Bishop Hill Richard Tol on Stern
Dec 23, 2009 Climate: WG2 Climate: WG3 Economics I'm very pleased to have had a comment by the eminent economist Richard Tol (even it is was to tell me that I was wrong about the Stern Report - the report was still flawed, but not for the reasons I had put forward).
Here's what he says:
Stern managed to focus the discussion about the Stern Review on the discount rate used. The issue is not that Stern argues for a particular discount rate. That is his right as a a citizen of a democratic country. The issue is that he used a single discount rate (without performing a sensitivity analysis) and that he used a discount rate that differs from the discount rate typically used by his own, democratically-elected government. And all without alerting the reader. Stern's use of the discount rate is a clear case of manipulation.
The sloppiness of the Stern Review is perhaps best illustrated with its assessment of an optimal climate policy. (By the way, the Stern Review concludes that the previously formulated long-term target of the UK government is exactly right.) Stern's "optimum" does not meet the first-order conditions. In the optimum, marginal costs should equal marginal benefits. Stern recommends that greenhouse gas concentrations be stabilized at 550 ppm CO2eq, but at that point his (faulty) estimates of marginal costs do not equal his (faulty) estimates of the marginal benefits.
When I pressed him over this, the paraphrased reply was that Newton and Leibnitz are so passe.
The subsequent discussion is very interesting too. In essence Stern is arguing that a philosopher king should tell us what is right, while Tol is making the libertarian case - that ordinary people should choose their own way. Global warming enthusiasts should be clear, both to themselves and to the public they seek to persuade, that this is their intention.
Which brings me back to my original point: Stern should be strongly criticised for not making this clear to his readers, and Ed Stourton, one of the most senior journalists at the BBC should hang his head in shame for precisely the same reason.
Reader Comments (19)
I just re-read Feynman’s speech on Cargo Cult Science…
http://calteches.library.caltech.edu/51/02/CargoCult.pdf
http://en.wikipedia.org/wiki/Cargo_cult_science
…and could not believe the parallels with the catastrophic AGW narrative.
Though related to an economic analysis, Stern's reply to Tol is a classic example of this and should set alarm bells ringing for anyone with even a modicum of interested in the true nature of this issue.
I've been meaning to read Leibniz...but if he's passe...I guess I have a lot of catching up to do. I'm pre-passe...
Note that Stern did not just argue in favour of a dictatorship by a philosopher-king, he put himself forward for the position.
Was this covered on the BBC radio or tv? I found this link, but when I clicked on the audio link, I got 404.
Isn't the stern the arse-end of a ship?
I posted the following on the Stourton thread, but it properly belongs here
The pure rate of time preference is not the same as the social discount rate. The usual idea - see for example a paper by Ulph and Pearce (curiously enough an East Anglia discussion paper)
http://www.uea.ac.uk/env/cserge/pub/wp/gec/gec_1995_01.pdf
is that the social consumption discount rate also depends on the marginal (social) utility of consumption and the long run growth rate of the economy. This captures the idea that future people may be richer than us and therefore an extra unit of consumption for us now is worth more than the same increment to someone in 200 years time who can be expected to be richer than us because of the economic growth that will occur. In addition we might want to add in something to account for the possibility that there will be no one alive in, say, a million years time, so there would be no point in our saving to look after them. Pearce and Ulph argue that the resulting rate is between 2 and 4 per cent.
Note that this is the rate at which future consumption should be discounted, not necessarily the rate that should be used in cost-benefit analysis. The reason for the difference is that the project being analysed may crowd out investment instead of consumption and this needs to be allowed for if we just look at project returns and not project effects on consumption.
What was odd about Stern was the choice of values for the marginal utility of income as much as the pure rate of time preference. As I said above it had the consequence of requiring poor people now to save to raise incomes of richer people in the future - hence the absurdly high savings rates it implied, as noted by Dasgupta.
There are many interesting things in this post but let's deal with what I find the most interesting, the issue of climate sensitivity. Since Stern did not couple climate sensitivity with the PAGE2002IAM and derive some multiplier like variable that could be fed in as an exogenous variable his conclusions about future harm are flawed methodologically. The PAGE2002 integrated assesment model treats sensitivity implicitly not explicitly. This makes the model cumbersome to adjustments given changes in research about overall climate sensitivity to a change in CO2 concentrations in the atmosphere. Sensitivity is simply treated as a value not a variable. A simple though elegant model has been developed by Weitzman. His model is about extreme low probability outcomes, fat tails, and what decision makers should do in the face of extreme uncertainty. The model curiously recommends that a risk averse agent should devote 100% of current income to avoid a far future catastrophe. Though unrealistic in its conclusions it is sound methodologically. Weitzman does not seriously recommend devoting 100% of current income to avoid disaster but rather points out that low probability events should not be ignored as they impact risk averse agents, collectively the human race, in his analysis.
By bringing climate sensitivity to the fore Weitzman has opened the door to discussion. Lindzen and Michaels can now enter and have a fruitful discussion without emotional outbursts from diehard adherents. Closing Weitzman's model and bounding it like he does with a statistical value of a life (life on earth or the human species in particular) can now be adjusted by adjusting the climate sensitivity parameter. By dealing with the fat tail problem Weitzman's dismal theorem can now be incorporated into convential CBAs and we can get some reasonable estimates for expected utility of a current action of mitigation.
