One of the most pressing difficulties for stateless government would seem to be national defense. While police protection can be well internalized, being primarily a service to individuals, businesses, or neighborhoods, defense against foreign enemies yields non-rival and non-excludable benefits to all, and thus would seemingly encounter a severe public-goods problem. But, in associating national defense with a military, I believe that we may be equivocating. If the nation can be made secure without the need for a military, there is no defense problem with anarchy.
I can see two reasons why one nation might invade another: preemption of threat, and ambition for land or resources. The first reason, preemption,1 need concern us little. Preemption is, by nature, against a threat; a nation that does not threaten others need not fear preemptive action against itself. If a nation has too weak an organized military to defend itself, it cannot have enough to pose a threat to another, and need not fear preemption; if it has a strong military, then concerns about its military weakness are unfounded. Therefore, in no case is defense against preemption a unique concern for an anarchal nation.
Thus, we are left with defense against territorial expansion. In this case, the aggressing nation does so with expectation of benefit; an expected cost greater than the expected benefit would suffice to deter the threat, even if it could not avert suffering if the attack were made.1 Furthermore, even conventional national defense does not prevent harms; consider England during WWII. Thus, all increases in the cost of a successful invasion equally serve national defense.
In considering this, I believe that Machiavelli’s argument from Ch. IV of The Prince is relevant: “[H]e who considers both of these states will recognize great difficulties in seizing the state of the Turk, but, once it is conquered, great ease in holding it…. But it is impossible to hold with such tranquillity states constituted like that of France. Hence arose those frequent rebellions against the Romans in Spain, France, and Greece, owing to the many principalities there were in these states, of which, as long as the memory of them endured, the Romans always held an insecure possession;” Although a decentralized state may not erect such a hard outer defense, and thus may be easier to defeat in pitched battle, it leaves no centralized mechanism with which to seize control of the country. Look at the Vichy government of France during WWII: once the military fell, the country on the whole (neglecting isolated private resistance, which would be if anything greater were occupied nation originally anarchist) served the purposes of the conquerors in subduing the country. If a decentralized nation is difficult to govern, how much more one with no institutions of government, with its people accustomed to freedom?
Iraq, too, presents a good example. Hussein’s regime should be the model of those who look to government for national defense: a centralized state lacking democratic hindrances to its warmaking and with a high military budget.2 Yet each time the United States invaded, the demoralized military of Iraq crumbled. Notably, the last time, when we sought to gain control of the government (rather than mere military defeat), a private resistance arose which we have not yet quelled, despite the support of Iraq’s government and much of the population. How much more problematic would be an invasion of a nation with no legitimacy and no domestic support on the side of the conquerors? Thus, it seems to me that far more effective at deterring invasion than a conventional military would be the assurance that an invader would have to suppress the population man by man, without support from the local state. A privately funded military is, I admit, difficult to conceive; but we must remember that a military is but a scarcely sufficient and wholly unnecessary means of achieving national defense, which could be as well or better achieved by other means under anarchy.
Thus, I think that national defense under anarchy would not consist of multi-billion dollar jets and a vast military-industrial complex, but rather the free ownership of weapons and the reluctance of free men to submit to tyrants. Even a committed minority of the population could make invasion prohibitively costly. Such a defense would be far less costly, with respect to both resources and rights. Gone would be the annual 650 billion dollar drain on our nation’s resources. Gone would be the ever-present excuse for violations of our rights that “wartime necessity” demands them. And, of course, if a group of people thought that an aircraft carrier would be a good thing to have, they would be free to fund one–but with their own money, and not, as at present, their neighbors’.
1 Exactly the same justification as for imprisonment of criminals: imprisonment does not seek to alleviate the original wrong, but seeks to alter the costs considered by the prospective criminal in order that he might not commit the crime. That imprisonment does not prevent the crime from occurring if it does occur need not influence our practical analysis of its effectiveness (although from a consideration of justice, I think that imprisonment and such defense as I describe are wholly incommensurate).
2 National defense being considerably more difficult in a democracy or under Constitutional government, why do not those who raise the objection to anarchy also raise it against limited government as a whole?
