Fun With Discount Rates

Cass Sunstein wasn’t entirely pleased with a previous post I did on his paper with David Weisbach about discount rates and climate change and with some justice. So I’ll try to give a clearer and more accurate account of the contours of the dispute.

One thing people often do when engaged in cost-benefit analysis is some discounting of future costs. In other words, we regard bearing $100 in costs today as worse than bearing $102 in (inflation adjusted) costs ten years from now. This kind of issue is relevant to the climate change debate. The costs of reducing carbon emissions in 2009 will be borne in 2009. But the main benefits of reducing carbon emissions in 2009 will be accrued in the future. Consequently, insofar as you use a lower discount rate the case for more dramatic near-term action looks more compelling. As a consequence of that, different economists using the same assumptions about the costs of climate change and the costs of mitigation measures come to different conclusions about policy prescriptions based entirely on the discount rate with some saying we should discount in much the way we would if making ordinary investment decisions and others saying that there’s no ethical justification for discounting the well-being of future people in this way.

Sunstein & Weisbach agree that you shouldn’t discount future people in that way. We should, in other words, adhere to a principle of intergenerational equity. But they say that this isn’t relevant to the issue of the appropriate discount rate to use for cost-benefit analysis. Rather, the right way to think about it is that all proposals to help future people should be subjected to cost-benefit analysis using standard discount rates and then we should do the thing that’s most helpful. Now of course all that still leaves the question of what the appropriate discount rate is to use, which is tied up with various other issues and I’m pretty sure that trying to give an adequate discussion to them would take more space than is appropriate for a blog and I’m probably not the best person to ask anyway. But check out this Brad DeLong post if you want to see some equations.