By last summer, what was briefly the independent republic of Bosnia-Herzegovina had been reduced to a few besieged enclaves and a seat in the United Nations. In the process, perhaps 200,000 Bosnians had been killed, and 2 million had been driven from their homes. Yet Western leaders were still dithering about whether to lift the two-year-old U.N. arms embargo that had prevented the Bosnian Muslims from effectively defending themselves since the civil war began in early 1992.
"With regard to the lifting of the arms embargo," President Clinton said, "the question obviously there is, if you widen the capacity of people to fight, will that help to get a settlement and bring about peace, or will it lead to more bloodshed?" For two years, the conventional wisdom of the world community has been that international diplomacy, abetted by an arms embargo, could revive the stability that Yugoslavia enjoyed in the years when it was ruled by the Croat strongman Tito, who had tens of thousands of his countrymen murdered to keep himself in power.
The current bloodletting in the central Balkans may be a modest affair compared to that of 50 years ago between the communists and proto-Nazi Ustashi storm troopers. But those massacres occurred out of the world's sight.
International conscience follows CNN's minicam crews, which are on the scene to record what war is like when waged between soldiers and civilians.
How did the poorly armed Bosnian Muslims, the chief victims of the war, come to be surrounded by well-armed enemies? Things went quite differently in the initially lethal skirmishes between Serbs and Croats in the north. There, after some fighting, Serbia cut a deal, settling for a relatively small share of Croatia. The Croats' military power was nowhere near enough to conquer the Serbs. It was merely enough to make the Serbs appreciate the advantages of peace. Guns did not so much win a war as avert one.
The Bosnian Muslims were not so fortunate. For the most part they were unarmed, and the arms embargo left them helpless against Croat and Serb enemies who wanted their land. Margaret Thatcher, as usual among the first of the world's politicians to discern the obvious, had warned for several months that the embargo spelled disaster for the Muslim people of Bosnia. Until last winter's "ethnic cleansing" proved her point, respectable opinion was very much against her views. Throughout Europe and in the United States, it had been a bipartisan article of faith that the only hope of peace in the Balkans lay in diplomacy aided by an arms embargo.
In the words of David Clark, the British Labour Party's shadow defense minister, "lift[ing] the arms embargo to the Muslims…has always seemed to us rather crazy," like trying to "douse a fire with petrol. It never works." President Bush embraced this premise, and President Clinton remains ambivalent.
The result has been gun control, writ large: a scheme aimed at limiting violence that instead encourages predators to take whatever they want. Both the U.N. arms embargo and domestic gun control are based on the notion that the accumulation of weapons as such tends to encourage violence. The chances for peace and security can thus be enhanced by limiting or reducing the total number of weapons.
This weapons-violence hypothesis has been part of the intellectual furniture of progressive politics for most of this century. It found expression as the fourth of President Wilson's 14 Points and appears in almost the same language in the Covenant of the League of Nations, whose Article Eight provided: "The Members of the League recognize that the maintenance of peace requires the reduction of national armaments to the lowest point consistent with national safety and the enforcement by common action of international obligations."
As the historian Alfred Zimmern pointed out, "The important word in the first paragraph is 'requires.' Peace, it says…cannot be maintained unless armaments are reduced, or, in telegraphic language: 'armaments mean war.'" The covenant was simply reflecting what was even then a superficial dictum of good-govemment progressivism. The notion that armaments mean war is "a favorite theme on pulpits as well as on platforms," Zimmern wrote. It is also "a mistake which, if allowed to pass uncontradicted, can do infinite damage to the cause of peace."
Modem strategic theory rejects the weapons-violence hypothesis, focusing on stability rather than stockpiles. It starts with the proposition that people tend to pursue the course of action that they believe will give them the maximum return. Hence the likelihood of violence depends on how the expected rate of return on violence compares with the alternatives.
If you can grab an island (or a purse) that belongs to someone else with complete certainty of getting away with it at zero cost, it does not automatically follow that you will grab it. But your probability of grabbing it will be greater than if you think you have a significant chance of getting killed in the attempt—especially if you think the current owner makes the same estimate of the odds. More generally, when the expected value of attacking falls to a value equal to or less than the expected value of doing nothing, rational people will do nothing.
