Economic Incentives

Review of Industrial Organization 12: 751–765, 1997. c 1997 Kluwer Academic Publishers. Printed in the Netherlands.

Raising Rivals’ Costs Strategies via Emission Permits Markets

EFTICHIOS SOPHOCLES SARTZETAKIS� University College of the Cariboo, Department of Economics and Finance, 900 McGill Road, P.O. Box 3010, Kamloops, B.C., Canada V2C 5N3

Abstract. In the present paper we examine the effect of emissions permit price manipulation within an oligopolistic model. We examine the effect that positioning strategies in permits markets have on the degree of competition in the product market as well as on social welfare. The analysis is based on the concept of raising rivals’ cost strategies. We find that competition in the product market can be lessened substantially. The welfare effect is ambiguous. If the leader expands its market share at the expense of a less efficient rival, or if it excludes a less efficient entrant, overall efficiency may increase despite the decrease in the industry’s output. When efficiency decreases, or when consumers’ protection is a policy priority, the initial distribution of permits can be used to control power in the permits market. Such interventions though, improve efficiency only when policy makers have substantial information on the technological structure of the industry, and thus, should be used with caution. Given the importance of information, sharing of information and coordination of actions between policy makers is very important.

Key words: Tradeable emission permits, raising rivals’ costs strategies, antitrust policy.

I. Introduction

According to the theory of externalities, a proper policy should provide economic agents with adequate incentives to undertake the right amount of the externality creating activity. In the case of environmental externalities, this is accomplished with the use of tradeable emission permits. Under ideal conditions, including perfect competition, this policy instrument achieves efficiency. Many regulated industries though, are not competitive, and the markets in which emission permits are traded may not be competitive either. The literature recognizes that, in the presence of product market distortions tradeable permits do not yield the optimum allocation of resources. However, little is still known about the impact of emission permits market distortions on product market structure, and on social welfare.

� I am grateful to Thomas Ross for his encouragement and support through all stages of this research project. I would also like to thank an anonymous referee of this journal, Donald McFetridge, Keith Acheson, Steven Ferris and Peter Tsigaris. Participants at the Sixth Annual Conference of the European Association of Environmental and Natural Resource Economists (Umea, June 1995) have also contributed with their comments. I gratefully acknowledge financial support from the Bureau of Competition Policy, Industry Science and Technology Canada.

752 EFTICHIOS SOPHOCLES SARTZETAKIS

Even though permits markets imperfections might not be an important concern when the number of regulated firms is large, it could be a serious problem in more localized permits markets. For example, if the sulphur dioxide emissions of electricity generating plants in Canada were controlled through emission permits regulations at the provincial level, the small number of participants in the relevant permits markets does not guarantee competitive behaviour. Similar concerns have been raised by von der Fehr (1993) and Fershtman and de Zeeuw (1996) concerning the U.K. electricity industry, and by Hanley and Moffat (1992) concerning Scotland.

In the present paper we examine whether power in the permits market can be used to reduce the existing and potential level of competition in the product market. Our analysis is based on the concept of raising rivals’ cost strategies.1

From its conception, the theory generated a vigorous debate on both theoretical and empirical level.2 Two main points of criticism are identified on the theoretical level. The first, questions the applicability of raising rivals’ cost strategies by arguing that such strategies are profitable only in limited cases. The second, argues that raising rivals’ cost strategies have an ambiguous effect on overall efficiency, and thus, do not necessarily require policy intervention.

Our objective is first, to contribute to the environmental policy literature by examining the efficiency of tradeable emissions permits regulations in the presence of market power. Second, to contribute to the theoretical evaluation of raising rivals’ cost strategies, by providing an example in which raising rivals’ cost strategies are profitable, and examine the overall efficiency effects of these strategies.

We examine an oligopolistic industry whose emissions are controlled under a tradeable emission permits regulation. Tradeable emission permits are considered as an input with fixed, exogenously determined supply. Each firm can substitute away from permits by engaging in abatement. Abatement depends on output and thus, decisions in the product and permits markets are linked. We assume that one of the firms, hereafter called the leader, has power in the permits market. The leader can exercise power in aggressive (exclusionary strategies) or moderate ways (positioning strategies). Although exclusionary strategies can be effective in forcing rivals to exit the market, we do not examine them because they can be challenged by existing antitrust policies.3 For the same reason we do not examine cases in which permits are used to stabilize merger agreements by securing profits against potential entry. We rather focus on moderate predatory (positioning) strategies and assess their effect on social welfare.

We find that positioning strategies are profitable. The more stringent the regu- lation and the more expensive the abatement is, the more profitable raising rivals’

1 See Salop and Scheffman (1981, 1983, 1987), and Krattenmaker and Salop (1985, 1986, 1987). Subsequent extensions of the theory include, Hart and Tirole (1990), Ordover et al. (1990), Gaudet and Long (1993).

