2nd World Congress of Environmental and Resource Economists
Monterey, California
24-27 June 2002
Tradable compliance credits for extended producer responsibility: market power and the allocation of initial property rights
Roger Salmons [1]
Department of Economics, University College London
Abstract
This paper considers how the allocation of initial property rights can affect the performance of a credit-trading scheme used to implement a recycling target under extended producer responsibility (EPR). The analysis is based on a simple, stylised model of the credit-trading scheme that has evolved in relation to extended producer responsibility for waste packaging in the United Kingdom. It is demonstrated that, when firms act strategically in one of the participating sectors, the choice of sector to receive the rights to the credits that are generated when waste material is reprocessed can affect the scheme’s cost efficiency and its environmental effectiveness. In doing so, the paper offers some new perspectives on the issue of market power in permit-trading systems, and provides guidance to waste regulators who may be considering the use of trading schemes in this policy area.
Keywords
Extended producer responsibility; Credit trading; Market power
1.Introduction
Waste management systems are typically subject to a number of market failures. In particular, municipal waste collection is often funded out of general taxation rather than by charging households directly for the service, while the gate-fees charged by landfill and incinerator operators do not properly incorporate the external costs of disposal. Together, these failures lead to inefficient levels for the amount of waste that is generated (too high), and for the amount that of material that is recovered from the waste stream (too low). In theory, these failures can be addressed directly – and efficiency restored – by a combination of externality charges applied to waste disposal, and unit-based pricing for household waste collection (Fullerton & Wu, 1998). However, there are a number of problems with this approach in practice. First, it may lead to an increase in the illegal dumping of waste by households (Fullerton & Kinnaman, 1996). Second, efficiency would require that the externality charges – and hence the collection charge – be differentiated between the various components of the waste stream, to reflect their respective toxicities (Dinan, 1993). Even if such as scheme were technically feasible, it would be prohibitively costly to administer.[2]
The potential limitations of the direct approach have led to various alternative policy responses being proposed, with the focus being shifted upstream to producers. These can be grouped under the collective heading of extended producer responsibility (EPR). Under EPR, producers are required to assume responsibility for the life-cycle environmental impacts of their products, with particular emphasis on disposal. Examples include the German Packaging Ordinance of 1991, and the 1993 Producer Responsibility initiative in the United Kingdom for six priority waste streams.[3] More recently, the EU Directive on End of Life Vehicles (2000/53/EC) has set recycling targets, and stipulated that producers must pay “all or a significant part” of the costs of take-back and treatment from January 2007.
Within the overall umbrella of EPR, a distinction can be made between price-based responses that combine some form of advanced disposal fee (including appropriate external costs) with a subsidy to recycling (Dinan, 1993; Palmer & Walls 1999), and quantity-based responses. Quantity-based approaches can take several forms, but there is commonly a requirement to recover (i.e. divert from the waste stream going to landfill) a given proportion of “end of life” products, or component materials.[4]
Whatever form they take, a common characteristic of these initiatives is a recognition that while individual producers should have an obligation to meet the cost of recovery, it is not economically efficient to require them to undertake the necessary recovery themselves. One common approach is to allow producers to discharge their individual obligations by joining a collective compliance scheme (or producer responsibility organisation). The scheme then assumes responsibility for ensuring that sufficient waste is collected and recovered to satisfy the aggregate obligation; with the cost of these activities funded by the membership fees. A good example of this approach is the Duales System Deutschland (DSD) for packaging in Germany. However, this is not the only approach that can be used. An alternative, which maintains the concept of individual responsibility, is to introduce a compliance credit trading scheme. Under this approach, “recovery credits” are created (on a one-for-one basis) for every product, or tonne of material, that is delivered to reprocessors, and producers are required to purchase sufficient credits to satisfy their individual obligations. An example of this trading approach has evolved over recent years in relation to waste packaging in the United Kingdom (Salmons, 2002a).[5]
This type of credit trading scheme has a number of important differences compared to traditional permit trading applications such as air pollution. In most traditional applications, the target market (i.e. the market in which the regulator wishes to intervene) is incomplete – i.e. the supply side of the market is missing. Furthermore, the underlying objective is usually expressed in absolute terms, which is translated into a fixed, ex ante “cap” on the total number of permits. In contrast, in a compliance credit trading scheme for EPR, the target market is complete (i.e. both the demand and supply sides exist), and the target is expressed in relative terms (i.e. as a recovery percentage). Consequently, the total number of credits is only determined ex post, as part of the market equilibrium.
