ONE
I begin with a simple argument, an argument that can be stated in the form of a statistical syllogism. Young men are prone to violence and disorder; America attracted unusually large numbers of young men; therefore America, or at any rate that part of it to which the surplus young men gravitated, was a more violent and disorderly place. Wherever and whenever young men have appeared in disproportionate numbers, there has been a disproportionate likelihood of trouble.
Few would doubt that young men are prone to violence and disorder. Given the choice, on an unfamiliar street, of walking past a knot of eighteen-year-old boys or an assortment of men, women, and children, most of us would choose the latter course. We know, both intuitively and experientially, that the odds of being insulted, jostled, robbed, or worse lie with the young men, not with the "normal" group of passersby. This is, to be sure, a prejudice, but not an unreasonable one. Young men's affinity for trouble is a real phenomenon. This is especially true of young American men, who are statistically among the most dangerous people on earth. The more interesting and controversial question is why this is so.
This question can be broken down into two parts: Why are young men generally prone to violence and disorder, and why have young men in America been particularly so? In this chapter I consider the first part of the question, looking at the violent and disorderly tendencies of young men as a statistical and biological phenomenon, rooted in the evolutionary history and biochemistry of the human species.
Men's Problematic Behavior
The male proclivity for violent and disorderly behavior is most readily apparent in criminal statistics, an imperfect but useful index of who did what to whom. Historians have reconstructed criminal statistics for selected American communities, colonies, and states in the seventeenth through nineteenth centuries; statisticians have compiled them for the nation as a whole in the twentieth century. Whether local or national, historical or contemporary, fragmentary or complete, these statistics tell the same story about gender and crime.
American men led women in nearly every category of crime as far back as 1636. The only exceptions were sex-related offenses like prostitution or the killing of illegitimate newborns, and these were not entirely exceptions, for their underlying cause was men's carnal use of women. The preponderance of male criminality has been observed in all regions and periods of American history, from Massachusetts to the Carolinas to California, from the colonial to the early national to the Progressive era. Drunkenness, assault, and property crime were overwhelmingly male endeavors, as were lethal acts against self and others. In nineteenth-century Philadelphia men were four to five times as likely as women to do away with themselves and more than ten times as likely to be accused of murder. They were also more than five times as likely to die accidental deaths, partly a reflection of their affinity for drink and reckless behavior.1
Drunken men who survived their misadventures often wound up in jail. Sober men did too, or at any rate they did so far more often than women. The finding of a study of Oakland, California, police records for 1872-1910 is typical: nine men arrested for every woman. "These differences are so great," the authors comment, "that they must reflect behavior. Women broke fewer laws, created fewer disturbances, than men."2
National statistics point to the same conclusion. Eighty-nine percent of those imprisoned for homicide in the United States in 1890 were men; ditto 88 percent of the homicide arrests recorded in this century. Male involvement in capital murder has been even greater. More than 98 percent of the convicts sent to Death Row have been men, a statistic that has remained practically unchanged since 1880, when it was first noted. Serial killing and mass murder, the most spectacular forms of homicide, have been virtual male monopolies.3
Those who have lived by the gun have often died by it. In the twentieth century American men have been four times more likely than women to be murder victims and between two and three times more likely to commit suicide, most often by pulling a trigger. The male-female ratio in the other leading category of violent deaths, accidents, has ranged from two-to-one among older adults to five-to-one among younger ones. Because accidental or homicidal death often conceals a self-destructive impulse—manifested by driving recklessly or picking a fight with a deadly antagonist—the male tendency toward self-annihilation may well be greater than suicide statistics alone would suggest."
