تخصیص مبتنی بر خروجی و بازیافت درآمد : مفاهیم برای طرح انتشار تجاری نیوزیلند

The New Zealand Emissions Trading Scheme (NZ ETS) is more comprehensive in its coverage of emissions than schemes introduced or proposed to date in any other country in that it includes agricultural greenhouse gases, which account for half of New Zealand’s total emissions. But, motivated by concerns for the international competitiveness of emissions-intensive, trade-exposed industrial and agricultural activities, current legislation provides for substantial ongoing free allocations to such activities, linked to their output. Here we use a computable general equilibrium model to analyse the impacts of output-based allocation, given the possibility of recycling net revenues to reduce prior distorting taxes. Unlike previous modelling studies of alternative NZ ETS designs, we allow for a more realistic modelling both of capital and labour supply. We find that, as suggested by theoretical results, interactions between the ETS and existing taxes are important. Given any level of output-based allocation, the negative macroeconomic impacts can be reduced by recycling net revenues as efficiently as possible. Less obviously, we find that there may be an optimal non-zero level of output-based allocation. This optimal level increases as the carbon price and/or factor supply elasticities increase, but decreases if revenues are recycled with greater efficiency.

Market-based instruments such as taxes or cap-and-trade schemes are widely seen as an efficient means of reducing global greenhouse gas (GHG) emissions. International emissions trading under the flexibility mechanisms of the Kyoto Protocol to the United Nations Framework Convention on Climate Change (UNFCCC) allows less costly emissions reductions than if each Annex I party met its obligations through domestic actions alone. Firm-level emissions trading schemes (ETS) and/or emissions taxes have been or are likely soon to be implemented in many developed countries. An issue of major political and economic significance is how permits are allocated in a domestic ETS. There are important trade-offs between objectives of global environmental effectiveness, promoting static and dynamic economic efficiency and equity. In September 2008, the New Zealand Parliament passed the Climate Change Response (Emissions Trading) Amendment Act (henceforth, 2008 Act).1 The 2008 Act provided for a comprehensive ETS, ultimately to cover all sectors (including agriculture and forestry) and all those GHGs covered under the Kyoto Protocol. Coverage of both agricultural emissions and emissions and removals from the forest sector was motivated by their large contributions (positive and negative respectively) to New Zealand’s emissions inventory.2 In November 2008, a new government instigated a review of the 2008 Act, citing “the weak state of the economy, the need to safeguard New Zealand’s international competitiveness, the position of trade-exposed industries and the actions of competing countries” (New Zealand Parliament, 2008). The legislation was subsequently amended in June 2009 (henceforth, 2009 Act3). The amendments included changes to short-term transitional measures, such as delaying the entry of agriculture into the NZ ETS from 2013 until 2015. More importantly, though they provided substantially greater protection to emissions-intensive, trade-exposed activities over a much longer period, it was also announced that the amended NZ ETS would be ‘fiscally neutral’, whereas the previous scheme was expected to raise significant revenue in the medium term (Smith, 2009). Under the 2008 Act, total free allocations to eligible agricultural and industrial firms would begin at 90% of 2005 emissions and be phased out linearly by 2029. Allocations to firms were not related directly to their output, although some provisions of the Act would have incentivised increased output (e.g. s81a made allocations conditional on continued operation while s81b provided for free allocations to new entrants). The 2009 Act increased the phase-out period for free allocations to 75 years. It also switched to an ‘intensity-based scheme’ for ‘emissions-intensive, trade-exposed activities’. Permits are allocated to firms carrying out individual activities in proportion to their output and to a benchmark emissions-intensity defined for each type of activity. That the NZ ETS might disadvantage New Zealand producers relative to their international competitors is a legitimate concern. Shifting of the existing or (more likely) new production to other countries with less stringent policies might have both economic and social costs for New Zealand, while reducing environmental effectiveness through emissions leakage.4 However, many modelling studies (e.g. Graichen et al., 2009) and recent empirical evidence from the EU ETS suggest that such concerns are often exaggerated. The levels and forms of free allocation in the first phases of the EU ETS have been widely criticised as excessive, leading to billions of euros of windfall profits for electricity generators (Point Carbon, 2008) and increase of emissions-intensive EU exports. In the context of the NZ ETS, Kerr and Zhang (2009) find that large changes in land use or intensity are unlikely to result at a price of