Discounting Sunstein

by John Quiggin on August 22, 2008

David Weisbach and Cass Sunstein (h/t Nicholas Gruen) have written a piece for the AEI Center for Regulatory and Market Studies[1] weighing into the debate about climate change and discounting, promising “A Guide for the Perplexed”. But their treatment of the topic is only like to add to the perplexity of anyone interested in resolving the issues.

Shorter Weisbach-Sunstein: If ethical considerations suggest that the most appropriate discount rate is 1.4 per cent, but the market rate of return is 5.5 per cent, it’s best to use the 5.5 per cent rate in evaluating climate change policies.

As a hypothetical statement this is broadly correct. Weisbach and Sunstein illustrate the point by considering someone choosing between an investment yielding a 10 per cent increase in value over 2 years and the alternative of putting money in the bank at 6 per cent, which is clearly superior.

The problems arise for the reader who tries to plug in real numbers. According to today’s New York Times the rate of interest on 10-year Treasury bonds is 3.84 per cent. Assuming (optimistically) that the Fed manages to hold inflation at 2.5 per cent over the next years, that’s a real rate of 1.34 per cent, almost exactly equal the rate quoted as justified on ethical grounds.

So to restate Weisbach and Sunstein with real numbers:If ethical considerations suggest that the most appropriate discount rate is 1.4 per cent, but the market rate of return is 1.3 per cent, it doesn’t matter which one you choose.

To be fair to Weisbach and Sunstein, all they are doing here is offering an uncritical restatement of the way the position is stated by people like Nordhaus, who should know better. Nordhaus, like nearly everyone else in the high discount rate camp, takes it for granted that the “market rate return” means something like “the average rate of return to capital”, even though this average is dominated by the risk premium for equity, which in turn is driven, in standard models, by the correlation between returns to equity and aggregate consumption. There are two big problems here
(i) The equity premium ‘puzzle’, that the risk premium for equity is much larger than standard theory says it should be. The same theory is used by Stern to estimate the appropriate discount rate This is why, if you use the average rate of return to capital, you get the gap Weisbach and Sunstein are talking about.
(ii) Second, even if the risk premium for equity is an appropriate market measure of the risk involved in public investments with return distributions similar to those of private investments, there’s no reason to think investments in climate change have these characteristics and plenty of reasons to think the opposite.

On reflection, I don’t think fairness requires this concession. The point that it’s the bond rate and not the average rate of return to capital that should be used in discount climate change mitigation ins’t new. William Cline was pushing it 15 years ago and it’s had a thorough airing in the debate over the Stern Review (I started early on in both blog pieces and journal articles and Stern made the same point in his Ely lecture, to which I can’t find a good link right now). It’s possible to make counterarguments, but Weisbach and Sunstein don’t even recognise there’s a problem here.

Coming back to the hypothetical, if the market rate of return is far above that suggested by economic analysis as socially optimal, the policy implication is that action is needed to increase savings and investment, so that the market rate and the socially optimal rate move closer together. Depending on your policy predilections, you might look for tax distortions to remove, externalities to internalise or failures of strong rationality assumption, requiring additional public savings and investment. There’s a big literature on all this, but I must admit reading it tends to increase rather than reduce the level of perplexity.

Finally, as I observed a while back, I’ve only paid attention to Sunstein only occasionally and over the last couple of years, and I’ve always been a bit puzzled by him. Despite his big reputation as a liberal thinker, everything I’ve seen from him has been both rightwing and weakly argued. Kathy G, with the benefit of a much closer acquaintance, has the same view, only more so.

[fn1]. This “builds on the success of the AEI-Brookings Joint Center which leads me to the discovery that this center is no longer in operation. Given the long descent of AEI into the worst kind of propaganda outfit, that’s not before time.

{ 20 comments }

1

Nicholas Gruen 08.22.08 at 9:22 am

Kathy G having the same views only more so is no surprise. That’s not a criticism btw. Kathy G is fun reading.

2

A. Y. Mous 08.22.08 at 11:00 am

This may seem naive, but is it that difficult to comprehend and accept life, lives, lifestyle, in that order, and work out the numbers for carbon emission and punitive damages?

3

Cranky Observer 08.22.08 at 11:45 am

> Nordhaus, like nearly everyone else in the high discount rate
> camp, takes it for granted that the “market rate return” means
> something like “the average rate of return to capital”, even
> though this average is dominated by the risk premium for
> equity, which in turn is driven, in standard models, by the
> correlation between returns to equity and aggregate consumption.