In the arena of international politics, notwithstanding the U.N.'s approach to the war in Bosnia, this way of looking at the problems of conflict and its avoidance has long been routine. Yale historian Michael Howard's views represent those of most researchers in this area: "Wars begin with conscious and reasoned decisions based on the calculation, made by both parties, that they can achieve more by going to war than by remaining at peace."
This model is quite versatile. Its predictions will hold whether we are dealing with two neolithic hunters eyeball to eyeball over a deer carcass, or a criminal and a victim, or two modem sovereigns, each with thousands of hydrogen bombs, toe to toe over Western Europe. In each case, the likelihood of violence depends upon how the antagonists view the relative probabilities of getting what they want, losing what they want, and being killed or injured either way. It also depends upon the availability of other options and the value, positive or negative, that each assigns to possible future interaction with the other. The pivotal question, whether for arms control or crime control, remains the expected return on violence, compared with the alternative.
The Cold War gave scholars and diplomats good reason to ponder such matters. The defining condition of postwar superpower competition was the maintenance of equilibrium. Almost as soon as both the Russians and the West had hydrogen bombs and credible delivery systems for them, the United States abandoned the doctrine of "superiority" in favor of the doctrine officially known as "assured capability for mutual destruction" but called by friend and foe alike "mutual assured destruction"—MAD.
Both sides recognized that the possession of hydrogen-bomb arsenals fundamentally changed the game. The element of surprise—striking an enemy first without warning—had always had its benefits, but never until the atomic age had it been possible to begin and end a war with the same stroke. This so-called first strike capability is everywhere recognized as a profoundly destabilizing condition. Two antagonist nations with first-strike capabilities represent the deadliest possible sort of Prisoner's Dilemma. Each is better off (and knows its adversary knows each is better off) striking first, no matter what its adversary decides to do. If both sides see things the same way, then the mutual first-strike situation is an OK Corral, where success is defined by which gunslinger is quicker on the draw.
Why, then, are we still here? A deus ex machina, in the form of the second-strike weapon, appeared before either country became confident about the once-and-for-all sufficiency of its nuclear arsenal. By the late 1950s, both the United States and the Soviet Union could see that the day of reckoning was at hand, and in 1958 President Eisenhower and Secretary Khrushchev agreed that high-level negotiators from each side should meet in Geneva to discuss the problem of surprise attack.
As economist and game theorist Thomas Schelling describes it, "the occasion was crucial in identifying what was to become pivotal in arms negotiations for the next decade and, more important, in the design of strategic forces….The idea that both sides could favor each other's strategic-force security was dramatized by Secretary of Defense Robert McNamara's testimony to Congress that he would prefer the Soviet Union to invest in secure, hardened underground missile silos, rather than soft sites above ground, because the latter both invited and threatened preemptive attack while the former would encourage patience in a crisis " The victim who can strike back at the aggressor from the grave has restabilized the situation by restoring to zero the expected payoff from mounting a preemptive strike. This point is not limited to arms-control negotiations and the currently suspended Cold War.
Consider a less well-known example of violence that should have happened but didn't. Purdue University economist John Umbeck has investigated the formation and initial distribution of property rights in the High Sierra gold fields in the middle of the 19th century. After gold was discovered at Sutter's Mill in 1848, many thousands of prospectors poured into the hills, staked out claims, and removed minerals with a value (in today's money) in the hundreds of millions of dollars. The gold fields comprised an area of some 30,000 square miles in the California mountains. There was no civic infrastructure there at all—no towns, no highways, no lawmen, and, perhaps most significant of all, no official law for them to enforce anyway. The military governor of California had recently nullified by proclamation the Mexican land law that had previously governed the region, without proclaiming any substitute for it, temporary or permanent. Virtually all the '49ers were carrying firearms or kept them handy. Umbeck was struck by an odd fact: There was very little violence.
Another story from the old West makes a similar point. In 1889 the great Oklahoma Land Rush began. Until the late 1880s the United States respected Indian claims to Oklahoma, which meant that white people couldn't legally own land there. Through a process that no one should be proud of, this deal was busted, and a portion of Oklahoma, about 20 percent of its territory at first, was opened to settlement, beginning on a certain day at a certain hour. Oklahoma real estate was valuable stuff, just waiting to be claimed. The people who went in to settle it, like the '49ers, were mostly armed. But according to Washington College historian Robert Day, who has studied the period, they accomplished their objective essentially without violence.
