|
Article Excerpt
1. INTRODUCTION
Around the world, interest is growing in designing and implementing mandatory, domestic, market-based climate change policies. The European Union launched the Emission Trading Scheme (ETS) in 2005 covering roughly half of all carbon dioxide (C[O.sub.2]) emissions in the EU and announced its intent to continue the ETS beyond the Kyoto Protocol's 2008-2012 commitment period (European Commission 2008). The EU has linked and is pursuing linking its trading regime with other domestic cap-and-trade programs, including those in Iceland, Norway, and New Zealand. With the election of a new government in late 2007, Australia is moving forward with plans for a domestic cap-and-trade program. In the United States, with twelve proposals for mandatory climate regulation in the 109th Congress and approximately that many again in the 110th Congress, momentum continues to build for federal action (Table 1). Trends at the state and regional level, including California, New England and the mid-Atlantic states, reinforce this effort through their own calls for mandatory policies. (1) Several governments have pursued carbon taxes, including Denmark, Sweden, Finland, Norway, and the province of British Columbia.
The design of domestic climate change policy has important environmental, energy, economic, and fiscal implications. Mitigating greenhouse gas (GHG) emissions is a critical element in addressing what is widely believed to be the most pressing environmental problem of the 21st century. Over the long term, climate change policy may radically alter how fossil fuels power industrialized economies. Climate change policy will affect more firms and households and impose greater costs and greater benefits than any environmental policy to date. The costs of domestic GHG mitigation policy--perhaps 1-2 percent of national income--may be roughly comparable to all other environmental policies combined. (2) Finally, market-based approaches to climate change provide the opportunity to generate government revenues of the magnitude comparable to other large streams of revenues, such as the corporate income tax. (3)
With the potential for such far-reaching effects, it is important to consider several key questions to frame the evaluation of various domestic cap-and-trade and emission tax proposals. Will these proposals promote efficiency by addressing climate change in a manner sensitive to costs and benefits? Will these proposals employ cost-effective implementation so that they achieve their stated emission reduction goals as inexpensively as possible? How will these proposals affect the distribution of benefits and costs across the U.S. economy?
The first question--will these proposals promote efficiency by balancing costs and benefits--is the most difficult for economic analysis. The significant uncertainty and long time horizons associated with mitigation benefits challenge underlying assumptions in conventional economic analysis. For example, a standard and sensible condition is that consequences further and further into the future, and/or with smaller and smaller probabilities, should not dominate our analysis. Otherwise, we find ourselves trying to model and forecast events sufficiently rare and/or distant that conventional tools are rendered useless (Weitzman 2007).
Even keeping the standard assumptions, putting the pieces together for a benefit-cost analysis is daunting. Nordhaus (2007) finds that an optimal global emissions pathway would result in a doubling of C[O.sub.2] concentrations. Supporting this conclusion are assumptions about various climate impacts, their valuation in multiple regions around the world, and the choice of discount rate to convert these estimates into present value terms. Yet, having employed benefit-cost analysis for the management of a global public good, we are still left with several more challenges in evaluating national policies. First, a global benefit-cost analysis does not provide guideposts on how to divide costs among countries; for example, will developing countries pay according to the same rules as industrialized countries? Second, national policies may involve a more provincial attitude to benefits, where those benefits accruing to other countries are not counted the same; this requires additional information than the global analysis. Finally, regardless of how any single country chooses to answer the first two challenges in developing a national policy, they must confront the fact that every other country will be doing the same thing--creating the potential for strategic behavior by some countries to free-ride on others' actions. (4)
Given the difficulties of applying economic analysis to the first question of balancing benefits and costs, we focus our attention primarily on the latter questions of cost effectiveness and distribution. We have identified the following six design issues to inform the consideration of these questions: (1) program coverage and scope; (2) cost containment; (3) use of offsets; (4) revenues and allowance allocation; (5) mechanisms to address competitiveness concerns; and (6) complementary R&D and technology policies. This paper synthesizes the literature on each of these design issues and highlights the implications for building a robust, efficient climate policy that can appropriately address distributional issues identified by policy-makers. We draw on the suite of proposals in the 110th session of the U.S. Congress to emphasize the practical nature and range of these choices (summarized briefly in Figure 1 and Table 1). Where relevant, we assess the need for additional analysis and research to better inform policy-making. We conclude by emphasizing the key messages that emerge from the synthesis in light of these design issues.