American men have continued to dominate crimes of property as well as those of violence. Arrests of women for serious property offenses like robbery and burglary gradually rose from 1934 to 1994 but remained below 10 or 11 percent of the total. Because violent and serious property crimes are most likely to lead to incarceration, the number of female inmates has also been low, averaging only 5 percent of those received in state and federal prisons and reformatories during the period 1910-1982.5
The low percentage of female prisoners could reflect a tendency to suspect, arrest, convict, and imprison men rather than women. However, survey research suggests that higher male arrest and imprisonment rates are not simply artifacts of prejudice. When guaranteed anonymity, men more often admit to having shot, stabbed, or seriously injured another person. They more often confess to alcohol, drug, and gambling addictions, as well as to lying, cheating, stealing, bribery, and tax evasion. In one survey the sole disreputable activity in which women exceeded men was surreptitiously checking on a spouse or lover.6
They had reason to check. Sex research dating back to Alfred Kinsey's work in the 1940s and 1950s has consistently shown a higher incidence of extramarital and premarital sexual activity among males, and nothing in the historical record suggests otherwise for earlier periods. American men report fantasizing more about sex, having more varied fantasies, and acting on their fantasies more often than women do. Whether or not men's sex drives are stronger and more central to their identities, as some researchers believe, they are certainly more apt to get men in trouble. Incest and pedophilia are largely male offenses, as are patronizing prostitutes and purchasing pornography. So is rape, although it is a crime of violence as well as sexual impulse.7
The consistent pattern of higher male rates of socially disruptive behavior throughout American history suggests that such behavior is rooted in biological differences between men and women. Suggests, but does not prove. The pattern might be due to a peculiarity of American society or culture, some quirk of the national character. When we look elsewhere, however, we discover that the male penchant for violent, disorderly, and sexually compulsive behavior is hardly an exclusively American trait. It can be observed throughout history, across cultures, and among different primate species.
The pioneering European social scientists and statisticians of the nineteenth century were among the first to systematically document the male tendency for violent and disorderly conduct. Lambert A. J. Quetelet, who analyzed French and Belgian records from the 1820s and 1830s, discovered that men were more than four times as often accused of any crime as women, the ratio increasing in proportion to the violence and seriousness of the offense. Francis Neison noticed virtually the same pattern in English criminal statistics from the 1830s and 1840s. Emile Durkheim, in his study of suicide, found that men were three to four times more likely than women to kill themselves in every European nation for which information was available. European historians have been able to trace the record of violent death back further—to the thirteenth century in the case of England, where nine out of ten murderers were male, as were eight of ten of their victims. Male-on-male is still far and away the commonest variety of murder in contemporary European societies; likewise North America, South America, Africa, Asia, and the Pacific.8
Social statistics are sketchy or nonexistent for most tribal societies, but anthropologists who have used ethnographic reports to make cross-cultural comparisons have found men to be the more aggressive, sexually active, and less responsible gender. Tribal men also report dreaming about sex, death, and weapons more often than women. Male violence has been a fact of hunter-gatherer life for a long time. Male skeletons exhibit more clubbing and stabbing wounds than female skeletons and are wounded more often on the left side, suggesting a right-handed opponent attacking from the front. The earliest known victim was a Neanderthal man apparently stabbed in the chest fifty thousand years ago. Among primate species, the earliest fossil evidence of gender-differentiated aggression—longer male canine teeth for threatening or attacking rivals—is fifty million years old.9
Age and Male Behavior
Male behavior is most likely to be socially problematic in the late teens and twenties. This is the age range in which American men have been likeliest to kill or be killed;10 to set fires, filibuster, riot, vandalize, rob, and steal;" and to abuse alcohol and other drugs.12 These are also the years in which reckless and intoxicated driving most frequently claim the lives of men. Young women die too, often as hapless passengers of negligent male drivers.13
The surest way to reduce crime, remarks the psychologist David T. Lykken, would be to put all able-bodied males between the ages of twelve and twenty-eight into cryogenic sleep. He has a point. Though the median age of arrest is subject to historical variation (it has gone down in the United States in the last century), the arrest bulge invariably occurs in the teens and twenties and declines rapidly from the thirties on. Figure 1.1 shows the rate of criminal offenders by age and gender for England and Wales in the years 1842-1844. Figure 1.2 shows the arrest rate by age and gender for the United States in 1990. The two societies could hardly be more different: one industrializing, racially homogeneous, and thinly policed; the other postindustrial, pluralistic, and heavily policed. Yet the shape of the curves is similar, with the highest levels for men (and women) in the teens and twenties. All societies, from Sweden to Samoa, manifest the same basic pattern. "Crime everywhere and throughout history is disproportionately a young man's pursuit," sum up Richard J. Herrnstein and James Q. Wilson. "Nowhere have older persons been as criminal as younger ones." It follows that any society or community with a surfeit of young men will have, other things being equal, a built-in tendency toward a higher level of crime.14
Evolution, Aggression, and Sexuality
Crime, a social construct, is not identical with aggression, a cluster of behaviors that involve intimidation of or taking from others by threat or deed. Nevertheless, crime and aggression are connected both logically (taking life and property is often illegal) and empirically, insofar as criminologists have shown that aggressive personalities are more likely to run afoul of the law. The fact that men, on average, are more aggressive than women goes a long way toward explaining why men engage in more lawbreaking and socially disruptive behavior. So does the more impulsive male sex drive, which helps to explain men's greater involvement in rape, pornography, and other illicit sexual practices."
