Bishop Hill Climate, ethics and democracy
Apr 18, 2013 Climate: WG3 Climate: other Economics The MIT Technology Review has a very interesting article about ethics and climate change and the knotty problem of discounting future costs:
Typically, economists calculate the discount rate by using money markets to determine the expected return on capital. The reasoning is that the market is the most democratic means of assigning value. But while that practice might work well to account for the value of commodities, Broome argues that calculating the discount rate for action on climate change is far more complex. For one thing, the conventional method doesn’t fully account for the possibility that even if people are richer in the future, climate change might reduce the quality of their lives in other important ways—and thus it underestimates the value of current investments. Broome ends up supporting a rate similar to Stern’s.
But his larger point is, more simply, that even such quantitative economic evaluations need to fully incorporate moral principles.
You can see where this is leading. Market discount rates - the ones that people choose of their own volition - are wrong. The answer is not that the public should be persuaded to adopt a different approach to discounting the future but that a different discount rate must be imposed by unelected "experts".
Reader Comments (26)
Trained in economics at MIT, Broome is particularly interested in assessing the ethical judgments made by economists. “Economists recognized, say, 50 years ago that economics is based on ethical assumptions,” he says. ...
Adam Smith was a professor of moral philosophy. He would turn in his grave to learn that "economists recognized, say, 50 years ago that economics is based on ethical assumptions."
If Broome knows so little about the history of his own subject, it is no wonder that he is drawn to climate science!
Interesting that a piece broadly sympathetic to the cause - at least the way I read it - should be quite specific that Stern used an oddly low discount rate. It seems to be another minuscule rowing back, from "Stern is right" to "Stern is a bit wrong but for the right moral reasons". Just a little ...softer.
And who is he to presume "..... climate change might reduce the quality of their lives in other important ways— climate change might reduce the quality of their lives in other important ways"?
It is just as valid to presume climate change might improve the quality of lives in other important ways.
Yeah funny that, with alarmists across the disciplines, from hard to soft science, it always seems incredibly important to redefine any and all metrics away from the more easily understood and agreed upon ideas, that are within a common ground, to something more obscure that strangely only seem to then have the virtue that the alarmist philosophy doesn't get challenged.
Understanding then becomes more of an abstraction that only fellow cult members can see and any criticism can be marginalised as coming from a denier out-group.
Reading the intro to Broome's book on Amazon I notice this genius is concerned about whether we all will be able to live in igloos in the future.
Oh sorry! of course, he doesn't mean *us* living in igloos - he is talking about the Inuit there. But he effectively has their value system today mapped out as a definition of all our future values - he just knows the future like this - he's a genius, he's been to Oxford. You get the picture ;)
Looking at the rest of the hype in his book's introduction, about Inuit finding animals and weather unpredictable today, he strikes me only as yet another over-qualified hyped up alarmist looking for a justification for their existence.
"Broome argues that calculating the discount rate for action on climate change is far more complex. For one thing, the conventional method doesn’t fully account for the possibility that even if people are richer in the future, climate change might reduce the quality of their lives in other important ways—and thus it underestimates the value of current investments. Broome ends up supporting a rate similar to Stern’s."
----------------------------------------------------
As Steve McIntyre says, watch the pea.
This is just a version of what happens when some pseudo-scientific "study" by Believers is proven to be trash. Common responses are that it doesn't matter because the science is settled, and "look, over here, a squirrel! I told you there were squirrels!"
The pseudo-economics which has grown up around climate alarmism always sneaks (or triumphantly parades) the injection of "values" into the numbers. No prizes for guessing whose values we are talking about here.
It is undeniable that trading in illegal drugs is very profitable. For those who do it, they have made a decision about risks and rewards. There are no moral "values" implicit in that statement. Comments on the desirability or effects of the drug trade are extraneous to the primary observation.
In the same way, the application of a discount rate is a empirical exercise, and it doesn't matter what you are applying it to.
"Green economics" is an oxymoron.
There is nothing democratic about the market. At the market, it is one dollar, one vote.
Quite right, Richard. Broome's statement about the rationale for the discount rate having something to do with democracy is illiterate nonsense.
Looks a fascinating article but as far as politics go I'm still with Winston:
Bruce Sharp in May 2008, quoting Arthur Schleshinger Jr in 1967. Churchill's repudiation of the desire of elites to feel morally superior was surely one of his greatest characteristics.
Hmm. There is actually a problem with using market interest rates as the discount rate.
