Thanks for taking the trouble to write such a lengthy and helpful comment. I do feel remiss in that I too often comment on things like the Stern Report and simply don’t have time to read the underlying documents. I’ve seen when I do read them that they are frequently mischaracterized, often not in bad faith, simply that a an article has to tell a story and journalist can’t be very nuanced in a short space.

The point you make about societal vs. individual preferences is very important, and unfortunately, most of the lay users of DCF approaches (finance and corporate types) don’t even consider the fact that there ought to be a distinction.

And a real discount rate of 1.4% is credible. I will try to put an addendum on subsequent posts on the Stern report to clear this matter up, since I didn’t have it right.

And thanks for your kind words! They are particularly welcome today, since I had a go with a couple of change deniers earlier.

Came across this a while ago but my proxy at work does not allow me to launch blogger comments. Wanted to correct this one thing though, seeing as it appears you are interested in this subject. Stern is not employing a .1% money discount rate. .1% refers to a parameter in the Ramsey formulation of interest rates known as the Pure Rate of Time Preference (and the applicable version in this case would be the Pure Rate of Social Time Preference). This parameter discounts utility not money/goods. The other parameter of Ramsey discounting is risk aversion. Taken together with his estimate of per capita income growth, Stern uses an effective discount rate of 1.4%. Keep in mind, this is a real, not nominal discount rate. That means for comparison purposes, you’d be comparing with real yields, and certain economists would argue for various reasons that the appropriate yield is the government bond yield. The long term data on that is not very far off Stern’s assumption, (though it is higher), and hence the numbers are not so outrageous (1.4% real vs. .1% nominal is a pretty large difference).

Saying all that, many economists criticize this assumption as being Philosopher King-esque, and indeed Stern alludes to moralist language in the Review to sustain to some degree these parameter choices. The intriguing thing with the PRTP is that if you set it very high, you are effectively saying that the welfare of a few generations hence is meaningless to people here and now (to society). Societal preferences are not like individual preferences- they require less than perfect assumptions to back out from individual behavior. There simply aren’t a lot of immortals running around out there whose revealed preferences tell us about pure rates of time preferences but that is precisely what society is. Even if people had settled on a particular ordained societal PRTP, it’s an highly simplified construct in the first place, and not necessarily a good one given that Ramsey discounting can’t even remotely explain why the risk free rate or the equity risk premium are what they are in the first place. In practice however, there are very good arguments to be made for very low or zero social PRTPs, and there are good arguments to be made for asymptotically declining discount rates applied to social projects.

Be that as it may, Stern is not off the hook. Some have pointed out, fairly, that The Review chose contradictory parameters for modeling in the IAM, (and, of lesser importance, high end estimates for damages), in particular with cross-sectional and intertemporal risk aversion, and based on these choices, many have accused him of reverse engineering a result. Something characterized like this isn’t likely to hold up to policy scrutiny. Weitzman as you point out is one of these who criticizes the review, but has gleaned some salient points from it, which can be summarized (roughly from the review)~ Averaging across probabilities conceals risks. That is an allusion to the impact of uncertainty and it is something Weitzman explores analytically, I think, to great effect (and no surprise this type of approach would more clearly argue for action). As a non-economist who has interacted with known climate change economists in the blogosphere, I can say that I have generally been highly unimpressed as to the way in which many of the estimates of damages from climate change have been estimated. The more you find out, the more scary the picture (one told me blithely that heat stress has no impact on drinking water- apparently he’s never been to the American Southwest or any of the major Asian nations that rely on Himalayan snow melt). In any case, you attention to this subject is welcome, and more broadly, your blog is superb. I would only warn you to beware the source, especially if it’s ever the WSJ editorial page!