2 For recent reviews of the debate see Ware (1994) and Coate and Kleit (1994). 3 For example, under Sections 77–79 of the Canadian Competition Act, these practices can be

clearly challenged for abuse of dominance position.

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 753

cost strategies are.4 The welfare effect of positioning strategies is ambiguous and it depends on the technological efficiency of the leader relative to its existing and potential rivals. 5 If the leader expands its market share at the expense of a less effi- cient rival, or if it excludes a less efficient entrant, overall efficiency may increase despite the decrease in the industry’s output. Thus, even when the market for per- mits is not competitive, tradeable emission permits regulations can be an efficient way to control pollution.

The literature on strategic behaviour in the permits market is limited. Misiolek and Elder (1987) present the first analysis of the use of raising rivals’ costs strategies in the permits markets within a dominant firm model. Von der Ferh (1993) analyses a Cournot duopoly regulated through a tradeable emissions permits system. Although our work bears some similarities in approach and results with von der Fehr (1993), it defers in two respects. First, we model firms’ abatement costs separately from all other production costs. Second, we focus on positioning strategies while von der Ferh is more interested in exclusionary strategies. Requate (1993) analyses the case of cooperation between firms. Fershtman and de Zeeuw (1996) examine the bargain process of emissions trading between two Cournot players. They find that once all possible sets of outputs have been identified, firms trade permits so as to maximize joint profits and thus, aggregate output is reduced.

Although strategic manipulation of permits markets might not be an existing problem,6 the potential application of permits systems to small numbers markets justifies a closer look at possible anticompetitive behaviour. Our welfare results support earlier views (Tietenberg, 1985, 1989) suggesting that the significance of permit price manipulation is limited. However, there are situations, such as the case of an inefficient leader, in which policy intervention is necessary. In such situations, the distribution of emission permits can be used as an instrument to control power in the permits market. Policy intervention can be welfare improving only if policy makers have substantial information on the technological structure of the industry. Given the importance of information, information sharing and coordination of actions between policy makers (environmental and competition policy) is crucial.

The rest of the paper is organized as follows: Section II develops the benchmark case of competitive permits market; Sections III and IV examine the cases in which the leader in the permits market decreases its rivals’ shares in the product market, and deters entry respectively. Section V contains the concluding remarks. All tables referred to in the text appear at the end of the paper. Proofs of the results discussed in the paper are available by the author upon request and they can also be found

4 Brown-Kruse et al. (1995) arrive at similar results conducting laboratory experiments in envi- ronments that resemble very closely our theoretical framework.

5 Assuming that profits are distributed to consumers, the term welfare denotes the sum of consumer and producer surplus.

6 Note however, that one of the explanations for the very low level of transactions in existing permits trading programs is that firms fear that buyers of emission permits could later use them to control the product market (see Hahn, 1989).

754 EFTICHIOS SOPHOCLES SARTZETAKIS

in Sartzetakis (1996a). In Appendix A we present the specifications of the model, while in Appendix B we present the initial values we used for the simulations.

II. The Competitive Benchmark Case

Assume a homogeneous Cournot duopoly facing linear demand and increasing mar- ginal cost of production. The production process generates emissions of a pollutant. Firms can reduce emissions by either reducing output or controlling emissions. The marginal cost of abatement is increasing both in output and abatement per unit of output. Firms’ production and abatement technologies differ.

Policy makers aim at reducing emissions to a certain level, which is not optimally decided.7 The regulatory intervention takes the form of tradeable emission permits. Firms are required to own emission permits in order to emit. Each permit specifies an amount of allowable emissions, assumed to be one unit, and is freely transferable. The number of permits issued is the maximum allowable units of emissions.

Emissions permits can either be auctioned or distributed free of charge (grandfa- thered) to the existing firms. Grandfathering of emission permits is based either on firms’ historical level of emissions, or it is arbitrary, for example equiproportionate allocation of permits. We assume that all permits are distributed to existing firms. Thus, the supply of permits is perfectly inelastic, and potential entrants have to buy all the permits they need from incumbent firms. This assumption simplifies the exposition of our results without limiting their generality.8 After the initial allocation of permits, trade is allowed.

Perfectly competitive markets for permits share the following two main proper- ties. First, minimization of abatement costs is achieved but overall efficiency cannot be achieved. If both firms act as price takers in the permits market, optimization implies that each firm trades permits to the point that the permit price equals its mar- ginal abatement cost. Equalization of marginal abatement cost across firms yields the efficient distribution of abatement effort, but due to the oligopolistic product market structure, it cannot achieve the efficient production allocation. Industry’s output decreases from its pre-regulation level, while firms’ shares are unaffected by the regulation. Thus, the more efficient firm reduces output relatively more than the less efficient firm. Trading of permits does not necessarily yield the first-best allocation of resources when product markets are imperfectly competitive.9

7 Emission ceilings are usually decided through a political process involving consultation with various interest groups and may be influenced by international agreements.

