At almost 5,000 deaths in 2015, the number of workers killed on the job is still too high, but it’s much better than the 20,000 deaths in 1929, when the population was less than two-fifths the size. Much of the savings is the result of the movement of the labor force from farms and factories to stores and offices. But much of it is a gift of the discovery that saving lives while producing the same number of widgets is a solvable engineering problem.
Figure 12-7: Occupational accident deaths, US, 1913–2015
Sources: Data are from different sources and may not be completely commensurable (see note 63 for details). For 1913, 1933, and 1980: Bureau of Labor Statistics, National Safety Council, and CDC National Institute for Occupational Safety and Health, respectively, cited in Centers for Disease Control 1999. For 1970: Occupational Safety and Health Administration, “Timeline of OSHA’s 40 Year History,” https://www.osha.gov/osha40/timeline.html. For 1993–1994: Bureau of Labor Statistics, cited in Pegula & Janocha 2013. For 1995–2005: National Center for Health Statistics 2014, table 38. For 2006–2014: Bureau of Labor Statistics 2016a. The latter data were reported as deaths per full-time-equivalent workers and are multiplied by .95 for rough commensurability with the preceding years, based on the year 2007, when the Census of Fatal Occupation Injuries reported rates both per worker (3.8) and per FTE (4.0).
Who by earthquake. Could the efforts of mortals even mitigate what lawyers call “acts of God”—the droughts, floods, wildfires, storms, volcanoes, avalanches, landslides, sinkholes, heat waves, cold snaps, meteor strikes, and yes, earthquakes that are the quintessentially uncontrollable catastrophes? The answer, shown in figure 12-8, is yes.
After the ironic 1910s, when the world was ravaged by a world war and an influenza pandemic but relatively spared from natural disasters, the rate of death from disasters has rapidly declined from its peak. It’s not that with each passing decade the world has miraculously been blessed with fewer earthquakes, volcanoes, and meteors. It’s that a richer and more technologically advanced society can prevent natural hazards from becoming human catastrophes. When an earthquake strikes, fewer people are crushed by collapsing masonry or burned in conflagrations. When the rains stop, they can use water impounded in reservoirs. When the temperature soars or plummets, they stay in climate-controlled interiors. When a river floods its banks, their drinking water is safeguarded from human and industrial waste. The dams and levees that impound water for drinking and irrigation, when properly designed and built, make floods less likely in the first place. Early warning systems allow people to evacuate or take shelter before a cyclone makes landfall. Though geologists can’t yet predict earthquakes, they can often predict volcanic eruptions, and can prepare the people who live along the Rim of Fire and other fault systems to take lifesaving precautions. And of course a richer world can rescue and treat its injured and quickly rebuild.
Figure 12-8: Natural disaster deaths, 1900–2015
Source: Our World in Data, Roser 2016q, based on data from EM-DAT, The International Disaster Database, www.emdat.be. The graph plots the sum of the death rates for Drought, Earthquake, Extreme temperature, Flood, Impact, Landslide, Mass movement (dry), Storm, Volcanic activity, and Wildfire (excluding Epidemics). In many decades a single disaster type dominates the numbers: droughts in the 1910s, 1920s, 1930s, and 1960s; floods in the 1930s and 1950s; earthquakes in the 1970s, 2000s, and 2010s.
It’s the poorer countries today that are most vulnerable to natural hazards. A 2010 earthquake in Haiti killed more than 200,000 people, while a stronger one in Chile a few weeks later killed just 500. Haiti also loses ten times as many of its citizens to hurricanes as the richer Dominican Republic, the country with which it shares the island of Hispaniola. The good news is that as poorer countries get richer, they get safer (at least as long as economic development outpaces climate change). The annual death rate from natural disasters in low-income countries has come down from 0.7 per 100,000 in the 1970s to 0.2 today, which is lower than the rate for upper-middle-income countries in the 1970s. That’s still higher than the rate for high-income countries today (0.05, down from 0.09), but it shows that rich and poor countries alike can make progress in defending themselves against a vengeful deity.64
And what about the very archetype of an act of God? The projectile that Zeus hurled down from Olympus? The standard idiom for an unpredictable date with death? The literal bolt from the blue? Figure 12-9 shows the history.
Yes, thanks to urbanization and to advances in weather prediction, safety education, medical treatment, and electrical systems, there has been a thirty-seven-fold decline since the turn of the 20th century in the chance that an American will be killed by a bolt of lightning.
Figure 12-9: Lightning strike deaths, US, 1900–2015
Source: Our World in Data, Roser 2016q, based on data from National Oceanic and Atmospheric Administration, http://www.lightningsafety.noaa.gov/victims.shtml, and López & Holle 1998.
