محدودیت های اقتصاد سیاسی در سیاست های قیمت گذاری کربن: مفاهیم برای بهره وری اقتصادی، اثر زیست محیطی و طراحی سیاست آب و هوایی چه هستند؟

Economists traditionally view a Pigouvian fee on carbon dioxide and other greenhouse gas emissions, either via carbon taxes or emissions caps and permit trading (“cap-and-trade”), as the economically optimal or “first-best” policy to address climate change-related externalities. Yet several political economy factors can severely constrain the implementation of these carbon pricing policies, including opposition of industrial sectors with a concentration of assets that would lose considerable value under such policies; the collective action nature of climate mitigation efforts; principal agent failures; and a low willingness-to-pay for climate mitigation by citizens. Real-world implementations of carbon pricing policies can thus fall short of the economically optimal outcomes envisioned in theory. Consistent with the general theory of the second-best, the presence of binding political economy constraints opens a significant “opportunity space” for the design of creative climate policy instruments with superior political feasibility, economic efficiency, and environmental efficacy relative to the constrained implementation of carbon pricing policies. This paper presents theoretical political economy frameworks relevant to climate policy design and provides corroborating evidence from the United States context. It concludes with a series of implications for climate policy making and argues for the creative pursuit of a mix of second-best policy instruments.

Climate change presents a pressing and wide reaching environmental risk requiring a proactive response. Yet policy and market responses to address the external damages associated with greenhouse gas (GHG) emissions have been piecemeal and insufficient to date. As such, the carbon intensity of the global energy supply has remained essentially static since 1990, despite increased policy action intended to address climate change (IEA, 2013). Economists have traditionally conceptualized climate change as a conventional environmental externality, albeit one with complex causes and damages that are particularly dispersed by timing, geography, and magnitude (Nordhaus, 1992, Stavins, 1997 and Stern, 2007). As such, the traditional economic prescription for climate externalities involves establishing a Pigouvian fee (Pigou, 1932) on the sources of GHG emissions that corrects for the un-priced externality, either via a tax on carbon dioxide (CO2) and other GHGs (a “carbon tax”) (Metcalf and Weisbach, 2009) or via a market-based emissions cap and permit trading mechanism (“cap-and-trade”) (Stavins, 2008).1 If these instruments successfully establish a carbon price equal to the full climate change-related external costs associated with emissions of CO2 and other GHGs (the so-called “social cost of carbon”), they will equalize the marginal social and private costs of GHG emitting activities, restoring a Pareto optimal level of emissions (see Online Supplement 1 for more). Considerable debate has been devoted to the relative advantages and disadvantages of carbon taxes versus cap-and-trade given the particular nature of climate-related externalities. Yet both market-based policy instruments rest upon a common economic foundation and in practice can be designed to yield equivalent results (Aldy et al., 2010). This paper thus refers to carbon taxes and cap-and-trade instruments collectively as “carbon pricing policies.” The economic literature on climate policy and instrument choice is substantial. The bulk of this literature has assessed single instruments or compared two or more instruments against one another, including carbon pricing policies (carbon taxes, cap-and-trade programs, or hybrid approaches), traditional “command-and-control” regulations, and production quotas and subsidies for low-carbon energy sources (Bennear and Stavins, 2007). These analyses regularly find carbon pricing policies to be the most economically efficient or “first-best” response to climate-related externalities or the most cost-effective way to accomplish a given emissions mitigation objective (Aldy and Stavins, 2012, Metcalf and Weisbach, 2009, Nordhaus, 1992, Nordhaus, 1994, 2008, Stavins, 1997 and Stavins, 2008). Most economists therefore typically favor carbon pricing policies over other prescriptions and generally argue against a mix of overlapping policy instruments, such as carbon pricing alongside subsidies or mandates for renewable energy sources or sector-specific emissions regulations (e.g., Fankhauser et al., 2010). See Lehmann and Gawel (2013) for a summary of economic critiques of overlapping climate policy instruments. Despite the substantial volume of economic literature arguing for carbon pricing as the “first-best” or optimal response to climate externalities, policy makers have in practice routinely implement a mix of multiple, overlapping instruments. This may include carbon pricing instruments, other energy or output taxes, subsidies, command-and-control regulations, as well as a variety of voluntary programs or information measures (Bennear and Stavins, 2007, Lehmann, 2012 and Sorrell and Sijm, 2003). For example, EU member states are subject to the EU Emissions Trading Scheme, a major carbon cap and permit trading program, as well as national targets for renewable energy adoption and energy efficiency, which have been implemented via a variety of domestic support policies (i.e., feed-in tariffs and other subsidies, standards, and production quotas). Furthermore, most EU member states have implemented additional carbon or energy taxes in a variety of sectors. Similarly, policy makers in California have responded to statewide GHG emissions reduction targets established by Assembly Bill 32 by establishing a portfolio of frequently overlapping policies, including renewable energy support schemes, renewable transportation fuels standards, energy efficiency incentives, an emissions portfolio standard for new power plants, and a cap-and-trade program for major emitters. The real-world prevalence of multiple policy instruments is by no means limited to climate policy and is in fact the norm in environmental and natural resource management domains (Bennear and Stavins, 2007). While the use of multiple, overlapping environmental policy instruments often seems to economists to be an unfortunate and inefficient departure from economic principles, a growing body of research has explored a