A serious program of life-extension and cryonics would be amiss to neglect the dangers posed from death by murder. In other essays I have discussed the dangers of death due to cardiovascular disease, cancer, dementia, aging and accidents. If aging & disease are eliminated and the world is made much safer (and/or surgical repair is vastly improved) so as to reduce the danger of death by accident, the major causes of death will be suicide and homicide. I believe that no matter how advanced the technology, people will always have the means of killing other people — and of killing themselves. In this essay I want to focus on murder (homicide), which is the tenth leading cause of death for males in the United States (much less common for females).
Gathering as much information as possible about the conditions under which murder can occur is a major step towards being able to take preventative action. A prudent step towards reducing one's chance of being murdered is to avoid being in the wrong place at the wrong time. So it therefore seems reasonable to begin by determining what some of the "wrong places" are.
Homicide rates are typically quoted as per 100,000 people per year. The rates I quote will be for VICTIMS unless I say otherwise. The 10-year average homicide rates (1987-1996) for Canada averaged 2.3 and for the United States averaged 8.8.
(return to contents)
Homicide statistics for much of the world are hard to come by and often very unreliable. The most comprehensive list I could find was from Interpol for the mid-1970s (International Crime Statistics). (Interpol presently only releases crime statistics to police organizations.) The top 10 countries for murder were:
|(5) Netherlands Antilles||12|
|(7) Sri Lanka||12|
|(9) Trindad & Tobago||10|
There are more recent statistics, which include fewer countries than the Interpol statistics: Nationmaster.com Map & Graph of Murders (per capita)
The top 10 countries for homicide conviction in 2003 were:
|(2) South Africa||51|
Both of these sources of statistics give as least as much evidence for the difficulty of getting accurate homicide data as they do of homicide rates. Among the top ten countries in the Interpol list, only Jamaica appears on the "Nationmaster" list. Neither list includes Brazil, which THE ECONOMIST (19-June-1999) cited as having a murder rate of about 23, with the highest percentage (88%) of murders being committed by firearms in the world.
Even in 2002 the statistics gathering for world-wide homicide rates shows huge gaps. The 2002 WORLD REPORT ON VIOLENCE AND HEALTH published by the World Health Organization (WHO, Geneva) lists murder rates for 75 countries. Eight of the ten countries included in the 1970s Interpol list do not appear in the WHO list. Such huge countries as India, Pakistan and Indonesia are omitted — as are all African countries. Jamaica appears in the list reporting a dubious 2 murders for 1991. Trinidad & Tobago are shown as having 11.4 murders per 100,000. I have extracted the ten worst countries for murder from the 75 listed by the WHO report:
|(2) El Salvador||50.2|
|(3) Puerto Rico||41.8|
|(7) Russian Federation||18.0|
I have extracted the ten safest countries for murder, based on reported homicide rates, from the 2002 report:
|(6) Hong Kong||2.4|
The figure for Israel is for 1997, which was before the rash of suicide bombings — although Israel was probably not as safe even in 1997 as the number might indicate. Hong Kong is counted as a country — it was a country in 1996, the year for which the statistic is reported. Northern Ireland is not included in the reported figure for Ireland.
An international chart summarizing world homicide and suicide
rates is available from the World Health Organization, despite the
fact that the data cannot be any better than the data-gathering
capabilities of the various countries. A somewhat confusing
distinction is made between the "South-East Asia Region",
which includes India, Indonesia, North Korea and Thailand (among
other countries) — and the "Western Region", which includes
South Korea, Laos, Viet Nam, China, Philippines, Australia and Fiji (among
International murder rates for cities are difficult to obtain outside the developed world. According to some reports Bagota (Colombia), Karachi (Pakistan), Lagos (Nigeria), Dhaka (Bangladesh) and Port Moresby (Papua New Guinea) have some of the highest murder rates in the world, but there are no reliable statistics and Interpol refuses to make its statistics public. Caracas, Venezuela reputedly has a murder rate over 100 per 100,000.
