Mobile Edition
Print
Get the Mag
Weekly Updates1. INTRODUCTION
Analyzing the future of oil in a climate conscious world is a delicate task and a highly contentious issue. From a climate policy point of view, one major challenge is that energy exporting countries, especially OPEC and its member countries Saudi Arabia and Kuwait, claim that their resource rents would be negatively impacted due to measures to reduce C[O.sub.2] emissions and that they should be compensated for this, see for example Barnett & Dessai (2002) and Aarts & Janssen (2003). This claim for compensation is formally based on article 4.8 of the United Nations Framework Convention on Climate Change (UNFCCC) and articles 2.3 and 3.14 of the Kyoto Protocol. OPEC has in part based its claim for economic compensation on energy-economic model studies, see e.g. Ghanem et al. (1999). Additionally, a range of other studies have also found that OPEC would lose rent due to policies to reduce carbon dioxide emissions (Berg et al., 1997; McKibbin et al., 1999; Bernstein et al., 1999; Bartsch & Muller, 2000; and Radetzki, 2002). Briefly summarized, these studies suggest that policies and measures aimed at reducing C[O.sub.2] emissions will reduce the consumption of oil, which in turn would force the producer price for oil down. If a C[O.sub.2] tax or a cap and trade system is used, this will impose costs for both the oil consuming and extracting parties, but entail a transfer of rents from the extracting countries to the consuming countries, see e.g. Amundsen and Bergman (2005).
The highest estimate of the impact of the Kyoto Protocol on OPEC (assuming trading among Annex B Parties and U.S. participation) is a 13 percent reduction in oil revenue below the reference scenario for 2010 (McKibbin et al., 1999). OPEC's own model (Ghanem et al., 1999) and Bartsch & Muller (2000) both estimate roughly a 10 percent reduction in oil revenue for 2010 compared to baseline revenues (still these lower revenues are significantly higher than the revenues in 2000; for instance, in the study by Bartsch & Muller the revenues are some 35 percent higher than in 2000). The studies by Bernstein et al. (1999) estimate the decline in OPEC's GDP due to the Kyoto Protocol to be 0.45-1.15%. Further, van Vuuren et al. (2003) estimate that the Middle East will lose about 35 percent of its oil export revenue by 2050 if the atmospheric concentration of C[O.sub.2] is to be stabilized at 450 ppm, and WBGU (2003) estimates the total abatement cost for the Middle East at approximately two percent of expected GDP by 2050 for the same stabilization level; reduced oil export revenues are the main cause. Berg et al. (1997) find that the net present value of OPEC's oil rent decreases by 23% due to a globally adopted carbon tax of about US$ 90 per ton carbon. A presumed loss in future oil revenues was also one key reason why some oil companies spoke out openly against the Kyoto protocol (van den Hove et al., 2002).
However, there are several reasons why OPEC would not stand to lose a substantial amount of oil rent due to climate policies. Actually, OPEC might gain rent with climate policies. The arguments for this are based on the following premises:
* Proven oil reserves and estimated ultimately recoverable reserves of conventional oil are smaller than the maximum allowable cumulative emissions over the century even when meeting low atmospheric stabilization targets (Grubb, 2001; Azar et al., 2003). For example, the proven reserves of conventional oil amount to about 1,200 billion barrels or 140 Gt C, while the estimated ultimately recoverable reserves left to be extracted are about twice as large. In order to stabilize the concentration of C[O.sub.2] in the atmosphere at 450 ppm, we may emit roughly 500 Gt C over the next 100 years.
* Given that most conventional oil is inexpensive to extract (extraction costs in the Middle East amount to a few dollars per barrel (Adelman, 1986; IEA, 2005; Brandt & Farrell, 2007)), and that oil is superior in terms of liquid fuel production, which is crucial for the transportation sector, with few real contenders, most--if not all--of the estimated ultimately recoverable reserves will likely eventually be used even if we opt for stabilization levels that many would consider very stringent, i.e., about 400 ppm C[O.sub.2] in the atmosphere.
* The cheapest liquid substitutes for conventional oil products are fuels derived from extra heavy oils, tar sand, coal, natural gas, oil shale, sugar cane, and perhaps ligno-cellulosic biomass. All these fuels, except biomass-based fuels and natural gas, have a higher net carbon to energy ratio in their primary form than conventional oil, and, in general, producing liquid fuels from them is energy-intensive (Brandt & Farrell, 2007). This implies that the production cost of these fuels, when carbon emissions are priced, would be more affected by the carbon price than would the cost of fuels from conventional oil. (1) This was observed by Manne & Rutherford (1994). If the OPEC member states behave strategically, they should be able to utilize this aspect so as to increase their rent. In the case of biomass-based fuels, the global supply potential is limited, and it has competing uses in other energy sectors (heat and electricity production) when C[O.sub.2] concentration is stabilized at low levels cost-effectively (Azar et al., 2003).
