the existing climate models, and suggests that they give about a 1% probability to truly catastrophic change, say a 20-degree centigrade rise in average temperature.)
… Suppose that there’s a 99% chance that Lomborg is right, but a 1% chance that catastrophic climate change will reduce world GDP by 90%. You might be tempted to disregard that small chance — but if you’re even moderately risk averse (say, relative risk aversion of 2 — econowonks know what I mean), you quickly find that the expected loss of welfare isn’t 0.5% of GDP, it’s 10% or more of GDP.
Well, ‘yes’, on the final point, but ‘no’ on every other point.
Indeed, a 20°C rise in average global temperature — which translates to perhaps 50°F warming over much of the inland U.S. — is “James Lovelock” territory where “the Earth’s population will be culled from today’s 6.6 billion to as few as 500 million.” Catastrophic climate change is anything significantly over 3°C, which is not a 1% chance, but a near certainty if we don’t reverse greenhouse gas emissions sharply and soon (see “Is 450 ppm politically possible? Part 0: The alternative is humanity’s self-destruction“).
Lomborg, of course, does not have anywhere near a 99% chance of being right that “climate change will reduce world GDP by less than 0.5%.” Indeed, if we actually followed Lomborg’s do-nothing prescription, then he has precisely a zero chance of being right. He is a pure disinformer (see “Lomborg skewers the facts, again” and “Debunking Bj¸rn Lomborg — Part III, He’s a Real Nowhere Man“).
Weitzman’s analysis is, however, very important for traditionally economists — and everyone else — to understand, so let me reprint my September post, Harvard economist disses most climate cost-benefit analyses, below:




Language Intelligence: Lessons on persuasion from Jesus, Shakespeare, Lincoln, and Lady Gaga
