A war, even a just war, becomes a reason or an excuse for your own state to increase its power over its own citizens--against you.
As the scientific study of the social consequences of rational or incentivized individual actions, economics can help us think about war and even about the morality of war. Here are a few ideas.
It seems that war has always existed, and that Rousseau's idyllic savage never did. In Constant Battles: The Myth of the Peaceful, Noble Savage, Steven A. LeBlanc provides archeological evidence of primitive tribes regularly fighting over resources. (I owe this reference to Daniel J. D'Amico, who, some years ago, directed a Liberty Fund conference on statelessness.) Our hope for the future is that cooperation through trade (if allowed to flourish), the limitation of power, and the enrichment of peaceful peoples will reduce the frequency of wars. But they are unlikely to disappear. Free and rich societies, if they still exist, will always be tempting prey. The empowerment of irrational religious ideologies only gives another justification for war.
Many theorists sympathetic to anarcho-capitalism believe that protection against foreign states is the chink in the armor of ordered anarchy. In Social Justice and the Indian Rope Trick, Anthony de Jasay suggests that anarchic societies may survive only "in relatively geographical remoteness that isolates them from other societies." He invokes David Hume's belief that quarrels between different societies could give rise to government. But even if the state should not exist at all, there are things it should still do to substitute for the private institutions it displaced. Defense against aggression by foreign states is the paradigmatic case, at least if the aggressor would violate individual preferences more than the current state. In this sense, the state is justified in waging defensive wars.
As always when evaluating coercive interventions (normally called "public policies"), moral judgments are ultimately required to define efficiency, because the latter concept is based on a starting status quo. One moral judgment underlying my reflections lies in the usual recognition of the value of individual preferences, which is a natural moral principle for economists. Ancillary moral judgments are required for thinking about war.
As philosophers would say, the war must be a "just war" (see Eric Mack's article "Rights, Just War, and National Defense"). A good illustration of the interface between economics and ethics lies in the issue of whether a just war incorporates a prohibition on hitting "innocent shields" (innocent people caught in the crossfire because they are unwittingly shielding aggressors). Such a prohibition would create perverse incentives: an aggressor could always shield its forces with innocent people, by installing its missile bases in populated cities, for example. Presumably, a moral war only requires to not intentionally target innocent civilians (see my Econlog post "Attacking Civilians in War"). I think that a just war also prohibits physical conscription--as opposed to indirect taxpayers' conscription. The Swiss model is appealing, if it could be made voluntary.
What is a defensive war is as much an economic as an ethical question, for the choice of different means of defense brings about different consequences. Legitimate self-defense cannot require that the bullet have left the barrel of your aggressor's gun before you shoot at him; otherwise, self-defense would never be morally allowed. The same applies to a group of individuals directly threatened by a foreign tyrant. In this sense, preventive strikes are not morally unacceptable, provided they are economically prudent and guided by the general principles of a just war.
The Latin dictum si vis pacem, para bellum (if you want peace, prepare war) also reflects the importance of incentives, and thus the interplay of economics and ethics in war. If a potential aggressor knows that you won't be capable of defending yourself, the expected cost of his aggression diminishes, and he is therefore more incited to attack you. This may justify a permanent army. It also justifies inter-state defense alliances. Such alliances, however, also increase the incentives of their nation-state members to bully other states, knowing that they would only shoulder part of the costs if the bullying degenerates into war.
The case for military intervention against some barbarian state that is not an immediate threat to "us" but may be gassing its own subjects, is more complicated. Economically--that is, considering the possible incentives, consequences, and available resources--the evaluation would depend on the probability that such a state becomes emboldened by its impunity or gains imitators, thus threatening us in the future. If your neighbor is regularly shooting his guests in his living room, you may fear that he will soon cross the property line and come into your own living room.
The ethical case for intervention against a rogue state is also complicated. On the one hand, one may argue, like Eric Mack, that your state does not have a mandate to protect you against hypothetical threats elsewhere in the world. On the other hand, perhaps we could read Chapter 2 of John Locke's Second Treatise of Government, as saying that, in the state of nature, anybody has the right to execute the law of nature, and that this principle applies to states because they are in a state of nature towards each other.
All this ignores what may be the main argument against war: what it does to "us." A war, even a just war, becomes a reason or an excuse for your own state to increase its power over its own citizens--against you. When the state has already, like today, accumulated extraordinary powers that cannot, except by a heroic stretch of mind, be reconciled with individual liberty, any new power, likely to long survive the end of a war, becomes a grave and immediate danger. The danger will only be intensified by the habits that the warriors will bring back home. The torturers will roam our streets. The danger is further intensified if the warriors don't know what they fight for except the country, the Nation, the leader, "us" against "them."
This domestic argument against war is attenuated, but not invalidated, when the conditions for a just war are realized. The consequences of foreign tyranny may be worse than those of domestic tyranny, but both sets of consequences must be somehow factored in. (Such consequences are of course difficult if not impossible to predict, but they must still be part of one's thinking.)
One would hope that our own state, before committing to a just war, would declare something like this: "We apologize to our citizens for having constantly infringed against their liberty with a long train of abuses and usurpations that have removed many of the reasons for fighting foreign tyrants. These abuses and usurpations go from occupational licensure to other bans or controls on domestic trade as well as ban or controls on foreign imports. In order to reinstate these reasons of liberty, we have decided to repeal a large number of measures of surveillance, control, and criminalization of peaceful behavior that have made the citizens fear us more than we fear the citizens."