The Allocation of Carbon Permits within One Country: a General Equilibrium Analysis of the United Kingdom

As part of the Kyoto agreement on limiting carbon emissions, from 2008 onwards an international market in auctionable carbon permits will be established. This raises the issue of whether trading should be simply between governments or between companies, or in the latter case how such permits should be allocated. Our paper uses the British section of a CGE model of the European energy sectors to evaluate the economics of various methods of allocating permits within a country, as discussed in Lord Marshall’s recent report to the British government. The option of allocation entirely by auction is similar to the setting of a carbon tax, and the recycling of revenues to reduce or offset other economic distortions could produce a potential net benefit to incomes and employment. 'Grandfathering' some of the permits free to large firms, according to their base year carbon emissions, would mean loss of the benefits of recycling auction revenues. This might be exacerbated if it created windfall profits repatriated by foreign shareholders. The third major alternative is to review the allocation regularly, awarding permits to all firms according to a ‘benchmark’ allocation, based on 'best practice' as estimated by outside experts. This would be similar in practice to recycling the revenue as an output subsidy to the industry, though it could be complicated to implement. Such a system could allow much of the potential ‘double dividend’ to be realised, though it might still be preferable to auction permits, with the revenues used to offset taxes across a wider spread of industry. Introduction. The growth of concern worldwide over the effects of various pollutants, and the high potential costs of control by quantitative means, has lead to increased interest in recent years in economic instruments for controlling certain pollutants. There is a considerable literature on both the theory and practice of such instruments: for example on the development of various trading schemes for sulphur emissions in the USA from the 1970s onwards, under the aegis of the Environmental Protection Agency, and over the potential for international trading of sulphur emission rights in Europe. Another example is the Swedish scheme for controlling NOx emissions (see Smith, 1998). Economic instruments are seen as particularly well-suited to the control of carbon dioxide, as it is a pollutant whose effects (in terms of climate change) are spread globally. Tietenberg (1990) stresses that economic controls are much more straightforward for such uniformly mixed pollutants. The early literature (see the review in Clarke et al, 1996) concentrated on the use of carbon taxes, and carbon or carbon/energy taxes have been introduced, or are under discussion, in many European countries, in the wake of the 2 A sample of US studies of command and control policies for sulphur, nitrogen, particulates, aerosols and noise pollution, surveyed in Tietenberg (1990), showed costs ranging from 1.72 to 22 times the least cost solution. Only one study showed costs 1.07 times the hypothetical least cost level. 3 See Baumol and Oates (1971) for a discussion on the relative merits of priceand quantity-based controls. 4 See, for example, Tietenberg, 1999 or the SEO report 1998 for discussion on the practice of such schemes. 5 See Klaassen’s study (1996). Sulphur trading schemes are more complicated in some ways than carbon permit trading, due to the fact that the location where the sulphur is emitted has significant effect on the costs it imposes. Kyoto agreement in 1997 on limiting greenhouse gas emissions. However, following pressure from the United States, the Kyoto Protocol also includes provisions for countries or firms to trade their agreed quotas in an international market, so following on an international scale the example set by sulphur emissions trading in the U.S.A., and making it possible to equalise marginal abatement costs across countries. For a global pollutant, such as carbon dioxide, a system of auctionable permits works in many ways like a carbon tax, although it is the total volume, rather than the marginal abatement cost, which is being fixed. However, a permit scheme has various advantages, particularly if it allows for international trading. In addition, unlike a carbon tax, a permit system where permits can be saved for future usewhich makes sense given that carbon is a long-lasting global pollutant allows carbon users greater freedom of choice over the intertemporal path of carbon consumption, and facilitates the development of a futures and options market in carbon use (see Cramton and Kerr, 1998). This paper examines various internal economic instruments for a country to control carbon emissions in compliance with its Kyoto targets: these include a carbon tax, a fullyauctioned permit system, or various systems of allocating tradable permits to certain energy users. The paper concentrates on the case of the United Kingdom, and employs numerical simulation using a computable general equilibrium (CGE) model of the U.K. 6 For example the UK government announced in the 1999 budget its intention of introducing a business energy tax from 2000 (see H.M.Customs and Excise, 1999). 7 Though Barrett (1998) points out there are drawbacks to the international trading of quotas. While the analysis is in principle applicable to other countries, the economic outcome will depend on the structure of the economy and pre-existing distortions. Carbon permit allocation in the United Kingdom: the Marshall Report. Lord (formerly Sir Colin) Marshall’s Report (1998) to the UK government on Economic instruments and the business use of energy reviews a number of the main arguments, and recognises that a major argument for a permit scheme is the potential for permits to be traded internationally, which in turn allows extra cutbacks in pollution to be made in those countries which have lowest marginal abatement costs, and then ‘sold’ to countries with higher marginal abatement costs. In theory, this equalises marginal abatement costs and reduces total abatement cost. The Kyoto Protocol allows for such international trading in greenhouse gas emissions to begin in 2008, when the commitment period for meeting emissions targets begins. The report argues (paragraph 46) that trading should be extended to company-to-company trading, on the grounds that ‘firms, not governments, are best able to spot abatement opportunities and best placed to decide whether it is most cost-effective in their interests to reduce their own emissions or buy permits’. It does, however, acknowledge (paragraph 55) that ‘participation in trading will probably never extend to the small business sector’, due to high costs of participation. As a result, there is likely to be a 8 The argument is debatable because firms also have freedom of action under a system where the government trades permits, and then sets a tax to equate total emissions to what the UK is allowed. 