I haven't read the book, but I find it concerning that this article criticizes Graeber's unrepresentative data collection, and then proceeds to arm-chair about why Graeber is wrong without really introducing any empirical findings.

There are surveys about both how much time workers actually 'work', and how much of that work they see as productive. They seem to suggest that a reasonable share of worker's time is not spent on actually doing anything:

https://www.inc.com/marcel-schwantes/new-report-only-12-perc...

https://www.worklife.news/talent/hours-in-workday/

But also, Hobart's arm-chairing here seems to take the perspective of whether something is useful to the firm, rather than if something is useful overall. Medical billing apparently charges typically 4-10% of the value billed. Clearly for the medical billing firm, having workers knowledgeable about CPT codes etc is worth it, and medical providers find it more economical to hire such a firm than to keep their own billing specialists on staff -- but if we just moved to single-payer health care, the role could disappear entirely. It's a bullshit job because it's solving a problem that doesn't need to exist.

https://neolytix.com/what-is-the-going-rate-for-medical-bill...

Similarly, Hobart mentions tax lawyers more than once. And certainly companies that having good tax advice can save them a lot of money. But I think most experts agree that a drastically simpler tax code would be more efficient overall, both because of the reduced need for tax lawyers and accountants, and because all the weird (inefficient) tax-minimization strategies could disappear. Being useful to a firm does not mean the job is not bullshit.