What this covers
Robert Pindyck of MIT joins Russ Roberts to discuss the economics of climate policy under uncertainty. Both speakers accept that human-caused global warming is real and directionally certain—the chain from emissions to atmospheric concentration to temperature rise to economic harm is qualitatively beyond serious dispute. The crux of the conversation turns on what Pindyck argues is a deeper problem: every magnitude in that causal chain remains deeply uncertain or unknowable, yet policy decisions require numbers. The discussion weighs whether this uncertainty should paralyze action or instead justify precautionary measures, and what discount rate is ethically defensible when weighing costs borne today against benefits accruing to future generations.
Pindyck contends that Integrated Assessment Models create a false sense of precision by treating poorly understood relationships as if they were known, a problem he connects to Knightian Uncertainty. He highlights specific blind spots: economists cannot assign monetary value to losses of biodiversity and aesthetic goods, making the true social cost of carbon unmeasurable. On discounting, he distinguishes between the rate of time preference and how the Ramsey Discounting Formula compounds to produce high discount rates, and discusses why different choices yield wildly different present values for future damages. Roberts pushes back on several fronts: he voices greater concern about the political costs of enforcing global carbon limits than about uncertain climate magnitudes, invokes the Free Rider Problem, and argues that future generations would not want present growth slowed by zero discounting. Pindyck defends a modest carbon tax as justified insurance against a low-probability catastrophic outcome, and briefly addresses geoengineering as an emergency backstop while noting its own ecological risks.
Pindyck argues that human-caused global warming is real and certain in direction, but the magnitude of both the temperature response and the economic damage is deeply uncertain (even unknowable), so the case for a modest carbon tax rests on precaution against a low-probability catastrophic outcome rather than on precise integrated-assessment model estimates.
- The qualitative chain (emissions → concentration → warming → harm) is certain, but every magnitude in that chain is unknown
- Integrated assessment models give a false impression of precision because the underlying relationships are poorly known
- A modest carbon tax is justified as a precautionary measure against a ~1-5% chance of catastrophic damage, not as an optimized policy
International carbon abatement requires unattractive control levels that pose worse political-economy risks than our climate uncertainty.
- Roberts argues he is more certain about the adverse political-economy consequences of empowering an international body to limit global carbon than about the climate magnitudes themselves: because billions of poor people need carbon-fueled growth to raise their living standards, serious global carbon abatement would require an unattractive level of international control, so he would be cautious about a serious carbon tax until worst-case scenarios appear more likely.
“what I'm not uncertain about is the political implications of giving some kind of international body the power to limit global carbon emissions... to limit growth around the world via carbon abatement is going to require a level of international control that to me is extremely unattractive, on what we call political economy or public choice grounds.”
Discounting future costs is necessary because economic growth benefits future generations tied to us by family.
- Because future generations are tied to us by family and benefit from economic growth, they would not want us to treat their costs and benefits as identical to ours in a way that ignores compounding; incurring costs today that slow growth imposes real costs on them, so they would want us to account for that — making a literal zero discount rate absurd.
“It would be absurd for us to say, well: Costs and benefits are the same 5 generations from now as they are today because it's an ethical thing; we are all the same; but they very much want us to take into account the fact that incurring costs today imposes costs on them.”