Abstract
This study examines the effect of waste and recycling policy on scrap prices and the importance of scrap price feedbacks as a determinant of policy costs. Price effects and direct and indirect channels of waste reduction are decomposed for deposit/refund, advance disposal fee, and recycling subsidies. Scrap price feedbacks decrease the cost of advance disposal fees, increase the cost of recycling subsidies, and have an ambiguous effect on the cost of deposit/refund. Simulation analysis finds that scrap price feedbacks substantially affect the costs of the policies and alter the ranking of instruments. (JEL Q53, Q58)
I. Introduction
The world has long engaged in private recycling activities, driven by the trade-off between the value of scrap materials and the costs associated with recycling. Only recently, in the past several decades, has recycling become a matter of public policy, due to concerns of declining landfill availability and upstream and downstream externalities in production and disposal processes. Efforts to increase recycling rates above private levels reflect a desire to internalize these costs. Importantly, however, these policy interventions may in turn affect equilibrium scrap prices, which could have important consequences for the cost of policy interventions in both waste and recycling markets. Thus, as international scrap markets become increasingly integrated (Lyons, Rice, and Wachal 2009), whether policymakers face endogenous or exogenous scrap prices may then affect the policy instruments selected and levels of those instruments.
A substantial theoretical and empirical literature has arisen examining the effectiveness and efficiency of waste and recycling policies (Dinan 1993; Fullerton and Kinnaman 1995;Sigman 1995; Palmer and Walls 1997; Walls and Palmer 2001; Kinnaman 2006; Ino 2011). The literature has generally identified two key channels of waste reduction: source reductionand increased recycling. By exploiting both channels of waste reduction, policies such as a deposit/refund achieve a given reduction in total waste at a lower cost than policies that exploit only one channel, such as advance disposal fees or recycling subsidies. It is typically assumed that the policymaker operates in a closed economy, such that scrap prices adjust to clear recycling markets.1 This also implies that scrap price is endogenously determined by the level of policy interventions, and thus an indirect effect of these policy instruments on waste and recycling levels also exists via changes in scrap prices. However, the literature does not provide much insight into the role these price feedbacks play in terms of the costs and rankings of policy instruments, nor guidance into how policymakers should choose instruments or instrument levels when faced with exogenous scrap prices.
There is evidence that scrap markets are becoming increasingly integrated at the global level, straining the closed economy (and thus endogenous scrap price) assumption noted above. Lyons, Rice, and Wachal (2009) characterize the substantial international trade in scrap paper, plastic, steel, and aluminum in the United States, noting a substantial increase in the globalization of trade in scrap products. For example, in 2005, nearly 25% of waste paper and aluminum produced in the United States was exported to over 160 countries, an increase from the 1990s when roughly 15% of waste paper and less than 10% of aluminum scrap was exported. The authors note similar trends in the integration of scrap markets around the world, particularly in Asia, where trade in secondary scrap exceeds trade in primary commodities.2 Thus, from the perspective of policymakers in small, open markets at the local, state, and even national level, their policy decisions are unlikely to affect international scrap prices. The contribution of this paper is to highlight the role scrap price feedbacks play in determining the costs of various waste and recycling policies, and to provide policy insight in an increasingly integrated world scrap market.
This paper explores the following three questions. First, what is the impact of waste and recycling policies on scrap prices? Second, how does the change in scrap prices affect waste and recycling levels? Finally, how do these scrap price feedbacks affect the cost of waste and recycling policies? In other words, given that policymakers are increasingly likely to face exogenous world scrap prices, how does this affect the costs of waste and recycling policies and thus the choice of waste and recycling instruments and instrument levels? In the analytical derivations below, the effect of each policy instrument on the price of scrap is derived, and the direct and indirect channels of waste reduction are decomposed for three instruments: deposit/refund, advance disposal fees, and recycling subsidies. In the simulation analysis, the costs of the three instruments with endogenous scrap prices are compared to the costs of the policy instruments with exogenous prices.