8 If more than one industry’s emissions are regulated under a tradeable permits system, and all firms participate in the allocation of permits, the supply of permits to one industry is elastic. To examine the effects of power in the permits market, we have to examine the impact on all industries. In the present paper we restrict our attention to one industry.

9 For extensive discussions of the efficiency of competitive permits markets when product markets are oligopolistic, see Borenstein (1988), Malueg (1990), Hung and Sartzetakis (1994) and Sartzetakis (1996b).

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 755

Second, the mode of permits distribution does not affect efficiency but only profits’ distribution. The competitive permits price is independent of the initial distribution of emission permits. This is not surprising since both firms are price takers in the permits market and each firm’s permits endowment is exogenous. Thus, firms’ optimal choices of output and abatement per unit of output are independent of the mode of initial permits distribution.

III. Permits Price Manipulation

1. EFFICIENCY AND WELFARE RESULTS

Assume that one of the two Cournot players has price setting power in the permits market. This case is modelled as a two stage game. In the first stage the leader chooses the permit price. In the second stage, both firms make their output and abatement decisions taking the permit price as given. To determine the subgame perfect equilibrium, we begin by solving the second stage of the game first.

In the second stage, firm 2 is a permit price taker and thus, it trades permits up to the point that its marginal abatement cost equals the marginal cost of permits, i.e. the permit price. Although the leader, firm 1, does the same, its marginal cost of permits is not equal to the permit price. This is because permits have an extra value for the leader; by manipulating the price of permits, the leader can influence its rival’s costs. Recognizing this, the leader trades a different than the competitive number of permits, adjusting its output and abatement such that the permits market clears at the permit price it has committed to in the first stage of the game. Thus, its marginal cost of abatement is not equal to the permit price.

In the first stage, the leader chooses the permit price that maximizes its profits, taking into account the Nash equilibrium of the second stage. The choice of a higher than the competitive permit price has a twofold effect on the leader’s profits. On the one hand, the rival’s costs increase and its output decreases resulting in an increase in the leader’s revenue. On the other hand, the leader’s costs increase, first because of the increase in the permit price, and second because the leader’s net demand for permits increases in order to clear the market. The leader commits to a higher than the competitive permit price if the benefits from raising its rival’s costs dominate the increase of its own costs. In such a case, the leader overbuys permits and thus, its marginal cost of abatement is lower than the permit price. Therefore, compliance costs are not minimized in the case of permit price manipulation.

Another difference with the competitive case is that the permit price under leadership is positively related to the leader’s endowment of permits. Thus, policy makers could control the leader’s power by manipulating the initial distribution of permits. Such an action though, requires full information and the willingness to base

756 EFTICHIOS SOPHOCLES SARTZETAKIS

Figure 1. Permits price determination.

the initial permits’ allocation on efficiency rather than distributional considerations. Since both requirements are strong we do not pursue this policy option further.10

Figure 1 illustrates the effect of permits price manipulation on permits market. Each firm’s demand for permits is given by its marginal abatement cost (MCAi). The horizontal summation of the firms’ demands is the market demand for permits. The supply of permits, �E, is perfectly inelastic. Under perfect competition, permits price is P" and each firm holds E"1 permits at the equilibrium. When firm 1 is the leader in the permits market, it sets a higher than the competitive permits price, Pm

and holds Em1 permits at the equilibrium, which are more relative to competition (Em1 > E

"

1 ). The leader’s demand for permits does not coincide any more with its marginal cost of abatement. The leader uses permits not only as a substitute to abatement but also as a device of manipulating its rival’s costs. Thus, the value of permits for the leader is higher and its demand for permits is higher accordingly.

Permits price manipulation also results in reshuffling of product market shares. Because of the increase in permit price, both firms’ marginal costs increase and

10 Hahn (1984) derives the cost-minimizing allocations of permits in the case that one firm acts as a monopolist/monopsonist in the permits market.

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 757

thus, industry’s output decreases. However, the increase in leader’s marginal cost is smaller and thus, its product market share increases relative to competition.

Thus far we dealt with the effect of permits price manipulation on the alloca- tion of abatement and production. We now move to examine the effect of permit price manipulation on industry’s profits. We compare industry’s profits under price manipulation to the competitive benchmark. Given that the allocation of produc- tion under competition is not efficient, could leadership improve industry’s profits despite the fact that industry’s abatement costs increase?

To illustrate the impact of leadership on industry’s profits we use simulations. The values of the parameters used in the simulations are such that the cost of abatement is a significant part of firms’ total costs. Table I presents the results of these simulations. We find first, that leader’s profits increase regardless of the technological structure of the industry (third column in Table I). Second, industry’s profits decrease when firms have similar production and abatement technologies, or the leader is less efficient that its rival (fourth column in Table I). Third, industry’s profits increase when either the leader is more efficient in production or the price taker in abatement (fourth column in Table I). Permit price manipulation implies that the price taker is forced to engage in higher abatement and lower production relative to competition, while the leader does exactly the opposite. Industry’s profits therefore, depend on firms’ relative efficiency in production and abatement. On the one extreme, if the leader is more efficient in production but less efficient in abatement, permits price manipulation enhances efficiency by inducing firms to exploit their respective advantages. On the other extreme, if the leader is less efficient in production and more efficient in abatement, efficiency is decreased.

