Unilateral Climate Policy: Can OPEC Resolve the Leakage Problem?

• Author: Bohringer, Christoph

1. INTRODUCTION

   In the absence of an effective global agreement to mitigate climate change, individual countries lead the way with unilateral greenhouse gas (GHG) emission reductions. The most prominent example is the European Union (EU), which adopted legally binding emission reduction targets until 2020 to be achieved through an EU-wide emissions trading system.

   One major problem with unilateral action, however, is the risk of emission leakage, i.e., the relocation of emissions to regions without climate regulations. Leakage mainly occurs through two intertwined channels. The first channel is the so-called fossil-fuel-price channel: Reduced fuel demand of emission-constrained regions depresses international fuel prices, which in turn stimulates fuel consumption and thus emissions in regions without emission constraints. The second channel is the so-called competitiveness channel for emission-intensive and trade-exposed (EITE) goods, such as steel, cement, or chemical products: Emission constraints raise production costs for these industries, reducing their competitiveness in the world market which triggers more production and emissions in unregulated regions.

   The policy debate focuses on leakage through the competitiveness channel, mirroring concerns of regulated EITE industries on adverse production impacts. Several anti-leakage measures for EITE sectors are implemented or discussed. For instance, the EU has decided to allocate free allowances to EITE industries under the provisions of its emissions trading system. Another possible supplemental instrument are tariffs on the carbon embodied in imported goods from unregulated regions.

   In the applied economic literature, leakage induced by unilateral emission constraints and the impacts of domestic policy measures to reduce carbon leakage have been analyzed in various papers including Bernard et al (2007), Mattoo et al. (2009), Bohringer et al. (2010, 2011, 2012a), and Fischer and Fox (2012). A common finding is that leakage takes place to a larger extent through the fossil-fuel-price channel. However, these studies are based on the assumption of competitive international oil markets, disregarding empirical evidence on non-competitive behavior of OPEC (see e.g. Griffin, 1985; Alhajji and Huettner, 2000a,b; Smith, 2005; Hansen and Lindholt, 2008).

   The purpose of this study is to investigate how different prescriptions of OPEC's behavior affect the extent of carbon leakage triggered by emission regulation of the EU or other industrialized countries. That is, are the leakage rates much changed when we replace the paradigm of competitive behavior with alternative assumptions about OPEC's behavior? Furthermore, how will the (cost-)effectiveness of anti-leakage measures for EITE industries such as carbon tariffs change when we alter our behavioral assumption about OPEC?

   Our analysis suggests that OPEC's behavior may have substantial impacts on overall leakage rates. This is especially the case if OPEC acts as a dominant producer maximizing profits subject to price responsiveness of the residual demand, i.e., global demand minus Non-OPEC supply.When the EU imposes carbon pricing, OPEC may in fact find it profitable to cut back supply to such an extent that the oil price increases. This is quite the opposite of what would happen in a competitive oil market setting. A higher oil price implies that oil consumption outside the EU declines, i.e., there is a negative leakage effect in the oil market. Overall leakage is still positive though, but the leakage rate declines from almost 20% in the reference analysis with competitive oil supply to less than 10% when OPEC acts as a dominant producer. If the EU also introduces carbon tariffs on EITE imports, overall leakage drops to 4% (vis-a-vis 14% in the reference setting with competitive OPEC behavior). Our findings are based on simulations with a large-scale computable general equilibrium (CGE) model of global production, trade and energy usage. The model due to Bohringer and Rutherford (2010) is extended to consider alternative assumptions on OPEC's behavior.

   Our paper contributes to an extensive literature on strategic behavior related to climate policies, especially with respect to the battle on carbon rents and oil resource rents between OPEC and countries that adopt emission constraints. This literature has not focused on carbon leakage though. Some of the early studies include Wirl (1994, 1995), Tahvonen (1995, 1996), Braten and Golombek (1998) and Rubio and Escriche (2001). Building on this literature, Liski and Tahvonen (2004) set up a differential game between a fossil fuel exporting cartel and a coalition of consumers who face a pollution externality. They show that the optimal emission tax consists of a Pigouvian part and a tariff element, which can be either positive or negative. If the cost related to pollution is not too high, consumers shift over resource rents through the tariff and can even be better off than without pollution at all. Dullieux et al. (2011) introduce a ceiling constraint for the carbon concentration in the atmosphere. They find that consumers are always able to save carbon rents and resource rents, while the exporting cartel is only able to shift carbon rents if the marginal damage under the ceiling is sufficiently small.