variety of potential rationales for a mix of climate policies (Bennear and Stavins, 2007, Fischer and Newell, 2008, Fischer, 2008, Jaffe et al., 2005, Lehmann and Gawel, 2013, Lehmann, 2012, Sorrell and Sijm, 2003 and Twomey, 2012). In general, this research justifies the use of multiple policy instruments in a “second-best” world in which one or more constraints within the general equilibrium system prevent attainment of the Pareto optimal conditions (Lipsey and Lancaster, 1956 and Bennear and Stavins, 2007). In the context of climate policy, this body of second-best theory implies that addressing climate-related externalities with a carbon pricing instrument alone may be suboptimal in the presence of one or more binding constraints. These constraints may include additional market failures, policy failures, institutional capacity limitations, prohibitive transaction costs, or political economy constraints (Bennear and Stavins, 2007 and Lehmann, 2012). To date, the literature extending second-best theory to climate policy design has focused predominately on the presence of multiple market and private coordination failures that may necessitate additional policy instruments alongside Pigouvian carbon pricing policies (Lehmann, 2012). For example, knowledge spillovers inhibiting low-carbon technology innovation may justify additional R&D subsidies or early deployment policies to induce learning-by-doing (Fischer and Newell, 2008, Fischer, 2008, Jaffe et al., 2005, Lehmann, 2013 and Nemet, 2013) while information asymmetries and principle agent failures may necessitate additional energy efficiency standards or labeling measures (Bennear and Stavins, 2007 and Jaffe et al., 2005). Other research has focused on institutional capacity limitations or transaction costs that may prohibit efficient implementation of first-best policy instruments (Bennear and Stavins, 2007 and Lehmann, 2012). While these constraints are critical factors in climate policy design, comparatively little research has focused on the presence of powerful political economy constraints that routinely impact efforts to implement carbon pricing policies.2 Social welfare can be maximized under an efficiently implemented carbon tax or cap-and-trade system and government revenues may theoretically be recycled in a manner that maximizes overall welfare (Goulder, 1998). Nevertheless, the imposition of a carbon price causes consumers and producers alike to experience both a private welfare loss and a transfer of surplus to government tax revenues (see Online Supplement 1). By design, pricing carbon will increase factor prices for carbon-intensive energy products and other intermediate and end-use products that involve GHG emissions during production or distribution. This increase in factor prices will cause a redistribution of economic resources as production and consumption shift to a new, less carbon-intensive equilibrium. While general equilibrium analysis traditionally assumes the transition from one market equilibrium to another is costless or “frictionless,” such transitions in reality can impose substantial private costs, even if social welfare is maximized in the end. Several industrial sectors possess a high concentration of assets that would lose considerable value under carbon pricing policies. These sectors are thus likely to mount vociferous opposition to such policies (Murphy, 2002). Capture of the regulatory process by industrial interests (Stigler, 1971) may also result in suboptimal instrument choice or design (Keohane et al., 1998). The diffuse nature of the climate externality also presents a classic collective action challenge (Olson, 1984) and principal agent failure (Eisenhardt, 1989), resulting in a relatively low willingness-to-pay (WTP) for climate mitigation by citizens (Johnson and Nemet, 2010). While several papers exploring second-best climate policy design have briefly discussed the importance of these political economy constraints (Bennear and Stavins, 2007, Lehmann and Gawel, 2013 and Lehmann, 2012), few have provided either a formalized theoretical explanation or real-world evidence of the binding nature of these constraints. This paper seeks to advance the literature on second-best climate policy design by providing several theoretical frameworks drawn from the political economy literature that are useful in predicting the expected behavior of both producers (i.e., commercial and industrial interests) and consumers (i.e., individual households or citizens) in response to proposed carbon pricing policies. In addition, evidence from the United States context is presented which corroborates these theoretical predictions. The presence of multiple constraints on climate policy design has important implications for policy making. In particular, while conventional carbon pricing prescriptions are designed to achieve the Pareto optimal equilibrium between present-value emissions abatement costs and benefits, political economy constraints can bind the climate policy making process long before this theoretical outcome is reached. Consistent with second-best theory (Lipsey and Lancaster, 1956), the presence of these binding political constraints therefore opens up a significant space within which alternative climate policy designs may achieve a second-best optimum with superior economic efficiency, environmental efficacy, and political feasibility relative to the politically constrained implementation of carbon pricing policies alone. This paper calls this the “opportunity space” for improvement in climate policy design and urges renewed creativity amongst climate policy makers, researchers, and advocates to explore this opportunity space in pursuit of policy formulations capable of improving economic and environmental outcomes. The remainder of this paper proceeds as follows: Section 2 briefly describes the materials and methods presented in this paper. Section 3 presents both theoretical political economy frameworks relevant to climate policy design and provides corroborating evidence drawn from the United States context. Section 4 discusses how political economy constraints can bind the implementation of a carbon price well below the range of estimates of the social cost of carbon and discusses implications for the economic efficiency and environmental efficacy of conventional carbon pricing policies. This section also introduces the “opportunity space” for improvement in climate policy design. Section 5 concludes with a discussion of implications for climate policy making.