Palmer, Sigman, and Walls's (1997) model provides an excellent framework for considering the importance of scrap price feedbacks both analytically and numerically. Using a calibrated mass-balance model of recycling and waste levels in the United States in 1990, they compare the costs of three price-based waste and recycling policy instruments: deposit/refund, advance disposal fee, and recycling subsidies. Several key results emerge from their analysis. First, deposit/refund is the least-cost instrument for waste reduction ($45 per ton for a 10% reduction in waste), followed by the advance disposal fee ($85 per ton) and recycling subsidies ($98 per ton). Second, they find that flexible deposit/refunds (in the sense that the deposit/refund is applied uniformly to all materials) achieve a 10% reduction in total waste at a substantially lower cost compared to the case where each material must be reduced by 10%.3 Finally, by comparing the marginal cost of deposit/refund with the marginal social damages of disposal, they suggest that a 7.5% reduction in total waste would have been justified. Following the rest of the literature, the above results assume a closed economy with endogenous scrap prices calibrated at the national level.4 As noted above, this endogenous scrap price assumption may not necessarily hold at the national level given the increased integration of world scrap markets, and it is very unlikely to hold at the smaller jurisdictional scale, where many waste and recycling policies are actually made.
The analytical model finds that advance disposal fees increase scrap prices and recycling subsidies decrease scrap prices. However, the effect of the deposit/refund is ambiguous: though the deposit decreases consumption and pushes up scrap prices, the refund increases the recycling rate and decreases scrap prices. The decomposition of the direct and indirect channels of waste reduction shows that the indirect scrap price feedbacks generate indirect source reduction and increased recycling effects for all three instruments. These indirect effects have opposing signs, implying that, analytically, the effect of price feedbacks has an indeterminate effect on the costs of the policy instruments. The simulation analysis addresses the numerical magnitudes of these effects and finds that indirect scrap price effects increase the cost of a 10% total waste reduction with a deposit/refund by 30%. The cost of the advance disposal fee falls by 10%, while the recycling subsidy is over 50% more costly due to the indirect effects. As a result, when scrap prices are exogenous, the instrument ranking is altered, with a 10% reduction in total waste costing $36 per ton under deposit/refund, $66 per ton under a recycling subsidy, and $96 per ton under an advance disposal fee. At a marginal social damage of $33 per ton for waste disposal (as used by Palmer, Sigman, and Walls [1997]), the optimal reduction in waste is roughly one-third larger when scrap prices are exogenous (10% compared to 7.5%). Thus, policymakers in small, open markets should be modestly more aggressive in their waste reduction targets.
II. Model Of Waste And Recycling
To begin, the model adopted by Palmer, Sigman, and Walls (1997) is reviewed. The basic mass-balance model is described by the following system of equations:
[1]
[2]
[3]
where W is disposed waste, Q is total consumption, and R is the amount recycled. The mass-balance equation [1] requires that all consumption is either disposed of as waste, or is recycled. Per Palmer, Sigman, and Walls (1997), supply of the final product Q is assumed to be perfectly elastic, while demand for the final product Q varies with the price of the final product pq, and if recycled, with the price net of scrap value pq − pr.5 The supply of total scrap goods varies with the scrap price (pr) and is equal to the recycling rate r(pr) times total consumption. Finally, demand for scrap goods is assumed to also vary with pr, such that 
. Thus, per equation [3], the market for recycled scrap clears at endogenous scrap price pr. The following intuitive assumptions hold: 
, 
, 
and 
.
Several assumptions of the original study should be noted. It is assumed that markets for the final material and recycled scrap are perfectly competitive. It is also assumed that there are no lags between when the material is purchased and when it is disposed of or recycled. Another important assumption is that the quantity of the consumption good Q does not affect the demand for recycled scrap Rd(pr). The final important assumption is that demand for the consumption good depends only on own price, effectively setting the cross-price elasticity across materials equal to zero. Realistically, increases in the price of one material would likely lead to substitution by other materials. While relaxing these assumptions would alter the quantitative results of Palmer, Sigman, and Walls (1997) and this study, they are unlikely to alter the qualitative comparisons between the two studies.
III. Policy Intervention And Price Feedbacks
Three price-based instruments—deposit/refund, advance disposal fees, and recycling subsidies—are examined. As in Palmer, Sigman, and Walls's (1997) study, it is assumed these policies are implemented at the producer level, and that the incentives provided by these policies are passed on to consumers.
