0جداسازی کربن، سیاست های اقتصادی و رشد

We present a model of endogenous growth in which the use of a non-renewable resource in production yields CO2 emissions whose accumulated stock negatively affects welfare. A CCS technology enables, via some effort, a partial reduction of these emissions. We characterize the social optimum and how the availability of CCS technology affects it, and study the trajectories of the decentralized economy. We then analyze economic policies. We first derive the expression of the Pigovian carbon tax and we give a full interpretation of its level, which is unique. We then study the impacts of three different second-best policies: a carbon tax, a subsidy to sequestered carbon, and a subsidy to labor in CCS. While all three tools foster CCS activity they generally have contrasting effects on resource extraction, carbon emissions, output and consumption. The carbon tax postpones resource extraction whereas the two subsidies accelerate it. Although the tax decreases short-term carbon emissions, the two subsidies can increase them, thus yielding a green paradox. The tax has a negative impact on the levels of output and consumption in the short-term, unlike the subsidies. The tax generally fosters growth whereas the subsidies reduce it; however, when the weight of the CCS sector in the economy is high, these impacts can be reversed.

The exploitation of fossil resources raises two concerns. The first is scarcity, because fossil resources are exhaustible by nature. The second is related to the emission of greenhouse gases associated with their combustion. Numerous models deal with this double issue; some of them in the context of partial equilibrium (e.g. Sinclair, 1992, Withagen, 1994, Ulph and Ulph, 1994, Hoel and Kverndokk, 1996, Farzin and Tahvonen, 1996 and Tahvonen, 1997) and others within general equilibrium growth frameworks (e.g. Stollery, 1998, Schou, 2000, Schou, 2002, Groth and Schou, 2007 and Grimaud and Rouge, 2008). A common feature of these papers lies in the fact that reducing carbon emissions necessarily means extracting less resource. Indeed, there is generally assumed to be a systematic link between resource extraction and polluting emissions, in the form of a simple functional relation (e.g. linear). In terms of economic policy, it is therefore equivalent to taxing either the pollution stream or the resource use itself. Nevertheless, it is now widely recognized that certain abatement technologies allow the reduction of emissions for a given amount of extracted resource. In particular, attention has recently been focused on the possibility of capturing and sequestering some fraction of the carbon embedded in fossil fuels, whether this capture occurs pre- or post-combustion. This has been reinforced by recent demonstrations of viability (for an overview, see IPCC special report, 2005). This process, often referred to as carbon capture and storage or carbon capture and sequestration (CCS), consists in separating carbon from hydrogen in the pre-combustion process or in separating carbon dioxide from other flux gases in the post-combustion process in an energy production plant. Once captured, the CO2 is injected into a reservoir1 for long-term storage. The availability of CCS technologies therefore means that the simple relation between resource extraction and carbon emissions is partially broken. Here we consider the availability of such an abatement technology in the context of a theoretical general equilibrium model with endogenous growth and a polluting exhaustible resource. We study how the socially optimal trajectories of the economy are modified by the availability of the CCS option, and how the first-best outcome can be restored in a decentralized economy. We also study the impact of three different second-best policies: a carbon tax, a subsidy to sequestered carbon and a subsidy to labor in the CCS activity. Endogenous growth allows us in particular to analyze the effects of the availability of CCS technology and the economic policy tools on growth, along the transition path and at the steady-state. Numerous uncertainties still surround the large-scale deployment of carbon capture technologies, especially with regard to the ecological consequences of massive carbon injection. The social acceptance of this abatement technique is also uncertain - for a survey on these issues, see for instance Jepma and Hauck (2011). Nevertheless, this technological option has become promising for the fossil energy extractive industry. For instance, Grimaud et al. (2011) show in an empirical model that, insofar as the right climate policy is implemented – a carbon tax in their model – the percentage of carbon sequestered can exceed 50%. We develop a Romer-type endogenous growth model in which the production of final good requires the input of an extracted resource, whose stock is available in limited quantities. This resource use generates polluting emissions, which we take to be CO2 emissions, whose flow in turn adds to the pre-existing stock of the pollutant – which features partial natural decay. Finally, this stock enters the utility function as an argument, making it possible to gauge how pollution accumulation negatively affects welfare. Here, we implicitly assume that the economy never reaches a critical level of carbon concentration that would yield an infinitely negative utility (for this type of assumption, see for instance Acemoglu et al., 2012). We then consider that a CCS technology is available. Via some effort, it allows for the partial reduction of the level of CO2 release. We thus distinguish between the total potential CO2 emissions associated with one unit of fossil resource (or equivalently the total carbon content per unit of resource) and the effective emissions, i.e. the fraction that remains after CO2 removal. Note that we do not account for geological CO2 