Stumbling and Mumbling

The "deserving rich"?

Sunny Hundal uses a phrase which troubles me - the “deserving rich.” My problem is: should the concept of desert have any role in our thinking about distributive justice?
The thing is, two of the most important books on this subject in recent years - Roemer’s Theories of Distributive Justice and Kolm’s Modern Theories of Justice - almost entirely neglect the notions of desert or merit.
There are, I think, two reasons for this.
One is that it’s impossible to tell what any individual really deserves. Do I, for example, deserve to earn more than the average worker? In one sense, no: my work is much less onerous or unpleasant than the average. But on the other hand, this pleasant outcome could be a just reward for years of effort earlier.
I don’t know which it is - or at least, I don‘t if I slough off the self-serving bias! - so I’m blowed if I can judge anyone else’s income. For this reason, I share the Devil’s consternation at the idea that public opinion should adjudicate.
Secondly, it’s possible that none of us deserve anything. This isn’t just the traditional Christian position that we are all miserable sinners. It’s also the Rawlsian one, that the distribution of talents - which include an appetite for hard work - is “arbitrary from a moral point of view.” And of course, none of us “deserves” the enormous good fortune of having been born into a liberal democracy in the late 20th century.
For me, these reasons suffice to disregard “desert” as a macro principle colouring our views about the distribution of income - though we might use it in other contexts, as when we say “he deserved that goal” or “he deserves to go to prison for that.”
This is not, in itself, a particularly leftist position. Intelligent libertarians share it. Hayek was wary of the idea of desert, and Nozick wrote of entitlements, not desert; there’s a difference.
All this said, I share Sunny’s view that there are distinctions within the “rich”. There is a difference between the talented footballer or musician or the innovative entrepreneur on the one hand and the rent-seeking exploiter on the other. But our tax system doesn’t - and probably cannot - make such distinctions. Which is, perhaps, one reason why the state alone cannot achieve social justice.

Cognitive biases in popular songs

Forget that guff about Rage against the Machine vs. X Factor - truly, a herd of independent minds. What’s more worrying is the large number of basic irrationalities contained in popular songs.
I was at the gym the other day - this finely chiselled physique doesn’t come naturally - and Alexandra Burke came on the TV, singing “The bad boys are always catching my eye.”
Well of course they are. Bad boys hang around on street corners and in malls where you can see them. Good boys on the other hand are working or studying and so are in offices and libraries where they’ll not catch your eye.
This is a sampling bias. It’s an elementary cognitive error.
But this is just the tip of the iceberg.
By this I don’t just mean over-confidence. Yes, this is ubiquitous, but in many cases it serves a rational purpose. When Tinchy Stryder sings “I will never leave you” to Amelle we all know the sub-text.
No, what I mean are more egregious errors. Take for example the Black Eyed Peas: “I gotta feeling that tonight’s gonna be a good good night.“
This, surely, is the base rate fallacy. Past experience shows us that tonight is rarely gonna be a good night; the best we can expect is to avoid too bad a fight and cop off with a girl who weighs less than us. But Will.i.am is ignoring this prior probability, and over-weighting his subjective feeling. This is a basic deviation from Bayesianism.
Let’s move to Cheryl Cole:

Anything that's worth having
Is sure enough worth fighting for
Quitting's out of the question
When it gets tough, gotta fight some more

The first two lines are acceptable. But the last two, surely, are not. Except in cases of severe duress, which Mrs C is not addressing, quitting can never be out of the question. Sometimes, when it gets tough, quitting is the right thing to do. To think otherwise is to commit the sunk cost fallacy.
Her protege Joe McElderry presents a more awkward case:

There's always going to be another mountain
I'm always going to want to make it move
Always going to be an uphill battle,
Sometimes you going to have to lose,
Ain't about how fast I get there,
Ain't about what's waiting on the other side
It's the climb
Keep on moving.