A 1998 BBC News Report of a UK Home Office survey compared murder statistics for select cities in Europe and North America. Although most of the cities are larger than one million in population, Geneva has only a couple hundred thousand and both Amsterdam and Belfast are well under a million in population. Washington, DC at half-a-million is much smaller than Detroit (which has an equivalent murder rate), yet Detroit is excluded. The worst 20 cities for murder rate listed were:
|(1) Washington, D.C., USA||69.3|
|(2) Philadelphia, USA||27.4|
|(3) Dallas, USA||24.8|
|(4) Los Angeles, USA||22.8|
|(5) Chicago, USA||20.5|
|(6) Phoenix, USA||19.1|
|(7) Moscow, Russia||18.1|
|(8) Houston, USA||18.0|
|(9) New York City, USA||16.8|
|(10) Helsinki, Finland||12.5|
|(11) Lisbon, Portugal||9.7|
|(12) San Diego, USA||8.0|
|(13) Amsterdam, Netherlands||7.7|
|(14) Belfast, N.Ireland, UK||4.4|
|(15) Geneva, Switzerland||4.2|
|(16) Copenhagen, Denmark||4.0|
|(17) Berlin, Germany||3.8|
|(18) Paris, France||3.3|
|(19) Stockholm, Sweden||3.0|
|(20) Prague, Czechoslovakia||2.9|
For the most up-to-date statistics on murder rates worldwide by country, see the Wikipedia page: List of countries by intentional homicide rate.
(return to contents)
In 2006 in the United States homicide was the second leading cause of death for infants. Homicide with a firearm was the second leading cause of persons between the ages of 10 and 24, the third leading cause of death for persons between ages 25 and 34 and the fourth leading cause of death for persons between ages 5 and 9 or between ages 35 and 44. For persons between ages 45 and 64 homicide with a firearm was the seventh leading cause of death. Homicide with a firearm or by any means was not among the top ten causes of death for persons aged 65 or older whereas there were at least two forms of homicide among the top ten causes of death for all persons under age 44. (See 10 Leading Causes of Injury Death by Age Group for details.)
To determine the States where homicide is most and least likely to occur I have extracted the most & least dangerous States based on FBI Uniform Crime Reports for the year 2003:
|(7) South Carolina||7.2|
|(2) South Dakota||1.3|
|(3) New Hampshire||1.4|
|(7) North Dakota||1.9|
|(10) Rhode Island||2.3|
States (and Puerto Rico) can be grouped by region with summary data for 2002:
|(1) Puerto Rico||20.1|
|(2) East South Central||6.8|
|(3) West South Central||6.8|
|(4) South Atlantic||6.7|
|(5) East North Central||5.8|
|(8) Middle Atlantic||4.7|
|(9) West North Central||3.2|
|(10) New England||2.4|
The regions above can be defined as:
East South Central:Alabama, Kentucky, Mississippi, Tennessee
West South Central:Arkansas, Louisiana, Oklahoma, Texas
South Atlantic:Delaware, DC, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, West Virginia
East North Central:Illinios, Indiana, Michigan, Ohio, Wisconsin
Pacific:Alaska, California, Hawaii, Oregon, Washington
Mountain:Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, Wyoming
Middle Atlantic:New Jersey, New York, Pennsylvania
West North Central:Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, South Dakota
New England:Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont
A Scientific American article (June 1999) accounts for the high murder rates in the South on the grounds of a "culture of honor". A white man living in a small county in the South is four times more likely to kill than one living in a small county in the Midwest. Southerners showed higher levels of cortisol and testosterone in response to an insult. Murder rates due to arguments are higher in the South and Southwest, but murder rates associated with felony (robbery or burglary) are lower.
