In Mutual Defense and The Free Rider, which in my opinion is a very well-crafted and thoughtful article, Goes has discovered, analysed and elaborated fruitfully upon some genuine difficulties with the scheme for Mutual Defence I previously outlined. However, I believe that these difficulties can probably be circumvented; in this reply to Goes I attempt to show how.

I entirely agree with Goes that MDAs do not solve the free rider problem in any mathematically rigorous fashion; it is logically possible that the process of combining into higher-level MDAs will break down before an adequate national defence is generated. However, I believe that the range of circumstances in which it might fail is far more restrictive than Goes suggests.

My first criticism is that I believe Goes greatly underestimates the natural belligerence of human beings, in small groups as well as large. If anything small groups get into fights more frequently than large ones. It doesn't take some remote politician to start a war; people are quite capable of doing so off their own bat. I'm not suggesting that MDAs would be for ever skirmishing at the drop of a hat - but conventional states don't either. I very much doubt that the internal relations between MDAs of various levels would be so much more peaceable than those between national states that the disjunction would render the topmost alliances ineffective against external threats.

I agree that higher-level MDAs could follow either centralising or decentralising strategies (in Goes's terminology). They could also combine the strategies, or some choose one, some the other. However, I disagree that centralising strategies need have the disadvantages Goes posits; indeed he indicates the solution himself: "the peaceful will ... join such an MDA ... only if it internalized the costs of being belligerent".

An MDA or alliance will pledge itself to a strategy of "all for one" subject to "one for all", that is, subject to the member MDA's agreeing to accept a set of rules or treaty obligations and an adjudicating authority. If a member starts a fight in defiance of the rest of the MDA it will not receive their military backing and indeed may face retributive action from them.

Consequently, centralising strategies will probably obtain even in the highest-level MDAs. This is not to claim that the topmost levels will be pure centralising alliances; it is likely that they may permit conflicts between their own lower-level members without interference (though they will probably offer dispute arbitration or mediation); this is technically known as a treaty of offence and defence (like a family that squabbles incessantly amongst itself but combines instantly against outsiders).

Empirical evidence for this is the existence of real-world centralising alliances on the largest scales (such as NATO, SEATO and the UN).

It takes very little additional effort to form high-level alliances, so long as most of the military hardware is affordable on the lower levels, which I believe in a free market economy it almost certainly would be. It is of course conceivable that threats might arise that could only be countered by a single massive investment at the highest level, in which case the demands upon such an alliance would be much greater, and medium-level MDAs would in all probability then combine to coerce compliance (still a solution to the public good problem, though not of the sort we've been hoping for). However, in most plausible scenarios almost all the materiel and weapons systems would be owned and deployed at fairly low levels (city level and below).

Goes points out that external enemy states might chose to attack the strongest rather than the weakest MDAs. This is of course quite correct, but slightly off-target. My argument is that MDAs will organise to counter all plausible threats, not merely those they consider the most likely. Since an enemy state might easily decide to attack anything from a city or large town upwards, MDAs need to consider the defence of their city, from other cities and MDAs as well as external states, as one of the scenarios they must be prepared to handle. Thus a contribution towards national defence is being provided from a fairly low level upwards. The incentives may not always be so strong on the lower levels, but they still exist.

One failure mode I should mention, because it is likely to prove fatal to a young free market society in a statist world. A libertarian minimal state, which can organise national defence directly, might face a coalition of statist enemies dedicated to its destruction and be overwhelmed. In similar circumstances an ancap society and its MDAs would also be overwhelmed, not through any failure to solve the free rider problem, but simply because it is too small and weak to stand against the rest of the world.

Contra Goes I believe such a scenario very probable. It is true that, as he points out, the enemies of the free market society face a "public good" problem of their own, and it would take time for them to put together a working coalition and screw themselves up to full-scale war. Yet we have seen through recent events that such coalitions can indeed be created, given the will - any country that defies the New World Order or the international community now faces the threat of annihilation. And do not be deceived, the very existence of a libertarian or ancap society would be seen as the greatest of threats to the statist order. Moreover, ancap society will almost certainly prove a magnet for terrorists, rebels, gun runners and drug smugglers who wish to operate beyond its territory, as well as ordinary criminal fugitives and refugees, thus providing all the excuse necessary for a combined statist attack; effective vilification will not be difficult.

I conclude that mutual defence will be practicable in almost all circumstances in which effective defence is at all possible (though not against unlimited and overwhelming odds) provided that a modicum of internal stability exists.

© Paul Birch, 18th Jan. 2003.