The second interesting notion here is about what generally is thought to be the problem with representative democracy, problems are not addressed until there is a crisis. This seems to scream out for an "engineering solution" and that is a benevolent rule by scientific elites. The argument here is that foresight is possible and that we understand enough about how the world works to make such rule possible. It all depends upon how risk averse we are. Certainly our genetic makeup makes us ill suited to paying serious attention to the future we are all about survival and social status. And it is those things that might be the problem. At least some futurists think they are. Robin Hanson is one economist who at overcomingbias blogs a lot about this. His main gripe is that these other futurists tend to make up there own economics which makes their proscriptions troubling. As an economist he thinks there should be more communication between disciplines particularly economics and other disciplines. In keeping other futurists or foresight specialists grounded in economic reality we might actually produce some decent policy proscriptions that overcome the biases present in today's policies that so rankle libertarians.
"Stern" is also German for "star". Perhaps that's where he gets his libertarian impulses from ;)
"The issue is not that Stern argues for a particular discount rate. That is his right as a a citizen of a democratic country."
Worrying sentiment, and also wrong. The discount rate is set in a free market economy by the ordinary price mechanism that balances supply and demand. That government can mandate it is lower, by printing money and lending it at that rate, but the difference is a tax on the money supply. The government cannot lower the real discount rate: only shift some of the costs onto other people.
".. the problem with representative democracy, problems are not addressed until there is a crisis. This seems to scream out for an "engineering solution" and that is a benevolent rule by scientific elites." The Aztec elite found that a stream of human sacrifice was adequate to ensuring a daily sunrise. Prob solved.
Mdc you state that the government cannot change the real discount rate. This is a confusion. The "discount rate" is these discussions is that rate at which future benefits SHOULD be discounted when making policy. It is inevitably a value judgement and does not necessarily have anything to do with actual market interest rates, real or otherwise. The discussion is all about how much we are prepared to sacrifice for people at various distances in the future. A zero "discount rate" (in this context a social time preference rate) implies that an extra unit of consumption for someone two million years in the future should (that word again) be valued at the same as an extra unit of consumption today.
Ah, MikeP, yes...but the argument here is in fact, should that discount rate be set by the philosopher king (Stern) or by everyone else in what they do daily, in the marketplace (roughly Tol's position)?
The market interest rate reveals, at best, private time preference given current income distribution. I'd go for the decision of democratically elected government. (And I reject Stern because his marginal utility of income is too high, so in practice for actual numbers I'd go with Tol)
Ah, Tim W, yes...but I thrashed this Philosopher King objection out with Richard T in a previous thread. In suggesting particular values for delta and eta, Stern is no more setting himself up as a Philosopher King that is Richard T by suggesting alternative values.
I know there are a number of anarcho-capitalist sympathizers at this site, but in the absence of an anarcho-capitalist utopia we are lumbered with the state and as Richard T and Mike P agree, it is the state that should ultimately set values for delta and eta.
@RichieRich
A properly elected and accountable parliament, rather than "the state", should make such decisions.
@Richard Tol
Fair point!
The lurking issue comes down to allocation of resources, from a central authority. the warmist political leaders want to move further into this nightmare--estimating total resources and then deciding how and where and by whom and for whom they can be used. false climate science is the lever.
Freedom is the casualty on that battlefield, regardless of what form of government makes the decisions.
Representative governments are limited by a constitution which establishes, among other things, boundaries on how much freedom can be reduced.
mikep (and indeed others in this debate),
You've totally ignored what I've said, as far as I can see. The discount rate (or the "interest rate") is an expression of time preference. Now the state can reduce the nominal interest rate, but it cannot reduce individuals' time preference. All it can do is tax people, and loan out this money where its original owners would not have. Since value is determined by peoples' subjective judgements, the loss to the economy is still the same.
You rejoinder that actually the state is the "correct" body to determine value and not individuals. This is a political point that is not related to my contention at all: from the point of view of the individuals whose money has been taken, they've still lost out, even if the state thinks they haven't really. Just like a highwayman may think the world is better off if he has your money; that doesn't mean that you do.
As to whether the state should do this: I won't answer in so far as it is a point of political philosophy, but there are certainly solid practical reasons why not. Principally, it becomes impossible to determine if any capital project is really worthwhile. Since almost no money is being spent by the people who directly benefit from it being a good investment, the interest rate that is set is totally arbitrary. Furthermore, taxing other people to spend even tangentially on one's self is always a better deal than not, at least in the immediacy, even if the investment loses money overall.
I see how state socialism appeals to some people. Or at least, how it could have done before we saw its catastrophic failure to provide anything like competent management or even an acceptable standard of living. But to argue that we should have essentially free credit to spend on anything, right after a huge banking crisis caused by people spending artificially cheap credit on bad investments, isn't just stupid, it is wantonly self-destructive.
@RichardTol/ RichieRich
Isn't there a respectable argument that under the principle of separation of powers, this is the Executive's responsibility and not the Legislature's? Given Stern's position within the Executive at the time of the Stern Report, and given that his report was accepted by HMG, does his report not therefore meet the requirements you have outlined?
Which all suggests to me that there is more to be said on this.