Killing the Competition

Economic Inequality and Homicide

Martin Daly

• Publisher: Routledge
• Year: 2017
• eBook ISBN: 9781351510158
• Topic: Social Sciences
• Subtopic: Criminology
• Index: Social Sciences

Criminologists have known for decades that income inequality is the best predictor of the local homicide rate, but why this is so has eluded them. There is a simple, compelling answer: most homicides are the denouements of competitive interactions between men. Relatively speaking, where desired goods are distributed inequitably and competition for those goods is severe, dangerous tactics of competition are appealing and a high homicide rate is just one of many unfortunate consequences. Killing the Competition is about this relationship between economic inequality and lethal interpersonal violence.Suggesting that economic inequality is a cause of social problems and violence elicits fierce opposition from inequality's beneficiaries. Three main arguments have been presented by those who would acquit inequality of the charges against it: that "absolute" poverty is the real problem and inequality is just an incidental correlate; that "primitive" egalitarian societies have surprisingly high homicide rates, and that inequality and homicide rates do not change in synchrony and are therefore mutually irrelevant. With detailed but accessible data analyses and thorough reviews of relevant research, Martin Daly dispels all three arguments.Killing the Competition applies basic principles of behavioural biology to explain why killers are usually men, not women, and counters the view that attitudes and values prevailing in "cultures of violence" make change impossible.

1
Why Homicide?

Crimes as serious as murder should have strong emotions behind them.

—George Orwell, Decline of the English Murder, 1946

Inequality, Competition, and Homicide

Homicides are intensely interesting. When there’s a murder, anyone who has even a remote connection to the case craves details. People want to know exactly what happened, and why. This morbid fascination is healthy and normal. The natural habitat of the human mind is a social universe of personal acquaintances, gossip, and hearsay. In such a social universe, it has always been important to gather all the information you can about who is dangerous, and in what situations. Lethality is a big deal.

As fascinating as the stories behind individual murder cases may be, homicides become even more interesting when we study their statistical properties dispassionately. When they are appropriately aggregated and summarized, the statistics of homicide provide us with a sort of “window” on human passions. What are the sources of interpersonal conflict that can induce someone to kill? The most important is apparently inequality: the degree to which the things that men desire are equitably or inequitably distributed.

Decades of research have shown that if you want to predict the homicide rate of a nation, or a state, or a city, the best single piece of information to help you make that prediction is the level of income inequality. Unfortunately, the criminologists who have confirmed this result again and again remain befuddled about what to make of it. One reason for their befuddlement is that, like many other social scientists, most criminologists believe that violence itself is a form of pathology. This is a mistake. Although some killers are indeed insane, violence itself is an organized, evolved capability of normal human beings, pursuing their interests in a world in which people’s interests conflict. The misunderstanding of violence as pathology has blinded analysts to a simple, intuitively appealing interpretation of the facts before them. A local homicide rate is a manifestation of the local level of competition for scarce resources, and economic inequality is a major determinant of the severity of that competition. One of my aims in this book is to lay out the case that this straightforward interpretation is correct.

To make this argument, I will have to familiarize readers with some little-known facts. First, it is not just killers who are overwhelmingly male, but their victims, as well. More specifically, in developed countries with stable governments and judicial systems, both those who kill and those whom they kill are primarily young, disadvantaged men. Second, the contexts in which these men kill one another are competitive contexts: sexual rivalries, turf wars, robberies, and, above all else, contests over the limited social resources of dominance, respect, and “face.” Third, these competitive killings make up not only the majority of all homicides, but the most variable component of the overall homicide rate as well. Together, these facts imply that if you wish to know why homicide rates are so variable, you should be looking at those aspects of society that determine the intensity of local competition and thereby affect men’s readiness to resort to dangerous tactics in their struggles with rivals.

In short, I will argue that the observed association between inequality and lethal violence is one of cause and effect. This cannot simply be assumed. In areas of science in which controlled experiments are impossible, it is a truism that correlation doesn’t necessarily imply causation. Life expectancy and the consumption of sugary soft drinks have both been increasing for decades, for example, and life expectancy furthermore tends to be low in those backwaters that sugary soft drinks have yet to reach. But it would be loony to infer that soda-pop extends lives! Other correlates of modernization presumably explain the association. In principle, the robust association that I will show you between economic inequality and the homicide rate may also turn out to be meaningless in this way. However, unlike the soda-pop example, there is a straightforward and highly plausible causal chain leading from economic inequality to homicide. When desired goods are more unequally distributed, there is more to be gained from being one of the winners and there is a higher likelihood of utter failure, raising the incentives for the competitors to resort to extreme tactics, including potentially dangerous tactics. An elevated homicide rate is just one predictable consequence of this escalated competition.¹

