I think the difference between an anthropologist like Graeber and an economist like Cowen is that they observe the system at different levels. Cowen judges the

I think the difference between an anthropologist like Graeber and an economist like Cowen is that they observe the system at different levels. Cowen judges the value of jobs from "inside the system" (with the system = contemporary capitalism), whereas Graeber takes a perspective outside the system and argues that the "zero/negative-sum game" (as Cowen said) is the problem. To illustrate with a drastic example: Suppose you have two workers, one who digs holes and one who shovels the same dirt back into the hole to close it. Outside the system, it's obvious that they're not doing any meaningful work. Inside the system, there's probably some argument why these jobs are useful. (Maybe government regulations require that their company dig at least X holes per month. Or you could argue that it's good because it increases GDP or reduces unemployment.)