The literature has identified two channels that can reduce total waste disposal: source reduction and increased recycling. Source reduction operates through reductions in consumption, or in terms of the model, by reducing D(pq, pq − pr). The recycling channel operates through increasing the recycling rate, r(pr). Palmer, Sigman, and Walls's (1997), discussion of these two channels focuses on the direct effects of the various policy instruments. However, through equation [3], the scrap price pr is endogenous and will thus be affected by the policy instruments. The derivations below determine the effect of the policy instruments on scrap prices and analytically separate the direct and indirect policy effects on waste reduction. Because the indirect effects arise through changes in scrap prices, they will occur only when scrap prices are endogenous. The magnitudes of these various effects are then examined in the simulation analysis to follow.
Advance Disposal Fee
Consider an advance disposal fee of f per ton. Consumers face this fee regardless if they choose to recycle, such that consumption prices for nonrecycling consumers are given by pq + f and pq + f − pr for recycling consumers. Equilibrium in the scrap market under this policy is given by
[4]
The effect of the advance disposal fee on scrap prices can be derived from the implicit function theorem:
[5]
The advance disposal fee discourages consumption, which reduces the quantity of scrap available and pushes up prices. The full impact of an advance disposal fee on waste can be decomposed into direct and indirect effects as follows:
[6]
where the signs indicate whether the effect increases or decreases waste. The advance disposal fee exploits one direct channel of waste reduction, through the source reduction (SR) effect. Though the policy does not directly target recycling rates, it does reduce consumption and the availability of recycled material, leading to price increases and stimulating the recycling rate through the indirect increased recycling (IR) effect. At the same time, the increased scrap price encourages consumption, increasing waste produced through the negative indirect SR effect.6 Thus, the magnitude of these countervailing indirect effects will determine whether waste reduction increases or decreases relative to the direct SR effect. In terms of the costs of the policy, if the indirect IR effect dominates the indirect SR effect, a smaller advance disposal fee would be needed to achieve a given waste reduction target relative to the case where scrap price is exogenous.
Recycling Subsidy
Next, consider a recycling subsidy of s per ton. Consumers who recycle receive a price of pr + s, and thus the effective price of consumption is pq − pr − s. Equilibrium in the scrap market under this policy is given by
[7]
The effect of the recycling subsidy on scrap price can be derived from the implicit function theorem:
[8]
The recycling subsidy encourages recycling, which increases the quantity of recycled scrap available and pushes down prices. The full impact of a recycling subsidy on waste can be decomposed into direct and indirect effects as follows:
[9]
Under a recycling subsidy, waste is reduced through the direct IR effect. However, by subsidizing recycling, the effective price of the consumption good for consumers who recycle is reduced, increasing consumption and waste through the negative direct SR effect. Two additional indirect channels of waste reduction are also generated by the recycling subsidy. Due to the decrease in scrap prices, recycling rates fall and waste increases via the indirect negative IR effect. On the other hand, lower scrap prices also raise the effective consumption good price for recycling consumers, decreasing consumption and waste through the indirect SR effect. Again, the relative magnitudes of these two indirect effects will determine how the cost of recycling subsidies with endogenous scrap prices compares to the case when prices are exogenous.
Deposit/Refund
Finally, consider a deposit/refund of d per ton. Consumers who fail to recycle experience a price of pq + d, while those who recycle the product receive the refund back, offsetting the deposit. Equilibrium in the scrap market under this policy is given by
[10]
The effect of the deposit/refund on scrap price can be derived from the implicit function theorem:
[11]
In contrast with the previous instruments, the effect of a deposit/refund on scrap price cannot be signed a priori. While the deposit reduces consumption, pushing up scrap prices, the refund encourages recycling, pushing scrap prices down.
The full impact of a deposit/refund on waste can be decomposed into direct and indirect effects as follows:
[12]
The direct effects of the deposit/refund are clear: the refund increases the recycling rate, leading to waste reduction through the direct IR effect, and the deposit increases the cost of the consumption good, decreasing consumption and waste through the direct SR effect. However, due to the ambiguity of the effect of deposit/refund on scrap prices (equation [11]), the indirect effects are correspondingly less clear. If scrap price increases, dpr/dd > 0, then the indirect IR effect will reduce waste, while the indirect negative SR effect will increase waste.7 If the scrap price falls, dpr/dd < 0, then the indirect effects flip signs. Thus, the change in the costs of the deposit/refund policy due to scrap price effects will depend both on the direction of the price change and the absolute magnitudes of the indirect effects.
