Bali Schmali?

<!– @page { margin: 2cm } P { margin-bottom: 0.21cm } –>

Is there a valid case for special treatment for poor countries (developing nations and some emerging markets) in global efforts to combat global warming?

Mr Munir Akram, spokesman for the 130-strong G77 group of developing nations believes he has fought the good fight in Bali, by resisting pressures on developing countries to accept obligations to make absolute cuts in their greenhouse gas emissions, under the successor to the Kyoto Agreement that is now supposed to be negotiated in the coming two years. In the Financial Times of Saturday December 15, he is reported as saying that the developing nations had come under pressure to agree to “commitments and obligations on mitigation which in their dimension we feel are unfair and unjust”.

There are three arguments Mr. Akram and other spokespersons for the developing world make to support their claim for special treatment for developing countries and emerging markets. Two are partisan, confused and invalid, and should be rejected.  The third is valid and can and should be accommodated.   

The first invalid argument is that, however large the aggregate volume of greenhouse gas emissions (and other manifestations of environmental vandalism) in developing countries such as China and India are today, their emissions are much lower than those of wealthy industrialised countries if measured per capita. Propositions can be both true and irrelevant.  The proposition that per capita emissions of greenhouse gases in China and India are low and therefore not a problem, is as relevant as would be the statement that the 6.9 percent inflation rate in China is less of of problem than the 4.3 percent inflation rate in the US, because per capita inflation in China is lower.  The fact that per capita emissions of greenhouse gases in China are much lower than in the US is  irrelevant as regards the damage done by these emissions, and irrelevant for the design of effective and efficient policies to reduce global greenhouse gas emissions. This is because the science and bio-engineering of greenhouse gas emissions make it abundantly clear, that a given volume of emissions damages the global environment regardless of how, why, by whom, and by how many it was emitted. The negative effect and the negative externality depend only on the volume of emissions.

An interesting implication of the fact that the benefits from a given reduction in the volume of greenhouse gas emissions is a pure public good – non-rival and non-excludable – is that  the countries with the largest populations like China and India in fact benefit more than countries with smaller populations from any reduction in greenhouse gas emissions anywhere. When there are more people, the per capita benefit from a reduction in greenhouse gas emissions does not decline. In fact, the value of the per capita benefit in many of the poorest countries, including India and China, may well be higher than in the richer countries, because the damage that will be caused in the poorest countries by global warming, by flooding, desertification etc. is likely to be much more severe than in the richer countries of the world.  (That would certainly be the case if you did not base the cost-benefit analysis on willingness to pay, which is contingent on ability to pay and therefore on the existing distribution of wealth and income, but instead on a more Rawlesian, status-quo-blind cost-benefit analysis).

The second invalid argument for special treatment for poor countries is that, regardless of whether China has already become the largest emitter of greenhouse gases, as some authorities argue, or will only overtake the US in a handful of years as the leading environmental bogeyman, the cumulative stock of past emissions of the old industrial countries is much larger than that of the developing countries and the emerging markets. This observation too is correct but irrelevant.  Since vacuuming past greenhouse gas emissions out of the atmosphere is not a cost-effective method for reducing atmospheric concentrations of greenhouse gases, the appropriate response to this second correct but irrelevant proposition ranges between: ‘and?’ and ‘so what?’.  It may induce sufficient guilt feelings among past atmospheric polluters (or among their descendants) to induce them to loosen the aid purse strings, but it has no bearing on what is the most efficient way to combat global warming.

The morally valid argument is that developing countries and emerging markets are poorer on average than the old industrial countries, and that it is  unfair to put an equal burden on the poor as on the rich. I agree. The question is, can we make this value judgement in a ‘green’ way?  It turns out we can.