This, I fear, is an example of time inconsistency. At the start of the climb, you might believe you should always keep moving. But what about when you face the third or fourth mountain? Will your preference be to keep moving then? Or will you prefer not to climb at all? If so, your later preferences will be inconsistent with your earlier ones. 
It could be that singing about the climb is a rational precommitment  strategy; if you publicly say you’ll keep on moving, you’re increasing the costs of subsequently stopping moving by exposing yourself to more embarrassment if you do. But it’s not clear that Joe is doing this.
Now, I don’t want to claim that all of this year’s most popular songs contain howling irrationalities. The mighty Lady Gaga passes muster. But could it be that such irrationalities - as distinct from plain nonsense - are more common now than years ago? If so, isn’t this yet more evidence of the decline of civilization? And what hope is there of people becoming properly educated if they are exposed to such cognitive errors? Will no-one think of the children?

Inequality & innovation

Why did the industrial revolution occur so late in human history?
It’s not simply because we had to wait for the science. The steam engine, for example, was invented in the first century AD. But it was another 18 centuries before it became widely applied. Why the wait?
One possibility is that the industrial revolution required huge stocks of coal, which only became accessible in the late 18th century. Another possibility, advanced by Gregory Clark, is that the revolution required people to have the right attitudes - hard work, patience, ingenuity, respect for property - and these took centuries to evolve.
However, this new paper suggests, though doesn’t explicitly articulate, another possibility - that inequality held back the economy.
Roughly speaking, there are two sorts of technical progress. One is the introduction of new products. This is helped by inequality - because this implies that there’s a pool of wealthy customers for the new goods. 
But there’s another sort - process innovation, the introduction of new techniques that allow goods to be produced more cheaply for a mass market.  And this form of innovation is retarded by inequality. After all, if the majority of people are on the breadline, or working in subsistence farming, producers have little incentive to finds ways of making new goods cheaper, because people still won’t be able to afford them.
In this sense, process innovation - which is what the industrial revolution was - requires there to be a middle-income market, which in turn requires at least a little equality.
Now, I’ll grant that it’s tricky to test this hypothesis against the evidence. But two things suggest that equality matters for process innovation.
First, Japan has for long been a relatively egalitarian nation. But it has also for years been the home of process innovation, at least in manufacturing, pioneering lean production and just-in-time methods.
Secondly, the mid-20th century was, for the west, a golden era of process innovation. Mass production and Taylorism allowed once-luxury items - cars, refrigerators, TVs - to become cheap enough to be owned by relatively poor people. But the mid-20th century was also a period of rising equality. Coincidence? Or is it instead that growing equality incentivized firms to invest in mass production methods?
As I say, it’s just a suggestion than inequality retarded the industrial revolution. (Of course, there‘s no reason why the explanation be mono-causal). But if there’s any truth in this, then it implies that the cost of inequality, historically speaking, has been colossal. Delaying the transition from zero growth to 2% per capita growth by 100 years means denying people a seven-fold improvement in their real incomes.

We're all anarchists now

The 193 governments that met at Copenhagen were unanimous about one proposition. And it’s a remarkable one - that whereas anarchy is a bad idea within national borders, it’s a good idea across borders.
The anarchist says: “We don’t need government. Private contractual agreements between individuals are sufficient.” No-one at Copenhagen agrees with this when they look within a national boundary. But they all agree with it, when it comes to supra-national matters. They think global government - in the sense of a coercive body standing above national governments - is inferior to agreements between national governments.
The failure to reach a meaningful agreement at Copenhagen, however, throws this view into  question.
What I mean is that there are clear reasons why anarchy within borders is thought undesirable.  If laws could only be reached by the unanimous agreement of all individuals, the rich and powerful would only consent to be bound by them on terms onerous to the poor. The problem of collective action means that people won’t agree to contribute to public goods, preferring that the cost of doing so falls upon someone else. And on top of this is the sheer difficulty of getting lots of people to agree to anything.
However, all of these problems were evident at Copenhagen.  Which poses the challenge to the 193 governments: if anarchy is a bad idea at local levels, why is it a good idea at an international one?
The question gains force from a key fact - people have evolved rules and institutions which can mitigate the problems of anarchy at local levels. But Copenhagen shows that such institutions are, perhaps, weaker at supra-national levels.
So, what defence do the 193 governments have?
It’s not good enough to hide behind “national sovereignty.” Why should this be a value, except insofar as it ecompasses individuals’ autonomy and self-determination - which, for countless governments, it does not?
The only answer can be that the costs of global government (within limited domains such as climate policy and protecting human rights) outweigh the benefits.
Such an argument can be made by global warming sceptics. But is it really tenable for those who claim that climate change poses “unacceptable risks” to the planet?

