Tuesday, December 12, 2006

How much do we really care about the future? Stern's critics assessed...

Time for another look at the Stern Review and what people have said about it. It's received much critical acclaim: Stern even lined up five Nobel laureates and other notables to applaud at the launch [see here] and the BBC canvassed largely favourable expert opinions [here]. And I think it is an incredibly important step forward - not least because of its politics... the messenger is at least as important as the message.

But objectors have criticised various aspects the report (see below). The criticisms cover several areas, including: the underlying science; optimistic costs of mitigation and adjustment; pessimistic views of adaptation and impacts; the discounting conventions used. The choice of discount rate makes a huge different to the weighing of costs and benefits which are separated by many years - our mitigation efforts on climate change only bear fruit late in the century. Any analysis of climate change is highly sensitive to assumptions about discounting - the chart above gives the present value of something valuable in 100 years time for different discount rates.

On science: Stern took the established scientifc consensus as an input to his review and was provided with good synthesis material from expert bodies such as the Hadley and Tyndall Centres. The main criticisms seem to rehash concerns that recent global warming is not that unusual given temperature anomalies like the Medieval Warm Period and Little Ice Age. The main scientific counter-argument here is that these were regional, not global, temperature anomalies. Stern provided a good discussion of some of the scientific criticisms levelled at them earlier in the process [World Economics article, June 2006].

On costs and difficulty of mitigation: Stern has made generous assumptions about future technology development and costs. Whether his assumptions are correct depends on how one believes technologies evolve. For environmental technologies, the evolution may itself be a product of the policy framework and the sharpness of the incentives its creates for investment and R&D - but the kind of response that would keep to below 550ppm would be quite dramatic and the incentive signals very strong (see Stern's launch presentation chart 4 requiring world emissions to peak by 2020). My view is that Stern made it all seem too easy. Not only is underlying economic growth currently driving emissions in the wrong direction (upwards!) at an unprecedented rate, there is also huge inertia in energy-using assets, production systems (like agriculture and vehicle manufacture), and in human behaviours, that stand in the way of slowing growth and starting to turn the trend downwards. All this means that the rates of change envisaged in Stern will require interventionist policies of unprecedented strength and global coverage to create the necessary changes in the next few decades.

On costs of impacts and adaptation: Stern has considered, rightly, costs of disasters and extreme events and major system failures. He has concluded there is a lot at risk and builds up his case carefully in chapter 6 showing thatearlier lower estimates had not included non-market impacts or reflected major system changes. My own view is that the GDP numbers (5-20% GDP loss in the absence of strong mitigation) are a bit misleading. I'm not sure what GDP in 2100 will actually look like - at current growth rates the world economy would be 36 times larger in 2100, and by then many other constraints will have kicked in. His higher damage estimates also include impacts on health, human life and the environment. Bravo! But rather than look into the future, if we look across the world today, the willingness to pay for protecting lives in far away countries doesn't seem that all that high.

On discounting: this reflects the need to compare costs and benefits that arise in the far future with costs that arise today. Stern used an unusually low discount rate and a first principles discounting regime - see excellent background paper for discussion of the economics and ethics of long-range discounting. This regime gives high weight to future generations and therefore high value to taking costs now to secure benefits much later. But I believe he is on weak ground here... not because it is wrong for Stern to believe this is right, but because he has made an ethical judgement about appropriate 'time preferences' that doesn't reflect what many or most people actually feel. Though discounting seems like a technical matter, it is fundamental in climate change. Here's a few observations:

Stated versus revealed time preferences
The discounting regime is central because it codifies the answer to the question: "how much do we collectively care about the future?" I suppose there are two ways of looking at this question. Firstly, an analyst could apply an ethical framework and ask 'how much should we care about the future?', or alternatively, she could look at how people and governments actually behave and ask 'how much do we care about the future?'. Stern has taken the former approach, but what governs how we and governments actually behave is closer to the latter. This dichotomy emerges in many aspects of policy & life - it is why we spend a fortune on acute health care and relatively little on long-term public health... the official discount rates say spend on public health, but the revealed time preferences say spend like mad on securing a few extra weeks life for the terminally ill - and that is what we do. [Update 17/12/06: But I think it means Stern should be clear that his economics of climate change are predicated on an certain ethical framework that is rarely found in real life - his economics are 'normative' and are based on political propositions. I think it is better to apply a 'positive' economics, looking at how the world is, and then build a consensus for a different politics and value-set based on based on conclusions drawn from where the real world appears to be taking us. In some ways, Stern's relatively upbeat conclusion is an artefact of his generous assumptions about how we value the future].