The differences between the two types of application are illustrated in Figure 1, which compares the trading system structure of a typical traditional application of trading – SO2 emissions by electricity generators, with that of a typical EPR application – waste packaging. In the first case, the obligation to purchase permits is imposed on one of the sectors participating in the target market – i.e. the “market” for environmental services provided by the atmosphere.[6] Of course, in this application the supply sector is missing, and so the only option is to place the obligation on the electricity generators. However, there is a choice regarding the initial property rights to the permits. They can be granted either to the generators under some form of grandfathering (as is shown in Figure 1), or to the regulator for them to auction off. In contrast, in the EPR application, the obligation is imposed on a sector – the producers – that does not participate in the target market. Furthermore, because this market is complete, there is a choice over which sector is granted the initial property rights when the waste packaging is received by the reprocessors – i.e. the credits can be given either to the demand side (the reprocessors), or to the supply side (the waste collectors).
Salmons (2002b) has shown that compliance credit trading provides a cost-efficient implementation mechanism for a recovery target under EPR, provided that households (or producers) are charged for the private cost of disposal.[7] However, this conclusion relies on the assumption that all markets are perfectly competitive. The objective of this paper is to investigate the implications of relaxing this assumption. In particular, it considers whether the presence of market power in the target market has any implications for the choice of sector to be granted the initial property rights to the credits that are created.
There is a growing literature on the issue of market power in permit trading systems. Hahn (1984) has considered the implications of monopoly power in the permit market; while Mauleg (1990) and Sartzetakis (1997a) have investigated the interactions between a competitive permit market and an oligopolistic product market. Finally, Sartzetakis (1997b) and von der Fehr (1993) have analysed the situation in which there is strategic interaction in both the permit and the product markets. These all conclude that cost efficiency is likely to be undermined if firms can act strategically in the permit or product markets. However, these analyses have all been conducted in the context of traditional permit trading applications, in which the issue of market power in the target market does not arise, and in which the objective is absolute. Consequently, this analysis provides some new perspectives on the topic.
The issue is analysed using a simple, stylised model of EPR for a single packaging material, in which there is only one reprocessor, that can exercise monopoly (or monopsony) power in any market in which it operates. The model captures some of the salient features of the trading scheme that has evolved in the United Kingdom for waste packaging.[8] A detailed description of the model is provided in the next section. While the focus of the paper is on the implications of market power, it is instructive to compare this situation with the perfectly competitive case, in which all participants are price-takers. Consequently, the perfectly competitive case is considered briefly in section 3, before turning to the monopolistic case in section 4. In each case, the market equilibrium conditions are derived under the alternative property rights regimes, and these are then used to draw some conclusions regarding the relative values of the system variables. In section 5, a specific numerical example is used to illustrate the results, and to investigate some issues that cannot be determined analytically. Finally, the implications of the analysis are discussed in section 6.
2.Stylised model of credit trading system for EPR
In this simple, stylised model of a credit trading system used to implement EPR for a single packaging material, there are three sectors: a waste reprocessing sector; a waste collection sector; and a production sector – i.e. the producers of packaged goods. It is assumed that there is complete separation between the three sectors, i.e. the production sector does not collect any waste packaging, and the waste reprocessing sector has no obligations under the EPR regulations.