But why are men more aggressive and sexually impulsive? Evolutionary biologists from Charles Darwin on have stressed reproductive asymmetry. Women have a finite number of eggs and, because of the long period from conception to weaning, must invest a large amount of time and energy in their offspring. Men, who produce billions of sperm, invest little time and energy in reproduction. A man who has intercourse with many women in the course of a month can conceive many children, but in the same interval a woman can become pregnant only once. Quality rather than quantity of partners is the best reproductive strategy for females, though not for males, who increase the chances of passing on their genes through intercourse with multiple partners.
This is where aggression comes into play. Over the millennia human males, like males of all mammalian species, have competed for access to the finite reproductive capacity of females. Those who were predisposed to aggressive behavior, and who thereby intimidated or eliminated their rivals, were able to sire more offspring and pass on more genes, including those which favored aggression. This phenomenon can be observed in many animal and human societies. Dominant males inseminate females; successful warriors and potentates often have many wives and concubines.
It may be objected that aggression should be selected against because it increases the prospect of early death. In fact, aggression does lead to a shortened life span. Women die violent deaths less often than men. Castrated males live longer than those who are not castrated. But, in the Darwinian scheme of things, the longevity of an individual organism does not matter as much as its reproductive success. A trait penalized by natural selection can win out by sexual selection. Males who fight hard, mate often, and die young have a greater genetic influence on posterity than passive males who seldom mate and yet live a long time.
Prowess and risk-taking in hunting pay similar reproductive dividends. Successful hunters in tribal societies frequently have more wives than unsuccessful ones. Though they increase their own chances of sudden death, aggressive hunters and warriors enhance both their own reproductive prospects and the fitness of the group by increasing the food supply and driving away competitors. Because hunting and war-making entail cooperation, traits like the ability to lead and plan are also selected for, not just raw aggression. Men kill one another as individuals, but they also kill one another and other species through concerted action in hierarchical groups.16
The more pronounced male sex drive, like aggression, is a consequence of reproductive asymmetry. Women do not require sexual arousal to conceive. Men do. As the anthropologist Donald Symons points out, selection favored the male tendency to be aroused sexually by the sight of females. Other things being equal, men who were easily and frequently aroused mated more often, thereby increasing the species' tendency toward male concupiscence. Because women necessarily invest more in reproduction, they are more circumspect about intercourse and, in any case, do not require sexual arousal to conceive and pass on their genes. These differences, magnified by selection over thousands of generations, explain biologically the greater male preoccupation with sexuality.17
But if selection favors more aggressive and highly sexed males, why hasn't evolution produced a completely violent, rut-and-run gender? Why do men love and nurture women and children? Why, in other words, do families exist?
One answer is that sexual selection is not entirely in the hands of men. Women (or, in traditional societies, their parents or guardians) generally do the choosing, and they do so on the basis of something more than toughness and virility. They look for the ability and willingness to make a long-term commitment to support children, a responsibility that women would otherwise face alone. Courtship is the process by which male capabilities and intentions are vetted. Men with the means and disposition to share in the task of raising a family are favored because they are more likely to be approved by prospective spouses and their relatives.18 Men who commit to family life also enjoy nutritional and emotional advantages not available to bachelors, whom they typically outlive. The male evolutionary legacy is thus paradoxical: aggression and sexual drivenness on the one hand, love and nurture on the other. Both sets of traits have been selected for; both are present in varying degrees in every man's genes. Social arrangements and cultural rules influence which set of traits predominates.
