Crime Is Not the Problem: Lethal Violence in America
by Franklin E. Zimring and Gordon Hawkins
(Oxford University Press, 259 pp., $35)
One of the more frustrating difficulties facing students of crime is our inability to compare crime rates across countries. Interpol gathers crime data from national police agencies, but it does so in a way that make its reports next to worthless. The agency fails to assess the quality of the accounts that it receives, and it presents them in a way bound to cause confusion. Thus, not long ago, someone published an op-ed essay in which the author claimed that the Netherlands had a higher murder rate than did the United States. That is, to put it mildly, an implausible idea. In his defense, however, he displayed the Interpol report. At first glance, the document seemed to confirm his view, until one noticed that every homicide reported for the United States was completed—that is, there was a dead body—but the homicides reported for the Netherlands included both completed and attempted (no dead body) homicides. The attempts, of course, far outnumbered the actual murders, and there was no explanation of how the Netherlands decided which actions were attempted murders and which were just everyday assaults. We do not know very much, in short, about how the characteristics of nations or their various criminal justice policies affect crime rates.
Franklin Zimring and Gordon Hawkins, two members of the Earl Warren Legal Institute at the University of California at Berkeley, have plunged into this thicket, fully aware of the snags that it contains, to sort out how American crime rates differ from those of comparably industrialized nations. No one will be surprised to learn that the United States has a far higher rate of violent crime, especially homicide, than Western Europe or Australia. But some may be astonished to learn that the rate of property crime here is similar to the rate of property crime elsewhere, and in many cases it is much lower. Zimring and Hawkins conclude that what is often described as the American “crime problem” is in reality a lethal violence problem, and that the main goal of public policy ought to be to reduce violence.
To do that, we must first understand why our rate of violence is so much higher than in England, Australia, France or Germany. The answer given by Zimring and Hawkins is that we kill each other more often (and engage in property crimes, such as robbery, that often have fatal outcomes) in large part because Americans are more heavily armed than are other societies. Opponents of gun control will reflexively object to this conclusion, but, if they are to prevail, they will have tough going against the arguments made here. Using data from the World Health Organization, a group that counts dead bodies instead of merely repeating police reports, and gathering facts from big-city police departments abroad, Zimring and Hawkins show that American cities are not very different from foreign ones of similar size with respect to theft or burglary, but they are vastly higher with respect to robbery and homicide. New York City has less theft and burglary than London but vastly more robberies and homicides. The same difference exists between Sydney, Australia, and Los Angeles.
Robbery involves the threat of violence; burglary need not involve violence, though violence may occur if the dwelling is occupied when the burglar enters. In neither crime is death likely. But thefts in American cities are more likely to lead to death than are thefts in other nations. In 1992, there were seven deaths in London resulting from a burglary or robbery; in New York City, there were 378, even though New York has fewer such crimes than does London. American property crimes are much more deadly than English ones, in large measure because our thieves are armed. And much the same story can be told about assault. When one Londoner attacks another, death occurs in less than one-half of 1 percent of the cases, but when one New Yorker attacks another, death is the result in over 3 percent of the cases. The reason in part is that firearms are used in 26 percent of all New York assaults but in only 1 percent of assaults in London.
STILL, THE USE of guns is not the whole story. If one looks only at robberies in which no gun was involved, the death rate in New York City is still three times as high as it is in London. Even in murder cases, guns are not essential: 30 percent of all American homicides did not involve a gun. This means that New Yorkers without a gun kill one another more often than do Londoners however armed. Obviously something more than weaponry makes New York a more lethal environment than London.
Since guns are not the whole story, we have extraordinary differences among our states in how frequently people are killed. Maine and North Dakota have the lowest homicide rates in the country, less than one-tenth of the rates in Louisiana and Mississippi, but the reason cannot be that no one in Maine or North Dakota owns a gun. Rural states are probably armed to the teeth, as anyone knows who has visited them during deer hunting season. The answer must be that personal encounters in rural states are more law-abiding and less productive of personal violence. North Dakota not only has the second-lowest murder rate, it has the second-lowest property crime rate.
Zimring and Hawkins suggest that many American communities are more dangerous not only because guns are more available, but also because personal conflicts are more frequent and more violent. In their words, firearms are “neither a necessary nor a sufficient cause of violent death,” but they are a contributing factor. If two men meet in a bar or on a street corner and have an argument, the result of that quarrel will depend heavily on what weapons might be available with which to manage any escalating violence. If there are only fists, only a fist fight can ensue; if there are guns, there may be a fatal shootout. Many years ago Zimring published articles suggesting that murder was often the consequence of an ambiguously motivated assault: at the outset, nobody intended the death of the other, but as the fight progressed and a gun was at hand, death was the result. To reduce deaths one must either reduce the likelihood of fights or disarm the fighters.
In their new book, Zimring and Hawkins largely reject other popular explanations for violence. They have little use for studies of the impact of the media, and I think that their rebuttals are essentially correct. Violence in the media is everywhere, in London as much as in New York, in Sydney as much as in Los Angeles, and yet those places differ dramatically in lethal behavior. When all cities are exposed to the same media, it is hard to see how the media can explain differences in violence. No doubt there are copy-cat killers, but their numbers are too small to explain why people in Tokyo almost never kill and those in Atlanta often do.
