Published in the Laissez Faire City Times, July 1, 2002
Standard economic and political theory grants to the state the irreplaceable role of protecting individuals against violence. There are reasons to doubt that only the state can render this sort of service, although we have no definitive proof that a system of private protection would work in a satisfactory way.[1] Of more immediate concern than this general issue is the question of what kind of state protection is economically efficient, i.e., more likely to satisfy diversified individual demands.
I propose to distinguish two broad types of public protection: defense and custody. Defense protection is meant to stop aggressions in progress, or to deter aggressors before they act; it focuses on aggressors. Custody protection aims at preventing aggression by controlling the aggressed so that they do not come into contact with potential aggressors; its focus is not limited to actual aggressors.
Defense protection can take three forms: (1) barrier protection, (2) reactive force (or self-defense), and (3) deterrence by punishment. Barrier protection is the erection of physical barriers to prevent aggressors from reaching their targets, and range from locked doors to gated communities. Stricto sensu, barrier protection cannot be offered in a pure form by the state, notwithstanding the image of customs posts or fences on the American-Mexican border, because state-erected barriers also stop the non-aggressor who is just returning home, as well as the peaceful visitor of the resident host. In fact, barrier protection is either a private activity (walls, doors and locks) or else, as a state activity, it is more akin to the custody type of protection considered below.
Reactive force is a neater form of defense protection, as it remains basically unchanged whether exerted by the aggressed individual, by an agent of his, or by the state. Reactive force means opposing force to an aggressor in order to stop the aggression. Reactive defense is nothing but self-defense, whether exerted by policemen or, to the extent that laws do not prevent it, by private citizens. In fact, reactive force always stands behind any kind of barrier protection for the simple reason that somebody behind the wall has to repel by force any aggressor who jumps the wall.
The third form of defense protection, punishment deterrence, shares with the other two (or, at least, with reactive defense and with private barrier protection) the characteristic that it is directed at aggressors only. As it name indicates, it aims at preventing aggressions by imposing punishments that increase their costs up to the point where they cease to be cost-effective. Punishment may be a more or less effective deterrent depending on the criminal, but even fanatics like Islamic terrorists, who believe that martyrdom gives eternal life in the company of 72 virgins could be deterred by burial in pig skins which, they believe, is a sure ticket to eternal damnation.
At any rate, there are other means of deterrence than punishment, such as reactive defense with a high probability of effectiveness. For example, if it had been known that some aircraft passengers were armed and that highjacking aircrafts was not feasible, the 9/11 terrorists might have dropped this specific strategy. Martyrdom is more attractive if it also brings practical results in addition to eternal life. And what I want to stress here is custody as opposed to defense.
Consider the following two idealized models. In the defense protection model, the state imposes a very harsh punishment to anybody who has aggressed another individual (deterrence variant), or provides each individual with an invincible bodyguard who immediately pulverizes aggressors with his laser gun (reactive-defense variant). In the custody model, the state protects individuals by confining everyone to a locked cell from the cradle to the grave, so that nobody can actually attack anybody else. Policemen are available to shoot anybody who tries to enter or leave a cell. In both models, we have a society with no aggression, but custody protection differs from reactive-force protection in that the former aims at prior control of everybody, while the latter only comes into play when an aggression has occurred, or has started to occur.
One striking characteristic of the "war on terror" is its strong custody-protection content. It is true that military operations in Afghanistan has been of the deterrence-by-punishment type, but anti-terrorist legislation nearly always aims to impose surveillance and prior controls on the potential victims of aggression. Just think of the heightened immigration and border controls, the increased police and army powers (searches, surveillance of communications, ID papers, etc.), the strengthening of "money laundering" controls, the turning of air terminals and airplanes into even more perfect zones of citizen impotence, dependency, and submission, the arbitrary (i.e., contrary to the traditional requirements of the rule of law) detention and expulsion of alien residents, etc.
We should note that there is another type of defense protection activities: preemptive strikes, which amount to striking and disabling conspirators before their aggressions can reasonably or uncontroversially be said to be in progress. The attacks in Afghanistan have often been presented as such, i.e., as a way to disable the terrorists' networks although, of course, the deterrence-by-punishment objective would have been sufficient to justify some kind of operation. The main difference between reactive defense and preemptive strikes is the negative time lag between attack and defense: it is obviously self-defense if you shoot the aggressor one second before he pulls the trigger, but it is pre-emptive strike if you shoot him when he is building the plant to make the steel with which he may manufacture his revolver. Intermediary cases are more open to debate.
If you replace "pull the trigger" by "drop the bomb," and "make the steel" by "build the reactor," you can see what is the economic (as opposed to philosophical) argument for (or against) preemptive strikes: the higher the expected loss from an aggression, the higher the expected benefits of a preemptive strike. Of course, when the cost in terms of lost liberty and destruction of the rule of law is (however imperfectly) factored in, the cost-benefit ratio of some preemptive strikes becomes higher than one. One illustration: preventive detention of all young black men would be a preemptive strike that would greatly reduce violent crime rates, but the costs of such a measure in lost opportunities for the innocents and in increased state power would be staggering. Thus, either preemptive-strike protection fits in the reactive force category (when the preemptive strike is made in response to a clear and present danger), or it is closer to the custody model and becomes more difficult to justify.