Listening to Steroids
The evolutionary account of male aggressiveness and sexuality fits well with paleontological, ethological, anthropological, and historical evidence. It is consistent with a growing body of genetic research showing that personality traits, like eye or hair color, can be inherited.19 And yet it cannot be directly tested, for we have no way of recreating the circumstances of the distant human past.
We can, however, establish the hormonal basis of differences between males and females, especially the role played by testosterone. Initially, the human fetus is sexually undifferentiated. Physical differences do not occur until minute genetic differences trigger the development of testes or ovaries. The former produce testosterone, which in turn organizes the development of the fetus's genitals and central nervous system. Testosterone is why boys are born boys, and why they later become men. In the absence of testosterone the fetus will develop into a female, nature's "default" body plan.
At the onset of puberty the testes flood the body with testosterone, raising blood levels to as much as twenty times those of women and prepubertal boys. This surge of testosterone in young men has anabolic effects, including increased muscle mass and bone density, as well as androgenic effects, including hairier bodies, deeper voices, and, what is of concern here, increased libido, impulsiveness, and aggressiveness.20
We know that testosterone is causally related to these changes because its presence or absence is easily manipulated. Castrated human males, even castrated criminals, lose interest in sex and fighting. All mammals react in the same way, which is why ranchers, herders, and farmers often clip the testicles of their bulls, stallions, and other livestock. Gelded animals, like gelded humans, are more manageable. But when testosterone is artificially replaced in castrated man or beast its androgenic effects soon reappear, proving that the hormone, not the missing gonad, is responsible for the physical and behavioral changes.21
The level of testosterone is a heritable trait subject to environmental modification, typically increasing with gains in status and declining with losses. A study of the Harvard wrestling team found that the victors had significantly higher testosterone levels ten minutes after their matches than the vanquished. Winning tennis players, students graduating from medical school, and successful skydivers react to victory or achievement the same way, with elevated testosterone levels.22
Statistical analyses which have controlled for variables like race and education have shown a significant link between higher levels of testosterone and verbal and physical aggression, drug and alcohol abuse, and violent offenses. Adolescents with high levels of the hormone are more prone to juvenile delinquency and to more serious and violent crimes later in life, particularly if they lack education, jobs, and spouses and are uninvolved in organized, prosocial groups. The effects of testosterone in humans are socially mediated. Married, college-educated males who have high serum testosterone levels are one thing; unattached, alienated drifters with abnormally high levels are quite another. Testosterone anomalies, in short, appear to be most dangerous among the least socialized men.23
Testosterone has masculinizing effects on females as well as males. Testicular grafts in hens will produce the comb and wattles of a cock. Human females exposed in the womb to synthetic androgenic hormones exhibit greater aggressiveness and other conventionally masculine traits later in life. Female-to-male transsexuals treated with these drugs show greater proneness to anger.24 Those who say that gender is constructed are speaking a biological and a social truth. Hormonal manipulation and socialization are both powerful means of influencing male and female behavior.