These findings should surprise anyone who believes that "weapons cause violence." Not only were the Oklahomans and '49ers heavily armed, they were poor even by the hard-scrabble standards of their time. They must have been ambitious for wealth, and they probably felt that they had little to lose. From such a mix of passions and motives and guns, there ought to be one gunfight after another, as always happens in the movies—until one fierce soldier of fortune survives to grab all the gold in California, and all the farms in Oklahoma, for himself. If weapons cause violence, it's strange that things did not work out this way.
Of course, not all frontier tales are of harmony and accord; sometimes armed populations did behave violently, and sometimes they still do. The important point is that sometimes they do and sometimes they don't. The fact that heavily armed people commonly do behave themselves throws down a gauntlet in front of the weapons-violence hypothesis. The theory definitely does not predict that a large, disorganized group of heavily armed men rushing headlong after the same treasure will ever behave in a consistently peaceable manner.
Umbeck offers an intriguing explanation for the low level of violence. He observes that violence is much more likely when there are large perceived differences in the ability of individuals to use force effectively. In the California gold fields and later in the Oklahoma Land Rush, everybody was about equally armed (as Umbeck notes, they didn't call the six-shooter the "equalizer" for nothing). To some extent this sort of equality is a matter of perceptions. But if perceptions tend to track underlying reality in the long run, then in the long run it is a matter of fact as well. A peaceable equilibrium, however tense, tends to prevail in a world where everyone reasonably fears retaliation from, or on behalf of, potential victims.
Why isn't the face-off between individuals with guns—each of whom can decisively preempt retaliation by the other—precisely analogous to that between two thermonuclear superpowers before they have acquired second-strike capabilities? It seems reasonable to suppose that the social structure in which the Oklahomans and '49ers were embedded offered a surrogate for a second strike. Historian Roger McGrath, who studied two now extinct Gold Rush boom towns, chronicled the activities of "vigilance committees" that punished wrongful violence when the duly constituted authorities failed to do so. These towns turned out to be passably lawful places—even though the police and the courts were notoriously shoddy. Very little violence against women was reported, and there was very little theft, robbery, or burglary—facts for which the official institutions of law enforcement could claim little credit.
McGrath writes: "Rarely were the perpetrators of these types of crimes arrested, and even less often were they convicted….The citizens themselves, armed with various types of firearms and willing to kill to protect their persons or property, were evidently the most important deterrent to larcenous crime." So a more or less peaceful equilibrium does seem to be compatible with very high levels of armaments, given a second-strike capability or some practical equivalent.
One sees the converse result, an equilibrium of violence and reciprocated violence, in many city neighborhoods where high levels of arms are combined with a weak social structure. Although we are accustomed to reading social disintegration from statistics that tell of high infant mortality, low graduation rates, or the bad condition of the housing stock, what is actually crucial is not that these conditions exist but that they persist, despite significant efforts to get rid of them.
It has been something of a puzzle for a generation of policy makers why, despite numerous, expensive efforts to fix them, these problems have been so hard to solve. From the viewpoint of strategic theory, however, there is no puzzle. What causes disintegration—the falling apart of things in such a way that they cannot simply be put back together—is a structure of incentives in which cooperation makes no sense.
When people do not believe that their own or anyone else's rights will be protected or wrongs rebuked, cooperation is the behavior of a sucker. In strategic terms, a disintegrated social world simulates a world of strangers. Dealings between strangers are precarious because a person cannot reasonably expect that cooperation will be reciprocated. In a world of strangers, self-interest is all on the side of selfish non-cooperation—vandalizing property, not flushing toilets after using them, cheating in transactions.
What all this suggests is that arms per se are not the issue, a conclusion that has obvious relevance to the gun-control debate. Gun-control laws usually aim to reduce the absolute number of firearms in circulation, but this is not at all important to the violence rate. What is important is the existence of a robust equilibrium between lawful and unlawful force—a second-strike weapon or some equivalent to make preemptive aggression seem a zero-retum proposition.
In theory, one way of achieving such an equilibrium would be simply to hire more police officers. But in practice there is no reasonable prospect of hiring nearly enough police officers to serve as an adequate second-strike surrogate in any American city. Do the math. Each new police officer adds about $60,000 per year to a city's payroll costs. If each officer works 2,000 duty hours per year, it will cost a quarter of a million dollars just to add a single additional officer to each shift. No serious student of public administration believes it is feasible to address existing shortfalls in security services this way.