Perhaps I am totally off the track here, but it seems to me that the concept of “return to capital” requires that there be firms that can generate that return. If the continued existence of those firms and their return is correlated with the events the rate is being used to analyze then there is a serious danger of overestimating the return and underestimating the risk. That is to say, if firms operating at high rates of return have a direct positive input to the probability of the coasts of the Americas being flooded to a depth of three meters…

Cranky

4

SJ 08.22.08 at 12:29 pm

Here’s a link to the Weisbach & Sunstein paper.

Haven’t read it yet.

5

a different chris 08.22.08 at 12:54 pm

I’m not even going to read the stupid paper.

Not to insult the quality practictioners of the art like the good Prof. Quiggin here, but consulting economists on the management of the very physical existence of our planet itself is like having astrologers running your space program.

I don’t care about discount rates that you can make give any answer you want. I don’t care about “prosperity” measurements, the assurance that a rising tide of something we print on paper (or even just enter into computers) is going to make everybody’s life better.

Economics, which is simply the slippery science of statistics with an overlay of the most claptrappy “just so” stories of anthropology (Rational Man) held together with an almost completely invented baseline of “worth” just isn’t up to

I am about to go to work. Shortly I will come across a sweeping downhill curve. If I cross over into the other lane I can go faster, and it will actually save fuel as well as time since it is, as I said, downhill. It’s a very lightly traveled road and I can therefore heavily “discount” the possibility that their will be somebody coming the other way, and even then I can count on their “rational” behavior to assist in swerving out of each other’s way if they actually do have the misfortune of being there at the wrong time.

So statistics say I could get away with this for years,.maybe, and since I charge by the hour then I can theoretically work more and have the bank type larger numbers into their computer for my account. And it is fairly likely that I will be working and/or living somewhere else in the not distant future so the I may stop using this road before the accident happens.

That’s how economics tells me to drive.

Me, I think I’ll stay on my side of the road. I wonder what Mr. Lomborg and his ilk would decide?

6

SJ 08.22.08 at 12:54 pm

OK, read it.

It’s not very good. Weisbach and Sunstein are lawyers rather than economists, fair enough, but the first half of the paper which discusses discount rates ignores risk completely. This is OK at a first-year level, I guess.

But the second half of the paper starts off with:

The second term is δ, the rate at which we discount future generations. The ethicists call it the “pure rate of time preference.” It reflects a lower evaluation of future generations simply
because they are distant from us.

This “future generations” stuff is just rubbish. δ is just the same discount rate they’ve spent the first half of the paper trying to explain.

7

Barry 08.22.08 at 1:39 pm

“Cass Sunstein (h/t Nicholas Gruen) have written a piece for the AEI “

Stop right there – *any* involvement with AEI beyond snagging some free food at their events [1] is a serious mark against somebody’s honesty.

Considering that Cass is also a U chicago law professor, this is pretty much proof that he’s one of the bad guys.

– Barry

[1] And that only if you sneeze on the food, to infect the AEI guys with something.

8

SamChevre 08.22.08 at 2:07 pm

δ is just the same discount rate they’ve spent the first half of the paper trying to explain.

It shouldn’t be. (I can’t read the paper right now.)

Normally, you have a discount for pure time preference (δ) and a discount for the uncertainty of the future.

9

David Weman 08.22.08 at 2:21 pm

Sunstei is also, in case you don’t know, a friend and important advisor to Barack Obama.

10

lemuel pitkin 08.22.08 at 2:48 pm

Quiggin’s critique seems right as far as it goes, but like other commenters, I think it could go a lot farther.

From the perspective of, say, the framers of the Versailles Treaty, what would an appropriate discount rate have been for the expected loss of GDP from a potential world war 25 years later? And more importantly, would such a calculation have shed *any* light on the importance of preventing that war?

11

lemuel pitkin 08.22.08 at 3:02 pm

… and suppose that, on balance, you found that WWII *raised* world GDP? (Maybe it did.) Or suppose — coming back to climate change — that you found that most of the costs were borne by people whose contribution to world GDP was minimal — Nairobi slum-dwellers, Bengali peasants, etc. — while the costs of prevention would fall on high-GDP folks like us. Probably better to do nothing, then, per Sunstein — and per Quiggin, or no?

12

noen 08.22.08 at 4:31 pm

Gavin Schmidt & Elisabeth Moyer in a commentary in Nature, “A new kind of scientist” call for greater interdisciplinary cooperation between climatologists and economists. They argue that:

“…the two fields must be combined to produce analyses that can usefully guide society’s response to a changing climate. For instance, regional changes in precipitation could have important impacts on agriculture and infrastructure needs. Only by combining detailed modelling of economic responses with spatially and temporally complex climate projections can we appropriately direct resources for mitigation and adaptation. Without the synthesis of these fields, policy responses can be incoherent and counter-productive, as in the case of the recent rush to biofuels.”

