This is also a common objection, but neglects to consider the scenario with all game theoretic incentives fully. It's covered in the video I linked earlier, but for the short and direct rebuttal; http://mises.org/etexts/longanarchism.pdf page 9.
> Which does not make it impossible to come up with theories, strangely enough.
Yeah, but you still don't get to call what really has been shown to happen wrong because it doesn't match the theory you came up with. (Well, I mean, you can if its a theory of morality rather than a predictive theory, but its important not to confuse the two.)
That would be valid if it had really been shown to happen the way you describe, it hasn't.
Warlords fighting over political authority is not a proof for what happens when the entire society rejects the concept of political authority and refuses to accept it as valid. Political aspirants wage war for political authority only because it has value. The point you raise is so common that it actually has a parody term; Argumentum ad somalia.
Now I grant that this still does not mean we are right, but this is no "provably effective" rebuttal, either.
In addition to nmrm's on-point comments, you appear to be referencing a rebuttal related to private security agencies battling, which is not a rebuttal to the observation that in the absence of established political authority, successful agencies of that type simply assume political authority.
There are two aspects of the "private security agencies will just become the government" question. If they clash for the Crown (page 9) or if they form a cartel (page 15).
