On checklists, again

A while ago I wrote that checklists are (i) underrated and (ii) underrated by people who might be expected to rate them highly.

One further note, which perhaps states the obvious: I suspect part of the resistance here is the tension between the rigidity and algorithmic nature of a checklist and the often very different nature of the thing the checklist is about. This blinds us to the possibility of using a checklist and tempts us to reject any such list on the ground that it must be missing something. It's not easy to reply to oneself: "Yes, it's surely missing something, but hopefully not too much.")

When you learn calculus and get to integration, it's standard to hear in the first lecture that although differentiation can be done algorithmically--with few caveats, if you learn a fixed set of techniques, you can calculate any derivative you'll see--integration cannot be done that way. In general it requires some insight. A while later, you're likely to hear that if it's tricky enough, you should really just try integrating by parts.

Anyway, here's a list I find useful:

What's going on when people are persistently behaving in a way I can't figure out

  1. They're signaling something.
  2. There's a feeling I don't understand, or a connection I don't understand between the behavior and some feeling.
  3. Taxes.
  4. The obvious alternative is less good than I think it is.
  5. The Coase theorem applies (am I conflating the formal locus of power with the practical balance of incentives?).

Home page