Fiscal squeezes, class war & growth

Guido says the economy can grow even if government spending is cut.  There is quite a bit of historical (pdf) evidence (pdf) for this, and not just for the UK.
The question, though, is: what’s the mechanism whereby lower public spending can raise GDP growth? Two obvious possibilities are blocked off.
First, tighter fiscal policy in the past has been accompanied by lower short-term interest rates. But this can’t happen now. Yes, the Bank of England could step up its QE policy, but the efficacy of this is uncertain.
Secondly, past fiscal tightenings came when the personal sector’s debt-income ratio was low, and when banks freely lent to households, so personal borrowing increased to offset lower government borrowing. A repeat of this is unlikely.
So, what mechanisms are left?
1. Lower gilt yields could crowd in corporate spending. But real yields  are already under 1%, so they can’t fall far. And aggregate corporate investment isn’t very sensitive to moves in the risk-free rate; if they were, it wouldn’t have collapsed in the last two years.
2. Ricardian equivalence; households respond to the prospect of lower government debt by borrowing in anticipation of lower future taxes. This, however, only predicts that a tighter fiscal policy  will not affect the economy, not that it’ll expand it. And even this requires that households be able to borrow freely, which is unlikely in an era when banks will have to increase their capital ratios.
3. Sterling could fall, as the FX market anticipates lower interest rates than would otherwise be the case, and this boosts exports. But as we’ve learned this year, the response of trade volumes to exchange rate falls is, at best, slow and at worst small.
There is, however, a fourth possibility.  Silvia Ardagna has shown (pdf) that fiscal tightenings can sometimes raise GDP - though not always - if they involve cutting the public sector wage bill. The mechanism here, I suspect, is not just the mechanical effect of lower wages tending to raise private sector employment. It could also be the case that capitalists’ confidence - profit expectations - improves if they see a government attacking workers.
In this sense, a fiscal contraction might boost growth, if it is part of a class war.
But how likely is a repeat of this? Yes, for many countries in the 1970s and 80s, capitalists’ investment intentions were depressed by a fear of worker militancy, which receded as governments successfully attacked public sector workers. But is this really the case now?

TV & happiness

According to Ofcom, the British are spending more time watching TV, in contrast to several other European nations, where TV viewing is declining.
Is this good for us? Research by Bruno Frey has suggested not. He shows that many people who watch lots of TV report lower subjective well-being. One reason for this is that TV raises our material aspirations and so increases our discontent.
This hints at a worrying possibility. Maybe people lack self-control, or are unable to predict well how activities will change their tastes. If so, people aren’t the best judges of their own well-being. And if this is true even of mundane, everyday things like TV, how much more likely to be true is it in less familiar contexts? Which suggests a case for some kind of paternalism.
However, a new paper (pdf) by Mitesh Kataria and Tobias Regner show that this view might be too hasty. They show that the relationship between TV viewing and happiness varies enormously from nation to nation. For example, in Germany, Australia, Finland and the US (to a lesser extent) people who watch no TV are happier than those who watch more than three hours a day. But the opposite is true for Spain, Norway and Sweden (their data doesn’t cover the UK).
This suggests we just can’t generalize whether TV is good for us or not.
The trivial implication here is that it depends what’s on TV.
But there’s a less trivial implication. This is that happiness research is not (yet) a settled field. It’s findings are subject to legitimate dispute. Might there be a danger that everyone - paternalists or not - are too hasty to find what they want in it?

