Targets' Targets, or Environmental Economics 303

When implementing emission targets, it's necessary to consider who the target of the targets is to be: that is, are the limits being implemented on a source-by-source basis, on an entity-by-entity basis, on an industry-by-industry basis, or on an aggregate basis? This is, in fact, crucially important to figuring out whether the targets are achievable and whether they're efficient.

Let's start by assuming that there's some cost to abating emissions from the level which would be chosen absent regulation (we'll call this the cost of abatement) and that there's some damage caused by emissions. The condition we need for an economically-efficient outcome is that the marginal cost of abatement should be equal to the marginal damage: that is, the cost of preventing a tonne of CO2 from being emitted should be equal to the damage that tonne would cause if it were emitted. We can argue about whether or not that's an appropriate way to determine how much we should pollute, but for the moment, let's accept it as given.

Carbon dioxide's a neat case, because the marginal damage caused by another tonne emitted is essentially location-independent: we don't really care whether that tonne comes from Canada, the US, or China, and we don't really care whether it comes from a car, trees decaying in a forest, or what have you. Since CO2 mixes quickly into the atmosphere and diffuses globally, a tonne's a tonne the whole world 'round. (This isn't true, for instance, with local (ground-level ozone, water contamination) or regional (sulfur dioxide) pollutants.) In principle, we can figure out how much damage a tonne of CO2 will cost the planet as a whole.