The ambiguity of the effect of permit price manipulation on industry’s profits is carried over to social welfare. The model used for the simulations assumes that permits price manipulation does not affect industry’s profits. Thus, the change in industry’s profits equals the change in social welfare. Similar results are expected in the case that output changes as a result of permits price manipulation. It is more likely that output decreases (except if the leader is much more efficient than its rival) and thus, welfare is more likely to decrease as a result of permits price manipulation.

One clear case that calls for policy intervention emerges from the above discus- sion. Namely, the case of a large, inefficient in production firm that can become a leader in the permits market. Given that this is not an atypical situation in regulated industries, such as electric utilities, policy makers should watch for overbuying of permits.

2. FACTORS AFFECTING THE LEADER’S ABILITY TO EXERCISE MARKET POWER

We now turn to examine the factors influencing the ability of the leader to implement raising rivals’ costs strategies. The leader’s choice of permit price depends on the effects on both firms’ marginal costs. It can be shown that firms’ marginal costs

758 EFTICHIOS SOPHOCLES SARTZETAKIS

Table I. Effect of permits price manipulation on industry’s profits, presented as a function of the leaders’ production and abatement efficiency relative to the price taker

c1 (c2 = 200) d2 (d1 = 250) (�

m 1 ��

" 1 )

�" 1

( P

�m i �

P �" i )P

�" i

200 250 0.0364 �0.01013 150 250 0.02918 �0.00552

100 250 0.02385 �0.00162 100 200 0.02038 �0.00089 100 150 0.01715 �0.00028 100 100 0.01417 +0.00021 100 50 0.01145 +0.00060 100 0 0.00901 +0.00087

80 250 0.0221 �0.00024 75 250 0.02169 +0.00009 50 250 0.01979 +0.00164 0 250 0.01663 +0.00432

Where P2

i=1 � j i = �

j 1 + �

j 2 , i = 1, 2 and j = ", m, are the

industry’s profits. Superscripts " and m denote equilibrium values under competition and leadership respectively.

of production and abatement can both be expressed as functions of the difference between the permit price under leadership and under competition. Thus, the permit price difference is a good indicator of the strength of raising rivals’ costs strategies. It can be shown that the permit price difference depends on the leader’s pre- regulation product market share, its share of the emission permits in the initial allocation, and its abatement efficiency relative to its rival. It is clear that the more efficient the leader is in abatement the more it can raise its rival’s costs. The effect of the leader’s share in the product market and in the initial permits allocation is discussed in some detail.

First, the permit price difference is negatively related to the leader’s pre- regulation market share. This result contrasts the predictions of the raising rivals’ costs theory.11 In the standard raising rivals’ costs model, it is assumed that the lead- er buys exclusionary rights of the input, and that the cost of exclusion is independent of the leader’s output. Thus, the higher the leader’s market share, the smaller is the effect of the fixed exclusionary costs on its average cost. In our model, overbuying strategies do not increase the leader’s fixed costs, they rather increase its marginal cost. Thus, the larger the leader’s market share, the higher on its marginal cost it operates, and thus, it is relatively more expensive to raise its market share further. Table II illustrates this result with numerical examples. All changes are reported

11 For instance, see Salop and Scheffman (1987), and Coate and Kleit (1995, p. 80).

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 759

Table II. Profitability of raising rival’s cost strategies as a function of the leader’s pre-regulation market share. Equiproportional distribution of permits ( �E1 = �E=2)

�s1 P m � P"

� @�m

1 @qm

1 �

@�" 1

@q" 1

�

@�" 1 =@q"

1

� @�m

2 @qm

2 �

@�" 2

@q" 2

�

@�" 2 =@q"

2

(� m 1 ��

" 1 )

�" 1

0.45647 65.7057 0.04145 0.13805 0.06189 0.47866 64.6653 0.04359 0.13427 0.04662 0.50 63.6249 0.04513 0.13058 0.0364 0.52053 62.5845 0.04621 0.12698 0.02918 0.54031 61.5441 0.04692 0.12345 0.02385

Where, �s1 is the leader’s pre-regulation market share; P j, j = ", m is the

permit price; �j1 are the leader’s profits; and � j i = di�iqi + e�

2 1q

2 1 + P

j(�qi � �iqi � �E), i = 1, 2 and j = ", m, is firm i’s total cost of compliance with the environmental regulation. Superscripts " and m denote equilibrium values under competition and leadership respectively.

in percentage terms. Column two shows that, as the leader’s pre-regulation market share, ŝ1 increases, the permits price difference P

m � P" is decreasing. Columns

three and four demonstrate that as the leader’s market share increases, its marginal cost of compliance is increasing, while that of the price taker is decreasing. As a result, the increase of leader’s profits resulting from raising rivals’ costs strategies decreases as its pre-regulation market share increases. This result is illustrated in the fourth column of Table II.