2. PROGRAM COVERAGE AND SCOPE
Greenhouse gas emissions occur as a by-product of virtually every form of economic activity. More than 80 percent of U.S. emissions arise from fossil fuel combustion (coal, oil, and natural gas) from an extremely wide range of sources: large power plants and industrial facilities; homes, businesses, and commercial buildings; agriculture and other small businesses; and automobiles and other modes of transportation. The remaining sources include fugitive emissions of nitrous oxide and methane from agriculture, and industrial releases of fluorinated gases and nitrous oxide (U.S. EPA 2008). Given this remarkable breadth of the cause of climate change, a cost-effective and efficacious emission mitigation policy should exploit emission abatement opportunities among as many of these sources as possible.
[FIGURE 1 OMITTED]
The economic literature has generally supported as broad a single-price policy as possible. This follows from application of Samuelson's (1954) basic result that a public good--or bad, such as GHG emissions--should be priced at its marginal social benefit. Numerous studies have empirically considered how non-price policies lead to much higher costs (Tietenberg 1985). A ton of C[O.sub.2] makes the same contribution to climate change regardless of the location of emissions in the world. For emissions of other GHGs, they will generally have different radiative forcings and different atmospheric lifetimes; however, their global warming potential (GWP) can be converted to "equivalent" C[O.sub.2] units (IPCC 2001).
The issue of non-C[O.sub.2] gases is not without controversy, however, as the GWPs are sensitive to assumptions about damages, discounting, and time horizon (Schmalensee 1993). For example, methane has an extremely high GWP according to the IPCC--23 times C[O.sub.2] by weight--but also has a very short half life. This raises the question: Are we comfortable trading off one ton of methane against 23 tons of C[O.sub.2], given the methane would have been scavenged from the atmosphere and have no discernible climatic effect several decades from now, presumably when we are really beginning to care about impacts? Meanwhile, the 23 tons of C[O.sub.2] would have decayed very little. Despite this question, most discussions of climate change economics and the design of policy ignore this issue and take the GWPs as an adequate measure of marginal benefit trade-offs--perhaps as an undesirable but necessary simplification. A notable exception is the report issued as part of the U.S. Climate Change Science Program where the relative price of gases changes in response to compliance with a target for radiative forcing (Clarke et al. 2007). Not surprisingly, methane comes up with a very low price in early years.
Returning to the issue of the 80 percent of emissions comprised of fossil fuel-related C[O.sub.2], the debate quickly turns to one of where to regulate. Traditional market-based regulation--the U.S. sulfur dioxide (S[O.sub.2]) and nitrogen oxides programs, the EU ETS, and most recently RGGI and the Alberta emission reduction regime--have focused on large point sources. (5) Such sources have reasonably low monitoring costs and--from a political perspective--are often easier to target for regulation. For traditional pollutants, the focus on smoke-stack emitters reflects the technological opportunities to pursue mitigation efforts through end-of-pipe treatment.
Carbon dioxide is unlike most pollutants because there are no end-of-pipe control technologies--it is the primary product of breaking down hydrocarbon chains. When a fossil fuel is mined, extracted, or imported, we can be relatively confident of the eventual C[O.sub.2] emissions--excepting efforts to sequester the fuel into products (plastics), exportation of fuels before combustion, or the potential for large emissions sources to capture and store C[O.sub.2] underground. (6) With the dispersed nature of mobile source and residential emissions, the idea of regulating C[O.sub.2] at or near the point of fossil fuel production has received substantial attention (Keeler 2002). Such regulation would have modest monitoring costs and would have to cover only about 2,000 to 3,000 facilities in order to control all fossil fuel C[O.sub.2] emissions (Stavins 2007; Hall 2007). More recently, many climate change proposals in the Senate have moved in this direction (see Table 1, where all economy-wide bills are at least partially upstream).