I find this causal interpretation of the association between income inequality and homicide rates compelling, but not everyone agrees. Skeptics who doubt that inequality plays a direct role in causing violence have put forward three challenges. The first is a claim that the low levels of lethal violence in egalitarian nations like Japan and the Nordic countries are due to the simple fact that very few of their citizens live in poverty. In other words, the relative disadvantage of inequality is itself supposedly harmless, but happens to be associated with a high incidence of absolute destitution, which is the real culprit. The second challenge resides in the fact that some small-scale societies studied by anthropologists are supposedly egalitarian and yet have very high homicide rates. The third challenge is that occasional episodes in which inequality and homicide are seen to be moving in opposite directions—such as in the United States during the 1990s, when inequality was rising and the homicide rate was falling—imply that the usual association between the two cannot be causal. One of this book’s major aims is to confront each of these challenges and demonstrate that they do not, in fact, undermine the hypothesis that links inequality to homicide through its effects on competition. Those who have advanced the first and second of these three challenges have simply got their facts wrong, and the third challenge withers when we recognize that any genuine effects of inequality must of necessity be “lagged” effects of past inequality.

Can well-chosen policies reduce violence? One pessimistic line of thought in the social sciences attributes violence to the existence of cultures in which the local norms reinforce and “valorize” violence, and further maintains that such cultures of violence are highly resistant to change. The relatively high rate of homicide in the southern United States, for example, is manifestly associated with a “culture of honor,” in which being willing and able to use violence is admired as a manly virtue, and many writers accept that this culture has perpetuated itself generation after generation for no other reason than that children are socialized to adopt the relevant values and that anyone who forsakes them suffers social costs. Another aim of this book is to refute this view. The glorification of violence in certain value systems is utilitarian and self-defensive, but it is not immutable. People are not the passive recipients of cultural indoctrination that this pessimistic theory would have them be. Attitudes, values, and behavior can all change quickly when incentives change, and rates of homicide soar and plummet with surprising rapidity. Moreover, economic inequality plays a clear role in the persistence of “southern violence” in the United States, and the alleged “inertia” of the local culture provides no justification for despair and inaction.

If I am successful in persuading you that inequality almost certainly has harmful effects, including that of exacerbating lethal violence, you might nevertheless question the wisdom of policies aimed at reducing it. A relentless message to which we have all been exposed is that economic inequality is a necessary evil if we are to maintain the rising tide of productivity and innovation that supposedly raises all our boats. Leveling policies such as a progressive income tax would kill the goose that lays the golden eggs, or so the story goes. In this book’s concluding chapters, I will make the case that this is mere propaganda on behalf of the privileged. What the science of economics actually indicates is that even if we cared not at all about suffering and injustice, but wished only to maximize productivity and growth, the economic inequality that now prevails in the rich countries is counterproductive and needs to be reduced.

The case against economic inequality as an important cause of lethal violence is strong. It has also been indicted by many researchers for its damaging health consequences. Economic inequality reinforces inequality of opportunity and other injustices, and economists now agree that it even damages the economy itself, by creating waste and inefficiency. No redeeming benefits to society have been identified. Economic inequality persists because of the disproportionate power that is wielded by a minority who are its beneficiaries and the disinformation that they cynically disseminate. Political action aimed at reducing it is overdue.

Getting Hooked on Homicide Research

When my late wife Margo Wilson and I began doing homicide research in 1978, it wasn’t because we were specifically interested in murder, nor even in human affairs. Our area of expertise was the behavior of other animal species. I was primarily interested in how ecological variables, such as the way that food resources are distributed in space, might help explain the diversity of animal social systems, and the particular animals that I studied were desert rodents: kangaroo rats in California and gerbils in the Algerian Sahara. As for Margo, most of her previous research had been concerned with the ways in which castration and hormone replacement therapy affect the behavior of monkeys.

We were teaching at the University of California, Riverside, where, in collaboration with a few colleagues from the psychology, biology, and mathematics departments, we had organized a weekly seminar for the graduate students in the area of “sociobiology.” One day, our topic was why animals sometimes kill infants of their own species. Harvard primatologist Sarah Hrdy and other researchers had recently proposed that in certain animals, such as the langur monkeys that Hrdy had studied in India, infanticide might be a normal response to a particular situation, namely that in which a new male establishes a mating relationship with a female partner who is still burdened with dependent young sired by her previous partner. The proposed utility of this horrific behavior—the reason why Hrdy thought it could have contributed to the reproductive success of the killers and thus have evolved by Darwinian selection—was that by disposing of his predecessor’s baby, the male monkey stops the mother from expending her time and energy raising the progeny of a rival male, and redirects her efforts toward reproducing with the killer and rearing his young instead.