Efficient greenhouse gas emissions control

If excessive CO2Eq (carbon dioxide equivalent) emissions are a problem, there are but two solutions. The first is command-and-control methods: limit the scale of the activities creating excessive CO2Eq emissions by administrative or regulatory fiat. In the limit, ban them. This was done with chlorofluorocarbons (contributors to the ozone hole over the Antarctic) which were phased out by 1996. It can be effective if something is to be banned completely. That is not possible with CO2Eq emissions. Any conceivable future will have continued emissions of CO2Eq. Bureaucrats are not very good at deciding who can produce how much CO2Eq in tens of millions of activities and firms. Last time something like that was tried we called it Central Planning.

From the perspective of an an effective policy to combat global warming by limiting and if possible reducing first the flow of greenhouse gas emissions and ultimately the atmospheric concentration of greenhouse gas emissions (the stock), it is necessary and sufficient to raise the opportunity cost of emitting an additional unit of carbon dioxide equivalent (CO2eq) and/or to raise the opportunity cost of not cutting back greenhouse gas emissions by an incremental unit.

Raising the opportunity cost of emitting an additional unit of CO2eq means taxing greenhouse gas emissions.  This can be achieved by imposing a uniform tax on each unit of emissions (as measured by say, a metric ton (Mt) CO2eq units). Let’s call it the uniform global CO2eq emissions tax or carbon tax for short. For efficiency reasons, this carbon tax should be the same, if the externality is the same, regardless of whether  the emissions come from the poor in Bangladesh or from the rich in St. Moritz.  That’s obviously unfair.

Fortunately, efficiency and equity can both be served once we recognise that the alternative to a common global tax per Mt of CO2eq emissions would be a common subsidy per Mt of CO2eq not emitted. The uniform tax rate per unit of CO2Eq and the uniform subsidy rate should be the same.  When emissions in a developing country increase above their allowance, it would not have to pay the tax, but there would be a progressive withdrawal of the subsidy.  So the incentive to reduce greenhouse gas emissions at the margin would be the same, but the distributional consequences would of course be quite different.

The subsidies for CO2eq emissions reduction would have to be financed out of a global fund, the Global Warming Fund, or GWF, say.  Taxes imposed by the richer countries on greenhouse gas emissions could go into the GWF, although there is no logical reason for earmarking them like that.  In principle, the GWF should be funded out of general revenues, in an amount sufficient to pay the subsidies required to achieve the desired reduction in greenhouse gas emissions in poor countries.

The carbon credits/ carbon offsets racket

A carbon tax is hard to implement, because it requires, as with any tax, that the authorities know and can observe the tax base.  In this case this means that the fiscal authorities must be able to monitor actual emissions of CO2eq. These measurement problems are hard, but not insurmountable in the advanced industrial countries.  With very weak tax administration and enforcement capabilities, many industrial countries will find this a formidable challenge.  But that holds equally for all alternatives, including subsidies and cap-and-trade. It would be helpful if  the carbon subsidy (or subsidy withdrawal) I propose could likewise be based solely on the actual volume of emissions, and not on both the actual volume of emissions and on the volume that would have been emitted in some hypothetical counterfactual parallel universe. That is the problem that has turned the carbon credit/carbon offsets generation and verification industry into such a cesspool of waste and corruption. 

A carbon subsidy (that is, a subsidy for not emitting a given volume of CO2eq), must be measured relative to some independently verifiable, measurable benchmark.  For practical purposes, this would have to be the most recent historical level of emissions, or that verifiable number plus or minus some amount or percentage representing ‘normal growth’.  The problem is that we really want to measure actual emissions relative to what they would have been this very same period if the subsidy regime had not been in effect.  However, determining what that benchmark is requires the observation, measurement and verification of a counterfactual number of CO2eq units that would have been emitted in the alternative state of the world where no subsidy regime was in effect but everything else would have been the same. Observing, measuring and verifying that counterfactual is impossible, even with the emergence, following the signing of the Kyoto Protocol, of a massive and still growing Verification of  Carbon Credits Industry. The information problems associated with the verification of the counterfactual CO2Eq savings associated with a typical carbon offset scheme is insurmountable. Just try answering the following questions: (1) how many Mt of CO2eq did you not emit in your aluminium smelter today? (2) How many scrubbers did you install that would not have been installed without the carbon credit incentive? (3) How many hardwood trees did you not cut down today? (4) How many times did you not beat your partner today? 