Discounting explained... Discounting used here has two main components. First, how we convert future consumption to future welfare; second how we value future welfare compared to our own... more precisely:
  • First component: something that reflects 'diminishing marginal utility of consumption' - which arises because we place less value on each extra £1 of consumption as we get richer. So if we assume economic growth will continue, future generations will feel less clobbered by, say, a £1 billion loss than we would today. Stern uses the familiar formulation that each doubling of consumption reduces the 'utility' of an extra unit of consumption by a half. So an extra pound provides half as much welfare to you when you earn £50K compared to £25K. Do the numbers, and this comes out at using the expected growth rate as this component of the discount rate. [update 17/12/06: I glossed over this as though there is only one answer - but this a distributional measure: capturing how much weight we give to the welfare of the poor. Some development economists argue that we should value marginal increases in the consumption of the poor more highly relative to rich- ie. an extra pound would be worth more than twice the welfare at £25K compared to £50k. This 'compassion' towards the poor has the effect of reducing our compassion towards the future, because we assume people will be better off in the future. [see discussion in The Economist] I don't think there is a right answer except that gained from observing how people and governments actually behave]
  • Second component: something that reflects 'pure time preference, which measures how much less we value future welfare or, more pointedly, a life lost or saved in the future. Stern uses 0.1% per year [see Stern FAQ q8] - arguing that there is virtually no ethical case for valuing the welfare of future generations below our own (the reason that he doesn't use 0% is that he factors in the possibility of a mass human extinction at a low probability and assumes that people will value welfare slightly higher today because in the future we may all be dead anyway. However, the Treasury uses 1.5% for time preference in its 'Green Book' on economic appraisal - and this makes a huge difference over 100 years (see chart again). There is evidence that real people can have much shorter term time preferences, with personal discount rates up to a hedonistic 30% [see American Economic Review / Google scholar search]. I think we instinctively discount the welfare of others the more remote they are from us - there's probably a evolutionary driver there somewhere. We see this as a kind of 'compassion exchange rate' for news reporting of deaths of people unknown to us depending on how far they are away - 1 person killed in the next street = 10 people killed in London = 100 people killed in Spain = 1000 people killed in India. Well, deaths in the future is a form of remoteness too and I think we would instinctively discount at a higher rate than Stern. I tested this proposition on a friend and found time preference discount rates for loss of life of 0.5% and 1.4% - 5 to 14 times Stern's assumption - and she cares!
Can a government's preferences be that different to the people that elect it?
It can be argued that governments should act differently to citizens, and take the long view. And indeed, it might well be that citizens have greater willingness to consider the long term if they are acting collectively - showing altruism without the annoyance of free-riders. This is right, but surely can only go so far - the Stern Review accepts the argument made in section 2 of the background paper on discounting that governments should use a different (lower) discount rate to that implied by the instinctive 'impatience' of the citizen - which manifests itself in very aggressive short-termist time preferences. But is it possible for governments to use time preferences that are much longer term than the typical citizen, even if polished technocrats believe it to be right? Governments are after all elected to express collective preferences, and if they depart too much from the preferences of the voter, they'll get the chop at the ballot box. This isn't theorising: look at fuel protests!

The bad news so far...
Coded into an arcane discourse about the discounting regime is probably the most fundamental economic question about climate change: how much do we care about the future? We protest that we care a lot, but we behave - individually and collectively - as if we don't care much. If the economics of climate change were recalibrated to the short-termism of our revealed time preferences we would act differently - placing much more emphasis on adapting to impacts over the next few decades and caring relatively less about controlling emissions because we know that reducing emissions only has benefits in reduced risks in the second half of the century, and only if everyone else does it too.

The better news...
Stern's report is part of the response to the counsel of despair in the paragraph above and human civilisation has progressed by development of altruism and common purpose. Stern's review should encourage us to think that modest sacrifices for the long term are worth making and manageable - and therefore we might put it into the category of those things we are prepared to be long-termist about. It also raises the question of whether using a welfare based global cost-benefit analysis is the right way to think about this problem - perhaps a human rights or pure justice lens would be better.

Examples of challenging commentary on Stern
Examples of positive commentary

6 comments:

Caspar Henderson said...

Thanks indeed for this sober and thoughtful analysis. I found this particularly useful after reading a 12 Dec article in the New York Times by Steve Lohr titled The Cost of an Overheated Planet.

It's clear that governments and societies (which are not the same think - thank you, David Cameron) will sometime spend on things they think will reduce the risk of big hazards. A Lohr points out, "In the late 1950s...American military spending reached as high as 10 percent of the gross domestic product and averaged about 4 percent, far higher than in any previous peacetime era. A Soviet nuclear attack was a danger but hardly a certainty, just as the predicted catastrophes from global warming are threats but not certainties."

The article then touches on the implications of spending 1% of US GDP to fight global warming (incidentally, says Lohr, 1% of US GDP is more than $120 billion a year, or $400 a person; it is about equal to the Bush administration’s tax cuts in 2001; and roughly the amount spent on the Iraq and Afghanistan wars in 2006).