Second, the larger the leader’s share in the initial permits distribution, the larger is the permit price difference. The leader’s marginal cost of abatement is lower than the permit price at the equilibrium. Thus, the leader’s compliance cost is lower the higher is its endowment of permits. Comparison of the simulation results in Tables II and III illustrates the impact of the mode of initial permits distribution on the leader’s ability to raise its rival’s cost. In deriving Table II we assume that permits are distributed equally among firms, while in Table III we assume that permits are distributed according to the firms pre-regulation market share. Comparison of each table’s column two reveals that, the higher the leader’s endowment of permits, the higher the permit price difference.12 The mode of initial allocation of permits affects the leader’s ability to raise its rival’s cost. When permits are auctioned, the leader pays the higher price for all permits it uses, and thus, its ability to raise its rival’s cost is reduced.

12 In contrast to the results in Table II, the permits price difference Pm � P" in Table III is larger the larger is the leader’s pre-regulation market share. This indicates that, for range of values we use, the impact of the endowment of permits on raising rival’s cost strategies is stronger than the impact of the pre-regulation market share.

760 EFTICHIOS SOPHOCLES SARTZETAKIS

Table III. Profitability of raising rival’s cost strategies as a function of the leader’s pre-regulation market share. Distribution of permits according to pre-regulation market share: ( �E1 = ��q1)

�s1 P m � P"

� @�m

1 @qm

1 �

@�" 1

@q" 1

�

@�" 1 =@q"

1

� @�m

2 @qm

2 �

@�" 2

@q" 2

�

@�" 2 =@q"

2

(� m 1 ��

" 1 )

�" 1

0.45647 61.0332 0.03529 0.12824 0.05704 0.47866 62.3291 0.04052 0.12942 0.04442 0.50 63.6249 0.04513 0.13058 0.0364 0.52053 64.9207 0.04924 0.13172 0.03087 0.54031 66.2166 0.05293 0.13282 0.02682

Where, �s1 is the leader’s pre-regulation market share; P j, j = ", m is

the permit price; �j1 are the leader’s profits; and � j i = di�iqi + e�

2 1q

2 1 +

Pj(�qi � �iqi � �E), i = 1, 2 and j = ", m, is firm i’s total cost of com- pliance with the environmental regulation. Superscripts " and m denote equilibrium values under competition and leadership respectively.

IV. Emission Permits as a Barrier to Entry

In this section we extend our discussion to cases in which raising rivals’ costs strategies blockade entry. We assume that the two incumbent firms of the previous section encounter a potential entrant who, in the absence of environmental regu- lation, would enter into the industry. The potential entrant has zero endowment of permits and thus, it has to buy all permits it needs. We assume that the potential entrant is efficient enough to make positive profits when permits market is compet- itive, but not as efficient as to rely solely on abatement. Thus, a permit price exists above which the entrant is forced to engage in a level of abatement that renders entry unprofitable. Under these assumptions, raising rivals’ costs strategies could lead to entry deterrence, since in most cases Pm > P". As we will show in what follows, raising rivals’ costs strategies lead to entry deterrence only in a limited number of cases.

In what follows we use simulations to examine: first, the effect of increased rivalry on leader’s ability to raise the permits price; second, the effect of entry deterrence on welfare; and third, the effect that the form of initial permits allocation has on entry deterrence.

First, as the number of firms in the industry increases, their market share and thus, their abatement decreases. All firms operate at a lower point on their abatement cost schedules. As a result, the leader has to hold more permits and rely less on abatement in order to increase the permit price and gain market share. Columns two and three of Table IV illustrate these points. The leader’s cost of raising its rivals’ cost increases the more competitive the product market is. Thus, the leader chooses a lower permit price the larger is the number of rivals i.e., Pm > P�, where the superscript � denotes equilibrium values under leadership when three firms share

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 761

Table IV. Leader’s ability to raise rival’s costs as the number of firms in the industry increases

d2 = d3 (d1 = 250) �A m 1 � �A

� 1 �(NE

d 1 )

m � �(NEd1 )

� Pm � P�

0 52.53590 �85.91165 59.1051 50 52.04524 �85.10928 53.8903

100 51.55458 �84.30690 48.6755 150 51.06392 �83.50453 43.4607 200 50.57326 �82.70215 38.2459 250 50.08260 �81.89978 33.0311 300 49.59194 �81.09740 27.8163

A j 1 = �

j 1q

j 1 , j = ", m, �, � is the leader’s level of abatement, and �A

m 1 � �A

� 1 =

(Am1 �A � 1) � (A

� 1 � A

� 1). Both parentheses are negative and thus, �A

m 1 � �A

� 1 >

0 implies that j�Am1 j > j�A � 1 j. (NE

d 1 )

j, j = ", m, �, � is the leader’s net demand for permits at the respective equilibrium and �(NEd1 )

m � �(NEd1 )

� = [(NEd1 ) m

� (NEd1 ) "] � [(NEd1 )

� � (NEd1 )

�]. The superscript � denotes equilibrium values under competition when three firms share the market and the superscript � equilibrium values under leadership.

the market. Column three of Table IV shows that this result holds regardless of firms’ abatement efficiency. Sartzetakis (1993) shows that as the product market tends to competition, the leader’s power in the permits market vanishes completely.