IV. Simulation Results
The previous section decomposed the direct and indirect effects of deposit/refund, advance disposal fees, and recycling subsidies on waste. In all cases, the indirect source reduction and increased recycling effects had opposite signs, and as such, the impact of scrap price feedbacks on the costs of the policies is ambiguous. To resolve these effects empirically, the simulation model presented by Palmer, Sigman, and Walls (1997) is used to compare the costs of the policies when scrap price is exogenous versus endogenous. In the simulation analysis to follow, the necessary level of the policy intervention for each instrument is determined in order to achieve a specific reduction in total waste.8 To ensure the accuracy of the comparison with Palmer, Sigman, and Walls's (1997) study, their simulation model with endogenous prices was recreated and checked against their published results.
Percentage Change in Scrap Price of Aluminum Beverage Cans versus Policy Intervention Level
Scrap Price Response to Policies
In the preceding section, it was shown that the advance disposal fee pushes up scrap prices, recycling subsidies push down scrap prices, and the effect of the deposit/refund is ambiguous. To determine the magnitude of these changes in scrap price and examine the change in scrap price under deposit/refund, as an illustrative example, consider the change in the price of a common recycled household good: aluminum beverage cans.
Figure 1 plots the change in aluminum beverage can scrap price under each of the three instruments for varying intervention levels. As expected, scrap prices rise under an advance disposal fee and fall under a recycling subsidy. At an intervention level of $100 per ton (roughly the cost for each instrument to achieve a 10% reduction in waste), the advance disposal fee increases scrap price by 2%, while the recycling subsidy pushes down scrap prices by 6%. While change in price under deposit/refund was ambiguous in the analytical decomposition, here the deposit/refund pushes prices down by 3%, implying that the refund is outweighing the deposit in terms of changing the quantity of scrap material available and thus the price. The other materials under consideration show similar trends to those for aluminum cans: advance disposal fees increase scrap prices a modest amount, recycling subsidies decrease scrap prices by a more substantial amount, and for all materials, deposit/refund decreases scrap prices.
Marginal Cost of the Deposit/Refund, Advance Disposal Fee, and Recycling Subsidy Necessary to Achieve Various Percentage Waste Reductions When Scrap Prices Are Exogenous
Policy Intervention Levels Under Exogenous Prices
Figure 2 displays the policy intervention levels required to achieve a given percentage reduction in total waste when scrap price pr is exogenous. The ranking of the instruments is clear: deposit/refund is the least-cost policy at all levels of waste reduction, followed by the recycling subsidy, and finally the advance disposal fee. This stands in contrast to the instrument ranking with endogenous prices, where the advance disposal fee was less costly than the recycling subsidy. A 10% reduction in total waste is achieved with a $36 per ton deposit/refund, a $66 per ton recycling subsidy, and a $96 per ton advance disposal fee.
Comparing these results with exogenous scrap prices against the results of Palmer, Sigman, and Walls (1997) with endogenous scrap prices, the cost of the deposit/refund and recycling subsidy have decreased, and the cost of the advance disposal fee has increased. The differences in costs for the advance disposal fee and recycling subsidy are easy to explain. From equation [6], the fact that the cost of the advance disposal fee is higher with exogenous scrap prices implies that the indirect increased recycling effect dominates the indirect negative source reduction effect, leading to a decrease in waste due to price feedbacks for a given advance disposal fee level.9 As a result, a smaller advance disposal fee is required to achieve a given waste reduction when prices are endogenous. Similarly, from equation [9], the substantial decrease in the cost of the recycling subsidy under exogenous prices also implies that the indirect recycling effects are larger than the indirect source effects, but in this case, the indirect increased recycling effect is negative (as scrap prices fall, recycling rates fall, implying an increase in waste). Thus, the recycling subsidy must be increased when prices are endogenous in order to counteract the waste increases from scrap price feedbacks.
