From what I can tell, the cause of this is that "number of papers published per unit of time" is the only metric institutions seem to care about when it comes to evaluating researchers' work.