For the United States I have also extracted the most & least dangerous large cities (having a population over 500,000) as reported by SafeStreetsDC.com for the year 2002:
|(1) Washington, DC||45.8|
|(9) Los Angeles||17.5|
|(2) El Paso||2.4|
|(3) San Jose||3.1|
|(5) San Diego||3.8|
|(8) New York||7.3|
|(9) San Francisco||7.3|
|(10) Oklahoma City||8.5|
The chance of being murdered in Washington,DC in 1990 was 3 times greater than the chance of an American soldier being killed in the Gulf War. The average American city with a population of 250,000 or greater has a murder rate of about 20, whereas cities in the 100,000 to 250,000 range have a rate of about 12. About as many Americans were killed (over 54,000) in New York City between 1962 and 2002 as died in the Vietnam War, but the murder rate in 2002 was only about a quarter what it was in 1990, when there were a record 2,245 murders. Mayor Giuliani is credited with the transformation.
(return to contents)
Homicide statistics for Canada can be found at the Statistics Canada website. The nation-wide average for 2003 was 1.73 homicides per 100,000. Ranking all of the Provinces and Territories by murder rate for the year 2003 results in the following table:
|(2) Northwest Territories||9.55|
|(6) British Columbia||2.24|
|(10) New Brunswick||1.07|
|(12) Nova Scotia||0.85|
|(13) Prince Edward Island||0.73|
The 1990 homicide statistics for the murder rate for large Canadian cities:
|(12) St. John's||0.00|
There were no murders in St. John's,Newfoundland in 1989 or 1990.
Overall, murder rates have declined between 1991 and 1996 in both the US (9.8 to 7.4) and Canada (2.7 to 2.1). Statistics Canada reports that murder rates continued to decline to a 1999 figure of 1.76, the lowest since 1967 (which was 1.66). Canadian gang-related homicides (drugs & revenge), however, doubled yearly from 1996 to 1999. Although aboriginal people represent only 3% of the Canadian population, aboriginals account for 20% of those accused of homicide and for one-sixth of all homicide victims.
(return to contents)
I find that statistics about homicide in the United States are the most readily available, so the rest of my essay will rely on American statistics, which often contain information regarding race. The richest source of homicide statistics is the US Department of Justice ( www.ojp.usdoj.gov/bjs/homicide/overview.htm). (Results of the September 11, 2001 terrorist attack on the World Trade Center are excluded from these statistics.)
In the 1976-1997 period, the average age of victims fell from 35 to 31 and the average age of murderers fell from 31 to 27. 16% of homicides involved multiple murderers, whereas 4% of homicides involved multiple victims in 1997.
The most commonly cited reason for homicide is argument (including arguments about money & property under the influence of alcohol or narcotics). One third of all homicides in 1997 were triggered by arguments. Felony (rape, theft, narcotics, etc. ) accounted for a fifth of homicides and gang killings accounted for one twentieth. About a third were of unknown motive and the other 10% were miscellaneous motives.
Between 1976 and 1994 the average age of murderers fell from 30.3 to 26.4 and the average age of victims fell from 35.2 to 31.3. By 2004 murderers were at least twice as likely to be in the 18-24 age group as in the 14-17 or 25-34 age group (the next highest age groups). By 2004 victims were about three times as likely to be in the 18-24 age group as in the 14-17 or 35-49 age group and about twice as likely to be in the 25-34 age group. The murder rates are lowest for the above 50 and below 14 age groups and the victimization rates for those age groups (although much higher than the murder rates) are the lowest for any age groups.
For the 1976-2004 period men committed 93.3% of felony murders and 85.5% of murders due to argument. Men committed 91.2% of gun homicides, 79.1% of arson homicides and 63.3% of poison homicides. The relationship of killers to murder victims classified by gender can be summarized as follows (rounding errors give a total of 99.9%):
|Male kills male||65.2%|
|Male kills female||22.6%|
|Female kills male||9.7%|
|Female kills female||2.4%|
In 2004 about a third of women were killed by intimates, whereas only about 3% of men were killed by intimates. For the 1990-2004 period two-thirds of spouse and ex-spouse victims were killed by guns. Detailed statistics for the 1976-2004 period is summarized in the following table (rounding errors give a total of 100.1%):
Intimate:Spouse or boyfriend/girlfriend
Family:Non-spousal family member,
For murders where the murderers and victims were classifed as black or white, the breakdown of murderers and victims for 2004 can be summarized by race & gender:
The relationship of killers to murder victims classified by race for 2004 can be summarized as follows:
|White kills white||46.1%|
|Black kills black||40.7%|
|Black kills white||8.5%|
|White kills black||3.5%|
|Other kills other||0.6%|
|Other kills white||0.4%|
|White kills other||0.2%|
|Black kills other||0.1%|
|Other kills black||0.0%|
(According to the US Census Bureau the US population in 2004 was 80.4% white and 12.8% black.)