How are policy makers to capitalize on this opportunity space for improvement and pursue creative climate policy instruments with superior political feasibility, economic efficiency, and environmental efficacy relative to the constrained implementation of carbon pricing policies? Careful attention to the multiple constraints on climate policy introduced in Section 3 yields several preliminary implications for policy design. First, policies should be compared to an “economically optimal” Pigouvian carbon pricing instrument to evaluate their performance relative to the first design constraint (that the optimal policy be welfare-improving). However, policies that meet this economic efficiency constraint without violating relevant political economy constraints should not be dismissed simply because they are not “economically optimal”; neither, in point of fact, are the politically constrained implementations of Pigouvian carbon pricing instruments most prevalent in the real world. In practice, numerous policy measures may exhibit superior economic efficiency and environmental efficacy relative to both the absence of policy intervention and the implementation of a constrained carbon price. Assessing the efficiency and effectiveness of alternate policies that exist in the opportunity space for improvement is therefore a critical research need for the climate policy and economics community. Second, there is some evidence that the choice of policy mechanism itself can affect consumer WTP. Consumers do not perform detailed cost-benefit calculations of policy measures and may in fact perceive the private costs and benefits of various policies differently. For example, Karplus (2011) discusses the way consumers respond to various policy mechanisms based on the visibility and distribution of compliance costs, as well as other factors. For example, a dollar increase in the price of a gallon of gasoline is much more conspicuous than a modest increase in the purchase price of a new vehicle at the dealership. Consumers pay for gas on a frequent basis and are well attuned to the fluctuations of prices at the pump, while they only infrequently purchase new vehicles and capital costs may be amortized over monthly payments. This difference in visibility helps explain the political durability of U.S. Corporate Average Fuel Economy (CAFE) standards for vehicle fuel efficiency relative to fuel taxes designed to provide market incentives to induce fuel efficiency. While CAFE standards may impose a higher net cost on consumers than a fuel tax (Knittel and Sandler, 2011), consumers greatly favor CAFE regulations over higher fuel taxes (Karplus, 2011). Careful attention to the way in which consumers perceive the costs of different policy measures is therefore essential, and instrument choice itself may either relax or tighten political economy constraints. Third, careful attention to industrial structure and related political economy dynamics will be critical to evaluate the likely reactions of producers to various climate policy proposals. For example: minimizing energy cost increases (by subsidizing low-carbon energy adoption rather than penalizing CO2-intensive fuels, for example) could neutralize opposition from energy-intensive manufacturers who do not directly emit CO2 during production; opportunities to expand markets for lower-carbon natural gas could win the gas sector to the side of policy action, undermining collective action within the oil and gas industry associations; or providing technology or transition assistance (and thus reducing asset specificity or providing compensation for lost asset value) to specific sectors may neutralize or weaken industry opposition. In contrast, ignoring these dynamics could easily undermine the political feasibility of any climate policy proposal. Fourth, pursuing mitigation strategies that link long-term avoided climate damages with near-term co-benefits (Nemet et al., 2010, Smith and Haigler, 2008, Bollen et al., 2009 and Knittel and Sandler, 2011) salient to consumers and citizens could help reduce the temporal and geographic mismatch between mitigation costs and benefits, potentially increasing the WTP threshold. For example, policies that can be credibly linked to perceived public health benefits, energy security benefits, or economic development and employment benefits for key constituencies could improve consumer WTP for such policies. Fifth, both the economic and political constraints on the optimal climate policy are not static, but rather dynamic over time, opening up temporal considerations for adaptive policy design. Reducing CO2 abatement costs through technological innovation can increase the CO2 reductions achievable without exceeding the marginal social benefit of mitigation or exceeding consumer WTP for climate mitigation, effectively relaxing both the economic and political constraints (Nemet, 2010). Likewise, strengthening industries that stand to benefit from climate policies (i.e., low-carbon technology sectors such as renewable or nuclear energy, energy efficient technologies, biofuels, etc.) before directly impacting incumbent industries could boost Stiglerian demand for climate regulation, potentially further relaxing political constraints. For example, clean energy deployment subsidies and innovation policies designed to effectively reduce the costs of low-carbon energy alternatives and build stronger