You'd have to solve the prisoner's dilemma though, and that's something Graeber certainly hasn't managed to crack.

You don’t have to solve it. There’s not really any problem to solve to begin with. People who are happy with the changed working conditions will continue to work there. People who do not want to tolerate the change will not continue to work there. Every person can precisely choose the outcome they want, without cooperation from anybody.

That's not what I was responding to though. This is:

> Imagine if almost everyone collectively at Salesforce said, “nope, this is dumb! We are going to stop working until this benefit is handed out to everyone at the company.” It’d be changed back / expanded within 48 hours.

That's a prisoner's dilemma, people holding out for the best they can get. If enough people defect then it won't work.

The prisoners dilemma doesn’t apply to this situation, because every single individual involved can, without any cooperation from anybody else, easily get the optimal outcome for themselves.

If you define “optimal outcome” as “people who don’t like having their benefits revoked quit”, yes. But I think a lot of people would define “optimal outcome” as “people who don’t like having their benefits revoked don’t get them revoked”.

The distinction is unimportant, and if you care about it very much then you’re just losing on purpose.

Why would you need to solve it? Because the prisoners dilemma suggests everyone should hold out for the holiday.

The reason the dilemma is of any interest is because people do defect.