Certain important prerequisites must be met in order for this argument to work:
* First of all, the climate regime has to be (close to) universal. If the climate regime is partial, then liquid fuels may be produced from coal in regions without a price on carbon. Thus, our paper does not attempt to invalidate results from models that have only looked at the impact of the Kyoto Protocol on OPEC's oil revenues.
* Second, countries must not protect synthetic fuel production based on unconventional oil, natural gas, and coal from a C[O.sub.2] price.
* Third, if climate policy is implemented through energy efficiency standards and subsidies to renewables, then demand for oil will drop, and so will the price.
Note that not all OPEC members will be equally affected by implementation of an international C[O.sub.2] mitigation policy. Most of the members are reported to have only small unconventional oil resources, but Venezuela is an exception. However, as will be seen below, because the carbon to energy ratio is higher for liquefied coal than for heavy oil, even extractors of unconventional oil might benefit from carbon prices.
In a previous paper, Persson et al. (2007), we analyzed whether OPEC would lose or gain rent from its export of conventional oil under the assumption that OPEC is a price-taker on the energy market, i.e., operates in a situation of perfect competition. In that paper we found that under some conditions OPEC gains rent, in the order of a few per cent, due to atmospheric C[O.sub.2] stabilization targets.
The aim of this paper is to see if this gain in rent can be observed when OPEC's market power is explicitly taken into account in a long-term modeling Framework (2). In order to perform this analysis we have developed a new numerical model, called OligOPEC, of the long-term liquid fuel market where resource constraints and OPEC's market power are taken into account.
Even if OPEC is often seen as the most famous cartel in the world, its behavior has been interpreted in divergent ways, see for example Alhajji (2004) and Smith (2005) for a survey of models trying to explain OPEC's behavior. The main conclusion from statistical tests of the validity of these models is that no model satisfactorily captures the behavior of OPEC over long periods of time; but that it is in general possible to find short time periods where most of the possible models cannot be rejected, see Smith (2005) and Kaufmann et al. (2008). Consequently, these statistical tests give at least some support for cartel behavior by OPEC (Smith, 2005; Kaufmann et al., 2008). Due to this limited cartel behavior OPEC has been described as a bureaucratic syndicate (Smith, 2005) or a clumsy cartel (Adelman, 2002). However, under all circumstances, the oil prices the past 30-35 years, maybe with an exception of a period in 1998, have been well above the reported production cost in most OPEC countries. This price/cost gap cannot satisfactorily be explained by the pure scarcity rent of oil, whereas rents generated due to market power offer a more satisfying explanation.
In section 2, we present our liquid fuel market model along with the parameter values used for calibration. Section 3 presents the results generated, and section 4 contains discussion and conclusions.
2. MODEL DESCRIPTION AND CALIBRATION
2.1 Model
The developed model is a dominant firm model where the dominant firm, representing OPEC, uses quantity-based strategies in order to maximize its net present value rent by taking into account the reaction of the fringe producers. The fringe consists of a large number of fuel producers that individually are so small that they cannot unilaterally affect the price; i.e., they are price-takers. For simplicity, OPEC is assumed to only have conventional oil resources. The fringe price-taking fuel producers, on the other hand, have resources of conventional oil, unconventional oil (extra heavy oils and tar sand), natural gas, biomass, and practically an unlimited amount of coal and carbon neutral hydrogen, e.g. produced by electrolysis using renewables.
A range of different dominant firm models of exhaustible resource markets have been developed and analyzed in the academic literature since the middle of the 70s. Usually, OPEC is either represented as a Cournot-Nash oligopolist, in which OPEC takes fringe production as given and hence not under the influence of OPEC's decisions (Salant, 1976; Salant, 1982; Ulph & Folie, 1980; Eswaran & Lewis, 1985; Berg et al., 1997), or as a Stackelberg leader where OPEC takes fringe reaction into account when deciding its production level (Gilbert, 1978; Newbery, 1981; Groot et al., 2003). The Stackelberg approach has a stronger theoretical appeal than the Cournot, since if OPEC is a dominant producer it should have the ability to take into account how its production level via the liquid fuel price affects the production level of the fringe, at least if it has rational expectations. However, the Stackelberg approach has serious problems in that the openloop version (3) of the model can be time inconsistent, see Newbery (1981), and that the Markov feedback version (4) of the Stackelberg model is unduly complicated to solve (although Groot et al., 2003 have recently succeeded in doing that). On the other hand, the open-loop Cournot-Nash approach is time consistent and can be used to construct defensible approximations of the Markov feedback Stackelberg version, see Newbery (1981) and Karp & Newbery (1993). Accordingly, the approach adopted in this paper is an open-loop Cournot-Nash approximation of the Markov-feedback version of the Stackelberg approach.
FEED