Expensive gimmicks

Perhaps a better plan to make these delinquents behave would be for NHS to offer a prize, say a million pounds, to the one who can find the missing 2.7 bn.  That would keep them busy.

Probably forever. These kids aren't exactly rocket scientists.

It's a nice joke, but I'm going to pretend it's serious for the sake of a bit of a discursion, as you've brought up the subject of prizes. The Economist this week carries an article on the efficacy of prize-giving as a form of motivation. The argument is that paying someone to achieve something typically delivers effort that is some fraction of the value of the pay, whereas prize-giving, by providing an open incentive, encourages multiple efforts that are, in combination, worth many times more than the value of the prize.

A recent example is Sir Richard Branson's offer (encouraged by Al Gore) to pay $25m to "the inventor of a commercially and environmentally viable method of removing greenhouse gases from the atmosphere" (in the words of The Economist). In a short article, they didn't have the space to provide the detail - that the winning scheme must be able to remove 1 billion tonnes a year from the atmosphere (leaving carbon accumulating in the atmosphere at a rate of around 2.2 bn tonnes a year, barring other significant reductions). If the scheme only worked for one year, that's a price for the carbon of around 2.5 cents per tonne. Assuming it were an ongoing process, then Sir Richard would be valuing carbon at a fraction of a cent per tonne. I do not accept the pseudo-economic calculations of the social cost of carbon, but if one took them as indicative of the order of magnitude of the value if climate-change theory is correct (which must be the assumption behind Branson's offer), typical estimations are in the range $100 - 150 per tonne of carbon. It looks like anyone going in for Mr Branson's competition would be seriously short-changed. Or put another way, no one in their right mind will be motivated by this because the value of the prize is not proportionate to the value (and probably cost) of the winning proposal. (One might think, cynically, that the objective of the prize is to produce, not a solution, but an improvement in the public perception of the environmental credentials of a man who really wants to keep increasing the emissions from his expanding airline, but perhaps that is being too unkind.)

Motivation by prize-giving has the same effect as IP protection by patent. In both cases, there can be only one winner. Those who think, like The Economist (which has gone unforgiveably wet in the last eighteen months) that favouring the first-comer leverages investment, make the classic mistake that Hazlitt and Bastiat warned against, of considering only the immediate and direct effects and not the broader ramifications that are not so immediately apparent. They look at what the winner delivers and the benefit to him and others of developing a winning scheme, without considering what the losers might have delivered and the broader benefit to society of a range of competing solutions. Not to mention the direct loss to the losers of having invested in a solution for which they now have no prospect of a return whatever the merits of their techniques (which may well be better than the one who rushed to win).

Though there can be more than one winner, the same criticism can be levelled at dealing with problems by grant-funding centrally-mandated innovation, George W Bush's preferred option. Who knows whether ethanol is the best (or even a good) solution to environmental issues? In a market where carbon were valued rationally, ethanol would win or lose according to whether it was sufficiently efficient at delivering carbon reductions relative to all other alternatives. With grant-funding, it is the government, rather than the market, that chooses the preferred solutions, by offering prizes to a few winners to deliver the pre-judged outcome. This approach discounts the possibilities (a) that the government's choices might be wrong (normally the case), and (b) that those who receive the grants may not be the best people to deliver the government's choices (also normally the case). All-in-all, it is not very different to the economic approach of the Soviet Union.

When people say that the current American government is too right-wing, I don't know whether to laugh or cry. It is authoritarian, interventionist and fiscally irresponsible. Those are hallmarks of left-wing, indeed socialist, governments. A true right-wing government would create property-rights frameworks, minimise government expenditure, balance the budget, and leave the rest to the market. Dubya is about as far from being a right-wing President as it is possible for a Republican to get. His social conservatism and faith-based politics are a red-herring. The same goes for the current British government, which is far more left-wing (i.e. interventionist and authoritarian) than the popular perception. Indeed, it is hard to think of a real right-wing government anywhere in the world at the moment.

