Cost-benefit analysis and public policy

Engineering design is, in no small part, the art of the trade-off. Including feature X means the cost goes up, excluding feature Y means sales go down. How do you balance considerations like these and find the sweet spot?

Engineers are taught to use cost-benefit analysis, to weigh and assess design alternatives preferring the most profitable financial outcome. Cost-benefit analysis got a severe black eye when it was learned that Ford engineers knew that the Pinto gas tank could fatally explode in certain rare collisions and proceeded with a design that they knew might kill some small number of people.

This was not so cynically amoral as it may sound. Engineers could easily design and build an automobile that would infallibly protect passengers in the event of any reasonably foreseeable collision — but nobody could afford to buy it; such calculations must be made.

There are two profound difficulties with cost-benefit analysis. The first is the difficulty of assigning a probability to a particular event, and the second is the difficulty of assessing the cost. What, for instance, is the statistical likelihood of an automobile collision that exerts a particular magnitude of force at a particular location on the automobile body? And what is the likely settlement cost of such an accident to the manufacturer? It’s one thing if the accident kills the parents of three young children, and it’s quite another if the accident kills the unmarried high-school dropout who delivers pizzas.

Again: It’s unpleasant to think about, but engineers do and must think about such things because nobody could afford an automobile that would infallibly protect passengers against any reasonably foreseeable collision — ‘reasonably foreseeable’ itself being a slippery notion.

I am thinking about these things because, at this exact moment, New York Governor Cuomo is conducting his daily coronavirus briefing and insisting that there is no point at which restoring the economy is more important than a life, contra Donald Trump, who seems to think that we’ve spent enough battling coronavirus and it’s time to get back to work. Cost-benefit analysis is all but impossible when it comes to social policy, because there are billions and billions of activities and transactions that can’t be foreseen and whose costs can’t be evaluated. Even so, cost-benefit analysis is what Trump and Cuomo are disagreeing about, and each is assuming a cartoon-character posture.

Ironically, millions of Americans are informally, intuitively, making such calculations a dozen times a day right now. Go to the store, and risk infection, or wait 2-days to have SuperMart put that item in the trunk of the car? Few of us would hesitate to go to the pharmacy to pickup a prescription for a sick child, but few of us would go out to buy a bag of potato chips.

Risk, statistics, cost-benefit, even when assessed informally, are at the center of all decision-making, and the idea that they can be evaded is naive.

This entry was posted in General. Bookmark the permalink.