Brown backstops versus the green paradox
Introduction
Well-intended climate policies may have perverse effects. Climate policies typically become stricter over time. Fossil fuel owners, deciding when to sell their scarce resources, may respond by speeding up extraction. This side effect can occur when fossil fuel reserves are limited and cheap to exploit: a reasonable characterization for conventional oil and natural gas, but much less for other important energy sources such as coal and unconventional oil. In this paper we ask whether climate policy has unintended consequences when there are two types of fossil fuels: one dirty and scarce, the other even dirtier and abundant.
Policies that reduce future dependence on fossil fuels might encourage suppliers, anticipating a future drop in demand, to bring forward the extraction of their resources. When present emissions are more harmful than future emissions, increasing carbon taxes can be counterproductive: a green paradox (Sinn, 2008a). Developing a carbon-free substitute for fossil fuels can cause a similar effect (Strand, 2007, Hoel, 2011). Cost reductions for the substitute decrease the scarcity value of fossil fuels, and thereby increase fossil fuel supply in all periods before exhaustion.1
The crucial feature that drives the above mechanism is the exhaustibility of the resource. This causes the tradeoff between current and future supply, and thus the effect of (expected) future policies on current supply and emissions. If the resource is fully abundant, resource owners supply the myopically optimal quantity in each period and the link between current and future markets is severed. Exhaustibility is a fair assumption for conventional oil and natural gas, which will be depleted in 50 to 70 years at current consumption rates.2 Coal and unconventional oil are much more abundant however. Coal reserves are sufficient to last another 250 years, and tar sand deposits in Alberta are estimated at 1800 bln barrels.3 The supply of these resources is primarily driven by costs rather than scarcity rents. Anticipated carbon taxes cause coal mines to shut down in the future, but do not increase near-term supply.
Coal and unconventional oil are significant from an economic and a climate change point of view. Coal satisfies a third of global energy demand and accounts for almost half of energy-related CO2 emissions,4 outranking petroleum in emission intensity by 30–40%. The IEA expects coal supply to increase by 60% in 2035 under business-as-usual policies5: twice as much as the projected increase in oil supply. Supply of unconventional oil, which is 20% more emission-intensive than petroleum (Charpentier et al., 2009), may increase fivefold to 11 mln barrels per day in 2035. These numbers suggest that in order to keep climate change within tolerable limits, it is imperative that coal and unconventional oil reserves remain largely unexploited (Gerlagh, 2011). A comprehensive assessment of the effectiveness of climate policies should take into account these dirty substitutes and their unique characteristics.6,7
In this paper, we develop a simple model with two time periods. We do not derive optimal policies, but present a descriptive analysis of the effect of future climate policies on emissions. We generalize assumptions in previous research along two important dimensions. Firstly, the model contains three energy types: a dirty exhaustible resource (e.g. oil), an even dirtier substitute (coal) and a clean substitute (solar). Secondly, we assume types to be imperfect substitutes for one another. Previous theoretical studies often assume perfect substitution (e.g. Hoel, 2011, van der Ploeg and Withagen, 2012a), which is unrealistic. We model climate policy as a future carbon tax or a decrease in the cost of the clean substitute. We calculate intertemporal carbon leakage as the increase in present emissions over the decrease in future emissions.
By virtue of the abundance of their resource, we assume that coal owners do not trade off present and future extraction. When faced with a demand reduction in the future, they will therefore not increase supply today. Oil emissions may leak away to the present, but the increase in current oil supply reduces demand for dirtier coal. Carbon taxes can cause negative leakage when the substitutability between oil and coal differs between periods. We may call this a ‘strong green orthodox’ (Grafton et al., 2012). Moreover, since carbon taxes decrease the price of oil relative to coal, a future tax delays rather than accelerates oil extraction when oil and coal are good substitutes in the future. Reducing the future cost of solar decreases present emissions when oil and coal are good substitutes or if the emission-intensity of coal is high.
Our contribution is twofold. Firstly, we offer a general theoretical framework that can make more accurate predictions than models that include only one or two energy types or assume perfect substitutability. The presence of an abundant dirty substitute reduces intertemporal leakage directly and indirectly, and may even cause negative leakage rates. By making more specific assumptions, we can obtain similar findings as in other papers on the green paradox. Secondly, our model is well-suited for empirical calibration. For carbon taxes, intertemporal leakage rates are less than 3%. For reductions in the future cost of renewables, leakage is between 7 and 8% for bioenergy, and 5% for renewable electricity. Bioenergy is a close substitute for oil, the most emission-intensive scarce fossil fuel, and hence more prone to intertemporal leakage than renewable electricity, which primarily competes with coal.