leakage – on this issue, see for instance van der Zwaan and Gerlagh (2009). In this economy, the crucial trade-offs are made between current consumption, future consumption, and current and future environmental quality. We model these trade-offs through the allocation of labor between its alternative uses: output production, R&D and CCS. This general framework has a straightforward implication in terms of climate policy: the first best outcome can only be restored by taxing pollution, i.e. emissions remaining after sequestration, and not by taxing the resource itself.2 However, for various reasons, it is likely that the tax cannot be set at its Pigovian level in the real world. Hence, we study second-best policies: a second-best tax on effective carbon emissions, a subsidy to sequestered carbon, and a direct subsidy to labor used in CCS activity. In this second-best world, such complementary policies can improve welfare. This analysis constitutes the main contribution of our paper. We show in particular that it is important to understand how these policies affect the time profile of the total price paid by resource users. This time profile determines the resource extraction path, and hence impacts the paths of CCS activity, carbon emissions, R&D and output. We first depict the socially optimal trajectories of the economy, and we study how such trajectories are affected by the availability of the CCS technology. Then we fully characterize the trajectories of the decentralized economy, we derive the expression of the Pigovian carbon tax, and we give a full interpretation of it. In the general case, at the social optimum as well as in the decentralized equilibrium, the economy is always in transition; we nevertheless obtain closed-form solutions. This allows us to study the impacts of the three different types of second-best economic policies. A strand of literature tackles the question of CCS within calibrated empirical models – see for instance Edenhofer et al., 2005, Gerlagh and van der Zwaan, 2006, van der Zwaan and Gerlagh, 2009, Golombek et al., 2011 and Grimaud et al., 2011, or Kalkuhl et al. (2012). The focus of our paper is on the theoretical side of the issue. Several authors have studied the links between carbon abatement, optimal climate policy and technical change in theoretical models. In particular, Goulder and Mathai (2000) show that the presence of induced technical change generally lowers the time profile of optimal carbon taxes. Moreover, efforts in R & D shift part of the abatement from the present to the future. In a similar framework, Gerlagh et al. (2008) study the link between innovation and abatement policies under certain assumptions, in particular the fact that patents can have a finite lifetime. In these studies, the authors use partial equilibrium frameworks in which baseline CO2 emissions are exogenous, and final (or effective) carbon emissions are endogenous as there is an abatement activity with dedicated technical progress. Hoel and Jensen (2010) show, in a two-period model, that if the climate policy is imperfect – that is, if it can only be implemented in the second period – cost reductions are more desirable in the CCS than in the renewable sector in particular because they postpone resource extraction. Many recent contributions take into account the availability of a CCS technology. Most of them consider the context of partial equilibrium frameworks: see for instance Lafforgue et al., 2008, Narita, 2009 and Amigues et al., 2011 or Rickels (2011). These papers mainly focus on socially optimal issues, and in particular they study the optimal time profile of carbon sequestration. Lontzek and Rickels (2008) and Ayong le Kama et al. (2009) study the same questions, but they also consider a decentralized economy. However, they do not study the impact of economic policies on the decentralized equilibrium. Most of these papers consider a carbon ceiling; in this case, Lafforgue et al. show that CCS is implemented only when the ceiling is reached. When the CCS cost function is convex however, as in Rickels, it is optimal to sequester carbon before the ceiling. Similarly, CCS activity has to start in the near term when there is no ceiling but a damage function, as in Ayong le Kama et al. Finally, technical progress is not explicitly considered in these studies. Our main results can be summarized as follows.3 At the social optimum, the amount of labor devoted to CCS activity is a constant proportion of resource use; in other words, the CCS effort per unit of carbon content is constant. Since resource use decreases over time, this implies that the abatement effort also decreases: the greatest effort in abatement should happen today4 – this contrasts with the result of Lafforgue et al. (2008) mentioned in the previous paragraph. We also show that the availability of CCS technology modifies the socially optimal trajectories of the economy. It speeds up the optimal pace of resource extraction, as it relaxes the environmental constraint. While it diminishes polluting emissions in the long run, it fosters them in the short run when the rise in resource extraction, and thus of potential emissions, is less than proportionally compensated by the CCS activity. Lastly, the availability of such a technology reduces the socially optimal growth of output as a result of the acceleration in resource extraction combined with a negative effect on R&D effort. Due to the availability of CCS technology, the Pigovian carbon tax is unique, which contrasts with the standard result obtained in a context without abatement, as in Dasgupta and Heal, 1979, Sinclair, 1992 and Groth and Schou, 2007 or Grimaud and Rouge (2008) for instance. In these models, there are an infinity of optimal taxes which have the same dynamics but differ in their levels. Here, the tax level matters, especially for setting the optimal