The steroid scandal underscores this point. Steroids, synthetic drugs that mimic testosterone, were tested clinically in 1937 and were reportedly used by German troops during World War II to enhance strength and aggressiveness. Steroid "doping" spread among athletes and weight lifters after the war, and today the drugs are commonly, if illegally, used in sports and body-building circles. The temptation is enormous: a male athlete can add as much as forty pounds of lean muscle in one year. Female athletes and body builders also use steroids to bulk up. They frequently experience masculinizing effects, among them growth of facial and body hair, deepening of the voice, enlargement of the clitoris, and increased libido and aggressiveness. Emotional effects reported by male athletes include increased aggressiveness, energy, explosiveness, self-regard, hostility toward others, tolerance of pain, and supercharged libido. Steroid use in male athletes is a chemical exaggeration of the hormonal changes experienced naturally by young men, with predictably exaggerated results: more sex, more violence, and more death.25
There are dissenting voices. The strength of the correlation between steroid use and violence has been questioned. Some researchers suspect that it is not steroid use alone but its interaction with alcohol that is especially dangerous. And there is certainly more to the biology of human aggression than testosterone and its synthetic equivalents. Neurotransmitters like dopamine, norepinephrine, serotonin, and GABA (the levels of which can likewise be affected by alcohol and drug use) are part of the complex neurochemical story that is only beginning to be unraveled.26 Even so, it seems beyond question that testosterone plays a gender-linked role in human aggression. It is also undeniable that its anabolic effects increase the potential for injury. Better to be hit by a boy weighing 70 pounds than a teenager weighing 140 pounds—or a galoot who has bulked up to 240 on steroids.
Young men, in short, are biologically primed for deeds that may have been reproductively advantageous in the distant past but are problematic in societies that have come to value disciplined competence over raw aggression. Not only are young men hormonally inclined to violent and socially forbidden behavior, they have a greater capacity for it. They can hit hard, run fast, break down doors, and climb in through windows. Though the conventions of the mystery drama invite us to think of crime as something carefully planned and rationally motivated, the thefts and outrages of everyday life are most often the opportunistic acts of young men impulsive or undisciplined enough to break the law and strong or quick enough to get away with it. Street crime is a young man's game. After the age of thirty both physical capacity and sexual drive begin a slow but sustained decline, another important reason why middle-aged and older men are less likely to be arrested for crimes like assault, robbery, and rape.
The decline of criminal involvement with increasing age is also partly attributable to the "selecting out" factor. The most reckless men, the James Deans and Sid Viciouses of the world, die young. In their graves they trouble no one, least of all the other members of their birth cohort as criminologists follow them through their forties, fifties, and so on. Not only do men lose physical capacity and sexual drive as they age, they lose their wildest contemporaries and companions. This too makes a difference.
So does individual character. While human behavior may be predictable in the aggregate, each person's actions are ultimately uncertain and indeterminate. "Individuals vary," Sherlock Holmes observed to Dr. Watson in The Sign of Four, "but percentages remain constant." The same is true of statistical generalizations about age and gender, which are best thought of as distributions around a mean. They do not determine individual destinies, nor do they lack exceptions. There have been continent young men of irenic temperament, just as there have been older men and women of opposite bent. Such exceptions do not disprove the generalizations, however, because the generalizations themselves are of a probabilistic nature.
American Population Imbalance
Statistically, then, we may say that violent and disorderly behavior occurs most frequently among men, particularly among men in adolescence and young adulthood. As it happened, America had both a surplus of men and an unusually large percentage of young people during its first three centuries.
Countries with high birth rates have youthful populations, as the most casual inspection of the world's developing nations will attest. America, which had a phenomenally high birth rate during its developing years, was no exception. Endowed with a relative abundance of land and natural resources, the mainland colonists married young and raised large numbers of children, who in turn married and raised large numbers of children, producing a continuous baby boom. That boom, combined with the fact that most colonists did not live past sixty, kept the median age of the population—sixteen years in the eighteenth century—very low. Family limitation by native-born Protestants slowed the boom somewhat during the nineteenth century, but America in 1900 was still a youthful nation, with a median age of only twenty-three years.27
It was also, relative to Europe, a disproportionately male nation. It was born that way. From Sir Walter Raleigh's attempted settlement of Roanoke Island in 1585 to the formal declaration of American independence in 1776, the British mainland colonies always had a surfeit of men. In 1608 Jamestown was home to but two women—as good a reason as any why the settlement had such a disastrous early history. Although the demographic imbalance of early Jamestown was extreme, the seventeenthcentury colonial male surplus was both sizable and persistent, particularly in the South, where a continuing influx of servants, convicts, and slaves kept the population predominantly young and male.28