More discussion at Real Climate – “Bridging the divides”

13

someguy 08.22.08 at 4:37 pm

I am on page 17. It is a nice read. It seems like pretty non objectional stuff.

I might change my mind but discounting at the market rate of return just seems
silly.

I mean why is the assumption that in the real world the alternative to the tax is money invested in the market?

I always wonder why progessives don’t worry about that more.

Surely the real world cost is the guess at real GDP lost.

Ok, now, I am on page 29 and I think the same thing.

Ok, I am done. Loose read on my part but I still think the same thing.

I don’t think they can really mean to say we should use the market rate of return.

Either I mis-read or they mis-spoke or both or is there something else I am missing.

I also missed how the paper was right wing.

14

Sebastian 08.22.08 at 4:55 pm

“Coming back to the hypothetical, if the market rate of return is far above that suggested by economic analysis as socially optimal, the policy implication is that action is needed to increase savings and investment, so that the market rate and the socially optimal rate move closer together.”

Maybe I’m totally confused by jargon or something, but how do you calculate a ‘socially optimal rate of return’? Does someone have a link to how the term is used in its technical sense? Because in a layman’s understanding of ‘rate of return’ if there were a regular and safe investment that produced very high rates of return, that wouldn’t be a bad thing. (For example if someone invented some cheap-to-produce robot that instantly quadrupled the manufacturing output capability of everything, that would be good, right?) Or maybe I’m confusing financial return with output on investment?

15

lemuel pitkin 08.22.08 at 6:29 pm

Sebastian-

The logic is that if the return on equity is the cost of capital to corporations, and if that return is well above what it should be based on expected income and risk from equity ownership, then the cost of capital is higher than it would be in a perfect market, and hence we are investing less than the optimal amopunt given current tastes, technologies, and endowments.

Of course, the first “if” is false — equity finances only a trivial fraction of investment. So the second “if” — the focus of Henry’s paragraph you quoted — is largely irrelevant.

16

lemuel pitkin 08.22.08 at 6:44 pm

er, John Q.’s paragraph.

17

noen 08.22.08 at 8:33 pm

Stern Review final report

Weisbach/Sunstein footnote 5
“A final important difference between Stern and Nordhaus is that Stern, at the end of the day, abandons economic analysis in favor of the intuition that concentrations of carbon dioxide over 550 parts per million impose an excessive risk on humanity. Nordhaus maintains a consistent economic analysis throughout his model.”

“Intuition”? The authors use this phrase at least three times in order to describe the Sern review. They are denialists. Maintaining CO2 levels at 550ppm means a rise of 4°C and would be catastrophic for humanity. CO2 concentrations must be reduced to 35oppm as quickly as possible, by any means necessary, and emissions reduced to zero by 2050 or sooner. Then we can start removing the excess carbon we put there.

The central theme of modern economics is the assumption of natural abundance. There is no model for how “externalities”, such as an ecological collapse, can lead to an economic collapse, because the basic factors are left out of the econometric models. Given this, big picture economic predictions are useless with regard to adapting to climate change.

18

John Quiggin 08.22.08 at 10:09 pm

In response to various points from Lemuel, I think (and have written in various places) that not all problems are amenable to benefit-cost analysis using discounting. For example, Megan McArdle had a very odd piece a while back trying to link abortion to discounting. But that said, it’s hard to see how you can talk about climate change policy without implicit or explicitly making assumptions about discounting, and, of the two, it seems to me that explicit is better.

As regards the treatment of people in poor countries, if you buy the full utilitarian package implicit in the economics treatment, attitudes to discounting and income redistribution are closely tied up. With the assumptions in Stern, a dollar of extra income to someone in Bangladesh with an income of $500 is valued the same as $100 of extra income to someone in the US with an income of $50 000. So that favours quite a lot of weight on the poor compared to maximizing GDP.

I cover this in more detail here (PDF)

19

drbiker 08.23.08 at 2:14 am

When will this continuing discussion realize that the equation for discounting is far more important than the discount rate itself. Behavioural economics argues that the simplistic power or exponential equations is biassed against the long-run, not matter what the rate used. If one used the time varying equations suggested in behavioural economics, much of the debate would be mute. To argue about the proper ex ante discount rate (which is an unknown), when applied to an imperfect equation, seems misguided to me.

20

terence 08.25.08 at 6:51 am

is, ought, Hume.

and if you want to read a good, kindof non-tech friendly discussion of the econo0mics of climate change try the following by Geoffrey Heal (a Columbia University environmental economist).
http://www2.gsb.columbia.edu/faculty/gheal/Climate-Change-Review.pdf

Comments on this entry are closed.