9 This conclusion is in line with Tietenberg’s (1990) conclusion that ‘emissions trading is probably not equally applicable to large and small pollution sources’, due to high transactions costs. two-tier system of carbon emission control, with large companies – perhaps those who currently participate in the Integrated Pollution Prevention and Control Directive (IPPC) trading in permits, while smaller companies (those too small to participate in the permit market) and households pay a carbon tax on fuel inputs. There are various possible schemes for allocating permits. Large companies, concerned about their balance sheets, favour a scheme of ‘grandfathering’ emissions permits, to be allocated free to companies in proportion to emissions in a base year, or number of years, and to be resaleable. Such a scheme would bring objections concerning its competitive effects, which give advantages to incumbents in industries against newcomers (and possibly undermining recent reforms of the electricity market), and concerning equity, as it involves handing out ‘windfall gains, in the shape of valuable permits, to failing companies or those who had been slow to act before’. A two-tier system under which large firms would be given trading permits, while small firms would pay a carbon tax, would discriminate against small firms, and favour manufacturing against services. An alternative to straightforward grandfathering is ‘benchmarking’, where firms are allocated permits according to a target for carbon emissions based on a regulator’s judgement of best-practice emissions for that industry (i.e. somewhat akin to the Best Available Technology principle used in the IPPC directive). This would be fairer in terms 10 IPPC is a directive covering large energy users in the UK, under which firms are set emissions limits for various non-carbon pollutants based on ‘Best Available Techniques’. Failure to meet BAT is a prosecutable offence. of ensuring large and small companies paid for carbon use, but more bureaucratic, and would involve and difficult to apply in less homogeneous sectors. The third, and possibly fairest, method of allocation would be a public auction, which would also raise revenues, which could be recycled in a variety of ways. This was the method Marshall preferred, although he believed political pressures from large energy users would result in a hybrid scheme. The economic effects of different methods of permit allocation. This paper aims to compare the effects of different methods of allocation of permits within the UK. For simplicity, it is initially assumed that no net international trading takes place (i.e. the system is simply an internal UK system). Allocation of the permits by means of a regular public auction has broadly similar economic effects to a carbon tax, certainly in the case of a competitive market. First, energy input prices rise, due to the need to purchase permits. For energy-intensive industries, this is the most important effect. In the short run, with capital immobile, the main effect is likely to be to drive down profits in such industries. In the longer run, 11 Tietenberg (1990) mentions that, in practice, it is rare for emissions trading schemes have allocate permits to new entrants. 12 This depends to some extent on the choice of auction method. Cramton and Kerr, 1998, argue that in the case of the USA, no single user, or small group of users, would have sufficient monopolistic power to rig the market, so a simple ascending clock auction would be efficient. The situation in the UK may be rather different, as it is conceivable the two largest electricity generators could exert some monopsonistic power to force down permit prices. however, assuming full mobility of capital, the capital stock, output and employment in energy-intensive industries will be reduced and the output price raised, restoring profits to ‘normal’ (assuming the industry is competitive). For less energy-intensive industries, the secondary general equilibrium effects will be just as important. Depending on how it is recycled, revenue from the sale of permits will compensate consumers for the rise in energy-intensive product prices, and lead to an increase in spending on less energy-intensive products. If the recycling reduces the tax burden on labour, its supply may rise, and real product wages may be reduced. In turn, the exchange rate and terms of trade may be altered by the introduction of the permits. Overall, a permit system is likely shift demand away from manufacturing towards services see the SEO study of the Netherlands (1998) for an example of this. Effects on profitability are assumed to be short-term – in the long-run it is output, capital stock, employment and output prices that will adjust. Farrow (1995) discusses these general equilibrium effects. The most frequently discussed system of grandfathering (e.g. in the SEO report) is to make allocations based upon companies’ emissions in a single year, or set of years, prior to introduction of the system. Allocations would be restricted to the initial group of recipient firms, gradually being reduced in volume in proportion to the national emissions quota. The permits would be resaleable. In a partial equilibrium analysis, with perfect capital markets and profit-maximising firms, the fact that some companies are being given an allocation of free permits rather than having to buy them in an auction would not affect decisions on input or output prices or volumes. The opportunity cost of fuels would be the same as under an auction system, since a company can gain revenue from selling the permits it does not use. However, profits of companies are higher when they receive permits free instead of having to buy them. This is analytically similar to making them buy the