Inflation & gold

Guido reckons double-digit inflation is coming, and says: “If you haven’t got any gold, stock up on baked beans.”
Now, as Giles points out, inferring that we have an inflation problem because of this week’s figures is just silly. And as I’ve said - along with MPC member Adam Posen (pdf)- there’s no reason to suppose that printing money, at least on the scale we’ve seen so far, will add much to inflation.
But let’s leave this aside, and assume Guido’s premise is right. This raises another issue: does gold really protect us against inflation?
My chart suggests maybe not. It shows annual changes in the gold price (in sterling terms) and annual inflation. You can see that high inflation is not always associated with good returns on gold. The gold price soared when inflation was high in 1980 (just before my chart begins), but it fell in 1981 when inflation was above 10%, and it plummeted in 1990-91, as inflation spiked up.
Yes, the correlation between annual changes in gold and the RPIX inflation rate has been positive since 1980: 0.29, R-squared = 8.3%. But this is entirely because of the 1980 experience. Looking at the data since January 1981, the correlation is negative.
One reason for this is simply that the gold price is astonishingly noisy. Another reason is that if global inflation rises, the world’s central bankers could reverse their slack monetary policies, causing gold to fall.
Now, I say this not to take a view on where gold’s going. I’m just  saying that fears about inflation are not sufficient to generate a bullish view of the metal.
Another thing. Guido says he’ll “bet a large amount of money that the Governor of the Bank of England will have to write to Chancellor Osborne next year telling him why inflation is over-shooting target.” He’ll not get great odds. The Bank itself thinks there’s a 36% chance of inflation topping 3% in Q1, and a 25% chance of it doing so in Q2.

Knowledge & ownership

Giles Wilkes gives this nef report (pdf) a superb kicking. What strikes me, though, is that nef is, in an important sense, on the same side as the bankers they try to demonise.
What I mean is that both pretend to have perfect knowledge. Nef thinks that all externalities are identifiable and quantifiable. Bankers thought that risks were identifiable and quantifiable. They are pretty much the same mistake, the same arrogant and spurious pretence to expertise.
But limited knowledge is, perhaps, the central problem in economics. The are lots of issues we just cannot understand without recognizing its importance: why does unemployment exist? Why are asset prices volatile? Why does the price mechanism work better than central planning? Why do firms exist? You can’t begin to answer these questions without appreciating the importance of limited information.
It is in this context that I welcome the launch of an ownership commission.
The thing is, part of the case for worker ownership is that it is a means of using dispersed, fragmentary knowledge. Workers sometimes (often?)  know better than bosses how to cut waste or improve efficiency. Giving them ownership might therefore be a means of improving efficiency.
There are two principles here we should make explicit.
First, ownership should, ideally, flow to where there is knowledge*. Not all ownership forms achieve this. For example, customer ownership did not stop Equitable Life collapsing, in part because customers didn’t really know what the hell it was doing. Similarly, shareholders’ failure to stop banks going bust wasn’t just a problem of collective action  - it was that they didn’t know what banks were doing.
Second, ownership must entail real power. Again, this is not always the case. Many employees at Lehmans were significant owners, and some  had an idea that it was heading for trouble. But they didn’t have the power to challenge Dick Fuld.
And herein lies my scepticism about the ownership commission.  There’s a danger that it will advance a form of pseudo-ownership, which gives workers and customers responsibility without power. A genuine change of ownership forms, by contrast, would involve disempowering bosses. Is it plausible that New Labour - which has been the party of managers - will really do this?
* I say ideally because in truth it’s impossible to implement this principle perfectly, as knowledge cannot be quantified or even, often, expressed.

