“Design for environment” takes place when firms explicitly incorporate environmental issues in their product design and manufacturing decisions (Joseph Fiksel, 1996). To reduce waste associated with consumer products, for example, firms can reduce product weight, including the amount of packaging that comes with a product, and make products easier and less costly to recycle. The latter can be accomplished by changing the type of material used in a product, avoiding mixing of materials in a single product, coding and labelling when different materials are used, and designing for ease of disassembly (U.S. Congress Office of Technology Assessment, 1992; Fiksel, 1996). An important debate centres around the question of whether design for environment (DfE) needs to be explicitly targeted using policy instruments focused “upstream” on producers or whether “downstream” instruments send sufficient signals back upstream. One important downstream instrument is a disposal fee (i.e., charging households a price per pound, per bag, or per can for their trash). Many observers feel that downstream instruments do not send appropriate signals back upstream. Some economists and others have suggested, on the other hand, that disposal fees can, in theory, send signals upstream to producers (Karen Palmer and Walls, 1998). We find that disposal fees provide incentives for efficient DfE only when there is a fully functioning recycling market (i.e., only when recyclers pay households a price for each of their recyclable items and that price varies with the degree of product recyclability). In the real world, transactions costs of paying households for each individual item recycled would be prohibitively high. In a more realistic setting in which households may place some items in a recycling bin but are not paid for doing so, a disposal fee by itself will not yield a social optimum. It encourages households to place items in the bin, but all items allowed in the bin are of equal value to the household (i.e., there is no reason to prefer an easily recyclable aluminium can to a difficult-to-recycle plastic milk jug). This means that there is no signal to producers to make products more recyclable. If the government can set product taxes that vary with product recyclability, the social optimum can be restored, but this option is not practicable. We argue that the infeasibility of (I) paying households for recycling and (ii) taxing products according to their recyclability means that a first-best outcome is no longer attainable. Instead, we set up a constrained second-best optimum and solve for policy instruments that achieve this outcome. We find that a disposal fee will not ensure the constrained optimum unless it is supplemented with upstream instruments