We know very well that human beings tend towards hyperbolic discounting. In plain English, that we tend not to put enough weight on the very long term. This is entirely understandable of course: we live lives of limited time span therefore it's not actually sensible for us to put great weight on something that might happen 100 or 200 years out.
Thus, when considering the very long term it's probably sensible for us to not use market interest rates. You'd have to look long and very hard to find an economist who would argue that we should use them without any adjustment at all.
Stern might go too far. But pure market interest rates is also probably wrong.
Of course, what we really need here is Richard Tol to tell us.
Broome is correct - to a point. Cost-benefit analysis can aid in making policy choices, but it cannot make them for us. A difficult (but excellent) read on this subject is Robert Formaini's book 'The Myth of Scientific Public Policy.' Political philosophy and ethics just also inform choices. To give a stark example, if we were to undertake a CBA of extermination camps in the Nazi era, Auschwitz might be the most efficient — but that does not mean it should ever have been built.
That said, the arguments over the discount — especially those of Stern et al — are highly dubious. Many economists would argue that the opportunity cost of capital (about 8%) should be used, otherwise public investment will crowd out the private sector. I'm not sure that much CBA was undertaken in Spain, but subsidies for coal and renewables crowded out private investment in CCGT. Now the Spanish economy is bust, the subsidies are unsustainable and the coal miners are angry.
There is a case to be made for adjusting the discount rate down because private individuals care less about the future than society perhaps should, and there is a case that can be made for (say) 4%. But Stern's was a fiddle that discounted the future hardly at all. And, of course, a discount rate is not solely about social time preference, but also about uncertainty: uncertainty over climate science, technological development, etc. We might have nuclear fusion by 2050, and future generations might rather we had invested in wealth creation rather than GHG mitigation.
I first grappled with these issues when the Hydroelectric Commission here in Tasmania wanted to build a dam and flood some wilderness. They used 4% rather than 8%, which favoured the dam over the thermal alternative, but they also assumed that coal would increase in price by 2% real pa, so the (substantial) fuel costs of the thermal option were discounted at only 2% over the 60 year comparison. As a colleague put it, if coal was going to escalate at 2% pa real, we'd be wearing it as jewellery after 60 years.
Always question results that are dependent on the discount rate chosen, or don't examine the sensitivity of the result to the selection of the discount rate and other assumptions.
Tim
Should people not be persuaded of the error of their ways rather than having an alternative imposed upon them?
@Tim Worstall
Indeed, for more distant future returns, the rate does need to be adjusted compared to the nearer term. If one in the hand is worth two in the bush, it's worth at least three in the trees over yonder.
Tim and Aynsley - the answer to both of your objections is that the discount rate, like all predictive tools, becomes less and less useful as the timeframe gets longer. In business, it is typically used for periods no longer than 10-15 years, and often much less.
No sensible economist gets involved in laying bets on scenarios which purport to predict the future in 50 years, let alone in hundreds of years. It is pure fantasy. There are too many "unknown unknowns."
However, if people like Stern choose to engage in hypotheticals, the least they can do is either use accepted tools, or provide a very strong justification for not doing it. Otherwise, they are just engaged in writing fiction while pretending that it is somehow empirically based.
Johanna,
Agreed. My financial calculator only ran to about 18 years - before it failed. (As Keynes would have put it, in the long run we - and our calculators - are all dead).
I always liked the critic of Stern (Richard Tol? Dasgupta? Nordhaus? — can't recall off-hand) who pointed out that half the costs that Stern's analysis suggested made immediate mitigation cost-effective occurred after the year 2,800. Who even knows what planet we might be on in 2,800!
What actually rubs salt into the wound for me is that public servants are seeking to misrepresent and distort the metrics which are being fed into the arithmetic of CAGW mitigation as we have seen looking at the polling results over in unthreaded over the last couple of days in the matter of DECC's "British Energy Challenge" consultation = contrived and deceitful misrepresentation on a stick.
One has to suspect that that they'd not give an honest reply to "how many fingers are on your left hand?" and leap straight to the matter of hypothetical prosthetic limbs yet to be designed, let alone manufactured.... The way things have been going frostbite might well become an issue for the U.K. ;-)
"In plain English, that we tend not to put enough weight on the very long term."
That may well be true in some cases/studies. But is it always so? And how would we know when we, collectively or individually, are not behaving thus, or are changing our behaviour?