A history of rising equality

Have we become a much more egalitarian society in the last 100 years? This is the question posed by the ONS’s survey of wealth.
Newspaper reports of this survey claim that it is the first official estimate of the distribution of wealth. This is not quite true. Back in the 1970s, there was a Royal Commission on the distribution of income and wealth, which Thatcher abolished.
It estimated that, in 1923, the richest 10% owned 89.1% of all wealth. By 1952, this fell to 70.2%, then to 60.6% in 1976*. This paper (pdf) reports that the proportion fell to 48% in 1993.
And the ONS estimates that it is now 41%, if chattels and pension rights are excluded; the data is beneath figure 2.2.
Now, these are ball-park figures; the RCDIW admitted that the data was ropey, and they draw upon different sources. So I wouldn’t stake much on the precise figures. But there does seem to be a downward trend.
I suspect there are four main reasons for this:
1. Big estates were broken up in the mid-20th century, under pressure of death duties.
2. Inflation in the 1960s and 70s eroded some financial wealth, such as gilts.
3. House price inflation, plus the spread of home ownership, mean those below the top 10% have got relatively richer.  But not the bottom 20%. These have negative net wealth.
4. Rising real incomes have allowed some people to start saving. In the 1920s, and in the late 40s, the aggregate savings ratio was under 2%
I don’t say this to imply that we should tolerate today’s inequality;  this is a question of philosophy that cannot be settled by any figures.
I say it merely to suggest that inequalities mightn’t be as great as they once were.
Or are they? Today’s unemployed might be much better off than were, say, the Jarrow marchers - certainly in absolute terms, and probably relative to the better off. But mightn’t they be poorer in terms of social solidarity, or is this to romanticize the past?
* Table 4.5 of report no. 7, Cmd 7595, July 1979.

Happiness in US states

Louisiana is the happiest state in the US, followed by Hawaii and Florida. The unhappiest states are West Virgina, Rhode Island and Pennsylvania, though California tops the list for most mental distress.
That‘s the finding of a new paper by Stephen Wu and Andrew Oswald.
You might object that this tells us more about people than about states. Older people are happier than middle-aged ones, educated ones happier than less educated, and whites are happier than blacks. However, the authors control for these things.
The differences between states are large. The numbers suggest that moving to Louisiana and from West Virginia is similar to moving from an income of $35-50,000 to one of $50-75,000.
What’s more, there’s no statistically significant link between a state’s GDP per capita and its happiness.
These differences raise a puzzle. Why is it that, in a nation whose people are famously mobile, are differences not eroded by market forces? In theory, if a particular state offers worse amenities - climate, congestion, whatever - people would tend to emigrate, which in turn would bid up wages to compensate for the disamenities. Across states, there should be compensating advantages.
But this doesn’t seem to happen - perhaps because people are poor predictors (pdf) of future tastes, and so mistakenly move to higher-income, lower-well-being states.
Nothing in Louisiana, it seems, can offset the fact that it is home to much of the world’s greatest music.

Crime, race & incentives

There’s a paradox raised by the reaction to “Rod” Liddle’s mostly incorrect claim that “the overwhelming majority of street crime, knife crime, gun crime, robbery and crimes of sexual violence in London is carried out by young men from the African-Caribbean community.”
The paradox is this. When it comes to tax, the right are keen to stress that people respond to incentives. And yet when it comes to crime they seem coy about incentives, and prefer to talk about “multiculturalism“ or genes.
The paradox is especially strong because economic theory is much clearer on the link between poverty and crime than between tax rates and tax revenue.
This is because in the case of taxes, the income and substitution effects work in opposite ways. The substitution effect causes people to prefer leisure over work when taxes rise, whilst the income effect causes them to want to work more to recoup lost income. However, with crime the two work in the same direction. The income effect causes a poor person to turn to crime to raise money, whilst the substitution effect means the unemployed have more time with which to commit crime, and lower penalties - no danger of losing one‘s job - for doing so.
Unsurprisingly, therefore, the empirical evidence is much stronger for a link between relative poverty or inequality (pdf) and crime than it is for a link between tax rates and revenues.
All this suggests an obvious reason why blacks seem to be over-represented among criminals*. Blacks are disproportionately likely to be poor, and the poor are disproportionately likely to commit crime. We’d expect, therefore, blacks to be over-represented in crime even if there were no causal link from race to crime.
From this perspective, what’s remarkable is how few blacks commit crime. There are over 800,000 blacks living in London, but in 2007 only 4435 - less than 1 in 180 - were found guilty in a crown court (table 6.4 of this pdf).
Now, I don’t say this to dismiss a role for race of culture in crime - though, remember, culture often has economic roots. For me, the claim that blacks are more likely to commit crime than others, after controlling for economic factors (and peer effects?), is a hypothesis just like any other. It’s something to be tested against evidence, not to be spouted willy-nillly.
Instead, I do so to point out that there does seem to be a blind spot on the right as well as the left. If the left is loath to accept that there are racial differences in criminal behaviour, the right seem loath to accept that poverty and inequality are causes of crime, even though the Econ 101 of which they are usually so fond predicts just this.
* I say “seem” because the data cited by Charlotte and Antonia Senior - which is taken from table 5.4a of that MoJ report - refer to arrests. But these data might tell us no more than that Constable Savage is still a serving officer.