permits at an auction (or pay a carbon tax) and reimbursing them in a lump sum for a fixed amount of those permits. In the short run, this would just offset losses from introduction of the permits. In longer run, prices would rise and output fall just as under an auction system to the point where the rate of profit on new capacity, or for a new entrant, who would have to purchase permits, is normal. This would give a windfall, supranormal profit to incumbents who receive the free permit allocation. There are some qualifications to the analysis of grandfathering as a lump-sum transfer. Firms may keep plants open which would have closed under an auction system, if closing them would mean losing allocation. If companies face imperfect capital markets or are not profit-maximising, the better cash-flow position of incumbents under grandfathering could affect behaviour. Perhaps more significantly, there are different general equilibrium effects, since the government no longer gets revenue from sale of permits, and cannot use them, for example, to cut taxes on labour. Higher distributed profits would, of course raise consumer spending, but there is a possibility, if company shares are foreign-owned, 13 For the systems to be analytically equivalent, it is necessary to assume if firms wanted more permits than the fixed allocation, they would be able to buy them at the auction. If their allocation exceeded the amount they would have demanded at the auction price, they could re-sell them at the same price. that some of the windfall from grandfathering would be repatriated to foreign owners, and have negative implications for the exchange rate and terms of trade. Parry et al (1997) estimated, using a CGE model of the USA, that the costs of reducing US emissions by 10 per cent were more than three times higher under a grandfathered permit system than with a carbon tax, due to the inability to recycle revenue. This is also backed by Smith (1998), who discusses the potentially large sums involved in carbon tax revenues, and the fact that grandfathering means loss of the possibility of recycling these in the form of reducing other tax distortions. Cramton and Kerr point to the likely costs of political lobbying and legal challenges to any system which allocates permits free to some firms but not others. There are alternatives to the lump-sum system of grandfathering. Objections from newcomers to the industry about the unfairness of favouring incumbents may lead to periodic reallocation of quotas. But where companies expect that under a rolling reallocation of permits, emitting more carbon now may well lead to getting a larger permit allocation in the future, there would be a counterbalancing incentive to continue emitting. It follows that, to have much effect on carbon emissions, a system with periodic 14 The effects would depend on the extent to which the profits are repatriated, or retained and reinvested in the country issuing the permits, which is not easy to guess a priori. 15 Smith quotes a study by Ballard et al, 1985, estimating that, at the margin, tax receipts in the USA had a deadweight welfare cost of 20-50 per cent. 16 Grandfathering makes more sense in some cases, where those participating in the scheme are the only ones who face economic penalties for polluting. There is a parallel, as Smith (1999) notes, with the case of Swedish controls on Nox emissions, where only large emitters pay a pollution tax, and the revenues are consequently recycled directly to participating firms, in order to maintain a level playing field vis-à-vis firms too small to take part. reallocation would need to be more of the benchmarking variety, where companies only receive allocations according to the assessment of outside experts on the amount a firm in their industry could reasonably be expected to emit. The effects of regularly revised allocations with benchmarking would be quite different to those of grandfathering as discussed above. Although companies would still be penalised for consuming more carbon-intensive fuels than best-practice for their industry and size, a company which expands in output would expect before long to receive a larger quota allocation, and newcomers could also expect to receive quotas. Consequently, a benchmarking system would not have the same disincentive effects on output in carbonintensive industries, with the result that output is higher and prices lower in these industries. If free permits are available to newcomers and there are no other major barriers to entry, the allocation of permits would no longer create windfall profits to incumbents, but would instead be reflected in lower output prices (analytically equivalent to recycling the revenue of permit sales by each industry as an offsetting output subsidy to that industry). This would help lower consumer prices, and could be seen as a form of revenue recycling, though the effects on labour supply are in theory ambiguous. 17 The actual present value of increasing carbon use now in order to obtain a higher permit ration in the next round of allocation depends on the length of time between reallocation, the rate of interest, the factor by which next round’s permit allocation is expected to be reduced compared to current ‘best-practice’ emissions, and the expected rate of increase in permit prices. If the reallocation is frequent, then these factors become less important, and the value is close to the current permit price. 18 Depending on whether the effect of cheaper consumer goods in raising the marginal value of income for work offsets the wealth effect, whereby people prefer more leisure as their real income (including income from non-labour sources) rises. Some papers on ‘grandfathering’ may have had more this kind of scheme in mind. An example is Boehringer, Rutherford and Voss’ paper on unilateral introduction of carbon permits by Germany, which discussed the possibility that free allocation of permits would prevent carbon-intensive industries from moving to countries not introducing the scheme. This assumes the permit allocation is not simply a lump-sum gift to incumbent firms, but affects marginal output decisions. In the following sections, we attempt to model the two methods of grandfathering (with quotas either based on existing emissions or made adjustable according to ‘benchmark’ estimates of needs), and compare these with the simulated effects of an auction of permits.