Smaller, disjointed entities have, by definition, higher overheads: if before you had 1 payroll accountant for 100 state employees, then you split duties over 5 nonprofits, now you have 5 payroll accountants.
The counter-argument is that the 5 orgs can be somewhat more efficient than the one monolith, but the key is the conditional: they can, but there is no guarantee, whereas the 4 extra overheads are guaranteed. That's one of the many systemic flaws of privatized/delegated systems.
It's true that larger organizations can be more efficient in theory. But it's largely just that: theory.
Because, as organizations scale up and enough time passes, the usual outcome is a gigantic web of inscrutable policies and procedures. And lots of people to write and enforce them. Getting something done that in a small organization can trivially be handled by the person originally seeing the need for it, will in a large organization often involve five other people in a messy chain of emails. Or it can't be done at all, unless one is willing and able to play politics to get it done.
Example: I work at a school/university with some 30 thousand students, that is the merger product of many smaller schools, each targeting different professions. These mergers were done for efficiency reasons. I'd estimate the teacher/support head count overhead for these originals schools to be somewhere between 30 and 50%. Currently, we're well over 100% overhead. So we employ more non-teachers than we employ teachers. (And on top of that, teachers handle a lot of non-teaching-related work, of course.) Somehow, the merger is seen as a success.
Small organizations are awesome. It's a shame that government rules and regulations often make them unsustainable.
The counter-argument is that the 5 orgs can be somewhat more efficient than the one monolith, but the key is the conditional: they can, but there is no guarantee, whereas the 4 extra overheads are guaranteed. That's one of the many systemic flaws of privatized/delegated systems.