Second, the welfare effect of deterring entry is positive in most cases. Table V presents the simulations when permits are distributed equally between the two incumbent firms. Column two presents the profits of the entrant under competition ��3. Firm 3 is assumed to be as efficient as firm 2, but firm 2’s profits are always higher than firm 3’s because firm 2 has a positive endowment of permits. Column three presents the profits of the entrant in the case of leadership, ��3 . Entry is profitable only if the entrant is more efficient in abatement than the leader, i.e. � �

3 > 0 only if d3 < d1. Raising rivals’ costs strategies lead to entry deterrence in the cases that the potential entrant is equally or less efficient in abatement than the leader. Firm 1’s profits are higher when it acts as a leader, and the increase in leader’s profits is larger when leadership deters entry as shown in column three of Table V. Entry deterrence results in a decrease in output, since competition in the product market decreases. In contrast, industry’s profits increase, since entry is blockaded only when the potential entrant is less efficient than the leader. The increase in industry’s profits dominate the decrease in consumer surplus resulting in higher welfare, as illustrated in column five.13 Furthermore, entry deterrence yields an improvement in environmental quality because firms are engaging in higher levels of abatement in order to deter entry,14 as illustrated in the last column

13 This result is more general. Dixit and Stiglitz (1977) and Spence (1976) have shown that there are cases in which there is too much entry and facilitating more does not improve economic welfare.

14 In the case of entry deterrence, the two incumbent firms make their decisions based on P�; the equilibrium in this case is denoted with a superscript 2�. Total emissions are lower than the required

762 EFTICHIOS SOPHOCLES SARTZETAKIS

Table V. Effect of entry deterrence on social welfare. Grandfathering according to pre- regulation market share: ( �Ei = �q̂i)

d2 = d3 (d1 = 250) � � 3 �

� 3 �

� 1 � �

� 1 W

� � W�

0 129,659.5 112,026.1 12,248.8 �8,001.5 50 106,456.1 87,033.4 13,841.2 �9,041.7

100 84,288.0 62,995.4 15,530.8 �10.145.5 150 63,155.1 39,912.4 17,317.7 �11,312.7 200 43,057.5 17,784.2 19.201.8 �12,543.6

� 2� 1 � �

� 1 W

2� � W� E2� � �E

250 23,995.2 �3,389.1 183,913.7 46,790.6 �131.746 300 5,968.2 �23.607.6 179,009.1 63,288.2 �174.799 350 �11,023.5 �42.871.2

Where, Wj, j = �, �, 2�, mon is the social welfare at the respective equilibrium. Social welfare is defined as the summation of consumer and producer surplus: Wj = uj(qj1 + q

j 2 ) �

P2 i=1

TC j i , where u

j is utility derived from the consumption of good q, and P Tc

j i is aggregate

cost of producing Q. The superscript 2� denotes values of the variables in the case that entry of firm 3 is deterred and the two incumbent firms are price takers at permit price P�.

of Table V. Raising rivals’ cost strategies may blockade the entry of even equally efficient firms, but at the same time ameliorate social welfare and environmental quality. Policy makers should take into consideration the welfare effects of raising rivals’ cost strategies before intervening to facilitate entry.

Finally, since the initial permits distribution affects the leader’s ability to manip- ulate the permit price, it also affects its ability to deter entry. Should policy makers change the initial permit distribution in order to facilitate entry? To answer this question, we compare the case that permits are equally distributed between the two firms to the case of auctioning. The latter case, i.e. zero initial endowment of permits for all firms, is presented in Table VI. Although incumbent firms have no advantage over potential entrants, entry deterrence is still feasible, but only when the leader is more efficient in abatement. Auctioning of permits reduces the leader’s ability to raise the permit price. Thus, permits allocation is an efficient instrument in controlling the leader’s power and facilitating entry. However, facilitating entry does not necessarily increase social welfare. This point is clearly demonstrated in the case that all three firms are equally efficient in abatement (bolded row in Tables V and VI). When �E1 = �E2 = �E=2 entry of firm 3 is deterred, but social welfare is larger relatively to the case that permits are auctioned, �E1 = �E2 = 0 and entry is facilitated. Changing the mode of permits allocation from grandfather- ing to auctioning, increases competitiveness in the product market but decreases social welfare. Although not shown in our simulations, permits distributions that could increase both competitiveness and social welfare exist. However, the spec-

ceiling �E, because P� > Pm implies that aggregate abatement is higher than what is required by the emissions ceiling.