Violence also accompanies drug dealing, but the proportion of murders that are connected to the drug trade is too small to make much of a difference. The best estimates are that no more than 10 percent of all killings are connected to the drug trade, though from time to time the percentage is much higher in a few cities. Moreover, the laws on drug-dealing are about as tough in Australia as they are here, but drug-connected deaths are about sixty times more common in Los Angeles than in Sydney. In the United States, drug dealing on a large scale has probably created an array of armed gangs that make violent encounters, and thus lethal ones, more likely. But why? That is like asking why the vast majority of drug users are in this country even though almost every country has similar laws.
THERE IS another contributing factor that the authors confront, but not, I think, quite adequately. They ask whether the very high rate of violence among African Americans explains the American homicide rate. There is no denying the core facts. Blacks are five times as likely to kill as are whites; black males are six times as likely to kill as are white males.Homicide is the leading cause of death among young black males, but it is the tenth cause for Americans as a whole.Zimring and Hawkins do not have much to say about why this is true, except to argue that it is probably because African Americans live disproportionately in urban “slum neighborhoods” and because less violent middleclass blacks live in “racial zones” that put them in close proximity to poor blacks.
This is not much of an explanation. Just limiting ourselves to big-city residents reduces the black-white difference in homicide from eight times nationally to only (only!) four times at the big-city level. Moreover, other equally poor and geographically isolated urban groups have much lower crime rates. Koreans, Vietnamese, and Chinese are often poor, and recent arrivals, and many of them live in similar “racial zones,” but they kill at a far lower rate than do African Americans.
Now, explaining these differences is not easy. I am not certain what it is, but I expect that it has much to do with the legacy of slavery, lynching, and past failure to enforce the law when blacks harmed other blacks. Oddly, Zimring and Hawkins write as if the explanation is either unimportant or obvious. It is, in fact, neither. If African American murder rates were the same as white murder rates, the national murder rate would drop substantially. The effect of lowering the black murder rate to equal the white one would not make America as safe as other industrialized nations, but it probably would have at least as big an effect as banning the existence of all handguns. Non-gun homicides in New York City are three times as common as all homicides in London, a number that is only a bit smaller than the difference in white-only homicide rates between the two countries.
In fact, Arnold Barnett of MIT has made some calculations that suggest that the homicide rate of adult black males has in fact been coming down much faster than the white homicide rate. No one is quite certain why this has occurred, though certain possible explanations—social progress, residential relocation—are obvious enough. We tend to forget these trends and to dwell instead on the great increase in juvenile homicide rates that took place between 1985 and 1992. Young people, white and black, were becoming much more lethal in the late 1980s, probably owing to the spread of gangs, their involvement in drug trafficking, and easier access to guns. The increase was greater for blacks.In the last few years, that rate has declined a bit, and this probably helps to explain why the homicide rate generally in the country has experienced so sharp a dip.
But this dip may prove to be short-lived. Census figures show that there will be an increase in the proportion of young people on the streets in the next few years, and there is no reason yet to suppose that those who now lead a life of no fathers, gangs for friends, and easy dollars in the drug trade have decided to abandon that life. Rescuing young people from those conditions, a frightfully difficult and expensive proposition, may be as effective as figuring out a way (none now exists) to deny them access to the knives and guns with which they can kill others.
Zimring and Hawkins neglect almost all of these issues in their desire to reassure us that there is no “black problem” in crime. I’m sorry, but there is. It is certainly not the whole problem, and solving it would certainly not solve America’s violence problem; Zimring and Hawkins are right to point out that equalizing racial differences in murder, desirable as that may be, would still leave America’s homicide rate at least twice as high as the rate in other major industrialized nations. An all-white America would be much more lethal than Italy, Canada, France, Germany and England, and vastly more lethal than Japan. But that is not the end of the story. It is impossible to deny that very high rates of violence among African Americans (rates that may have been coming down of late among black adults) not only contribute mightily to the problem of life in our cities, they also disfigure and polarize any effort to deal with our most serious domestic problem. The authors at least acknowledge this effect. As long as black violence is at so high a level, they observe, it will reinforce “white fear in ways that palpably contribute to the exclusion of blacks from the social mainstream.”
BY THIS POINT the reader expects that Zimring and Hawkins will offer some remedies for murder. Given their analysis, there are only two such remedies: reduce the availability of guns or lower the frequency of hostile encounters. But they suggest neither. Though they devote two long chapters to “Prevention,” reading them reminds me of watching Mike Hargrove getting ready to bat. He comes to the plate. He stretches his shirt, tugs at his glove, pulls at his pants, shifts his cap, adjusts his grip. He gets in place. Then he backs out and does this all over again. To watch Hargrove at bat was like killing time during a rain delay. Will this ever end?
In this book, no. Zimring and Hawkins write that a “book of this kind would be a terrible place to posit a detailed and comprehensive program of loss prevention from violence....” A terrible place? Franklin Zimring has devoted much of the last thirty years of his professional career to studying the impact of guns on violence, and he still has nothing to say about what we should do? If not now, when?
