The other day Clark expressed some ambivalence about cost-benefit analyses in the realm of environmental policy.

It so happens Dan Phaneuf at the Environmental Economics blog has some thoughts on that very matter.

For my part, I think Phaneuf — whose comments are generally quite sensible — underplays the risks of CBA. He acknowledges:

The benefits of environmental regulations can include, for example, reduced human and wildlife mortality, improved water quality, species preservation, and better recreation opportunities. The costs are usually reflected in higher prices for consumer goods and/or higher taxes. The latter are market effects readily measured in dollars, while the former are nonmarket effects for which dollar values are not available. In addition to complicating the practice of cost-benefit analysis (dollar values for the nonmarket effects must be inferred rather than directly observed) this raises ethical issues. Should we assign dollar values to undisturbed natural places? To human lives saved? To the existence of blue whales and grey wolves?

But my concern is less about the ethical propriety of assigning dollar values to these things — dollars are, for better or worse, how our society expresses value — than the practical difficulties of doing so in the real world.

The problem is not just that it’s tricky — of course it is.

The problem is that there’s no conceivable way the result will be, as Phaneuf says, "unbiased." The dollar values of consumer goods are themselves the expression of biases — albeit the fairly stable and easily identifiable biases of large groups. There is no independent "fact" about what a particular widget is worth, only the aggregate judgments of millions of people about what they’re willing to pay.

There is no such large-scale consensus on the value of ecosystems and their sub-elements. People haven’t had to pay for them. To put a dollar value on them involves the judgments of a relatively few people rather than the judgment of "the market." Phaneuf says economists "infer" the value, but it would be as accurate to say that they "invent" it.

Whatever might be produced by this process, whatever use it may serve, it will not be "objective" or untainted by bias.

The question is what kind of bias will be involved. My sense is that there will be constant pressure to undercount the benefits ecosystems provide, and due to the complexity and degree of arbitrariness involved in assigning dollar values to places and species and the like, that pressure will manifest in a thousand small ways that are subtle and difficult to pinpoint.

I don’t know any way around this. Saying that ecosystems have "absolute" value or something of the sort is just inane, both philosophically and economically. Phaneuf is right that somebody’s going to decide whether and how to use ecosystems, and they’re going to use information of some sort, so we might as well try to produce information for them that accounts for the value of the natural world.

But let’s not pretend it’s something it’s not.