Returning to carbon-valuation, let's also consider what a successful winner of Branson's prize would mean for cap-and-trade markets, the preferred option of naive free-marketeers. Under this sort of scheme, arbitrary caps are placed on allowed emissions within a period (typically a year), and emissions rights within those caps are allocated equally arbitarily (in the best case auctioned but more commonly handed to polluting incumbents). If combined emissions look like they will exceed the cap, the marginal value of techniques that reduce emissions is very high. If it looks like emissions will be less than the cap, the marginal value of techniques that reduce emissions is very low.

These schemes do not generally accommodate carbon-absorption very well - they are designed to encourage emission reduction rather than absorption. But assuming they were adapted to include a major absorption scheme like that desired by Branson, what effect would that have on the carbon market? The caps in every carbon cap-and-trade scheme are set on a very gentle downwards path (sometimes even an upwards path, justified on the basis that the upwards slope is less steep than is anticipated by the scheme's designers' crystal ball) in order to "avoid shocks to the economy". A winning entry to Branson's competition would result in massive oversupply of carbon absorption capability relative to that required by the market. The marginal value of carbon-reductions (within the mechanism) would fall to zero. All other measures to reduce carbon emissions would be made uneconomic, even though concentrations of carbon in the atmosphere were continuing to accumulate at around 2.2 bn tonnes a year. Quite likely, this rate of accumulation would increase as other efforts to keep carbon emissions down were abandoned as uneconomic. For instance, with a minimal carbon value, electricity generators would be encouraged to use the cheapest fuel (rather than the fuel that offered the best balance of cost, efficiency and carbon emissions). As this is currently coal, and is likely to remain so for the foreseeable future, a low carbon-value (such as is being delivered currently by the absurd European cap-and-trade mechanism, EU-ETS) would likely encourage increased adoption of coal-fired generation and therefore accelerated carbon emissions.

Many people who support markets think that any market is a good market, and that you can justify any scheme, however daft, if you can dress it up as a market. They do not have a sufficient understanding of the philosophical basis of value to realise that only markets that allow reasonable access to all participants, without preference to technique, and that accurately reflect subjective valuation, risk, time-preference and above all property rights are good markets. A market in stolen goods, for instance, is not a good market. Neither is a market where access is partial, some techniques are preferred over others, and valuations are established by arbitrary rules unrelated to property rights and personal preferences. At root, cap-and-trade is such an irrational market.

Whether or not you believe in climate-change theory, neither prize-giving nor cap-and-trade delivers a rational price for carbon and an appropriate incentive to action. (The test of whether a carbon market is rational for those who do not accept anthropogenic global warming as a real problem, is whether, should they succeed in persuading most people that they are right, the valuation of carbon within the market would fall towards zero, and whether in the meantime they can back their belief that carbon emissions are not a problem and potentially profit from doing so if they are proved right - and pay up if they are proved wrong.) As usual, what is needed is not some trendy, externally-imposed, arbitrary, partial incentive, but a real market that allows values to be discovered and all-comers to compete to service demand. Negotiations are commencing for what will follow the Kyoto Treaty when its provisions come to an end in 2012. Rather than persisting with the current broken model, we should start again and design a real market.

An offhand joke earns a serious economics lecture -- I do appreciate  this blog!

 While I largely agree with your analysis, I think the comparison of Branson's apparent willingness to pay for carbon removal vs. the "social cost" of carbon is off-target.  There's no reason why they should correspond.  Also, why "Economist's" logic is faulty, there's nothing objectionable about private parties offering a prize as an incentive for something they wish to see done.  OTOH, the idea that some particular solution to the problem (if there is a problem) will be the politically favored one is very problematic.

 I am likely to have more thoughts on your comments, but as I have students waiting to see me I suppose I should briefly exit the blogosphere for a bit.

Yes, prizes don't have to match prices, and private parties can offer whatever they want for whatever they want, but:

(a) If you are going to hold them up (as The Economist did) as an efficient method of delivering outcomes, you need to take into account the whole picture and not just the cost and value of the winner.

(b) If they are more than just a bit of fun and are to be taken seriously as a way of incentivising outcomes, it is fair to compare them to alternative ways of incentivising outcomes. Their effectiveness at valuing their objective is a fair part of that comparison.

(c) If you offer a prize that is worth very much less than the value of the winning entry, you may be subject to some legitimate questioning as to whether your intent is genuine or whether the offer is just a PR exercise. In Branson's case, not only is it inconceivable that someone will be able to win his prize for a cost that makes it worth entering, but if they did, the value of the carbon savings would so much exceed the value of the prize that the latter is almost an irrelevant incentive. And if that weren't true (i.e. if carbon were worth so little that an award worth a fraction of a cent per tonne were the dominant incentive) we should look to the carbon-valuation mechanism to consider whether it was creating a rational and effective market. If it were, people's concern about the risk of anthropogenic global warming would be so low that Branson should save his money. If not, we should create a better mechanism, not try to cover up the flaws in the current one with philanthropic gestures. 

(d) How is the winning entry to be judged? Is it just that it could hypothetically deliver the billion tonnes of carbon savings per year? Or should it also be economically feasible? If the latter is not a consideration, there are dozens of solutions one could come up with right now. If economics is relevant, which it must be:

  (i) it is very dependant on rational carbon valuations. Which rather begs the question....

  (ii) the judges will be caught as much by the Hayekian informational constraints as any central-planning government department.

This problem needs a proper carbon-pricing mechanism, not prizes, grants, cap-and-trade, regulation, and all the other mechanisms of which big businesses and politicians are so fond. At root, I suspect we would both agree that this is a property-rights issue. My problem with alternative mechanisms, is that they not only skew the market and take the pressure off politicians to avoid introducing a rational valuation mechanism, but often in the process are abused to protect preferred businesses and technologies. So although prizes may seem like a soft target, it becomes necessary to deal with them when a well-respected journal starts suggesting that they are an effective way of delivering outcomes. A bit of fun, yes. But effective economic instruments, no.

I greatly admire Richard Branson in most regards. But this is just a cosmetic exercise to try to gloss over the fact that he wants to fly more and more of us around the world or into orbit. It is a piece of greenwashing, little more. He wouldn't let his competitors get away with it, and we shouldn't let him.

I've managed to drive off the students and can return to more interesting matters.

Here's a devil's advocate argument for carbon cap & trade (CT) and Branson's contest...only it's a partial devil's advocacy, and partial advocacy -- the desirability depends on what alternatives are available, and must assume there's really a problem.

So let's assume that carbon emissions are indeed a problem.  A CT scheme with non-binding caps is superfulous, except that it prepares market participants for the day when caps become binding. But even then, the value of abatement technology isn't zero, since firms must place some value in avoiding binding caps.  Under a binding cap, abatement clearly has value, since it allows firms to expand production w/o purchasing emission permits.  But if the CT scheme is adapted to cover net emissions, then carbon removal technology also takes on value in both cases.

I don't see any reason why the Branson winner would necessarily crowd out other technologies for abatement and removal, which I believe is what you are suggesting would happen.  A properly designed CT system would have firms buying emission permits, abatement, and absorption, in a cost minimizing mix that's balanced against the value of output.

The trick is setting "the right" cap.  The cap is, in a sense, pulled out of the air; likely isn't what we'd have with a true market.  And if reduction in total carbon is called for, the cap would be negative, which seems very problematic.  One might develop a scheme for negative net emissions, but I can't see how this would work.

But given a goal of reduction, the following could work: suppose we wish to remove x units of carbon per year.  Establish a cap and trade system for y units of carbon annually.  Then, separately, offer contracts for bid to remove z units of carbon, where z - y = x, and low bidders get the contracts.  I suspect that doing this on an annual basis wouldn't be quite right, but the time period isn't important here (although from the standpoint of climate science it's probably crucial, but no one has any good answers on this.)

Who would pay for carbon removal?  I suppose governments will.  But private associations of people harmed by supposed global warming would have some incentive to pay.  There are costs to organizing these groups, but there's nothing inherently problematic here.  Maybe tougher is the public good problem, since beneficiaries can free ride rather easily.

Contrast this with a "complete" market: those who suffer damage from climate change would either pay to avert damage, or else polluters would be liable for damages inflicted.  This solution is problematic, since transaction costs are quite high -- defining the property rights is problematic.  Of course, governments have worked diligently to ensure that this is so.  http://www.perc.org/perc.php?subsection=6&id=653

The real market is superior to my scheme above.  It would better take into account the tradeoffs between reducing emisssions and producing goods; the lack of clear property rights renders it difficult.

So the CT & carbon removal scheme beats command and control regulation; it also beats the pure market, *if* the pure market is unattainable.  (Maybe I should submit it to Branson's contest!)

My analysis ignores all sorts of difficult problems, such as how to determine whether there's a problem in the first place -- something a real market would solve.  For what I've written to make any sense, it has to be assumed there is a problem and that reducing carbon emissions is desirable, that costs of defining property rights in the relevant dimensions are exhorbitant, and that government can overcome the usual knowledge/calculation problem and public choice problems so as to not make things worse.

In the end, it sounds to me not much better than Post-Science.

The Branson contest rules are very vague, enough to give me great hope that my blackboard scheme should be a winner.

But when I look at the panel of judges, my expectations are quickly dashed upon the rocks of despair.  There's not a judge among them (save Branson) who would have a clue as to what "commercially viable" might mean.

Still, if the market process isn't a "commerically viable process," nothing is.

I need to get home, so it's my turn to dash off a quick reply with hopefully more to come later.

When you say, "Under a binding cap, abatement clearly has value", have you considered the scenario in which the level of the cap exceeds the level of emissions? That is the situation under the largest cap-and-trade scheme in the world - EU-ETS. As any first-year economist could have predicted, prices have collapsed (to less than one euro per tonne of carbon). Does that mean that carbon savings amongst the sectors of European industry affected by EU-ETS are almost worthless? Are they worth more or less than carbon savings from other sectors, and twenty times less than they were a year or two ago? Our emissions haven't fallen, in fact they're still rising, so how come the value of savings has collapsed?

You might say, that is the fault of the design of EU-ETS. But it is inherent in the cross-border political process that trust and common-interest is insufficient to motivate people to play straight, and certainly not enough, in a grouping of independent democracies, to provide a stable long-term framework about which people are sufficiently confident that they are prepared to make long-term investments. Yet a stable, long-term, cross-border (not just European, but worldwide) mechanism is exactly what is needed (as you quite rightly point out in your excellent Penguins and Conservatives post over at Unforeseen Contingencies).

In short, cap-and-trade always creates an unrealistic market because marginal valuations change abrubtly at an arbitrary threshold (I've got an old paper on this I might try and dig out), but it is particularly inappropriate as an international mechanism. That is why the first implementation (for sulphur dioxide in the USA) worked tolerably well (because it suffered only from the lack-of-realism problems and not from the trust problems, and the former are not very visible if the caps are set sufficiently aggressively that there is never an excess of supply), whereas EU-ETS is a dog of a scheme that needs to be put out of its misery.

Frustratingly, I couldn't find that paper (see previous comment below), but I found the figures that accompanied it, so I'll have a go at precis'ing the argument. Much of this will be teaching grandma to such eggs, but bear with me, as I hope we won't be the only ones ever to read this.

I believe it is fairly widely accepted in economics that the more one has of a good the greater its utility to its possessor, but that increases in utility relative to volume are not linear - that the marginal increase in utility attributable to each extra unit is less than the marginal increase attributable to the previous unit. This is a rather long-winded way of expressing the concept of ordinal utility, which Rothbard depicted simply and graphically by the following list, ranking the expected order of preference for a number of eggs:

5 eggs
4 eggs
3 eggs
2 eggs
1 egg
2nd egg
3rd egg
4th egg
5th egg

We cannot say that the utility of any item from this list is in any specific mathematical proportion to any other item (e.g. the utility of two eggs is not necessarily double that of 1 egg, because we cannot say that the utility of the first egg is equal to the utility of the second egg, that is a subjective judgment depending on the circumstances), but we can say that they are normally in this order. This is the insistence that utility is ordinal not cardinal, which puts Austrian economics on a collision course with most other flavours of economics.

Whilst one cannot put numbers to the amount of utility yielded by given volumes of a good, I believe one can depict the concept graphically. For marginal utility (the additional utility of each extra unit of a good), it is a falling curve tending towards zero marginal utility as volumes tend towards infinity. For total utility, it is a rising curve, tending to flatten off as volume tends towards infinity. It will be easier to see what I mean by looking at the images below, but I want first to describe the assumptions about utility that are embedded in a cap-and-trade scheme, and how they differ from what we would view as normal preference.

In a cap-and-trade scheme, we are talking either about the disutility of a good (or "bad" in the horrible modern parlance, or "externality" in the economist's parlance, and typically some form of pollution in practice), the utility of a reduction in the volume of that good (pollution savings), or the utility of an increase in an alternative good (air purity). It is really the same thing expressed in different ways. Because it tends to be viewed that way, let's stick with the notion of the utility of a reduction in the good.

The utility of a given reduction of a good (such as carbon emissions) under a cap-and-trade scheme is established, not just with reference to the total amount, but more strongly with reference to the threshold. Where the total amount of the good is above the threhold (e.g. emissions exceed the cap), there is a significant utility to each unit of reduction. Where the total amount of the good is at or below the threshold (e.g. emissions are less than the amount allowed), there is no significant utility to each unit of reduction. At the point where a unit of reduction causes total emissions to fall to the level of the cap, there is a dramatic change in the marginal utility of each unit of reduction. This does not reflect what most people accept to be a normal preference scale. A graphical depiction of the way utility changes with volume according to the classical model and the targeted cap-and-trade model is shown below ("Compliance" refers to a particular cap-and-trade scheme - the x-axis simply measures units of a good).

The point at which the target-based curve falls to zero is obviously the level of the cap. To the left of that, an insufficient quantity of the good (e.g. pollution savings) means that each additional unit of saving has a high utility. To the right of that, a surfeit of the good relative to the level of the cap means that each additional unit of saving has no utility. The classic curve demonstrates how one would expect the utility of the good to change in reality. It can be seen that a target-based mechanism, such as cap-and-trade, distorts the utility and therefore the value of reducing goods at all but two points where the curves cross. When no reductions have been made, the utility of each unit of reduction may be undervalued compared to a normal utility curve. (This is not necessarily the case - the targeted mechanism may create a value that exceeds the value of a more rational mechanism, but in this case, the curves would cross only once, and the discrepancy with the rational utility of the good will be increased as the volumes tend towards the cap. In other words, the upper plateau of the purple curve may be so much higher that it may be higher than the starting point of the orange curve.) At some point before reaching the level of the cap, the target-based marginal utility of additional savings will be higher than one would expect the marginal utility to be on a falling curve. And then, once one reaches the level of the cap, the marginal utility under the target-based scheme collapses and is lower than would be rational.

The impact on total (or cumulative, as opposed to marginal) utility is shown below.

Again, the target-based mechanism does not create a utility curve that approximates to a rational model. At some point (the level of the cap), the target-based curve hits a point where, no matter how much more of the good we have, there will be no further increase in total utility. At this point, everyone stops trying to produce more of the good (e.g. pollution reduction). On a rational curve, people try less hard to produce more of a good the more that we have of it, but they continue to be motivated to produce more of the good providing it can be done sufficiently cheaply (i.e. with sufficiently little disutility).

Cap-and-trade is not the only example of a target-based mechanism, but it is one of the most popular versions nowadays. Whatever the detail, a targeted mechanism is inherently wrong, by its very nature. The many who talk up such mechanisms as somehow virtuous have been seduced by the suggestion that it is a market-based mechanism, which it may be, but it is not a sensible market.