I am very aware of these problems, because in a previous life, as Chief Economist of the European Bank for Reconstruction and Development, I was involved in assessing the ‘additionality’ of projects the Bank was considering undertaking. A project was additional if it would not have been undertaken but for the EBRD’s financing being available.  As we were unable to run history twice, once with the EBRD and once without it, the determination of the additionality of a project required a host of untestable ‘identifying assumptions’. No matter how plausible these assumptions were (from my perspective), someone would always propose another counterfactual for which additionality was present when I believed it to be absent and vice versa.

It may seem that the subsidy withdrawal scheme does not require the verification of a hypothetical counterfactual. The subsidy rate is s per Mt of CO2Eq.  Each year the country is set a benchmark level of emissions, say, B  Mt of CO2Eq. This could be last year’s emissions plus some increment for ‘normal’ emissions growth.  Actual emissions during the year are E Mt of CO2EqThe amount of subsidy received, S, is given by the larger of zero and  s(B – E) .  Thus, if actual emissions fall below the target for the year, that is, for B > E , the subsidy is proportional to the emission savings, with a marginal subsidy rate of s.  It would be possible to turn this into a combined tax and subsidy scheme by extending it to actual emission levels in excess of the benchmark, B, but this would run into the familiar fairness problems.  However, we have merely hidden the counterfactual problem in the benchmark B.  Ideally, B would be the volume of emissions the country would engage in with a zero marginal tax or subsidy.  Basing an estimate of what that would be on history plus some estimate of ‘normal’ growth amounts to constructing an observable proxy for the unobservable counterfactual.  It may seem plausible, but it won’t be rocket science.

It is important to note that the possibility of manipulation of the choice of the benchmark does not in itself undermine the efficiency-enhancing properties of the subsidy-withdrawal scheme.  Its incentive effects depend only on  the marginal subsidy rate,  s,  and not on the benchmark level of emissions, B, as long as the benchmark, B, is independent of the actual volume of emissions, E.  The benchmark could be set at an arbitrarily high level, so there would always be a large subsidy, without this affecting the incentive for economizing on emissions.  Manipulation of the benchmark would, however, undermine the distributional legitimacy of the scheme.

Under the compromise reached in Bali (which actually was no more than talks about talks or at best talks about negotiations), developing countries will take on commitments to curb the growth of their emissions in exchange for financial incentives.  This rather sounds like the benefit withdrawal scheme outlined above.  This can only work, that  is, be efficient and fair, provided a reasonably neutral and non-manipulable mechanism is created for setting the annual or multi-year benchmarks B,  this could be part of an efficient and fair solution.

So is there a tax-cum-subsidy solution?  There is, but it involves uncoupling the efficiency and equity parts of the equation.  Every country, rich or poor, would impose a uniform CO2Eq emission tax on the actual volume of emissions.  No need for a counterfactual.  Then the poor countries would get, from the rich countries, a subsidy, or rather, aid, to compensate them fully or in part for the welfare losses caused by the imposition of the tax.  Two problems with this ‘solution’ are, first, that the rich countries are better at making aid commitments than at following through on these commitments (and they are not even very good at making the commitments) and, second, that in many poor countries, governments and public administrations are so corrupt, incompetent and/or repressive, that aid often does more harm than good.



Cap-and-trade (a global ceiling on greenhouse gas emissions, implemented through the issuance of a fixed global volume of CO2eq emission permits for a given period) is to all intents and purposes equivalent to imposing a uniform global carbon tax combined with a mechanism for distributing the carbon tax revenues, as long as a uniform global price is established for the permits in an efficient secondary market.  The only difference concerns the distribution of the scarcity rents. In the case of a carbon tax, the revenue would go to the national fiscal authorities. In the case of cap-and-trade, the distribution of the rents would depend on the way the permits are initially distributed or sold. The efficiency-enhancing effect of the cap-and-trade scheme will be there regardless of how the permits are distributed, as long as the overall amount is chosen well and there is a well-functioning secondary market.  It is the presence of the market that creates the opportunity cost for the would-be polluter/ emitter.