political interests around clean energy sectors can potentially launch a self-reinforcing cycle: stronger industries and lower technology costs yield greater demand for low-carbon policy and lower compliance costs which in turn yields even stronger industries and lower costs, and so forth. Sixth, the presence of constraints on carbon pricing instruments makes the creative use of resulting revenues critical to maximizing the economic efficiency and environmental efficacy of these instruments. Uses for revenue can be pursued that both offset pre-existing market failures or economic distortions and foster additional GHG emissions abatement without violating either economic efficiency or political economy constraints. Two important climate-related market failures could be deserving of attention. First, as many economist have demonstrated, knowledge spillovers, asymmetric returns, and the non-rival nature of new knowledge lead to under-investment in technological innovation (Arrow, 1962 and Jaffe et al., 2000). This market failure is not fully addressed by carbon pricing instruments alone (Fischer and Newell, 2008, Fischer, 2008 and Jaffe et al., 2005), even if they are unconstrained by political economy factors (Nemet, 2013). This market failure can impact the pace of low-carbon technical change and reduce the economic performance of carbon pricing instruments (Acemoglu et al., 2009). Using revenues to reduce low-carbon innovation-related market failures could therefore improve the economic and environmental performance of carbon pricing instruments, particularly when they are politically constrained (Fischer, 2008 and Nemet, 2010). Furthermore, use of revenues to procure additional emissions abatement beyond that secured by the carbon price itself would further reduce economic distortions associated with the remaining externalized costs of GHG emissions and could improve the net economic and environmental performance of the policy.15 To illustrate the opportunity to achieve much greater abatement than accomplished by a politically constrained carbon pricing instrument alone, consider what would be required to reduce annual U.S. energy-related CO2 emissions, now at roughly 5.3 billion tons (EIA, 2013), by 1 billion tons (in rough terms, what the Waxman–Markey bill aimed to accomplish by 2020). Assuming, for sake of example, a marginal abatement cost of $30 per ton of CO2 in 2020 (consistent with EIA, 2009) and a roughly linear abatement supply curve, we can infer that the average abatement cost across those 1 billion tons is roughly $15 per ton. If abatement were accomplished by establishing a Pigouvian carbon price equal to the marginal cost of abatement, it would require a $30 per ton carbon price to reduce CO2 by 1 billion tons, well in excess of apparent WTP thresholds discussed in Section 3.2.2. In contrast, the same emissions mitigation level could be accomplished through direct procurement of clean energy adoption and emissions mitigation on the order of $15 billion (i.e., 1 billion tons at an average cost of $15 per ton). The requisite procurement of emissions abatement could be financed by spreading the $15 billion total abatement cost across the remaining 4.3 billion tons of U.S. emissions by establishing a carbon fee of $3.50 per ton of CO2 in round terms.16 The targeted abatement could thus be achieved at less than one-eighth the carbon price necessary through a straight Pigouvian carbon tax approach and within the range of consumer WTP. As subsidies for low-carbon energy sources or other mitigation options can distort markets and lead to overconsumption of these energy sources relative to consumption levels at their true cost, this tax and procure approach can result in additional economic costs relative to a first-best implementation of a Pigouvian carbon tax. However, such a policy may in fact generate substantially greater economic welfare improvements than a Pigouvian tax approach constrained by real-world political economy considerations (i.e., within the $2–$8 per ton range evident in the U.S.). This simple example should indicate the significant potential for second-best policy designs capable of substantially improving the economic efficiency and environmental efficacy of climate policy in the presence of binding political economy constraints. In conclusion, addressing climate change externalities by pursuing an “economically optimal” or “first-best” Pigouvian carbon price is likely to yield suboptimal outcomes in practice when such policy efforts collide with real-world political economy constraints discussed in Section 3. As such, there may in fact be any number of viable second-best policy strategies that reside within the opportunity space for improvement discussed in Section 4. These policies could result in superior economic and environmental performance relative to either policy inaction or to the real-world implementation of politically constrained carbon pricing policies. There thus remains a considerable space for creativity in climate policy design and implementation. Full consideration of both economic and political economy constraints can yield several important implications for climate policy makers and help advance the design of more effective climate policy instruments. Finally, further research should be directed towards better characterizing the nature and operation of political economy constraints on carbon pricing.