When bad economics and climate science collide

One of my favorite economists, Stephen DeCanio, has a good piece at Real Climate Economics, reprinted below. In addition to his many awards and publications, DeCanio was Senior Staff Economist at the President’s Council of Economic Advisers from 1986 to 1987.

When bad economics is applied to climate change, the result is often at odds with climate science. This tendency was on display at a recent series of panel discussions, “The Bloody Crossroads of Science and Policy” held by the American Enterprise Institute. During the Q&A following the panel on climate, a question from the room about discounting and the time scales of climate change elicited the following responses from two of the economists on the panel, Robert Mendelsohn and David Montgomery (direct quotes):

Mendelsohn: “It turns out that this problem with discounting — this problem with looking into the future — I think is a failure by scientists to understand why society has a value on time. The scientists just basically don’t understand that you can invest things today and that they’ll be worth more in the future because of it. That actually is the reason why we discount — is because a dollar today is worth more than a dollar in the future. So therefore something in the future is worth less than it is today. That concept is a basic principle of economic analysis. And to ignore things like that is the same thing as an economist going back and saying, ‘Well I find it inconvenient that greenhouse gases trap heat, so I’m going to assume that they don’t.’ You just can’t assume things that are fundamental to a field away just because you don’t like the outcome.”

Montgomery: “On discount rates, it’s not a conflict between science and economics. It’s a conflict between a rather amateurish application of utilitarian ethics to argue that discount rates should be zero versus kind of an effort in positive economics to explain what happens when you invest.”

The appropriate choice of discount rate in climate economic models has been widely discussed by economists, and important new insights have emerged in recent years. The low quality of the economic analysis exhibited in the quotations by Mendelsohn and Montgomery is notable. Consider the following points which have been established in the economics literature:


(1) In a world of multiple goods, there is no single discount rate, but rather many discount rates. If “the environment” and “produced goods” are not good substitutes, the discount rate for produced goods can be much higher than that for the environment.

(2) In the simplest standard growth model without uncertainty, the discount rate that emerges depends on the rate of growth of consumption. Positive economic growth leads to a positive discount rate because the future generations will be richer than we are. However, environmental disaster could lead to a decline in consumption, and thus even possibly a negative discount rate.

(3) In overlapping-generations models with only one good and without uncertainty, multiple equilibria with different discount rates are possible, even under the highly restrictive assumption of “rational expectations.” There are even more possibilities for discount rates under other assumptions about expectations, and there is no scientifically agreed-upon theory of expectations.

(4) Going a bit deeper, there is no good reason to believe that simple growth models without uncertainty are appropriate for analyzing a world in which uncertainty is intrinsic. With uncertainty, the discount rate(s) depends on the riskiness of the asset(s) being examined. Stocks have a higher expected rate of return (or “discount rate”) than T-bills, because stocks are riskier than T-bills and the higher return is required to induce investors to hold stocks rather than T-bills. Insurance-type assets have lower (or even negative) “discount rates” compared to investments in ordinary capital goods.

(5) With the possibility of catastrophic losses, risk avoidance dominates discounting even in the most basic growth model because expected future losses can be very, very large. (This has been demonstrated unambiguously by Martin Weitzman in recent path-breaking work.)


[JR: For a discussion of Weitzman’s work, see Harvard economist: Climate cost-benefit analyses are “unusually misleading,” warns colleagues “we may be deluding ourselves and others.”]

(6) All of the above points have been made within a very specific and limited utilitarian framework that includes the implicit assumption that it is possible, at least in principle, for the beneficiaries of a policy to compensate the losers and still having something left over. But for an intergenerational problem like climate in which such transfers are impossible (because time travel is impossible), there is no escaping the need to make moral judgments to balance the interests of those whose lifetimes do not overlap. The “amateurish” application of utilitarian ethics is the common presumption in economics that the material standard of living is the sole yardstick of well-being or right action. Many other moral standards have an equal or greater claim to our adherence. For example, the four classical cardinal virtues””wisdom, justice, courage, and moderation””are not based primarily on the material standard of living at all.

(7) Finally, as a purely practical matter, the future of the climate is linked to the path of economic progress taken by the least-developed countries, and to the foreign policy interests and actions of the great powers. Economic analysis that ignores these elements fails to incorporate the first requirement of good policy””that it be feasible in the real world.

Mendelsohn and Montgomery were speaking at a panel, so allowance should be made for the fact that spoken arguments are usually not as tight as written ones. However, the problem with their statements goes beyond style. Their arguments are simply incorrect

Stephen DeCanio

Related Posts: