> It seems to me that a lot of desk jobs that feel like bullshit to the employee are really part of the nervous tissue of the organization, and the value they provide to the organization is still larger than the employee's salary

I'm not so sure that would actually count as a bullshit job in Graeber's eyes. He works towards a definition of a bullshit job over multiple pages. Here is a bit, I think, that would rule out the job above:

> This allows us to refine our definition. Bullshit jobs are not just jobs that are useless or pernicious; typically, there has to be some degree of pretense and fraud involved as well. The jobholder must feel obliged to pretend that there is, in fact, a good reason why her job exists, even if, privately, she finds such claims ridiculous. There has to be some kind of gap between pretense and reality. (This makes sense etymologically: “bullshitting” is, after all, a form of dishonesty.)

I read Bullshit Jobs and enjoyed it, yet it kinda nagged at me that I was missing the subtleness of the argument. I'm glad to see here that it hasn't crystalized for many others, too.