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 763

Table VI. Effect of entry deterrence on social welfare. Auctioning of permits: ( �Ei = 0)

d2 = d3 (d1 = 250) � � 3 , �

� 2 �

� 3 , �

� 2 �

� 1 � �

� 1 W

� � W�

0 129,659.58 122,331.15 1,834.8 �12,822.1 50 106,456.17 97,647.42 2,480.9 �15,136.7

100 84,288.04 73,918.56 3,224.2 �17,514.7 150 63,155.17 51,144.56 4,064.9 �19,956.4 200 43,057.58 29,325.42 5,002.8 �22,461.5 250 23,995.27 8,461.15 6,037.9 �25,030.3

� 2� 1 � �

� 1 W

mon � W�

300 5,968.23 �11,448.25 652,130.7 �11,009.1 350 �11,023.54 �30,402.79

Where, Wj, j = �, �, 2�, mon is the social welfare at the respective equilibrium. Social welfare is defined as the summation of consumer and producer surplus: Wj = uj(q

j 1 + q

j 2 ) �

P2 i=1

TC j i , where u

j is utility derived from the consumption of good q, and

P TC

j i is aggregate cost of producing Q. The superscript mon denotes values

of the variables in the case that entry of the firm 3 is deterred, firm 2 exits the industry and firm 1 monopolizes the product market.

ification of such distributions requires a vast amount of information on the part of policy makers. Intervention in the case of incomplete information can be detri- mental to both competition and welfare. An example is presented in row seven (d2 = d3 = 300, d1 = 250) of Table VI. Firm 2 that has zero endowment of permits makes negative profits under leadership and exits the market. The leader buys all permits at price P�, and monopolizes the product market. The equilibrium is denoted by a superscript mon. Both the level of competition and social welfare decrease. Thus, informational requirements and the ambiguity of the welfare effect render policy intervention problematic.

V. Conclusions

We examine cases in which emission permits markets are used as vehicles in lessening competition in the product market. We focus our attention on moder- ate cases of predatory behaviour in the form of raising rivals’ cost strategies. In these cases, the leader in the permits market employs positioning rather than exclusionary strategies; i.e does not attempt to exclude existing rivals. Positioning strategies though, can lead to entry deterrence under certain conditions. We find that although in many instances competition is lessened substantially, overall effi- ciency may increase when the market share of the more efficient firms increase. However, there are situations that should be examined closely by policy makers. For example, the case of a large entrenched inefficient firm that overbuys permits calls for policy intervention. According to our analysis this is a case of an inefficient leader and can lead to large decreases in social welfare. In such cases as well as

764 EFTICHIOS SOPHOCLES SARTZETAKIS

when consumers’ protection is a policy priority, policy makers can use the initial distribution of permits to control power in the permits market. Such interventions though can succeed only when policy makers have substantial information on the technological structure of the industry, and thus, should be used with caution. Given the importance of information, sharing of information and coordination of actions between policy makers (environmental and competition policy agencies, or any other type of regulators involved) is very important.

Appendix A. Notation of the Model Used in the Simulations

Inverse demand: p = a � b(q1 + q2), where qi, i = 1, 2, is firm i’s output. Cost of production: CPi = qi +�q

2 i , where c, � � 0 are technological parame-

ters and k = k1 = k2. Emissions generation: ei = �iqi, where � is the rate of emission per unit of

output, with �1 = �2. Cost of abatement: CAi = d1�iqi + e�

2 i q2 i , where � is the abatement per unit

of output, and d, e � 0 represent technological parameters, with e1 = e2 = e. Supply of emission permits: �E = �Q̂, where Q = q̂1 + q̂, is industry’s pre-

regulation level of output, and � 1. Methods of permits distribution examined: (1) according to firms’ pre-regulation

level of emissions: �Ei = �qi; (2) equiproportional: �Ei = �E=2; (3) auctioning: �Ei = 0; where �Ei is firm i’s endowment of permits.

Appendix B. Initial Values of the Parameters Used in the Simulations

a = 1,500; b = 0.20; k = 0.07; e = 0.30; � = 0.60; = 0.40; d1 = d2 = 250; c1 = c2 = 200; and F1 = F2 = 333:000. We start with the assumption that both the leader and the price taker have identical production and abatement technologies, in order to demonstrate that leadership in the permits market translates into raising rivals’ cost strategies without requiring technological asymmetries. We then allow the technological parameters to vary. Technological parameters and the required percentage reduction in emissions are chosen such that the cost of compliance is a significant part of firms’ total cost.

References

Borenstein, S. (1988) ‘On the Efficiency of Competitive Markets for Operating Licenses’, Quarterly Journal of Economics, 103, 357–387.

Brown-Kruse, J., S. R. Elliot, and R. Godby (1995) ‘Strategic Manipulation of Pollution Permit Markets: An Experimental Approach’, Paper presented at the Canadian Economic Association Meetings, Montreal, Québec, June 1995.