It is always possible, however, to make a bad situation worse. Gun-control laws discourage a private alternative to hiring more police officers by making it harder for the average citizen to obtain a firearm. Indeed, gun control has a disproportionate impact on people who want firearms for legitimate reasons. Both potential victims and criminals seek guns for essentially the same purpose—to get tactical dominance in a confrontation with another person. But criminals know for certain that they'll need their guns, because they plan to have hostile interactions with other people. Law-abiding people, on the other hand, will need their guns only if confronted with a situation in which threatening to use lethal force is both legal and feasible.
So even if the good guys and the bad guys each assign identical values to dominating a hostile encounter, bad guys will still value guns more, because on average they will be more certain of having such encounters. So there is some wisdom to the old NRA slogan, "If guns are outlawed, only outlaws will have guns." Gun control tends to put potential victims at a disadvantage relative to criminals.
This will remain true as long as guns are available in significant numbers.
Whatever the obstacles to gun ownership, criminals will have a stronger incentive to overcome them. But what if guns were eliminated completely, or nearly so? A society with no guns to speak of might possibly be safer and less violent than the one in which we live. A different sort of equilibrium might prevail—the equilibrium of doves rather than the equilibrium of hawks.
This dream may or may not appeal to you. The crucial point is that it's not likely to be realized. It would require the government to confiscate some 200 million privately owned firearms and prevent future production or smuggling.
In the real world, security would be enhanced by encouraging the distribution of more arms rather than less. Certain categories of city dwellers have a very low statistical probability of engaging in predatory behavior. Social-security pensioners, virtually all adult females with clean criminal and psychiatric records and no history of substance abuse, and most employed men over 40 with similarly clean backgrounds are all essentially invisible in the crime statistics. Any such person who is prepared to learn what is necessary in order to handle a sidearm safely and appropriately ought to be encouraged—not merely permitted—to acquire that knowledge and carry the weapon, as police officers do, wherever they go.
Everyone appreciates that the presence of armed police officers in a neighborhood makes it a more secure place than it would be in their absence. Armed civilians of equally good character and with equivalent firearms training would be useful in a similar way, and tens of thousands of such people live and work in every big city in the country. One should think of them as auxiliary peace officers, not vigilantes, for there is no reason to believe that they would act beyond the law. Common law has always allowed self-help when regular legal remedies have been inadequate. Self-help means individuals acting under legal sanction but on their own initiative to defend important interests that court officers and police cannot protect.
Self-help is not "taking the law into your own hands." It is the law currently in every state and has been a part of the Western legal tradition practically from biblical times. Any gun-control measure that makes weapons harder to get for good guys than for bad guys certainly complicates self-help to some extent and would likely produce more violence rather than less. Lawmakers would be better advised to consider how to help people organize to defend themselves from violence that the police cannot possibly stop.
The advice offered here runs counter to conventional wisdom, to say the least. But conventional wisdom has generally been wrong about arms policy. It predicted that keeping weapons out of Bosnia would lead to peace. It overestimated the law's ability to get guns away from predators and keep them away and overestimated the ability of the police force to protect disarmed civilians. It has made its futile pursuit of first-best solutions (that is, universal disarmament) the enemy of achievable second-best ones. It has thereby made a mess that will not get better by continuing to wish for a return to some imaginary square one where nobody had a gun or a reason to use one.
While the daydreaming goes on, conditions continue to stagnate or deteriorate in city neighborhoods that went from bad in the '60s to worse in the '70s and have stayed as bad even as legal gun controls have become ever more stringent. Political scientist John Dilulio has persuasively argued that the paramount problem of the inner cities is a crime problem. Where crime and violence flourish, nothing else will. Crime either causes or makes it prohibitively difficult to alleviate almost every other kind of social ill. If the foregoing analysis is right, then the public policy of conventional wisdom has a great deal to answer for in the present situation, which it did much to create. It is time to consider some expedients that have, according to our best understanding of what makes people tick, a reasonable prospect of doing some good rather than making matters worse once again.
Daniel D. Polsby is Kirkland & Ellis Professor of Law at Northwestern University.