NZ$25/tCO2e. They argue that international emissions leakage from New Zealand agriculture may be relatively limited, and that preventing loss of output through free allocations may be excessively costly, especially if other negative environmental impacts of agriculture in New Zealand (e.g. on water quality) are accounted for. In theory, the best way to address these concerns would be to impose comprehensive border adjustments, imposing an emissions cost on imports and removing the emissions cost from exports. However, calculating the required adjustment rates for specific goods (especially imports) would be difficult and costly, and their design would be severely constrained by the international trade law (e.g. van Asselt and Biermann, 2007). In this context, output-based allocations may be a more effective means of reducing competitive disadvantage and associated emissions leakage (Fischer and Fox, 2007 and Fischer and Fox, 2009). Greenhalgh et al. (2007) consider these two options in the specific context of the NZ ETS. Granting free allocations means forgoing revenue from selling permits, by auction or other means. If net revenues were generated by the scheme, these could be ‘recycled’ to reduce distorting taxes, leading to a ‘double dividend’. However, interaction of the ETS with pre-existing distortionary taxes will increase its economic costs (Bovenberg and de Mooij, 1994). The double dividend is qualified as ‘strong’ if the former effect outweighs the latter, or ‘weak’ if it does not. Bovenberg (1999) provides a review of the relevant theoretical literature. Both theoretical results and empirical studies of emissions taxes or trading schemes suggest that weak rather than strong double dividends should be expected. Strong double dividends may nevertheless be obtained if particularly distorting taxes are reduced. For example, in a study of carbon pricing in Switzerland, Felder and Van Nieuwkoop (1996) find strong double dividends when high marginal tax rates are reduced. We assume that ‘fiscal neutrality’ of the NZ ETS means that there will be no revenues to recycle. However, it is unclear from official documents (e.g. Emissions Trading Scheme Review Committee, 2009) if or how the government has determined that this will actually be the case. Moreover, forgoing potentially significant revenues from the NZ ETS potentially conflicts with the government’s stated interest in pursuing substantial tax reforms to increase economic productivity and growth (Dunne, 2009). While there have been four general equilibrium modelling studies published since 2007, only NZIER (2008) considers the current situation in which substantial free allocations will be maintained in the long term (2025). In studies conducted for the government’s Emissions Trading Group, Infometrics, 2007 and Infometrics, 2008 models output-based allocation only in the shorter term (2011/12) while most recently, NZIER and Infometrics (2009) assess the impacts of different levels of obligation and international emissions trading under full auctioning. These previous studies of the NZ ETS give little consideration either to the scheme’s interaction with pre-existing distortionary taxes or to the gains or losses associated with more or less efficient use of net auction revenues. The scope of gains from revenue recycling is in fact very limited in the Infometrics model, because it assumes fixed long-run supply of both capital and labour. In the NZIER model, long-run labour supply is fixed, but long-run capital supply adjusts to meet an exogenously fixed rate of return. However, the NZIER studies do not consider preferential lowering of taxes on capital income. In this paper, we examine the effects on the New Zealand economy of different rates of output-based allocation to emissions-intensive trade-exposed industries within a comprehensive ETS open to international emissions trading. We account for the fact that net revenue from auctioning permits may be recycled more or less efficiently or, if free allocation is excessive, existing distorting taxes will have to be increased. These policies reflect the broad emissions coverage (ca. 2020) of the NZ ETS and (somewhat more loosely5) its provision for free allocations linked to output. Our aim is not to provide a specific assessment of the NZ ETS or specific design changes, but to highlight in more general terms important trade-offs that merit more extensive analysis and policy consideration. We assess the macroeconomic impacts of the modelled policies using a large-scale CGE model: the New Zealand Climate Economics Model (NZCEM). NZCEM is comparable in many respects to the aforementioned CGE models, but includes a labour–leisure choice of households, unemployment and a long-run elastic supply of capital (via changes in the level of foreign investment). It also includes a more detailed treatment of land-based production. We focus on effects of the policies in 2020. This has become a milestone year in the context of international climate negotiations and, in that context, represents the ‘mid-term’: substantial international progress may be made, but a high degree of international harmonisation of mitigation policies is unlikely. In the context of domestic adjustment to the introduction of the NZ ETS, we may generally consider it the ‘long term’.

