I think it's easier to have such an exploitative regime in sciences that depend on "big lab" research: a higher body of graduate-student subordinates trying to get PhDs makes it easier to treat them as interchangeable components; the success of the lab depends on maximisation of (competent) labour hours put in; the lab equipment is seriously expensive and a high return on the costs must be shown to justify the next round of research applications; the graduate students are rather more dependent on the PI's goodwill to get the research done for their PhD.