the official site of Michael Shermer

top navigation:
Scientific American

The Decline of Violence

published October 2011 | comments (10)
Be skeptical of claims that we live in an ever more dangerous world
magazine cover

ON JULY 22, 2011, a 32-year old Norwegian named Anders Behring Breivik opened fire on participants in a Labour Party youth camp on the island of Utoya after exploding a bomb in Oslo, resulting in 77 dead, the worst tragedy in Norway since World War II.

English philosopher Thomas Hobbes famously argued in his 1651 book, Leviathan, that such acts of violence would be commonplace without a strong state to enforce the rule of law. But aren’t they? What about 9/11 and 7/7, Auschwitz and Rwanda, Columbine and Fort Hood? What about all the murders, rapes and child molestation cases we hear about so often? Can anyone seriously argue that violence is in decline? They can, and they do—and they have data, compellingly compiled in a massive 832-page tome by Harvard University social scientist Steven Pinker entitled The Better Angels of Our Nature: Why Violence Has Declined (Viking, 2011). The problem with anecdotes about single events is that they obscure long-term trends. Breivik and his ilk make front-page news for the very reason that they are now unusual. It was not always so.

Take homicide. Using old court and county records in England, scholars calculate that rates have “plummeted by a factor of ten, fifty, and in some cases a hundred—for example, from 110 homicides per 100,000 people per year in 14th-century Oxford to less than 1 homicide per 100,000 in mid-20th-century London.” Similar patterns have been documented in Italy, Germany, Switzerland, the Netherlands and Scandinavia. The longer-term trend is even more dramatic, Pinker told me in an interview: “Violent deaths of all kinds have declined, from around 500 per 100,000 people per year in prestate societies to around 50 in the Middle Ages, to around six to eight today worldwide, and fewer than one in most of Europe.” What about gun-toting Americans and our inordinate rate of homicides (currently around five per 100,000 per year) compared with other Western democracies? In 2005, Pinker computes, just eight tenths of 1 percent of all Americans died of domestic homicides and in two foreign wars combined.

As for wars, prehistoric peoples were far more murderous than states in percentages of the population killed in combat, Pinker told me: “On average, nonstate societies kill around 15 percent of their people in wars, whereas today’s states kill a few hundredths of a percent.” Pinker calculates that even in the murderous 20th century, about 40 million people died in war out of the approximately six billion people who lived, or 0.7 percent. Even if we include war-related deaths of citizens from disease, famines and genocides, that brings the death toll up to 180 million deaths, or about 3 percent.

Why has violence declined? Hobbes was only partially right in advocating top-down state controls to keep the worse demons of our nature in check. A bottom-up civilizing process has also been under way for centuries, Pinker explained: “Beginning in the 11th or 12th [century] and maturing in the 17th and 18th, Europeans increasingly inhibited their impulses, anticipated the long-term consequences of their actions, and took other people’s thoughts and feelings into consideration. A culture of honour (the readiness to take revenge) gave way to a culture of dignity (the readiness to control one’s emotions). These ideals originated in explicit instructions that cultural arbiters gave to aristocrats and noblemen, allowing them to differentiate themselves from the villains and boors. But they were then absorbed into the socialization of younger and younger children until they became second nature.”

That second nature is expressed in the unreported “10,000 acts of kindness,” as the late Stephen Jay Gould memorably styled the number of typically benevolent interactions among people for every hostile act. This is the glue that binds us all in, as Abraham Lincoln so eloquently expressed it, “every living heart and hearthstone all over this broad land” through “the mystic chords of memory” that have been touched again by these better angels of our nature.

topics in this column: , , ,

10 Comments to “The Decline of Violence”

  1. Rob Says:

    Surely you don’t really mean that 0.8% of Americans died from domestic homicides and wars in 2005? That would be something like 2.5 million people. I suspect you mean that 0.8% of deaths in America were homicides or war deaths?

  2. Glennzaah Says:

    I did a double-take at that 0.8% figure too. It makes the USA sound roughly as bad as 14th century Oxford! (ie. 110 homicides per 100,000 people )

  3. ron Says:

    i am sympathetic to this premise, but to truly claim that violence has declined one would need to take into account advances in medical technology. surely the vast majority of acts of violence that would have resulted in death 100 years ago are not recorded as homicides today because the victim has a hugely greater chance of surviving?

  4. Samuel Álvarez Says:

    It says “eight tenths of 1 percent” not “eight tenths of 100 percent”.

  5. BillG Says:

    Modern times has the potential that yester-year has not:
    weapon technology – causing a de-evolution in Pinker’s trends and percentages almost instantly.

  6. Glennzaah Says:

    ron and BillG: Good points. Measuring “deaths resulting from violence” is not the same thing as measuring “acts of violence”. It’s just the easier thing to measure. If there was a constant relationship between the two throughout history then there’d be no problem here. But the “Deaths Per Violent Act” ratio is going to be either decreased or increased by different factors (eg. medical technology on one hand and weapons technology on the other).

  7. Glennzaah Says:

    Samuel Álvarez: Just to clarify – “eight tenths of 1 percent” means 0.8%, not 80%. I don’t think anyone talked about a figure of 80%. I think that Rob interpreted the article correctly to mean: “0.8% of American deaths were from homicides or wars” rather that “0.8% of Americans died from homicides or wars” as it seemed to say in the text.

  8. James Babb Says:

    They seem to be overlooking “legal” state violence. In Philadelphia, police assault 200,000 innocent people a year for warrantless “stop and frisk” searches. How about the millions of gropings by TSA agents at US airports? What about all the highway robberies conducted by state police? Are these acts of violence counted? How about the millions of kidnapped pot smokers?

    A mugging or rape is still violence when the victim doesn’t physically resist. It’s still violence when the aggressor has a badge.

    Homicide may be down, but assault, kidnapping and robbery seem to be rising thanks to the state.

  9. maxina Says:

    what branch of science would this article be?

  10. Jamie Taylor Says:

    Ron,

    As a physician, having worked in urban trauma centers in the United States, I would like to agree with you about the marvelous advances in the art and science of modern medicine. It may well be that they have made significant impacts on the death tolls. However, I would caution against assuming that in “the vast majority of acts of violence that would have resulted in death 100 years ago…..the victim has a hugely greater chance of surviving” today. The old saying that “an ounce of prevention is worth a pound of cure”, probably still holds a lot of water, even, and perhaps especially for trauma. Bullet rittled gang bangers still arrive in emergency departments every day, beyond hope of salvation, to say nothing of those who die in the streets.