Ambulatory Sesquipedalian's Blog 
I think you’ve neglected an important aspect of human nature; if the military is privately funded, how can people who constantly quarrel with their own neighbors hope to defend themselves? Without some kind of leadership (limited, of course), no one would be able to fight against enemies because they’d be fighting among themselves. Also, you have left out one other significant reason for war. If there were no centralized military, then the Islamic Jihad would be much easier to wage on our people. And if any other nation decides that we are doing wrong and it is morally necessary for them to take control of us, what is to stop them since the citizens who should be organizing the military would be divided?
Here I treat only international problems, assuming a functioning domestic legal system (not to say that the proper functioning of a domestic system is trivially obvious, but the incentives are different and thus I believe that protection from domestic and foreign threats are best treated separately.
Ultimately, I fail to see what a centralized government can do against Islamic terrorism. It seems to me that our past intervention has been counterproductive (witness all the isolationism examples), while terrorism cannot really be prevented by an organized military (although it can be largely prevented by sound security measures, and private security (such as that on charter airlines) tends to be better than state (such as that by the TSA).
I do not believe that divided control of the military would be a problem because I doubt that there would be a military, per se, but rather more of a militia. Sundry insurgencies have proven that one does not need modern aircraft and AFVs to stall an invasion.
It would, however, be easier to win a fight against a “milita.” While terrorism may not be stoppable either by anarchy or leadership, I think our country would be stornger and less likely to be conquered if there were leadership. Secondly, you cannot assume a functioning domestic system. The laws would be clear, but they would be abused by citizens.
I seem to be still communicating poorly. A militia would be easier to defeat in open battle (what many people forget about the Battle of Lexington and Concord is that it began with two sharp defeats for the militia on the open field), but it would be very difficult to actually rule an anarchal society. In every successful invasion that comes to mind, the invading government coopted existing governments. Government cannot function without the consent of the people; while that consent has, historically, been given to the co-opters of an existing government, would it be given to a purely imposed government? I place little weight on leadership per se, but only as it is necessary to the functioning of a military. Since the provision of defense I here outline does not depend on a military, it does not depend on leadership. Look at Iraq–how much leadership do the insurgents have? Yet if we were invading for selfish gain, surely we would have abandoned the effort in frustration by now (despite the fact that we have the support of the domestic government, assistance unavailable to the invaders of an anarchy).
I am not assuming a functioning domestic system epistemologically, but hypothetically. Here, I attempt to answer the question of whether an anarchal society that did settle domestic problems be able to avert invasion. If such hypothetical assumption is illegitimate, then the only way one could speak about anything would be a complete proof from foundational principles, which is clearly absurd. This is actually the same principle as that behind fiat power: fiat power allows a debate to consider whether a policy would be beneficial if enacted without considering whether it would be enacted, making the same hypothetical assumption you here condemn.
No, you don’t have to completely prove every point; but you must, if speaking realistically, factor in the precedents of human nature. Yet I will concede that hypotheticals are a different matter.
That I shall grant. It seems to me that an anarchal society can provide for domestic order (in the sense of preventing violations of rights, not in the social engineers sense) better than government, so that is the hypothetical assumption I use here. I do not wish to assume away any item of human nature, but I do not believe that my analysis rests on any particular interpretation of human nature (except a modicum of rationality, which seems to me to be well confirmed), but rather a plausible statement regarding conditions under anarchy.
First of all, I disagree with the notion that humans are by nature bad and will exploit each other. I think that is a product of capitalism and hierarchical society. Furthermore, even if that is the case, then that makes a government that much more dangerous because it allows people a mechanism and power structure in which to exploit others.
As far as the ability of an Anarchist society to defend itself, the Black Army was a very competent force during the Russian civil wars. Their one mistake was that they aligned themselves with the Red Army to fight a common enemy (the White Army) and were subsequently betrayed by the government.
I believe that government does not change the character of men, but changes their incentives in such a way that their existing character faults are magnified. Thus, I believe that selfishness is an integral part of fallen human nature. At what time have men ever demonstrated perfect altruism and community since that time? However, the invisible hand of the free market ensures that cooperative action is mutually beneficial (c.f. Bastiat’s Harmonies of Political Economics), whereas government sets men’s interests against each other, thus forcing conflict and discouraging cooperation. Selfishness is inescapable, but government worsens its effects.
That is an excellent example–perhaps an anarchist society would have greater power to mobilize an army than I had thought. Public goods theory predicts that it would never happen, but the predictions of that theory rarely seem accurate with respect to the private market (interestingly, it does a much better job of explaining government action in a democracy).