Cost Ratio of the Deposit/Refund, Advance Disposal Fee, and Recycling Subsidy for a Given Reduction in Waste
The impact of price feedbacks on the deposit/refund is less clear. As noted in equation [12], the indirect effects have opposing signs and themselves vary in sign depending on the effect of the deposit/refund on scrap price (equation [11]). The simulation results find that for a 10% reduction in total waste, the change in price due to the deposit/refund is negative dpr/dd < 0 for all materials. Therefore, from equation [12] and the fact that the cost of the deposit/refund falls under exogenous prices, one can conclude that the indirect price feedbacks are such that the indirect negative increased recycling effect dominates the indirect source reduction effect. As a result, when scrap prices are endogenous, a larger deposit/refund is needed to offset the scrap price-induced waste increases in order to achieve a given reduction in total waste.
Magnitudes of the Direct and Indirect Effects
Above, the net direction of the indirect effects was deduced from the changes in the costs of the instruments. In the analysis that follows, the magnitudes of the direct and indirect effects are considered in more detail. Figure 3 displays the ratio of intervention levels for each instrument against the percentage change in total waste reduction, where the numerator is the intervention level under endogenous prices (indirect and direct effects) and the denominator is the intervention level under exogenous prices (direct effects only). A value greater than one indicates that the indirect effects increase the required intervention level (price feedbacks increase waste), while a value less than one indicates that the indirect effects decrease the required intervention level (price feedbacks decrease waste).
The ratio for the advance disposal fee is invariant across waste reductions and is roughly 0.9. Thus, price feedbacks lead to a modest 10% reduction in the marginal cost per ton of the advance disposal fee. By contrast, the costs of the deposit/refund and recycling subsidies are increased by price feedbacks, with a more pronounced effect at lower levels of total waste reduction. For the deposit/refund, the indirect effects are initially 50% as large as the direct effects but decline to roughly 20% as total waste reductions increase. The effects of price feedbacks are even larger for the recycling subsidy: the indirect effects are initially 140% as large as the direct effects, before declining to roughly 30%. Thus, indirect price effects do play a substantial role in the costs of the instruments, particularly for deposit/refund and recycling subsidies.
Direct and Indirect Channels of Waste Reduction: 10% Total Reduction in Waste
To further explore the importance of indirect price effects, the magnitudes of the direct and indirect channels of waste reduction (measured in million tons of material) are broken down by material and policy in Table 1 for a 10% reduction in total waste.10 While the terminology used in Table 1 is similar to that used in Palmer, Sigman, and Walls's (1997) Table III, there is an important distinction to highlight. Palmer, Sigman, and Walls (1997) calculate source reduction and increased recycling as simply the change in total consumption (ΔQ) and total recycling (ΔR). However, because total recycling is a function of total consumption, these calculations are difficult to interpret. For example, while a recycling subsidy increases total recycling by 6 million tons, one cannot determine how much of that increase is due to increases in recycling rates compared to how much of that increase is driven by increases in consumption. Thus, in the table presented here, the terms source reduction and increased recycling refer to the decompositions derived in Section III: source reduction is the change in waste solely due to changes in consumption, while increased recycling is the change in waste solely due to changes in the recycling rate.11 These definitions have been adopted to more closely match the analytical derivations and more accurately reflect the channels of behavioral change in response to the policies.
Across policies in Table 1, the magnitude of physical changes in materials driven by the indirect channels provides insight into the cost ratios of the policies discussed above. Per the analytical decomposition, the advance disposal fee has no direct effect on recycling rates, and thus the only direct effect of the policy is through source reduction. The indirect increased recycling effect dominates the indirect source reduction effect, which drives the previously discussed result that the cost of the advance disposal fee falls due to price feedbacks. However, the magnitude of the net indirect effects is relatively small, leading to the modest 10% reduction in the costs of the advance disposal fee when prices are endogenous. For the recycling subsidy, the reductions in waste from the direct increased recycling effect are substantially offset by large increases in waste from the indirect increased recycling effect, particularly for steel, increasing the cost of recycling subsidies when price feedbacks are present. A similar story arises for the deposit/refund, where the indirect recycling effects erode the waste reductions from the direct increased recycling effect. Furthermore, under deposit/refund the direct recycling channels are typically larger than the direct source reduction channels (except for plastics). Across instruments, the indirect effects are, in general, smaller than the direct effects but, nonetheless, are substantial in magnitude.