Figure 2 provides a schematic representation of the model, which is developed at the sector-level.[9] The tonnage of waste packaging diverted from the waste stream (w) is a function of the amount of capital devoted to waste diversion activities (kd). The resultant capital requirement function kd = g(w) is assumed to be convex, with g//(w) + w g///(w) 0.[10] Diverted waste packaging is transformed into reprocessed material according to the production function y=f(w,kr), where kr is the level of reprocessing capital. It is assumed that f(w,kr) is strictly concave with fkw = fwk > 0. The production sector’s inverse demand function for packaging (z) is denoted by Pz(z). It is assumed that Pz(z) is continuously differentiable, with Pz/(z) < 0, 2Pz/(z) + z Pz//(z) < 0, and ( Pz(z) + z Pz/(z) ) > pz.[11]
The target recovery rate for waste packaging is denoted by [0,1], and this is used to define a “recovery obligation” (in tonnes) for the production sector – based on the amount of packaging produced.[12] However, the sector is not required to undertake the necessary diversion of waste packaging itself (i.e. incur the cost of diversion directly). Rather it is required to purchase sufficient recovery credits (n) to satisfy its obligation, i.e. nz. Recovery credits may only be created – on a tonne-for-tonne basis – when the diverted waste packaging is received by the reprocessor, i.e. nw. Two alternative scenarios are considered regarding the assignment of the initial property rights to these credits. In the first scenario (A) the rights are assigned to the reprocessing sector; in the second scenario (B) they are granted to the waste collection sector.
The market prices of delivered waste (qw) and recovery credits (qn) are determined endogenously within the model, while the prices of capital (pk), packaging (pz), reprocessed material (py) and landfill (pl) are all assumed to be exogenous (i.e. fixed).
The model is static, with the values of all variables being determined simultaneously. Thus, the amount of waste packaging that is sent to landfill (l) by the waste collectors is equal to the difference between the amount that is produced and the amount that is diverted, i.e. l = z – w. The model is also open, in that it is assumed that the reprocessed material is not used for packaging.[13]
3.Market equilibrium under perfect competition
In this section the outcome is determined under the assumption of perfect competition in all markets – i.e. none of the sectors can exercise market power in any of the markets in which they operate.
Scenario A:Property rights assigned to the reprocessing sector
Given exogenous prices py, pk, pl, and pz, a simultaneous market equilibrium comprises prices qw* and qn*, and non-negative quantities y*, w*, n*, z*, kr* and kc*, such that these values solve the three sectors’ optimisation problems, and satisfy the respective market clearing conditions for diverted waste packaging and recovery credits. When the initial property rights are allocated to the reprocessing sector, the optimisation problems for the three sectors are:
ReprocessorsMaxpy f(wr,kr) + qnnr qwwr pkkr
s.t.nr - wr 0
CollectorsMaxqwwc pkkd+ plwc
s.t.g(wc) kd 0
Producers[14]Max pzz qnnp
s.t.z np 0
and the market clearing conditions for the two markets are:
wc wr=0
np nr=0
The Kuhn-Tucker sufficiency conditions are satisfied for all three optimization problems. Consequently, the Kuhn-Tucker first order conditions for the three problems, together with the two market clearing conditions, are necessary and sufficient for a unique simultaneous market equilibrium. Assuming an internal solution, it is straightforward to derive a set of market equilibrium conditions expressed solely in terms of the real variables and market prices (i.e. eliminate all shadow prices). These are given in the first column of Table 1.