White (usually European) immigrants to the US are 2.1 times as likely to be homicide victims as native whites, whereas black immigrants are only 60% as likely to be homicide victims as native blacks ( www.ph.ucla.edu/sph/pr/wr038.html). I speculate that white immigrants are naive of the dangers of the new country, whereas black immigrants would not be associated with gangs or ghetto-culture.
The 15-24 year age group had the highest homicide victimization rate, and were less likely to be murdered while under the influence of alcohol than those between the ages of 25-54. August was the most popular month to be murdered, followed by March and October. Saturday was the most popular day-of-the-week to be murdered, followed by Friday. Murder rates are higher in the afternoon than in the morning, but are highest at night — climbing steadily from 6 pm, peaking at 11 pm and declining thereafter.
During the 1980s, homicide was the leading cause of occupational death for American women and the third leading cause of occupational death for American men (www.cdc.gov/niosh/homicide.html). During the 1980s the average annual occupational homicide rate was 0.7 (out of 100,000, as usual). Occupations with the highest homicide rates were:
|Law enforcement officer||9.3|
|Gas station worker||4.5|
The most dangerous workplaces were:
|Justice/public order establishments||3.4|
In the 1997-2000 period 80.1% of work-related homicides were due to shooting, 9.0% due to stabbing and 6.2% due to hitting, kicking or beating.
Homicide-suicides involve cases where a perpetrator commits one or more homicides before committing suicide. Peculiarly, the rate for homicide-suicide is between 0.2 to 0.3 for most countries, despite large differences in homicide rates in the different countries [AMERICAN JOURNAL OF PSYCHIATRY 155(3):390-396 (1998)].
Within the United States, states with a high unemployment rate have a high homicide rate, but there is a negative correlation between suicide rate and unemployment. Suicide is 3 times more highly correlated with divorce than is being a murder victim [AMERICAN JOURNAL OF DRUG & ALCOHOL ABUSE 21(1):147-150 (1995)].
In a Memphis, Tennessee study, 85% of murderers and 75% of murder victims were intoxicated during the murder. A review of 331 American medical examiner (coroner) studies published between 1975 and 1995 found that victims tested positive for alcohol in 29% of suicides, 38.5% of unintentional injury deaths, 39.7% of motor vehicle deaths and 47.1% of homicides. Half of those murdered by drowning were intoxicated at the time, but only 16% of those murdered by strangling or suffocation were intoxicated [ANNALS OF EMERGENCY MEDICINE 33(6):659-668 (1999)]. This could mean that it pays to be able to put up a fight! However, the vast majority of murder victims were killed by gunshot or stabbing/cutting of which 30.6% and 43.0% of the victims were intoxicated, respectively. The higher rate of intoxication for stabbings may indicate a greater impulsiveness for crimes committed with this kind of weapon (grabbing for a handy knife or pair of scissors in a moment of drunken rage).
Resisting an assailant may not be futile and dangerous. A very large survey by the US Justice Department of survivors (!) of violent crimes found that 73% resisted by trying to attack/capture the offender, by running-away/hiding or by trying to persuade/appease. Of those who resisted, 63% felt their resistance was useful and 7% said their resistance was harmful. It would be valuable to know the full statistics — ie, of all those who were attacked, not just the survivors, but technical difficulties...
The 28-July-2000 (Vol.289,No.5479) issue of SCIENCE had a special coverage of violence. One article (p.575-579) noted that animals & humans with lower brain serotonin tend to be more impulsive & aggressive — whereas the opposite was true for those with higher brain serotonin. This suggests Prozac as a more selective anti-violence drug than the sedatives & antipsychotic drugs usually used for violent mental patients. Higher vasopressin in the hypothalamus also correlated with increased antisocial aggression.
