Cost-benefit analysis is big these days. In broad strokes, it makes perfect sense – policies have benefits and costs, and ceteris paribus the former, properly measured, should outweigh the latter, properly measured. But the words “properly measured” are doing a lot of work. Actually computing costs and benefits are a thorny business. An example I’ve been musing on lately really demonstrates this.

A few blocks from my house, there is a “Wings to Go,” and naturally as a consequence of this chicken bones can often be found strewn about the neighborhood. This is, of course, a huge choking hazard for dogs, such as my dog:


So let’s say I proposed a regulatory ban on the sale of wings to go. Well, a good regulatory agent would go about calculating the costs and benefits. Primary among the costs would be loss of consumer surplus, and if our agent were a god agent she would also tabulate the disparate impact this might have on people based on their socioeconomic status. In terms of benefit, though, you would have to measure “dog lives saved.” How do you value a dog’s life?

Turning to her trusty cost-benefit manuals, our agent would find a very consistent answer – when such information is available, use price data. Therefore, the amount paid for the dog should be a good proxy for it’s market value.

The fee to rescue our dog was $300. You know what else cost $300?

Apple Introduces Two New iPhone Models At Product Launch

Therefore, a market-oriented cost-benefit analysis would judge the death of my dog to impose costs on me equivalent to the death of my iPhone. But if you ask anyone who owns both an iPhone and a dog, that’s obviously insane. I would rather see orders of magnitude more of my physical possessions destroyed then see harm come to my dog. Yet that’s not how cost-benefit analysis would view it.

A lot of criticism of cost-benefit analysis comes from the difficulty of imputing monetary values onto costs and benefits that are totally outside markets, or at least distant enough from markets so as to make assigning them a single price ambiguous, such as the cost to health from pollution. But we should also be skeptical of  using prices to imply value.

And if you were wondering to what extent this post was an excuse to post pictures of my basset hound:


Mystery solved.