In this study, we have shown how different levels of output- based allocation and different methods of recycling any net revenues can significantly alter the macroeconomic impacts of an ETS in New Zealand, which is open to international permit trade.Our model differs from those used in previous studies of the New Zealand ETS in that it allows not only for the long-run elastic supply of capital but also for a labour–leisure choice of house- holds, making long-run aggregate labour supply responsive to changes in the real wage. Moreover, by considering the rate of free allocation as a continuous policy variable, we show more clearly the effects of output-based allocation to emissions-intensive trade-exposed industries and the interactions with existing taxes. Our results suggest that at moderate to high carbon prices, revenues from pre-existing taxes may be significantly eroded. Consequently, a significant number of permits backed by the government’s initial endowment of AAUs must be auctioned to avoid negative fiscal impacts. If additional permits are auctioned rather than freely allocated, it is important that the resulting net revenues are recycled as efficiently as possible – namely to reduce the most distorting taxes – if macroeconomic impacts are to be minimised. With a fixed carbon price, output-based allocation is very effective in increasing aggregate output and hence demands for labour and capital. However, it also reduces domestic mitigation and thus increases the cost of purchasing emissions permits abroad to meet a given emissions reduction obligation. Thisnegatively affects aggregate income and although reducing erosion of tax revenues, reduces net revenues available for recycling. Recycling revenue to reduce distorting income tax wedges increases factor supply and consequently output and factor incomes. It has negligible effects on domestic emissions. The positive effect on factor incomes reduces revenue erosion, creasing a positive feedback effect, i.e. the marginal efficiency of revenue recycling increases with the level of revenue recycling. Considering a wide range of factor supply elasticities, we find that output-based allocation tends to be most effective in preserving aggregate output. However, impacts on aggregate income are generally minimised by low or zero rates of output- based allocation. Low rather than zero rates of output-based allocation yield smaller impacts on aggregate income if factor supplies are relatively elastic. This may be explained by the importance of negative interactions between carbon prices and pre-existing taxes. Such optimal rates of free allocation are higher if revenue is recycled less efficiently. However, we find no case in which it is beneficial to raise income taxes to further increase the rate of output-based allocation. For a given carbon price, changing the QELRO has nearly linear impacts on output and income. However, the method of revenuerecycling significantly alters the marginal benefit to New Zealand of relaxing its obligation. It is approximately three times as great if net revenues are recycled to capital rather than being returned lump sum. This result is largely independent of the rate of output-based allocation. Marginal impacts on GDP, labour supply and the capital stock are actually slightly less than zero with lump sum recycling. We have not estimated net impacts of the policies on global emissions. This would require accounting for at least (1) leakage of emissions associated with lost output in New Zealand and (2) leakage of emissions associated with foreign emissions permits used to meet the QELRO. Other things being equal, if the latter effect is larger than the former, preserving domestic output will come at the cost of environmental effectiveness. Consideration of the economic benefits of output-based allocation as described above should therefore be tempered by the possibility thatoutput-based allocations may in some cases reduce environmen- tal effectiveness. Our results are consistent with the standard economic theory on optimal taxation and the more specific theoretical results concerning double dividends and environmental taxation in a second-best setting. However, in the protracted and often heated public and political debates over the introduction of emissions trading schemes in New Zealand, these considerations have been peculiarly absent. This is all the more surprising, given that the New Zealand government is conducting a wide-ranging review of taxation with a view to significant reforms that would boost economic productivity and growth. A similar situation can be observed in Australia. From this perspective, our modelling can only be considered a preliminary indication of what is possible. More detailed modelling of labour supply, taxes and impacts ondifferent categories of households is needed. Such modelling should, however, have regard to general equilibrium effects, such as that relating to labour productivity noted above. The policies modelled here do not correspond specifically to the current NZ ETS. Limitations of the model have precluded our modelling the response of the forestry sector under the NZ ETS. Furthermore, we have modelled a rather generous eligibility for free allocation, while eligibility rules in the NZ ETS have still to be determined and seem likely to be more restrictive. Nevertheless, eligibility will have to be very much narrower than modelled here if the high rates of intensity-based allocation under the NZ ETS (90% of 2005 emissions phased out at 1.3% p.a.) are to be at all consistent with our findings on the levels of output-based allocation that are optimal from a macroeconomic perspective.