Problems arise when numbers have to be adjusted because they "may not be quite right." Of course this is honed to fine-art in climate science, but the touchy-feely approach to numbers has long been favoured by the greenshirts.
Speaking as an economist who has been forecasting for the last 30 years - and with a happy client base to boot - there are areas where you can forecast out further: demographics and structural shifts. Demographic projections, while not strictly speaking economics, aren't even forecasts out for the next 70 years because the long-term actors are already born (a child today is the adult of tomorrow). Add to these the kind of structural shifts we are seeing world-wide (the development of a Third-World middle class, for instance) and you can indeed do long-term forecasting for some very interesting numbers.
But GDP 30 years out? At best a guess.
On the other hand, you can buy 30 year bonds and in some countries bonds with even longer time horizons. While there are some who like to sit on these and earn their interest, that's really only institutional investors such as life insurance companies who plan in decades, rather than years. Mining companies - as Tim well knows - generally also operating in time-frames of multiple decades to ensure that their business doesn't decline when the copper in a mountain is mined out to the point of feasibility, and are more than willing to buy mineral rights 100 years in advance both to keep competitors away and to retain resources.
The "proper" discount rate doesn't exist: this is thanks to the fact that interest rates are currently administered, rather than leaving it up to markets (they're administered in the sense that the central banks prevent them from finding their true, market-driven position), meaning that to use whatever long-term government bond rate would be to underprice actual discounted costs significantly.
Stern et al use an imaginary rate that ensures that their political needs are met. Nothing more, nothing less. As such, useless.
If you want to find the proper rate to use, given the highly speculative nature of what the AGW advocates claim, you should look at what a 10-year junk bond, issued by a mid-ranked company, costs. You'd find rates between 3% and 10%, depending on the company, country and length of time involved, and my gut feeling is that 7% is really a more appropriate discount rate to use.
Which means, of course, that based on discounted costs, it's vastly more sensible to wait and see, rather than apply an expensive precautionary solution that remains highly speculative at best and downright harmful at worst, but I dare say I don't need to point that out here. :-)
Oh, and the author of the article linked to makes another astounding error: markets aren't by any sense at all "democratic," but rather are by far simply the most efficient way of finding prices.
Calling them democratic is more than slightly mind-boggling: even a 100% vote (monopson) on what we are willing to spend on an iPhone - call it 10 quid - doesn't make it possible to sell at that price point. Rather, we do without if we aren't willing to pay the price point where a supplier is willing to provide the volume demanded.
Someone wasn't paying attention in their Intro to Microeconomics course.
John F. Opie,
"Speaking as an economist who has been forecasting for the last 30 years - and with a happy client base to boot - there are areas where you can forecast out further: demographics and structural shifts. Demographic projections, while not strictly speaking economics, aren't even forecasts out for the next 70 years because the long-term actors are already born (a child today is the adult of tomorrow). Add to these the kind of structural shifts we are seeing world-wide (the development of a Third-World middle class, for instance) and you can indeed do long-term forecasting for some very interesting numbers."
==========
Interesting.
As a thought experiment, going back to say 1900, how much could have been forecast, with even loose accuracy, 70 years ahead?
We saw enormous political changes, medical advances, agricultural advances, technological advances. We saw things like the rise of the motor industry, air travel and things arising from electronics, e.g. communications and computing.
John Broome was Professor of Economics at Bristol before he became Professor of Moral Philosophy at Oxford. Whatever he is, he is not ignorant. It is fairly obvious that what he meant was that 50 years ago (when I was just about to become an undergraduate!) economists knew that economics was intimately bound up with ethical judgements. He is right, and he also seems to be right that a lot of economists these days seem to have forgotten that. The appropriate discount rate for society to use in evaluating investments is exactly one such ethical judgement. The market discount rate (which one?) roughly speaking reflects individuals' marginal rate of consumption time preference (how individuals value present as against future consumption). But the issue for long lived investments is how these impact on people yet to be born. Should we take their preferences into account - how can we? The social discount rate reflects how we as a society value consumption now as against that of future generations. There is no reason why this has to be the same as the private marginal rate of time preference.
Indeed the standard welfare economics treatment is to argue that the social time preference rate (using I hope obvious terminology) should reflect factors such as whether we expect there to be any people around (the risk of catastrophe) and how well off people in the future will be (the richer relative to us the less a given amount of future consumption matters in a simple utilitarian approach). Stern's answers seem implausible because he seems to want to value consumption of people in the future very highly compared to consumption of poor people now. There is a good discussion by Partha Daasgupta called "Dscounting Climate Change" which can be found on the web.