At first sight one might think: "Case closed - if [we] are willing to spend 1% on war surely [we] are willing to match this in the struggle to save the planet", or some such worthy statement (1% being also the figure Stern recommends for expenditure to fight climate change).

As you rightly point out, however, revealed preferences for things that seem to be remote tend to be smaller, even among those who self-define as altruistic (which takes us back to Adam Smith, one's little finger and the Chinese empire http://en.wikiquote.org/wiki/Adam_Smith#Far-away_disasters).

Climate change is perceived as a much more remote, and lesser threat than nuclear war was in the 1950s and 60s.

It may be that that perception is wrong. The probabilities of a nuclear exchange between the USSR and the USA, which occasioned the latter to spend up to 10% of GDP to fight it, may for much of the time, including in the 1950s, have been smaller than thought (although, if Robert McNamara is right about Castro, we got incredibly close in 1961). And the possibility of "runaway" climate change is not, as far as I know, something that serious scientists dismiss completely (this would necessarily be as bad as nuclear war, just that it could be very bad).

I'm not suggesting one should try and create a "duck and cover" hysteria about global warming. I am suggesting we need to recalibrate our understanding of respective risks.

Here's a further thought. As you say, in Britain (and other rich industrialised countries) we are willing to spend disproportionate sums on immediate near term problems. You give the good example of health care in the last few days of life.

So what if societies such as ours started to place greater value on the near term impacts of climate change (for example, the likely loss of tropical coral reefs at less than 2C global average temperature rise)?

I feel out of my depth using a term like "the construction of value", but isn't that what we are talking about? Those concerned with local and global social and environmental justice may seek to reshape values so that more people cherish such things (yes, even in our preferences) a little more with respect to what we believe to be our own near term utility. (By the way, perhaps there is a case for dialogue/education with those facing terminal illnesses and their families regarding the case for a less drawn-out death -- ouch! a difficult but valid debate).

So to your final point: "perhaps a human rights or pure justice lens would be better". Well yes! The comparison may be a tendentious, but imagine a political economist of the 1770s looking at the economic case for and against the abolition of slavery. That would provide useful data, but it would only take him so far. The battle in that case was political.

Also, this time the people on the receiving end may have other ways of expressing their grievances or otherwise making themselves felt - through, for example, large scale migration from West Africa to Europe.

Reading it back, this comment comes across as impossibly idealistic. As Jonathan Glover points out in his extraordinary book "Humanity: a Moral History of the 20th Century", we have hardly begun to tackle the social psychology and practical ethics of warfare, still less issues such as environmental protection.

But as he is flavour of the week, let's cite Barack Obama - let us have the audacity to hope.

Caspar Henderson said...

sorry - at least two error in my previous post.

I meant to write that "runaway" climate change "would *not* necessarily be as bad as nuclear war".

And on reshaping values, I meant to write "even in our *revealed* preferences".

Clive Bates said...

My colleagues have responded in characteristically high-minded fashion by suggesting the following link:

Futerra's Stern Review Smack 'Em

Use the weighty Stern Review to thwack prominent sceptics...

Caspar Henderson said...

Nice game, but I got a lousy score. Boo hoo

Henry Leveson-Gower said...

In terms of action on climate change, I think that considerations of our preferences is not morally appropriate in a similar way the they are not appropriate in a court. Basically justice and preference/utility max seem different moral criteria and the decision as to which should apply in particular situations is a moral decision in itself. Hence in a court situation we do not think preferences/utility are relevant (even some-one we hate, we would give a fair trial, no?). So the only question is whether overall we bequeath future generations a world that is at least as good as the one we got or better (duty to improve world during generational tenure?).

I would then link this with setting up an EU (world later?) organisation with a mandate to provide justice to future generations - this faces directly the problem of current short term NAP decisions based on current generational interests. This involves a major political statement that CC is about justice for future generations. In terms of organisation funding, powers, capacities etc, my initial thoughts are:
- should have independant income to ensure political independance from current generation (top slice auction of EU ETS allowances?)
- Income can be both for costs (including significant research capacity into CC impacts) and investing for future generations inc developing world (spillover here to sustainable finance agenda and 'bank for the future' idea)
- should have power to influence (& set?) EU ETS caps based on a 2050 target for Carbon concentration in atmosphere; this is its point of political leverage/responsibility. It alone should be responsible for setting the long-term target which will effect future generations, its central mandate.
- how this feeds through to the current EU ETS cap is more complicated in terms of balancing current and future generations and getting the whole thing off the ground (give this organisation too much direct power and it won't be politically possible to get it agreed). It could be in the form of setting the next cap but one, which still provides some flexibility to governments but influences the broad direction and helps make the intergenerational trade-off explicit. Another option is that they set all caps, but allow borrowing by governments/businesses from future generations and they set the 'interest rate', which is a model more akin to a central bank, but may be a bit complicated to sell politically.

Dan said...

http://climatechange3000.blogspot.com