Another article (p.580-581) noted that "Hitting, Biting and Kicking" behavior was observed in nearly half of 2-year-olds, but declined somewhat steadily to the age of eleven (the last point graphed). Personality trait "risk factors" for violence (which twin studies indicate are genetic) include impulsivity, low IQ and a temperament predisposed toward anger, vindictiveness and blaming others. Physical correlates include lower skin conductance and slower brainwaves — indicative of a low autonomic arousal (ie, the person is less anxious & inhibited and requires more extreme behavior to experience stimulation).
A graph on p.582 shows that since 1988 the 18-24 year age-group has accounted for the most murders in the US, followed by the 14-17 year age-group and then by the 25-34 year age-group. (The fact that the youngest age-group is a 3-year span, the middle age-group is a 7-year span and the oldest age-group is a 10-year span probably results in an understatement of the homicidal tendencies of the youngest ages.) The same graph shows homicide peaking in 1980, falling until 1985, peaking higher in 1992 and then falling again. The 1992 peak has been blamed on crack cocaine. One explanation for the declines is increased incarceration rates, which did not change in the 1925-1970 period, but has quadrupled since 1970. A study of dangerous neighborhoods in St.Louis & Chicago showed that every 10% increase in number of people incarcerated reduced the number of homicides 15-20%. Another researcher, however, estimated that 50% of the drop in crime is due to the 1973 Roe vs. Wade decision which allowed for the abortion of many unwanted children who could have become killers.
From 2000 to 2010 the number of "justifiable homicides" (called acts of self-defense) in the USA increased from 176 to 326 while the overall homicide rate declined slightly. The increase in justifiable homicides only occurred in states that had passed "stand your ground" laws — which give people more leeway to attack and kill someone who is threatening them.
(return to contents)
Two thirds of all 1992 US murders were accomplished with firearms. Handguns were used in about half of all murders. Sharp instruments were used in 17% of murders and blunt instruments in about 6%.
Gun control laws are stiffer in Canada, and many claim this accounts for the murder rate being lower in Canada than in the United States. 65% of US homicides were committed with firearms, versus 32% in Canada. However, a large American study indicated that liberalized laws for carrying concealed weapons reduced murder rates in the US by 8.5%. US homicide rates in the year 1900 were an estimated 1 per 100,000 — at a time when anyone of any age could buy a gun. Statistics-gathering may have been less thorough at that time — and few people had the money or interest to buy guns. But American gun supply (including handguns) doubled from the 1973-1992 period, during which homicide rates remained unchanged (WALL STREET JOURNAL, 4-Aug-2000, p.A10).
Politicians in Massachusetts have cited the State's tough gun control laws as the reason for its low murder rates. However, the adjacent states of Maine, New Hampshire and Vermont have some of the least stringent gun control laws in the US, yet the first two have lower murder rates than Massachusetts and the murder rates in Vermont are comparable to those in Massachusetts. Murder rates in Boston increased 50% in 2004 over the previous year, while murder rates in Los Angeles, Miami, Washington and many other major cites saw murder rates decline.
(return to contents)
About half of the countries of the world still have the death penalty.
On average, two countries per year have abolished the death penalty since
1976, the year it was abolished in Canada. Since then the Canadian murder
rate has dropped from about 3 to about 2, undermining the argument that
capital punishment is a deterrent. China, Iran and Nigeria
accounted for 87% of reported executions in 1994, although there is debate
about the number of executions in Iraq. An estimated 23 people proven
innocent were wrongly executed in the United States between 1900 and 1995
From the gangland era of the 1930s to 1963 there was a gradual decline in both murders & executions in the United States. In 1963 the US Supreme Court imposed rules on confessions & searches that accompanied a popular sentiment increasingly opposed to capital punishment — and in 1972 struck down capital punishment laws as being "arbitrary and capricious". There were no executions in the United States between 1967 and 1977. Murder rates soared to levels not seen since the 1930s and remained at that level until the late 1970s when sentiment changed and execution began to be increasingly reinstated. As executions rose, the murder rate declined through the 1990s. In 2002 the Supreme Court ruled that the mentally retarded cannot be executed and that only juries can impose the death penalty — two rulings that affected nearly a quarter of death-row inmates. (See The Death Penalty in the U.S. for a more detailed history.)