Of course, he does have a few things to say, but mostly by way of criticizing other people’s ideas. Zimring and Hawkins dislike many of our prison policies because they think that, under the impact of those policies, we send too many nonviolent offenders to prison. They argue that, in California, the “three strikes” law has had no connection to the recent reduction in the rate of violent crime, but they leave the explanation of this controversial judgment to a document that they do not bother to summarize. (You will have to look it up. But I warn you, it will be a waste of your time.) They attack people who support various popular anti-crime programs for making absurd predictions and failing to evaluate the results.
They are probably right about this. But what programs do they favor, and how should we evaluate them? They speculate about regulating hand guns, but they offer no idea as to how it might be done better. They ruminate about violent encounters, but they suggest no way to reduce their frequency except to suggest that victims be “as cooperative as possible” if they are threatened by a robber. They note that some people are trying to teach violence avoidance in the schools, but they conclude that there are “insufficient data to form a judgment” as to whether these plans work.
Perhaps Zimring and Hawkins are vague because they do not have any good ideas. That is not an embarrassing predicament. Very few people have good ideas about this subject, and for good reason. Eric Monkkonen, after years of careful digging in historical records, has been able to show that the homicide rate in New York City has exceeded that of London by a factor of at least five for the last two hundred years. Similarly, Roger Lane has shown that in the early nineteenth-century Philadelphia had a high homicide rate. Big-city Americans were killing each other at a far higher rate than were Londoners long before the invention of radio and television, and long before the introduction of semi-automatic weapons (and automatic ones) or the sale of any drugs (other than alcohol). It is very hard, I think, to devise an easy way to reduce a homicide rate that has been so high for so long. The hostility of American encounters is at least as important as the presence of American guns. If New York City can have a non-gun homicide rate that is three times larger than the total homicide rate in London, then removing all guns from the United States (which is impossible) would still leave us in a troubling condition.
Suppose we take Zimring’s and Hawkins’s analysis of the problem as correct, and then try to imagine what might be done. We must begin with the fact that the private ownership of guns cannot be substantially reduced. There are no point-of-sale restrictions that will reduce this huge stock by very much. Moreover, point-of-sale restrictions overlook the fact the most guns used in crimes are stolen or borrowed. And no government can do very much when people believe, with some empirical support, that having a gun makes you safer.
Using the data compiled by the National Crime Victimization Survey (NCVS) of 56,000 families, scholars have estimated that there are, at a minimum, between 65,000 and 80,000 defensive gun uses per year. Some estimates based on private polls suggest much higher defensive uses, ranging up to 1.5 or even 2.5 million. The data supplied by private polls are controversial, since so much depends on inferring society-wide effects from the answers of a tiny number of respondents. (If, to take a recent study, only 54 people out of 2,500 surveyed said they used a gun to defend themselves, then each of the 54 represents 68,000 Americans. Reporting errors—lies, exaggerations, poor memory—on the part of just a few people can have huge effects on the total number of defensive gun uses.) So consider instead the much larger and more reliable NCVS, conducted by the Census Bureau, according to which defensive gun uses in America are not trivial: 65,000 to 80,000 uses each year. No democratic government can afford to say that, while it is having its own trouble protecting people against crime, it wants to deprive these 65,000 people of the means to protect themselves. Under such conditions, you don’t need the National Rifle Association to defeat a government effort to disarm Americans.
There are more desirable and less controversial forms of gun control. The most important is to reduce the chances that a person will carry concealed on his person an unlicensed weapon while he walks about town. With a bit of new technology that is now being developed, it may become much easier for the police to spot and to question such gun carriers. Doing this may reduce the rate at which guns will cause angry encounters to escalate into lethal violence.
We also might wonder a bit about the magnitude of our penalties for homicide. They are about the same here as in Europe—that is to say, they are short in both places. Nationally, the median homicide inmate is released from prison after only about six years, while in California the release comes after about three-and-a-half years. Even many offenders sentenced to prison “for life” spend much less time there. Some inmates, of course, spend a lot of time in prison. But the small number of years the median (and the average) offender serves suggests the low price that we generally place on the average victim’s life. These sentences should be made longer.
And much remains to be done, finally, to lead children away from a life on the street. We are still trying to learn how best to do this, but a growing body of evidence suggests that early intervention in the lives of very young, at-risk children and their mothers (often there is no father) can make a lasting difference. It will take another generation to learn whether these plausible guesses will bear lasting results for large number of children, but the nation’s perpetually high homicide rate suggests that it might be time well spent.
Above all, we will have to learn to think about our crime problem historically. It took England several centuries of tough rule, brutal punishment and the inculcation of class-based values to achieve a low homicide rate. America has spent less time at the task, and it has sought to inculcate different values. As someone once said, the low murder rate in England is produced the same way you produce good lawns: plant good seed and then roll it for three hundred years. Zimring and Hawkins offer some sensible data on violent crime rates, but they plant no seeds and they roll no lawns.
James Q. Wilson is the author most recently of Moral Judgment (Basic Books). This article appeared in the August 25, 1997 issue of the magazine.