abatement effort level. The optimal carbon tax is equal to both the sum of discounted social costs of one unit of carbon and the cost of sequestering this unit. We study its properties, and we show that, under specific assumptions on preferences and technology, it is proportional to output.5 The second-best carbon tax fosters CCS activity and postpones resource extraction. It also lowers short-term CO2 emissions. However, this leads to lower levels of output and consumption in the short term. The impact of the tax on economic growth is more complex. As polluting emissions stem from the use of non-renewable resources, if no carbon abatement technology were available, a more stringent environmental policy would generally enhance economic growth, since it leads to postponing resource extraction (see for instance Groth and Schou, 2007 and Grimaud and Rouge, 2008). We show that this does not always hold true when CCS technology becomes available. Indeed, if the weight of the CCS sector in the whole economy is high, that is if the CCS sector employs an amount of labor which is significant, this effect can be reversed: the tax can reduce the growth of output and consumption. The subsidy to sequestered carbon is a perfect substitute for the carbon tax with regard to its impact on CCS activity. However, the effects of the two policy instruments on resource use are opposite: with the subsidy, extraction is faster. A kind of (weak) green paradox can therefore occur here in the sense that this climate policy can increase short-term emissions (on the issue of green paradox in other contexts, see e.g. Sinn, 2008, Gerlagh, 2011 and van der Ploeg and Withagen, 2012). Short-term resource use and carbon abatement being both promoted, this green paradox occurs when the former effect exceeds the latter. The impacts of the subsidy and the tax on short-term output and consumption are also opposite: the subsidy prompts greater output and consumption levels in the early periods. This point should be considered when taking into account public acceptance issues. Indeed, this subsidy could be seen as a good complementary tool to a second-best carbon tax since it alleviates the burden of climate policy in the short term. Finally, the effect of the subsidy on economic growth is basically also opposite to the effect of the tax: the subsidy generally reduces growth, but, if the weight of the CCS sector in the whole economy is high, it can promote long-term growth. Another result is that the subsidy to labor in CCS alone does not trigger any CCS activity. This tool has an effect only when it is used jointly with a carbon tax or a subsidy to sequestered carbon. In this case, its impact on CCS activity, resource extraction, carbon emissions and the level and growth of output and consumption are similar to those of the subsidy to sequestered carbon. The remainder of the paper is organized as follows. We present the model and we portray the social optimum in Section 2. We characterize the equilibrium of the decentralized economy in Section 3, and we study the first-best economic policy and the impact of the second-best policies in Section 4. Finally, we conclude in Section 5.

We have developed an endogenous growth model with climate change that features CCS technology. Such an abatement technology can be used to endogenize CO2 emissions for a given use of fossil fuel. We have fully depicted the socially optimal outcome of this economy and we have shown that the greatest effort in CCS should be undertaken today. Moreover, the availability of CCS technology can produce a rise in CO2 emissions in the short run since it speeds up the pace of resource extraction, which can offset the CCS activity. We have computed the first-best carbon tax, which is unique and generally increasing over time. We have fully characterized the decentralized economy's trajectories and, when the Pigovian carbon tax cannot be implemented, we have studied three types of second-best economic policies. The first one is a standard unit tax on carbon emissions. The second and the third are subsidies to sequestered carbon and the effort in CCS, respectively. The latter has an impact on the economy only if it is implemented together with at least one of the two other climate policies (the carbon tax or the subsidy to sequestered carbon). If so, the two subsidies have similar impacts. The three tools foster CCS activity. However, their impacts on resource extraction are opposite: the carbon tax postpones resource extraction whereas the two subsidies accelerate it. Therefore, while the tax unambiguously decreases short-term carbon emissions, the two subsidies can yield a green paradox in the form of a rise in short-term emissions. The carbon tax has a negative impact on the levels of output and consumption in the short term contrary to the subsidies. In this sense, the subsidies can favor the public acceptance of a carbon tax. The effects on growth are more complex. The carbon tax generally fosters growth whereas the subsidies reduce it. However, when the weight of the CCS sector in the economy is high, that is, if the rise in CCS drains an important amount of labor from research activities, the carbon tax can reduce output growth while the two subsidies can foster it in the long term. It remains necessary to move away from a carbon economy by switching to renewable or non-fossil fuel based energy sources (Gerlagh, 2006). In order to keep the model tractable, the availability of a clean and renewable energy source has not been introduced. This so-called backstop would not drastically alter the qualitative properties of our results. Nevertheless, it would be interesting to study the impact of the CCS option on the adoption timing of these alternative sources of energy. We can infer that the possibility of sequestering carbon emissions would delay the introduction of renewable energy.