Among left leaning economists a new term has crept into the lexicon: "externalities". The idea is that the free market results in transaction prices that fail to account for the true social cost of the transferred item. The canonical example they point to is cigarettes. Cigarettes cause cancer. Cancer imposes costs on the public (in the form of increased medical costs) and those costs are not reflected in the free market price for the cigarettes. Therefore (wait for it) the government must attach taxes to the cigarettes in order to cover the social cost (the externality).
The public has generally accepted this idea as it related to cigarettes. Now politicians, seeing an opportunity to raise new revenue, are applying the idea to all sorts of other situations. In the US politicians are talking about externality taxes for guns (a uniquely American issue), sugary drinks, and non-politically-correct sources of energy.
But as it related to energy, why is it that only the negative effects of energy production are considered when talking about externalities. Surely cheap and abundant energy has resulted in significant health *benefits* when compared to the alternative. There is no mechanism in free market pricing to account directly for these benefits. Why wouldn't the government provide credits to energy companies for the benefits of cheep energy? Why only consider the costs?
I'm extremely dubious of this whole subject.
"Externalities" is hardly a new term. It goes back at least to Pigou's Economics of Welfare published before the first world war. It's been on undergraduate economics courses for years, and externalities can be positive as well as negative.
Apr 18, 2013 at 5:54 PM | mpaul
I find one good thing about observing climate alarmism over the years is that I've picked up stuff on a wide range of topics. I have changed my initial thinking on things like externalities. It is a concept that makes sense. I was persuaded by reading about a guy called Coase, and his examples e.g. the effect of a steam railroad that passes through a a farmers wheat fields, are quite good. This is a link I found:
The Coase Theorem: Controlling Externalities through assigning property rights
I think the issues start to get blurred when you talk about scaling up and arbitrary adjusting the description of harm in ways that are not easily agreed upon across cultures, location and time. For instance claiming to know about the time and people in the future where the harm is supposed to happen is the trick being played by the voodoo economists and philosphers in climate here I think.
Because we know they have had such a good track record up to now I guess we are supposed to be impressed by their predictions /sarc ;)
@ mikep
"But the issue for long lived investments is how these impact on people yet to be born. Should we take their preferences into account - how can we? The social discount rate reflects how we as a society value consumption now as against that of future generations. There is no reason why this has to be the same as the private marginal rate of time preference."
----------------------------------------------------------------------------------------------
That's a pretty controversial assumption, Mike. Your construction of the "social discount rate" is exactly the opposite of how the discount rate used in economic planning and decision-making works. The conventional discount rate assumes implicitly that the best way to care for the interests of future generations is to look after their parents and grandparents - that would be us, now. This maximises the chances of bringing healthy, well educated offspring into the world of the future.
I am curious to learn what kind of investment decisions you are thinking of here. Do you refer to babies that will be born tomorrow? I rather suspect that like other "economists" who witter on about future generations, you are attempting to impose some sort of crystal ball generated "social" cost on much longer-term scenarios. By all means do that with your own money, but not with mine, thanks.
Your term "long lived investments" is also mysterious. No investment lasts very long without further investment - that's why we have the notion of depreciation. You make it sound as though investment decisions are irrevocable, unmodifiable imprints that our oft-cited "grandchildren" will have to live with. This is a gross oversimplification and misrepresentation of how investment works.
If you,re talking Climate Change in the Future and Economics.
The alarmist are predicingt that the Oceans will rise by a few feet.
So what are Property Prices doing on the Coasts
Anyone been in touch with any Estate Agents in Brighton Southend not Margate ,London Docklands
Sydney, San Francisco, a nice Beach Front Timeshare Villa in the Maldives.
@Cosmic:
I'll leave it up to as an Gedankenexperiment to forecast from 1900 to 1970: we didn't have any of the forecasting tools we have now, let alone anything like today's statistical base, back then. Meaningless exercise, to be honest: you can't predict wars and other national calamities. If you could, it'd be psychohistory (see Asimov).
Further, I was speaking of demographics and structural shifts, not forecasting technological developments: barring massive die-offs and wars, I can tell you today which countries will have severe structural problems and which will be doing well. Tanzania, for instance, will show the largest individual increase, measured as a percentage of current population, of all countries through 2050, as they have a very young population and will reach close to 300mn in 2050.