The discussion was lively. Hrdy’s “sexually selected infanticide” hypothesis is now well established, but in those days, it was controversial. Some class members recoiled from the very suggestion that anything as repugnant as infanticide could possibly be a normal behavior that had been actively favored by the evolutionary process. Other students countered—correctly, in my view—that their classmates were guilty of a version of the “naturalistic fallacy”: the naïve assumption that anything that is “natural” must be nice, and that anything we find distasteful must be pathological. What could not be denied was that Hrdy’s hypothesis had pinpointed a recurrent feature of the circumstances in which infanticide had been observed or inferred in her monkeys, in lions, and in a few other species of animals.²

It was in the context of this discussion that one of the students in the class, Suzanne Weghorst, raised a question: Are human stepchildren disproportionately mistreated in reality, as they are in folklore? (Think of Cinderella. Think of Snow White.) No one in the room knew the answer, and so, in a pre-internet world, we suggested that Suzanne should go to the university library, look up the relevant research literature, and report back to the class. A week later, what she had to report was a surprise: No relevant research could be found. Child abuse had been a topic of intense investigation since 1962, when Colorado pediatrician Henry Kempe and his collaborators announced the discovery of a “battered child syndrome.” But somehow, no-one had thought to ask whether children who are living with stepparents are abused at higher rates than other children. If we wanted to know the answer, we would have to do the research ourselves. And so we did.³

Margo, Suzanne, and I were soon able to demonstrate that stepchildren are indeed heavily overrepresented in child abuse case files. In 1976, there were laws obliging health care professionals and others to report suspected cases of child abuse in jurisdictions comprising about half of the population of the United States, and within those jurisdictions, we found that one of every 160 children under three years of age who dwelt with a birth parent and a stepparent had been a victim of physical abuse that had been validated by some sort of follow-up investigation subsequent to the initial report. The corresponding incidence of validated abuse for same-age children living with both birth parents was much less than one in a thousand. However, we realized immediately that this statistical fact could, in principle, be an “artifact”: an illusion arising from biases in the detection of abuse rather than a reflection of a genuine elevation of the rate at which stepchildren are actually mistreated.⁴

To see how this might be so, imagine that you live next door to a family in which you have noticed that one small child keeps showing up with bruises. Is it not possible that you would be just a little bit more suspicious if you happened to know that the mother’s partner was a new boyfriend and not the child’s father? And is it not possible that your likelihood of contacting child welfare authorities would therefore be just that little bit greater? Okay, maybe not you, but some people might react in this way, and for that reason, we placed special emphasis on the fact that the contrast between rates of abuse at the hands of stepparents versus birth parents got larger when we made our criterion of “child abuse” more and more stringent, and that the greatest difference was seen in the relatively rare cases of fatal abuse. If biased detection were an important source of apparent differences in abuse risk, then that bias, and hence the overrepresentation of stepparents, should have shrunk or disappeared when we confined our attention to those least ambiguous lethal cases. However, what we actually observed was precisely the opposite. Surely biased detection could not explain away a hundred-fold difference in a toddler’s probability of being beaten to death by a co-residing stepfather versus a genetic father!⁵

Inspired by these dramatic results—and perhaps also by discovering that research on human violence attracts a larger audience than research on squirrel monkeys or kangaroo rats—Margo had a further brainwave: Killings might provide a good “assay” of conflict in any relationship category. Because homicides are more reliably recorded than nonlethal manifestations of antagonism, she reasoned, they may be especially revealing of both the circumstances that inflame conflicts of interest and the circumstances that alleviate them.

We could think of many questions and hypotheses, derived from our interest in social evolution, which looked as though they might usefully be addressed by analyses of homicide data. Is infidelity a uniquely potent sour...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contents
• Preface
• 1 Why Homicide?
• 2 Homicide and Economic Inequality
• 3 Competition and Violence
• 4 Inequality, or Just Poverty?
• 5 Jockeying for Position
• 6 The Arena of Competition
• 7 Culture of Violence?
• 8 Lags and Lifetimes
• 9 Too Much Inequality
• 10 What Keeps Competitive Violence in Check?
• References
• Index