Though we focus on intertemporal leakage, our framework can also be used to analyze spatial carbon leakage, by relabeling the two time periods as two countries and setting the interest rate to zero. Calibrating a spatial version of the model, we find leakage rates ranging from 9% to 16%, within the range of estimates from computable general equilibrium models (Böhringer et al., 2012, Di Maria and van der Werf, 2012). These findings suggest that the green paradox is a small concern relative to spatial carbon leakage.
The rest of this paper is organized as follows. The section “Model” outlines the model. The section “Emission taxes” analyzes intertemporal and spatial leakages when carbon emissions are taxed in the future. The section “A cheaper clean backstop” studies the impact of reductions in the future cost of a clean substitute. We calibrate the models in the section “Empirical calibration”. The section “Spatial carbon leakage” discusses implications of the model for spatial carbon leakage, and calibrates a spatial version of the model. The section “Conclusion” concludes. All proofs are relegated to the Appendix.
Section snippets
Model
Consider a model with three types of energy: an exhaustible resource, a dirty backstop and a clean backstop. The backstops are inexhaustible, supplied competitively and have constant marginal costs.8 Though the word ‘backstop’ is sometimes used to denote a perfect substitute for an exhaustible resource, we explicitly allow for imperfect substitutability. The exhaustible resource is
Emission taxes
Regulators who want to reduce carbon emissions may not be able to do so immediately. Swift implementation of climate policies is often impeded by political and technological considerations. Announcing carbon taxes or caps in advance reduces compliance costs: it gives firms the opportunity to purchase abatement equipment and adjust their production processes, and allows consumers to make informed decisions about durable good purchases (Di Maria et al., 2008). The European Commission notes that
A cheaper clean backstop
In addition to implementing a carbon tax, climate-conscious policymakers may opt to reduce emissions by stimulating the development of clean alternatives to fossil fuels. To model such a policy, we analyze the effect of a reduction in the period 2 price of the clean backstop PC on emissions. The development of alternative energy sources requires resources to be committed well before the new technologies can be put to use, so exhaustible resource owners anticipate the lower period 2 clean
Empirical calibration
In this section, we explore the magnitude of intertemporal carbon leakage by calibrating Propositions 2 and 5.
In line with the integrated assessment literature on climate change, we impose a nested CES production function for composite energy Y (see Fig. 1) from five individual types. We differentiate between energy that is used for electricity generation E and energy used for non-electricity purposes N, including transport, heating and industrial uses, but the model arrives at a single,
Spatial carbon leakage
Climate policies may also have unintended side-effects in a spatial setting. When a group of countries reduces emissions unilaterally, pollution might move to other countries. This spatial carbon leakage occurs through two main channels (Felder and Rutherford, 1993). Firstly, dirty industries relocate to countries with laxer regulation. Secondly, a stringent environmental policy in environmentally conscious countries causes the world market price of fossil fuels to fall, increasing their use in
Conclusion
We employ a general model to analyze intertemporal carbon leakage in the presence of an abundant dirty backstop such as coal or unconventional oil. The green paradox literature overstates the adverse consequences of imperfect climate policies by not taking into account their potential to reduce emissions from coal and unconventional oil. It is important to consider these fuels as they already account for 50% of energy-related emissions, and will become even more important in the future.
A carbon
Acknowledgments
I would like to thank Ujjayant Chakravorty, Reyer Gerlagh and an anonymous referee for helpful comments on earlier drafts of this article, as well as seminar participants at University of Oxford and the 18th Annual EAERE Conference.
References (51)
- et al.
Tax distortions and global climate policy
J. Environ. Econ. Manag.
(2003) - et al.
Economic growth and interfactor/interfuel substitution in Korea
Energy Econ.
(2004) - et al.
Assessment of China׳s climate commitment and non-fossil energy plan towards 2020 using hybrid AIM/CGE model
Energy Policy
(2011) The competition between coal and natural gasthe importance of sunk costs
Resour. Policy
(1996)- et al.
Unilateral CO2 reductions and carbon leakagethe consequences of international trade in oil and basic materials
J. Environ. Econ. Manag.
(1993) Efficient climate policy in the presence of free riders
J. Environ. Econ. Manag.
(1994)- et al.
Depletion of fossil fuels and the impacts of global warming
Resour. Energy Econ.
(1996) - et al.
Border adjustment for European emissions tradingcompetitiveness and carbon leakage
Energy Policy
(2010) - et al.
China׳s energy economytechnical change, factor demand and interfactor/interfuel substitution
Energy Econ.
(2008) A dynamic analysis of japanese energy policiestheir impact on fuel switching and conservation
Energy Policy
(1994)