The economics of the Peter principle

The Peter principle isn’t just folk wisdom - it’s founded in economic theory. A new paper by Matthias Krakel shows how:

Competitive careers systems have an inherent tendency to promote the least productive individuals, thus leading to mediocracy.
The intuition for this result comes from the fact that more productive people have better fall-back positions than less productive ones when failing in the competition for top positions. Hence, highly productive people have only moderate incentives to win the competition for top jobs, whereas individuals with low productivity have strong incentives to avoid their rather unattractive fall-back positions.

This doesn’t just explain basic explain basic facts about your own workplace. It also explains why politicians are so mediocre.
However I also fear it also justifies not only wage inequalities, but increasing inequalities. Maybe companies have to pay CEOs fortunes to offset productive people's lack of incentive to chase top jobs.
Worse still, if ordinary economic growth means that productive people over time can earn increasingly good wages, then CEOs’ wages have to increase even more in order to offset the tendency for such people to happily do good lower level jobs.
There are, though, perhaps more basic messages here. Competition needn’t necessarily yield optimum outcomes; “survival of the fittest”  mightn’t be the most efficient principle. And the organization of corporate hierarchies might be inefficient, not (just) from heterodox perspectives, but from the perspective of conventional neoclassical economic theory itself.

Darling's spending plans: tough or not?

The Pre-Budget report envisages current government spending growing by just 0.4% a year in the three years to 2014-15. Is this a lot or a little?
In two senses, it’s not a great squeeze. There are two precedents for spending growing more slowly - the late 80s and mid-90s. Also, even by 2014-15, spending is envisaged to be 39.6% of GDP - well above its 1999-2000 level of 34.4%.
In other senses, though, it is a tight squeeze.
First, current spending is only part of the story. It’s net investment that’ll be clobbered. Darling sees this falling by more than half between now and 2014-15. And he foresees gross investment falling to 2.6% of GDP - less than in any year under the Thatcher government.
Secondly, those two precedents for such a spending squeeze also suggest that spending can’t be squeezed for very long: after them, spending surged. The trouble is, with net borrowing in 2014-15 forecast to be 5.6% of GDP - don’t even ask about the error margin around this forecast! - such a surge looks infeasible.
Thirdly, the fact that spending has been a smaller share of GDP even quite recently is meaningless; it’s much harder to reduce spending as a share of GDP than to increase it, especially if GDP isn’t growing much.
One question raised here is: is it possible to reform the public services with spending growth so slow? Two things suggest not:
1. Low spending rises - which imply low pay rises - antagonize public sector workers and so increase hostility to reform. People who are scared for their jobs are rarely enthusiastic about innovation. When Nye Bevan introduced the NHS he famously overcame doctors’ resistance by “stuffing their mouths with gold.” This won’t be an option.
2. Some desirable reforms require higher spending, at least initially. For example, if parents are to be given genuine choice among schools, there must be an excess supply of school places; if not, schools will choose pupils, not vice versa.  But how can this be achieved with real spending almost frozen?