   In a similar vein, Wirl (2012) sets out a differential game between a fossil fuel exporting cartel and a coalition of consumers who face external costs from global warming. In comparing price to quantity instruments for both the cartel and the consumers' coalition, he finds that each player can save more resource rents as well as carbon rents when implementing a price instead of a fixed quantity. Strand (2013) employs a static theoretical framework, where a "policy bloc" imports fuel from a welfare optimizing fuel exporter, and introduces either a tax or a cap-and-trade scheme to reduce emissions from fuel consumption. In each case, there is a fringe of competitive importers, as well as an efficient CDM-like offset market. He shows that the "policy bloc" prefers the tax over the cap-and-trade scheme, as the fuel export price is lower due to less strategic scope for the fuel exporter. Wie et al. (2011) conclude that climate policies in fuel consuming countries make it optimal for OPEC to increase subsidies to domestic oil consumers. Haurie and Vielle (2010) show with a CGE model that under different (global) carbon taxes it is optimal for OPEC, when behaving as a dominant producer, to cut back on supply substantially and lose market share in order to not letting the oil price drop too much. This is in line with our conclusions.

   The remainder of this paper is organized as follows. In Section 2, we put forward various formulations of OPEC behavior, which have been discussed and to some extent empirically tested in the literature. In Section 3, we present the CGE model underlying our quantitative analysis and the policy scenarios we are investigating. In Section 4, we present and discuss simulation results. In Section 5, we conclude.

2. ALTERNATIVE FORMULATIONS OF OPEC BEHAVIOR

   The characterization of oil markets, which is commonly used in CGE models, assumes price-taking (competitive) behavior. In this section we describe alternative non-competitive formulations of OPEC behavior in the oil market, which we later implement in the CGE model.

   2.1 Fixed Production or Fixed Price

   OPEC's total production of crude oil is regulated through country-specific production quotas, which are updated about twice a year (depending e.g. on the market situation). However, there is no secret that production quotas are every so often exceeded. The additional volume though is relatively small. Thus, at least in the short run, one alternative formulation of OPEC's behavior is to keep output fixed.

   From 2000 to 2005, OPEC had a price target of $22-28 per barrel of oil, meaning that OPEC intended to adjust its supply in order to keep oil prices within this price band. Yet, the strong growth in oil demand put more and more pressure on the oil price, making it very difficult for OPEC to keep prices below the upper bound of the price band. As a consequence the price band was abandoned in early 2005, when prices on OPEC oil were around $40 per barrel. Since then, representatives of OPEC countries have regularly suggested new price targets, but no official targets have been expressed. Nevertheless, it seems reasonable to argue that OPEC might regulate supply in order to attain an oil price level which the producer group thinks is profitable in the medium to long run.

   2.2 Target Revenue

   An alternative to targeting production or price is to target revenue (e.g., Alhajji and Huettner, 2000b). The assumption is that OPEC countries target a certain revenue to balance their budget, i.e., they choose a level of output that generates the targeted revenue. Let the exogenous revenue target be given by R, which we interpret as gross revenues. Moreover, let [x.sub.OPEC] denote OPEC supply, p the worldwide oil price, and p = [p.sub.R]([x.sub.OPEC]) the residual inverse demand function for OPEC, which accounts for price responses in both demand and Non-OPEC supply. Then we have:

   [p.sub.R]([x.sub.OPEC])[x.sub.OPEC] = R (1)

   If maximizing gross revenues has an internal solution, say at [x.sub.OPEC] = x* with R = R*, it follows that any [x.sub.OPEC] below or above x* must give a lower R. Thus, if we let x(R) denote the output level that gives revenue R, it follows that x(R) cannot have any solution for R > R*, x(R*) = x*, and x(R) have at least two solutions for R

   2.3 Dominant Producer

   The alternatives to competitive behavior discussed so far postulate that OPEC targets either its production or gross revenue, or the oil price. From an economic perspective, it is more appealing to assume profit-maximizing behavior. The latter is reflected by the dominant producer hypothesis, where the dominant producer--in...