Coate, M. B., and A. N. Kleit (1994) ‘Exclusion, Collusion, or Confusion?: The Underpinnings of Raising Rivals’ Costs’, Research in Law and Economics, 16, 73–93.

Dixit, A. K., and J. E. Stiglitz (1977) ‘Monopolistic Competition and Optimal Product Diversity’, American Economic Review, 67, 297–308.

COSTS STRATEGIES VIA EMISSION PERMITS MARKETS 765

Fershtman, C., and A. de Zeeuw (1996) ‘Tradeable Emission Permits in Oligopoly’, Working Paper, Department of Economics, Tilburg University.

Gaudet, G., and N. V. Long (1993) ‘Vertical Integration, Foreclosure and Profits in the Presence of Double Marginalization’, Working Paper, Université du Québec à Montreal.

Hahn, R. W. (1984) ‘Market Power and Transferable Property Rights’, Quarterly Journal of Eco- nomics, 99, 753–765.

Hahn, R. W. (1989) ‘Economic Prescriptions for Environmental Problems: How the Patient Followed the Doctor’s Orders’, Journal of Economic Perspectives, 3, 95–114.

Hanley, N., and I. Moffat (1992) ‘Efficiency and Distributional Aspects of Market Mechanisms in the Control of Pollution: An Empirical Analysis’, Working Paper, Department of Economics, University of Stirling.

Hart, O., and J. Tirole (1990) ‘Vertical Integration and Market Foreclosure’, Brookings Papers on Economic Activity: Microeconomics, 4, 205–286.

Hung, M. N., and E. S. Sartzetakis (1994) ‘Is Tradeable Emission Permits Systems Superior to Command and Control?’, Working Paper, GREEN, Université Laval, Cahier 94-13.

Krattenmaker, T. G., and S. C. Salop (1985) ‘Anticompetitive Exclusion: Raising Rivals’ Costs to Achieve Power over Price’, Yale Law Journal, 96, 209–295.

Krattenmaker, T. G., and S. C. Salop (1986) ‘Competition and Cooperation in the Market to Exclu- sionary Rights’, American Economic Review, Papers and Proceedings, 16, 109–113.

Krattenmaker, T. G., and S. C. Salop (1987) ‘Analysing Anticompetitive Exclusion’, Antitrust Law Journal, 56, 71–89.

Malueg, D. A. (1990) ‘Welfare Consequences of Emission Credit Trading Programs’, Journal of Environmental Economics and Management, 17, 66–77.

Misiolek,W. S., and W. Elder (1987) ‘Exclusionary Manipulation of Markets for Pollution Rights’, Journal of Environmental Economics and Management, 16, 156–166.

Ordover, J. A., G. Saloner, and S. C. Salop (1990) ‘Equilibrium Vertical Foreclosure’, American Economic Review, 80, 127–142.

Requate, T. (1993) ‘Pollution Control in a Cournot Duopoly Via Taxes or Permits’, Journal of Economics, 58, 255–291.

Salop, S. C., and D. T. Scheffman (1981) ‘Strategic Interaction in Multiple Markets: A Beginning to a General Theory of Dominant Firm Industries’, Mimeo. Presented at the Summer Meetings of the Econometric Society. San Diego, June 1981.

Salop, S. C., and D. T. Scheffman (1983) ‘Raising Rivals’ Costs’, American Economic Review, 73, 267–271.

Salop, S. C., and D. T. Scheffman (1987) ‘Cost-Raising Strategies’, Journal of Industrial Economics, 26, 19–34.

Sartzetakis, E. S. (1993) Emission Permit Trading and Market Structure, Ph.D. Dissertation, Depart- ment of Economics, Carleton University, Ottawa, Canada.

Sartzetakis, E. S. (1996a) ‘Emissions Permits Trading as Vehicle for Raising Rivals’ Costs Strategies’, Working Paper, Fondazione Eni Enrico Mattei, Milano, Italy, Nota di Lavoro 59.96.

Sartzetakis, E. S. (1996b) ‘Interaction of Competitive Markets for Tradeable Emission Permits with Oligopolistic Product Markets’, Environmental and Resource Economics, forthcoming.

Spence, A. M. (1976) ‘Product Selection, Fixed Costs and Monopolistic Competition’, Review of Economic Studies, 43, 217–236.

Tietenberg, T. H. (1985) Emissions Trading: An Exercise in Performing Pollution Policy. Washington DC: Resources for the Future, Inc.

Tietenberg, T. H. (1989) Environmental and Resource Economics, Harper Collins Publishers, Second Edition.

von de Fehr, N.-H. M. (1993) ‘Tradeable Emission Rights and Strategic Interaction’, Environmental and Resource Economics, 3, 129–151.

Ware, R. (1994) ‘Understanding Raising Rivals’ Costs: A Canadian Perspective’, Canadian Compe- tition Record, March 1994, pp. 9–17.