A number of recent theoretical studies have shown that, in the presence of illegal disposal opportunities, an optimal policy for reducing solid waste is a combination of a final product tax and a recycling subsidyi.e., a deposit/refund system. A deposit/refund is equivalent to a charge on disposal: the deposit is refunded if the product is recycled, so the consumer only bears a cost if the product is discarded. As a result, the deposit/refund ensures that the least-cost method of reducing disposal is used, whether it be through source reduction or recycling. By contrast, other policies take advantage of only source reduction or only recycling as opportunities for waste reduction.
This study provides empirical estimates of the costs of deposit/refunds, along with advance disposal fees (ADFs), recycling subsidies, and recycling rate standards. The estimates are generated through calibration of a simple two-equation model of the supply and demand for secondary materials and a mass balance condition describing waste disposal. The calibration is performed using supply and demand elasticity estimates from the extant economics literature and 1990 baseline prices and quantities of materials in the municipal solid waste stream. These empirical estimates have two kinds of policy implications. First, they allow a comparison of the relative costs of the different approaches. Although previous research has shown that the deposit/refund provides the best incentives, other policies may be more politically or administratively feasible. Our comparison of the relative costs of the policies will allow policy- makers to assess the trade-off between political and administrative considerations and the incentives the policies create. Second, the paper evaluates the absolute magnitude of the costs of waste reductions. Thus, these costs may be compared with the benefits of waste reduction to decide if any policy intervention is desirable, and if so, how much waste should be reduced.
We find that, indeed, the deposit/refund is the least-cost approach. For a 10 percent reduction in waste disposal, the deposit/refund has a marginal cost of only $45 per ton of waste reduced. By comparison, the ADF at $85 per ton and the recycling subsidy at $98 per ton are almost twice as costly as the deposit/refund. Setting recycling rate standards to achieve a 10 percent reduction has a marginal cost of $61 per ton of waste reduced. The larger the desired percentage reduction in disposal, the greater the absolute divergence in costs between the deposit/refund and the other approaches. Even allowing for uncertainty in the elasticity estimates, through a Monte Carlo simulation approach, the deposit/refund is still found to be cheaper than the other three alternatives. A remaining concern, however, is the administrative costs of the deposit/refund. Existing estimates suggest that traditional "bottle-bill" states have high administrative costs--so high that the ADF, which is essentially a deposit without the refund, would appear to be cheaper. Estimates of the administrative costs of the more centralized California system, on the other hand, are low enough to make the deposit/refund still the preferred approach. Nonetheless, we feel that more research into administrative costs by material is needed.
We find that a policy that requires a uniform percentage reduction in each material in the solid waste stream is substantially more costly than an approach where materials are allowed to be reduced in the least-cost fashion. A 10 percent reduction in each material in the solid waste stream, for example, would be achieved with a set of deposit/refunds that cost $70 per ton of waste reduced.
We argue that the principal benefits of waste reduction are the avoided landfill disposal costs. In this case, we find that a deposit/refund of approximately $33 per ton would be optimali.e., would provide a reduction in waste disposal to the point where the marginal benefits of that reduction just equals the marginal costs. A $33 per ton deposit/refund yields about a 7.5 percent reduction in waste disposal for the wastes that we address in our model. We qualify this finding, however, by noting that our model is calibrated on 1990 data. There have been significant changes in recycling markets in the mid-1990sprices for many secondary materials have risen as demand for those materials has increased. As a result, further increases in demand (and reductions in waste disposal) could be more costly than predicted by our model. This would imply a smaller optimal reduction in disposal. Updating the data is a topic for future research.