Updated Simulation Results
The analysis in the previous sections relied on 1990 data from Palmer, Sigman, and Walls (1997) for comparison purposes. In this section, prices and quantities are updated to 2010 levels using data from Acuff and Kaffine (2013). Unfortunately, updated elasticity estimates are unavailable, and therefore sensitivity analysis is conducted on the demand and supply elasticities for recycled goods. Qualitatively, the 2010 results are consistent with the 1990 results, in that scrap prices fall under the deposit/refund and recycling subsidy but rise under the advance disposal fee. Thus, the indirect effects from scrap price feedbacks raise the cost of the deposit/refund and recycling subsidy but lower the cost of the advance disposal fee.
Relative to 1990, the most important quantitative change in the price and quantity data is the increased recycling rate for all materials except aluminum. As expected, this increases the cost of achieving waste reductions via a recycling subsidy, which primarily exploits the increased recycling channel, as larger interventions are required to achieve a given waste reduction. After adjusting for inflation, the cost of achieving a 10% reduction in waste with the recycling subsidy increases by roughly 60% under both exogenous and endogenous scrap prices. By contrast, the cost of the deposit/refund has roughly kept parity with inflation, despite its reliance in part on the increased recycling channel. This is consistent with the fact that the cost of the advance disposal fee has fallen by roughly 30% with both exogenous and endogenous scrap prices, suggesting the cost of achieving waste reductions through the source reduction channel has fallen since 1990.
Next, a sensitivity analysis on the demand and supply elasticities for recycled goods is conducted. The aim here is to model potential across-the-board changes in elasticities and ascertain whether indirect effects through scrap price feedbacks continue to play an important role currently. Each recycled good demand elasticity given by Palmer, Sigman, and Walls (1997) is scaled up or down, and the cost ratio (cost with endogenous prices over cost with exogenous prices, as in Figure 3) is calculated for each instrument for a 10% reduction in waste at 2010 prices and quantities.
Figure 4 plots this cost ratio as a function of the scale parameter for each instrument. A scale parameter of one corresponds to the original elasticities used in the previous analysis, and Figure 4 shows that the importance of scrap price feedbacks has increased in 2010 for the recycling subsidy and the advance disposal fee. Scrap price feedbacks increase the cost of the recycling subsidy by 77% (up from 48%) and decrease the cost of the advance disposal fee by 20% (up from 10%). As expected, increasing the magnitude of the demand elasticities reduces the magnitude of scrap price changes and, thus, decreases the magnitude of the indirect effects. However, demand elasticities would have to increase substantially before indirect effects cease to be an important driver of the cost of policies. A similar exercise was conducted with the supply elasticities for recycled goods; however, in this case increasing the supply elasticities magnifies the importance of scrap price feedbacks. Again, even with substantial changes in the supply elasticities reported by Palmer, Sigman, and Walls (1997), scrap price feedbacks continue to play an important role.
Cost Ratio of the Deposit/Refund, Advance Disposal Fee, and Recycling Subsidy for a 10% Reduction in Waste as a Function of the Scaled Demand Elasticity for the Recycled Good
V. Conclusions
This study examines the effect of waste and recycling policies on scrap prices, and, consequently, the role of these price feedbacks as a determinate of the costs of waste and recycling policies. Though these feedbacks were present in an original study by Palmer, Sigman, and Walls (1997), their role was ultimately unremarked upon. In a deeper exploration of their model, the analytical derivations in this study find that advance disposal fees increase scrap prices and recycling subsidies decrease scrap prices, while the effect of the deposit/refund is ambiguous: though the deposit decreases consumption and increases scrap prices, the refund increases the recycling rate and decreases scrap prices. The decomposition of the direct and indirect channels of waste reduction shows that the indirect price feedbacks generate indirect source reduction and increased recycling effects for all three instruments. These indirect effects have opposing signs, implying that, analytically, the effect of price feedbacks has an indeterminate effect on the costs of the policy instruments.