Opponents of capital punishment generally hold that capital punishment is inhumane and has a "brutalizing effect" on society. They will often also say that capital punishment is applied in a haphazard manner — if not systematically racist. They deny that anyone commits a crime having a concern about the consequences of getting caught. And many assert that the execution of even one wrongly convicted person is too high a price to pay, while others assert that execution is too high a price to pay whether the convicted person is innocent or not.
Abolitionists have pointed to the fact that states with the highest execution rates have the highest murder rates, whereas proponents have suggested that high murder rates had forced the adoption of execution. On March 1, 1847 the State of Michigan became the first English-speaking territory in the world to abolish the death penalty. It may be no accident that Detroit rivals Washington, DC as the city with the highest murder rate among American cities having a population over half-a-million. Texas, the state with the highest number of executions, dropped from being the state with the second highest murder rate to the 15th in the 1990s after beginning lethal injection in 1982. Thousands of murders are committed yearly by murderers released from prison — a problem which could be eliminated by ensuring that convicted murderers are never released from prison.
In the early 1960s the vast majority of murder victims were acquainted with the murderer, but by the year 2000 nearly half of murder victims were strangers. This may undermine the argument that murders are impulsive crimes of passion wherein the threat of execution is not a deterrent. Murderers who kill their victim during a pre-meditated rape or robbery may well have enough familiarity with the criminal justice system to realize that the chance of escaping by killing a victim-witness may be worth the risk if execution upon capture is unlikely. Persons already habituated to prison life may not regard possible return to prison as much of a deterrent. If this argument is true, then humanitarian abolitionists must reconcile the 100,000 lives of American homicide victims who might have survived the 1963-1997 period against the lives of murderers who were not executed.
(return to contents)
A good piece of advice would seem to be, "Don't hang-around with (or marry-into) a bad crowd." That includes occupations that require associations with dangerous people, such as taxi-driving, liquor store work and police work, but it also includes country & city of residence. Other adages would be, "Don't be the guardian of valuables, even if those valuables are your own" and "stay sober".
Yet another adage is, "Don't make enemies." I believe this last adage will be the ultimate challenge. In the long run, if human lifespans begin to span centuries, this would be the most crucial, since I believe that random, stupid, felonious and impulsive violence would decline sharply. The most pre-meditated, technically sophisticated and well-planned murders would be the ones which would succeed. Such murders would likely not be for material gain, but to exterminate an enemy for whom there is bitter animosity. And in the future, ethnic or racial motives are likely to become less common, so the attack would be very PERSONAL.
Assassination attempts against heads-of-state has a long history. More recently attacks using anthrax-laced letters have targeted a wider range of political figures. Celebrities of all kinds face an increasing risk of murder or attack by obscessed persons. John Lennon's 1980 shooting marked the advent of the recent phenomenon of celebrity stalking. A stalker wanting to slit Madonna's throat scaled the walls of her estate and Steven Spielberg was targeted by a man who wanted to rape him. Movie stars now employ professionals to review their mail for evidence of stalkers. So public prominence is definitely associated with increased risk.
To end on a more upbeat note, Stephen Pinker makes a convincing case for the decline of violence with the progress of civilization. The human slaughter that accompanied the first half of the twentieth century stands in sharp contrast to the second half of the twentieth century. Pinker argues that the trend of decreasing violence has been a continuing aspect of human history. Murder rates are estimated to have declined in Western Europe more than an order of magnitude from the Middle Ages to the twentieth century. Non-state societies (tribal societies) have had murder rates estimated to be nearly another order of magnitude worse than Western Europe since the Middle Ages.
For details about the risk of death from all causes other the murder — see my essay Causes of Death .
For a history of mafia killers and government assassination attempts see my book Schemers in the Web.