Let's then go back to defense and custody. It is tempting to argue that custody protection is legitimate because it would also be resorted to by private protection agencies in an anarcho-capitalist society.[2] I don't think this is correct. From a philosophical viewpoint, it may perhaps be argued, à la Nozick, that activities that create general apprehension and fear can be prohibited (with compensation), and that non-compliance calls for a preemptive strike,[3] but this still does not justify prior control of peaceful individuals, i.e., it does not justify custody protection. It can also be shown that custody protection is not a legitimate option in a society based only on private property rights.
In order to see this, consider a little formalized model. Imagine that Robinson Crusoe and Friday each occupies his own part of the island. Their properties share a common border, and are elsewhere limited by the open ocean on which, we further assume, they cannot travel. In this setting, custody protection has no meaning: property limits keep each on his own land. If Friday leaves his property, he immediately trespasses on Crusoe's, and the latter's use of force is defense protection. Imagine that a state appears on the island, and wedges a public street between Crusoe's and Friday's pieces of land. Now, there is a difference between preventing Friday from leaving his land (which is custody protection), and preventing him from intruding into Crusoe's land (which is defense). It is the existence of public spaces and the state that creates a distinction between the custody and the defense types of protection. Without the state, all protection (as opposed to aggression) is of the defensive type.
Consider the case of airports, which terrorists (and doubtlessly other aggressors) can use as a base for carrying out their aggressions. The state could protect individuals against terrorists by defense means, i.e., by stiff penalties for aggressors (deterrence protection) or by making sure that threats of force will be immediately be met with self-defense (reactive-force protection). Strangely, before 9/11, reactive force was virtually impossible on flights because honest citizens were forbidden from carrying guns, and no state agent was on board. This problem has only been partly corrected since. What the state has done very efficiently, though, has been to increase custody protection by nationalizing airports even more thoroughly and controlling more tightly citizens who enter this public zone (similar to the street in our Crusoe-Friday model). In an anarcho-capitalist society, it may be that some airport or airline owners would impose similar measures on their property but, if there is a demand for non-custody travel, any attempt to impose such measures to other airports or to other aircrafts than their own would be likely to meet armed resistance.
Why does the state seem more and more interested in providing custody protection instead of defense services? It could be argued that custody protection is more efficient (i.e., cheaper for the same level of protection), especially in fighting international terrorists. This claim ignores one important lesson of economic theory: we must not generally expect the state to be efficient, in the sense of its being able to satisfy all individual preferences; it is only efficient in the sense of buying off the clienteles that are necessary for the rulers of the state to maintain their tenure. Why then does the state like custody protection? A likely hypothesis is that this form of protection is especially useful to protect the state against its own citizens or, at least, against a portion of its citizens. Indeed, when the state was less powerful and public opinion more libertarian, terrorism was not fought with custody-protection (or, at least, such means were less resorted to), as illustrated by the fight against anarchist terrorists in the late 19th and early 20th century.
In the world as it is now, we do expect the state to protect us against terrorists, but not at any cost. At the point where our own state would become as dangerous as international terrorists (in terms of expected net long-term costs for our life, liberty and property), we should stop granting it additional power. Individual valuations of these matters will of course vary, but this is an argument for keeping the state as small and as consensual as possible, not for overriding individual preferences blindly. Since the state was already, before 9/11, much too much powerful, since it already had too many custody-protection powers, there is no reason to grant it even more. This argument is related to what Auberon Herbert wrote in 1894:
"If we cannot learn, if the only effect upon us of the presence of the dynamiter in our midst is to make us multiply punishments, invent restrictions, increase the number of our official spies, forbid public meetings, interfere with the press, put up gratings — as in one country they propose to do — in our House of Commons, scrutinize visitors under official microscopes, request them, as at Vienna, and I think now at Paris also, to be good enough to leave their greatcoats in the vestibules ... I venture to prophesy that there lies before us a bitter and an evil time."[4]
[1] See my "Explaining the State," Laissez-Faire Electronic Times, Vol. 1, No. 17 (June 10, 2002); reproduced at www.pierrelemieux.org/artstate.html.
[2] A related argument is presented in Sarah Fitz-Claridge, "War, Free Trade and Liberty: Strange Bed-fellows," speech given at the Youth for Liberty Summer Camp, Orono, Ontario, May 25, 2002, available www.fitz-claridge.com/node/1.
[3] Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), Part I.
[4] Auberon Herbert, "The Ethics of Dynamite", Contemporary Review, May 1894; reproduced in The Right and Wrong of Compulsion by the State, and Other Essays by Auberon Herbert (Indianapolis: Liberty Classics, 1978), p. 226.