The distributional objectives achieved by taxing CO2eq emissions in rich countries and subsidising the non-production of CO2eq emissions in poor countries can be achieved in the cap-and-trade scheme    without the need for dodgy counterfactual scenario constructions, by allocating the poor countries larger shares of the stock of emission permits.  Distributive justice would be served if the poorest countries got emission permit allocations well in excess of their likely domestic emissions under efficient abatement policies.  They could then sell the surplus permits on the open market.  Both poor and rich countries would receive the right marginal incentives for emission control from the uniform price established for the permits.  With cap-and-trade, the Global Warming Fund would have to be replaced by a body (the Global Warming Body, say) deciding both the overall size of the global emissions of CO2eq during a given period, and the initial distribution of the permits, either to the national authorities or to the would-be emitters themselves.  The initial allocation of the permits could be sold rather than given away, which would turn the Global Warming Body into a Fund.  Such a Fund could have priorities other than greenhouse gas emission management.

The main weakness of the cap-and-trade scheme is that it stands or falls with the degree of efficiency of the secondary market for permits that would be established.  If that market were a textbook efficient market, I could only cheer it on.  All kinds of derivative products (futures markets, options of various kinds) would soon be created and the opportunities for trading and sharing risk would be enhanced.  However, the events in global financial markets since the beginning of the year, and especially since August, may lead one to doubt the starry-eyed stories of market efficiency that still dominate the teaching of finance in economics departments and business schools.  Taxes and (subject to the counterfactual problem) subsidies are a viable alternative to cap-and-trade.  The fact that politicians prefer cap-and-trade because it is an invisible tax-cum-subsidy scheme, that is, a quasi-fiscal construction, may be a further reason for preferring taxes and subsidies.



Like so many before him, Mr Akram tries to leverage indignation, feelings of moral outrage and perhaps guilt about poverty and extreme inequality, into an argument for allowing poor developing countries and emerging markets to avoid their share of the burden of implementing an effective global strategy for combating climate change. Should he be successful, the effectiveness of the strategy could be severely impaired.   The notion that poverty gives you a permit to pollute is a nonsense, and insults the poor. Efficient policies to combat global warming require a uniform increase in the opportunity cost of emitting greenhouse gases.  This increase in opportunity cost should be the same for the poor as for the rich.  It can be implemented through a tax on emissions by the rich and through the the withdrawal of a subsidy  for emission reduction by the poor.  The construction of the benchmark relative to which emission reductions are to be measured is deeply problematic, but has to be faced through an open and transparent process.

Cap-and-trade on a global scale, with a single, integrated secondary market for CO2eq emission permits would solve the efficiency problem.  The initial global allocation of permits can address many distributional concerns. 

There ought to be no preferred treatment as regards the need to internalise the externalities from greenhouse gas emissions, for countries like China, which by some measures has already overtaken the US as the world’s leading producer of greenhouse gases and which on any measure will be in that position in a handful of years, or India, which seems determined to catch up with China in the area of environmental vandalism.  The fact that the rich industrial countries have had a couple of centuries to add to the concentration of greenhouse gases in the atmosphere, unencumbered by Kyoto Protocols or Bali Declarations  does not mean that it’s now the developing countries turn to pollute with impunity.  However, there is an obligation on the rich countries (as on the rich living in poor countries) to ensure that the necessary internalisation by developing countries of the global environmental externalities from their production and consumption activities does not place excessive burdens on the poor in these countries.  The tools are there.  Let’s hope the usual excuses for not using them won’t be.

Maverecon: Willem Buiter

Willem Buiter's blog ran until December 2009. This blog is no longer active but it remains open as an archive.

Professor of European Political Economy, London School of Economics and Political Science; former chief economist of the EBRD, former external member of the MPC; adviser to international organisations, governments, central banks and private financial institutions.

Willem Buiter's website