The colonial population achieved a more normal balance in the course of the eighteenth century. By 1790 the gender ratio (the number of males per 100 females) stood at just under 104 for the white population of the United States. The national gender ratio remained in the 102-105 range for most of the nineteenth century, peaked at 106 in 1910, and thereafter began to decline. It was not until 1946, however, that the country finally acquired a female majority. That female surplus, largely a product of changing immigration patterns and the widening difference in life expectancy between men and women, has been a fact of American life ever since.29
European nations developed and sustained female majorities much earlier than the United States. Most of them, in fact, had female surpluses during the entire period of America's colonization and development as an independent nation. The primary reason was migration. Forced or voluntary, American immigrants were heavily male. The proportion of men to women among transported convicts was four to one; among slaves, imported legally or otherwise, upwards of two to one. The proportion among indentured servants, the largest group of colonial immigrants, was three to one during the seventeenth century, increasing to nine to one during the eighteenth.30
The nineteenth-century successors to the indentured servants, Chinese laborers indebted for their passage, had the most unbalanced gender ratios of all. In 1890 there were twenty-seven Chinese men for every Chinese woman in the United States: "more monks than rice porridge," as they described the situation. Women were at such a premium that Chinese laborers were said to line up for a block and pay an ounce of gold merely to gaze upon a famous courtesan's face.31
Nineteenth-century European immigrants who came without legal or financial obligations were far more likely to come in familial groups, but even among them there was a surplus of prime-age male workers. Only post-famine Ireland, where economic and marital prospects for young rural women were particularly bleak, furnished more female than male immigrants, and then only by a small margin. Male majorities were the norm for immigrants, and they helped to keep the national gender ratio well above parity.32
The differences between European and American gender ratios that grew out of migration were magnified by the effects of warfare. With the single exception of white southerners during the Civil War, Americans, unlike Europeans, never suffered catastrophic military losses. It is true that intermittent warfare against the Indians and the French from 1675 to 1760 made widows of many New England women, and that both the American Revolution and the Civil War produced appreciable casualties relative to the total population. Yet these sacrifices were small compared to those made by European nations. France lost a million soldiers during the Revolutionary and Napoleonic wars and more than a million in World War I. During World War II the Soviet Union, Poland, Germany, and Yugoslavia each suffered more deaths than the United States endured in its entire military history from the Revolution to Vietnam. Because most of the casualties were male, the gender ratios of European nations were noticeably lower after major conflicts. The first Soviet census taken after World War II, in 1959, counted nearly 21 million more women than men. That year there were only 82 men for every 100 Soviet women, compared to 98 men for every 100 women in the United States." Warfare may well have been more important for its indirect than for its direct effect on the American gender ratio. During the Revolution, the Civil War, and World War I immigration, which would ordinarily have been 60 to 70 percent male, fell off precipitously. The Civil War reduced immigration by half a million, World War I by three million or more. In time of war European nations were reluctant to permit emigration, above all the emigration of young men. One can even think of the maritime impressment that bedeviled the young republic during the Napoleonic wars as a type of enforced reverse migration, brought about by the Royal Navy's ability to siphon off American sailors into its depleted and desertion-prone crews. The impressment of several thousand men into British service did not significantly reduce the overall American surplus, however. Supernumerary males remained a peculiarity of American life throughout the early nineteenth century and beyond.34
Circumstances thus conspired to endow America with an abundance of young men. From the standpoint of economic growth this was undoubtedly a good thing. Young adult males have an enormous potential for labor, especially when the labor in question requires vigorous physical exertion. The developing nation needed field hands and stevedores and ironworkers. The demand was initially met by the sellers of servants and convicts and slaves, who received higher prices for male workers than for female, and later by wave after wave of voluntary male immigrants.35
The problem, from the standpoint of social control, was that young men who worked hard also lived and played hard, often to the point of causing serious violence and disorder. Endocrine research, bolstered by historical, criminological, and cross-cultural studies, suggests that this tendency is universal and that it has a biological basis. But not that the problematic behavior of young men is biologically determined, for their innate potential for trouble is everywhere mediated by cultural norms and contained, however imperfectly, by social institutions.