Costs, Benefits, and Distribiution

Whenever people say they’re “against” cost-benefit analysis as a method for evaluating policy initiatives or regulatory schemes, they appear to be talking in paradox. To say that you think something is a good idea more or less just means that you think the benefits of doing it would outweigh the costs of doing it. So pretty much any proposal for changing the way these things are evaluated amounts to a proposal to “mend but don’t end” the practice of cost-benefit analysis. That said, the current way of doing things has a number of very serious flaws. Mark Kleiman offers up three here but let me just site the most egregious one:

Formal benefit cost analysis counts everyone’s gains and losses equally. But common sense and the principle of diminishing marginal utility agree that a dollar’s worth of gain is more valuable to someone with few dollars than it is with someone with many. Obviously, taking $1 each from 900,000 poor people to give $1 million to a hedge-fund billionaire doesn’t reflect a social gain, but a formal benefit-cost analysis will show that it does: after all, the net benefit is $100,000. Thus gains and losses should be adjusted by (at least) dividing each gain or loss by the income or wealth of the person bearing it, so that a $20 gain to a family with an income of $20,000 weighs as a heavily as a $10,000 gain to a family with an income of $1 million.

This is a very common pathology of economic analysis. As Brad DeLong points out in this Socratic dialogue what passes for “value-neutral” positive economics in fact embeds some very strong and perverse ideas about value:

Agathon: “That means that the market system, in weighting utilities and adding them up, gives you a much lower utility than it gives Richard Cheney. In fact, if marginal utility of wealth is inversely proportional to the square of lifetime wealth, the market system gives Richard Cheney about 400 times as big a weight as it gives you.”

Glaukon: “That’s sick.”

Agathon: “And it gives Bill Gates a weight about 400,000,000 times as big a weight as it gives you.”

Glaukon: “That’s sicker.”

Agathon: “But it gives you about 40,000 times the weight it gives your average Bengali peasant, who thus has about 1/16,000,000,000,000 the amount of the market system’s concern as Bill Gates has. Will you teach that?”


Glaukon: “We are value neutral economists! We don’t care about distribution! We care about efficiency!”

Agathon: “But claiming that you don’t care about distribution is implicitly saying that shifts in distribution are of no account–which can be true only if the social welfare function gives everybody a weight inversely proportional to their marginal utility of wealth.”

Glaukon: “You’re introducing politics into a value-neutral technocratic social science.”

Now as it happens it’s not 100 percent clear what alternative rule you should use. Which I think is one reason economists remain attracted to the “distribution doesn’t matter” point of view. It’s false to say that distribution doesn’t matter. But if you choose to believe that distribution doesn’t matter, that provides an unequivocal answer to how you ought to build distribution into your analysis. If you decide, accurately, that distribution does matter you’re left with the tough problem of specifying exactly how it matters. Much easier to just pretend it doesn’t matter, and then pretending that the fact that you’re pretending it doesn’t matter doesn’t matter either because it’s a “value-neutral” point-of-view. But it just isn’t/