There are several points to note from the market equilibrium conditions. First, the reprocessors equate the value of the marginal product of diverted waste packaging to the net price of diverted waste (i.e. after subtracting the value of the credit that is generated). Put another way, the waste packaging has value to the reprocessors both in terms of the material that they can recover, and in terms of the credit that they can sell. Second, all of the credits that are generated when the waste is received are sold to the producers (i.e. w* = n*).[15] Finally, the requirement for producers to purchase credits in order to meets their obligations has the effect of raising the price of packaging material, by an amount equal to the credit price (qn) multiplied by the target recovery rate (). Thus, from the producers’ perspective, the trading scheme is equivalent to a tax on packaging. Indeed the same overall outcome could be achieved if the government were to levy a tax on packaging material (z=qn), and use the revenues to subsidise the price of diverted waste packaging (w = -qn).[16]
The market equilibrium conditions yield the following expression for the inverse demand curve for credits; giving the amount that the producer would be willing to pay for a credit, conditional on particular values of and pz.
Qn(n; , pz)=
From this expression it is clear that the demand curve for credits will be downward sloping, and that if the performance target is set at the rate achieved pre-regulation (i.e. 0 = w0/z0) then qn = 0 when n = n0 = w0. It is also clear that the impact of a change in the target recovery rate () on the demand curve for credits will depend critically on the shape of the packaging demand curve. This is illustrated in Figure 3, which shows the derived demand curves for credits for two different functional forms for Pz(z), together with two possible supply curves for credits – a “high cost” curve S1, and a “low cost” curve S2. With a constant-elasticity demand curve for packaging, an increase in the performance target causes the demand curve for credits to shift upwards. Consequently, the equilibrium price of credits increases irrespective of the slope of the supply curve. In contrast, when the demand curve for packaging is linear, an increase in the target recovery rate causes the demand curve for credits to rotate. The impact on the price of credits now depends on the slope of the supply curve. When the supply curve is relatively flat (S2), the price of credits increases. However, when it is relatively steep (S1), the price of credits decreases as the performance target becomes more stringent!
Scenario B:Property rights assigned to the waste collection sector
When the initial property rights are allocated to the waste collection sector, the optimisation problem for the producers and the market clearing conditions are unchanged.[17] However, the optimisation problems for the reprocessors and the waste collectors are now:
ReprocessorsMaxpyy qwwr pkkr
s.t.y f(wr,kr) 0
CollectorsMaxqwwc + qnnc pkkd + plwc
s.t.g(wc) kd 0
nc wc 0
The derived market equilibrium conditions are given in the second column of Table 1. Comparing these equilibrium conditions with those derived for scenario A, it is clear that the equilibrium values of the physical variables and the price of credits will be the same in each case. However, the price of diverted waste packaging will differ between the two scenarios, with qw** = qw* qn*.[18] Thus, the difference between the price of delivered waste in the two cases is exactly equal to the credit price. The total cost of meeting the target recovery rate is the same in each case, as is the distribution of costs and benefits.
4.Market equilibrium with monopoly power
In this section it is assumed that, while waste collectors and producers are price-takers in any market in which they operate, there is a single reprocessor, which acts as a monopsonist in the market for diverted waste, and as a monopolist in market for recovery credits.
Scenario A:Property rights assigned to the reprocessing sector
When the initial property rights are allocated to the reprocessor, the optimisation problems for the waste collectors and the producers are the same as for scenario A under perfect competition, as are the market clearing conditions. However, the optimisation problem for the reprocessor is now:
ReprocessorMaxpyy + Qn(nr) nr Qw(wr) wr pkkr
s.t.y f(wr,kr) 0
nr - wr 0
Unlike the case of perfect competition, the second constraint may not always be binding in the solution, even with a positive price for recovery credits (i.e. it is possible to have n* < w*). As can be seen in Figure 4, for values of below a particular threshold value (), the reprocessor’s marginal revenue curve for credits will intersect the marginal cost curve to the left of the point n0 = w0 (i.e. the number of credits generated at the pre-regulation level of delivered waste). If this is the case, then the reprocessor can divorce its two decisions, and set a monopoly price for credits based on its zero marginal cost. In contrast, when the target recovery rate exceeds this threshold value, the reprocessor must make a joint decision. The value of will depend on the functional form of Qn(n), which in turn depends on the functional form of Pz(z), but it will always be strictly greater than the pre-regulation recovery rate (0).