The simulation analysis addresses the numerical magnitudes of these effects and finds that indirect price effects increase the cost of a 10% total waste reduction with a deposit/ refund by 30%. The cost of the advance disposal fee falls by 10%, while the recycling subsidy is over 50% more costly due to the indirect effects. As a result, when scrap prices are exogenous the instrument ranking is altered, with a 10% reduction in total waste costing $36 per ton under deposit/refund (versus $45), $66 per ton under a recycling subsidy (versus $98), and $96 per ton under an advance disposal fee (versus $85). At a marginal social damage of $33 per ton for waste disposal, the optimal reduction in waste is roughly one-third larger when scrap prices are exogenous (10% compared to 7.5%).
In addition to providing further insight into the behavioral channels that drive waste reduction and the corresponding costs of policy instruments, this study also may provide guidance for policymakers. While the model is calibrated for the national level, where scrap prices are likely endogenous, policy is typically made at subnational scales, where scrap prices are likely exogenous. Furthermore, increased integration in world scrap markets suggests that even at the national level, policymakers may increasingly face exogenous scrap prices. This paper finds that the absence of price feedbacks will substantially decrease the costs of recycling subsidies and deposit/ refund, which may make them more appealing at city and state levels than the original Palmer, Sigman, and Walls (1997) study would suggest. Of course, this also suggests that waste and recycling policy made in one jurisdiction can have spillover effects on the costs of policy in other jurisdictions, which may be an interesting topic for further exploration.
Acknowledgments
The Alcoa Foundation provided generous support. Rod Eggert, Saumya Rana, Beth Schmitt, John Tilton, Greg Wittbecker, Casey Wagner, and Brian Batson provided valuable feedback.
Footnotes
The author is associate professor, Department of Economics, University of Colorado, Boulder.
↵1 One exception is Copeland (1991), who considers the policy choices of a small, open economy facing exogenous world prices in waste disposal services, which, though not formally considered, could include recycling activities.
↵2 Markets within the United States are also reasonably integrated. Gruver and Giarratani (2005) examine the market for scrap steel in the United States, finding three or four large submarkets with relatively uniform prices.
↵3 This result is driven in part by the fact that, by assumption, each type of material has equal marginal social damage. Acuff and Kaffine (2013) show that a uniform deposit/refund is no longer the least-cost policy when material-specific upstream greenhouse gas externalities are considered.
↵4 It is also important to note that despite the presence of scrap price effects in the simulation model, their role is unremarked upon. For example, the reported recycling rates under an advance disposal fee increase, despite the lack of a direct channel affecting recycling markets.
↵5 There may also be a marginal effort cost or marginal psychic benefit associated with recycling the good, but as in Palmer, Sigman, and Walls's (1997) work, these considerations are abstracted from in this study. It should also be noted that the private price of waste disposal is assumed to be zero, and the demand for the final product is assumed to depend only on prices.
↵6 The use of the term negative to describe the indirect source reduction effect under the advance disposal fee is potentially confusing, as this effect actually increases waste. However, it is used to denote the fact that the change in waste is opposite the desired goal of decreased waste.
↵7 Note that the first sign on the indirect effects corresponds to this case.
↵8 The simulation model uses Palmer, Sigman, and Walls's 1990 data on baseline consumption and recycling by material, as well as own-price elasticities for demand and supply in the consumption and recycling markets. At the end of this section, updated price and quantity data from 2010 is used to verify the continued importance of scrap price feedbacks, and sensitivity analysis on the elasticities of the demand and supply elasticities for the recycled good is conducted.
↵9 Recall that the indirect negative source reduction effect implies an increase in waste.
↵10 Note that for each policy, the sum of all effects across all materials sums to 9.163 (million tons), corresponding to a 10% reduction in total waste.
↵11 One important caveat, helpfully pointed out by an anonymous referee, is that the elasticity estimates used for the calculations in Table 1 may include adjustments along the product-redesign margin (Calcott and Walls 2000). The incentive for firms to redesign their goods to decrease the use of materials will be diminished under local policies as opposed to national policies, which would imply the source reduction channel will be less effective.