Climate change vs global justice

Here’s another thing that puzzles me about Copenhagen: why is the issue of climate change being conflated with that of global justice? For example, the Copenhagen editorial says:

Social justice demands that the industrialised world digs deep into its pockets and pledges cash to help poorer countries

Justice demands it, but tackling climate change needn't. As anyone with a passing acquaintance with the Coase theorem knows, the payments required to internalize externalities can go in either direction.
It could be, in theory, that the best way to reduce emissions would be for countries to pay China and the US to do so, without paying Africa a bean.
One reason for this is simply that their emissions are huge whilst Africa’s are negligible.  For example, a cut in US carbon emissions of just 0.4% would suffice to offset a doubling of Zambia’s emissions. Also, the very fact that US emissions are so big could be a sign that they can be cut more easily than other countries’ - the low-hanging fruit here is perhaps in the US. And because China is a relatively poor country - its GDP per head is only one-sixth the UK’s - it could be easily bribed; of course, African countries are poorer still, but their emissions are too small to matter.
I don’t say this to deny the importance of global justice; there are countless arguments for helping Africa, of which helping it adapt to climate change is just one. It’s just that climate change is one issue, and global justice another.
I don’t see how either cause is helped by conflating the two.
Another thing: Norm says there’s a “quick answer” to the question I asked yesterday - that “even if global government is what is needed, it isn't available at the moment and it isn't easy to envisage its becoming available in short order.” This is true, but it relocates my question rather than answers it. It just raises the question: why, in the 15-plus years that climate change has supposedly been a priority (Kyoto was way back in 1997) has there been so little movement towards global government? I don’t think it’s as simple as saying that national governments are unwilling to give up powers, as European governments have done this to the EU.
I say this not because global government is a practical possibility, but rather to show how sharp is the contrast between our actually-existing politics and the institutions and options are are available in theory.

Copenhagen: a Buchananite puzzle

There’s something about the Copenhagen summit that‘s puzzling me. To see it, approach the issue from the perspective discussed by James Buchanan in The Calculus of Consent.
There are, he said, three broad ways of organizing our affairs, each with costs and benefits; he was writing about individuals, but for our purposes, we can replace individuals with governments. These ways are:
1. Simple individual action. In our context, this means national governments setting their own individual climate change policies.
2. Voluntary contractual agreements.

3. Enforced collective action. In the case of the private individuals Buchanan was discussing, this means government action. In our context though, it means a global government with the ability to coerce national ones into reducing emissions.
Now, here’s my puzzle. Politicians seem to think that climate change is best tackled by (2) - hence the summit. Why?
It’s obvious why (2) is superior to (1) - because stopping climate change is a public good. But why is (2) superior to (3)?
There are some obvious possibilities. Maybe climate change isn’t so catastrophic, so the costs of failing to reach an agreement will be too small to justify losing national sovereignty. Or maybe the costs of reaching an agreement are in fact small, so we are in the world of the Coase theorem, where private agreements can be optimally efficient without government action.

However, much of the rhetoric surrounding the summit seems inconsistent with these possibilities. That editorial says:

Unless we combine to take decisive action, climate change will ravage our planet, and with it our prosperity and security.

And Brown says:

If by the end of next week we have not got an ambitious agreement, it will be an indictment of our generation that our children will not forgive

But if the costs of failure at Copenhagen really are so catastrophic, and are a serious risk, then perhaps  mode (2) is the wrong way of addressing the problem, and we might need global government instead. And yet, AFAIK, none of the summiteers is suggesting this possibility.
Let's contrast this with smoking in pubs. In theory, private contractual arrangements could control this - either by non-smokers paying smokers to desist, or smokers paying non-smokers for the right to do so. But the UK government decided against this, and imposed a ban. It decided that mode (3) is superior to mode (2). But if this is true for smoking, why is it not true for an issue with allegedly vastly greater costs of failing to reach a private agreement? 
So what’s going on? Is their rhetoric just wind? Are governments ignorant of the basic Buchananite principles for deciding where decisions are best taken? Or do they instead care more about preserving their power than saving the planet, and see Copenhagen as merely a legitimation ritual? Could it be that those calling for radical alternatives to the summit do, in fact, have some logic on their side?