Despite the potential practicality of broad coverage through upstream regulation at or near the point of production, there have been a number of primarily anecdotal concerns (Pizer 2007). First, some have advanced the concern that producers cannot pass on the cost of allowances or taxes to consumers. In some cases, this reflects existing institutional constraints. For example, natural gas pipeline tariffs may be regulated in ways that make it difficult for pipeline companies to pass on costs. Market power by the railroads may make it difficult for coal mines to pass through costs to coal users. Despite these concerns, however, most seem surmountable. Second, some have voiced the concern that firms only change their behavior in response to direct regulation, and will not adjust in response to changes in the prices of carbon-intensive inputs such as fossil fuels. (7) While there may be some "awareness effect" of forcing end users to think about their fossil fuel use in not just cost but also pollution consequences, the magnitude of this effect would seem to have some practical limits. More importantly, a substantial empirical literature has characterized induced technological change in the U.S. economy, showing that higher input prices induce firms to invest in factor-conserving technology. For example, industrial firms invested in energy-conserving technologies in response to the energy prices shocks of the 1970s and early 1980s. Finally, there is often pressure from some sectors that believe their situation warrants special consideration--competition from abroad, vulnerability to price volatility, or security.
The renewed interest in broader, upstream regulation partly reflects concerns about the inefficiency of narrow policy coverage and the difficulty reaching more aggressive targets. First, excluding some sources from regulation reduces the set of low-cost abatement opportunities to be exploited. Second, incomplete coverage of the economy's sources may spur emission leakage. The regulation of large sources may drive economic activity towards smaller sources, e.g., home use of natural gas and heating oil, instead of electricity. It may also cause unregulated sources to generate power on-site instead of purchasing power from regulated firms. Third, U.S. EPA (2007) shows that under one proposal uncovered sources constitute 20-25 percent of reference case emissions, but they eventually comprise almost half of emissions after imposing regulations on covered sources. Yet, the idea that there would be an eventual broadening of coverage--starting with a few sectors with the aim of expanding to cover other sectors over time--may produce strong, concentrated special interests opposed to such expansion, making it increasingly difficult to further reduce emissions if such sources are not included from the start.
Exactly what is the economic cost of narrow versus broad coverage? There has been considerable numerical analysis focused on this question. Until recently, the vast majority of studies examined the cost savings from broad coverage among countries rather than within countries. In perhaps the most influential study of the former sort, Weyant and Hill (1999) summarized the results of Energy Modeling Forum 16, where modeling teams compared the cost of implementing the Kyoto Protocol targets unilaterally, with Annex I trading, and with developing country participation. They found that costs were cut by more than half going from autarky to Annex I trading, and at least half again when including developing country participation. Considering a more relevant dimension for the domestic regulatory discussion, EMF-21 considered the effect of including non-C[O.sub.2] gases in mitigation policies, and again found costs were roughly cut in half with such broader scope of coverage (Weyant et al. 2006). (8) This suggests that even though these gases are often a relatively small part of emissions, they are a disproportionately important part of low-cost emission abatement. The same result has been observed in studies of U.S. mitigation costs (U.S. EIA 2005).
Only recently has attention shifted to quantifying the effect of partial coverage of C[O.sub.2] emissions within a given country. Pizer et al. (2006) consider both the question of coverage and inefficient policies, examining the consequences of excluding different sources from a cap-and-trade program as well as using policies such as fuel economy standards and renewable portfolio standards. They found that limiting the policy to the power and transport sectors--excluding the industrial sector--doubled costs. Using inefficient policies, however, raised costs by a factor of ten. (9) Excluding relatively small sectors--direct emissions from residential and commercial buildings--had a negligible effect. This is consistent with EIA (2007) which similarly found a very small effect of excluding the commercial building sector in a domestic policy. The contrast could be quite stark between legislative proposals for effectively economy-wide caps (e.g., Lieberman-Warner) and proposals for utility-sector-only caps...
|