* * *
Humanity’s conquest of everyday danger is a peculiarly unappreciated form of progress. (Some readers of a draft of this chapter wondered what it was even doing in a book on progress.) Though accidents kill more people than all but the worst wars, we seldom see them through a moral lens. As we say: Accidents will happen. Had we ever been confronted with the dilemma of whether a million deaths and tens of millions of injuries a year was a price worth paying for the convenience of driving our own cars at enjoyable speeds, few would have argued that it was. Yet that is the monstrous choice we tacitly made, because the dilemma was never put to us in those terms.65 Now and again a hazard is moralized and a crusade against it is mounted, particularly if a disaster makes the news and a villain can be fingered (a greedy factory owner, a negligent public official). But soon it recedes back into the lottery of life.
Just as people tend not to see accidents as atrocities (at least when they are not the victims), they don’t see gains in safety as moral triumphs, if they are aware of them at all. Yet the sparing of millions of lives, and the reduction of infirmity, disfigurement, and suffering on a massive scale, deserve our gratitude and demand an explanation. That is true even of murder, the most moralized of acts, whose rate has plummeted for reasons that defy standard narratives.
Like other forms of progress, the ascent of safety was led by some heroes, but it was also advanced by a motley of actors who pushed in the same direction inch by inch: grassroots activists, paternalistic legislators, and an unsung cadre of inventors, engineers, policy wonks, and number-crunchers. Though we sometimes chafe at the false alarms and the nanny-state intrusions, we get to enjoy the blessings of technology without the threats to life and limb.
And though the story of seat belts, smoke alarms, and hot-spot policing is not a customary part of the Enlightenment saga, it plays out the Enlightenment’s deepest themes. Who will live and who will die are not inscribed in a Book of Life. They are affected by human knowledge and agency, as the world becomes more intelligible and life becomes more precious.
CHAPTER 13
TERRORISM
When I wrote in the preceding chapter that we are living in the safest time in history, I was aware of the incredulity those words would evoke. In recent years, highly publicized terrorist attacks and rampage killings have set the world on edge and fostered an illusion that we live in newly dangerous times. In 2016, a majority of Americans named terrorism as the most important issue facing the country, said they were worried that they or a family member would be a victim, and identified ISIS as a threat to the existence or survival of the United States.1 The fear has addled not just ordinary citizens trying to get a pollster off the phone but public intellectuals, especially cultural pessimists perennially hungry for signs that Western civilization is (as always) on the verge of collapse. The political philosopher John Gray, an avowed progressophobe, has described the contemporary societies of Western Europe as “terrains of violent conflict” in which
“peace and war [are] fatally blurred.”2
But yes, all this is an illusion. Terrorism is a unique hazard because it combines major dread with minor harm. I will not count trends in terrorism as an example of progress, since they don’t show the long-term decline we’ve seen for disease, hunger, poverty, war, violent crime, and accidents. But I will show that terrorism is a distraction in our assessment of progress, and, in a way, a backhanded tribute to that progress.
Gray dismissed actual data on violence as “amulets” and “sorcery.” The following table shows why he needed this ideological innumeracy to prosecute his jeremiad. It shows the number of victims of four categories of killing—terrorism, war, homicide, and accidents—together with the total of all deaths, in the most recent year for which data are available (2015 or earlier). A graph is impossible, because swatches for the terrorism numbers would be smaller than a pixel.
Table 13-1: Deaths from Terrorism, War, Homicide, and Accidents
US
Western Europe
World
Terrorism
44
175
38,422
War
28
5
97,496
Homicide
15,696
3,962
437,000
Motor vehicle accidents
35,398
19,219
1,250,000
All accidents
136,053
126,482
5,000,000
All deaths
2,626,418
3,887,598
56,400,000
“Western Europe” is defined as in the Global Terrorism Database, comprising 24 countries and a 2014 population of 418,245,997 (Statistics Times 2015). I omit Andorra, Corsica, Gibraltar, Luxembourg, and the Isle of Man.
Sources: Terrorism (2015): National Consortium for the Study of Terrorism and Responses to Terrorism 2016. War, US and Western Europe (UK + NATO) (2015): icasualties.org, http://icasualties.org. War, World (2015): UCDP Battle-Related Deaths Dataset, Uppsala Conflict Data Program 2017. Homicide, US (2015): Federal Bureau of Investigation 2016a. Homicide, Western Europe and World (2012 or most recent): United Nations Office on Drugs and Crime 2013. Data for Norway exclude the Utøya terrorist attack. Motor vehicle accidents, All accidents, and All deaths, US (2014): Kochanek et al. 2016, table 10. Motor vehicle accidents, Western Europe (2013): World Health Organization 2016c. All accidents, Western Europe (2014 or most recent): World Health Organization 2015a. Motor vehicle accidents and All accidents, World (2012): World Health Organization 2014. All deaths, Western Europe (2012 or most recent): World Health Organization 2017a. All deaths, World (2015): World Health Organization 2017c.
Start with the United States. What jumps out of the table is the tiny number of deaths in 2015 caused by terrorism compared with those from hazards that inspire far less anguish or none at all. (In 2014 the terrorist death toll was even lower, at 19.) Even the estimate of 44 is generous: it comes from the Global Terrorism Database, which counts hate crimes and most rampage shootings as examples of “terrorism.” The toll is comparable to the number of military fatalities in Afghanistan and Iraq (28 in 2015, 58 in 2014), which, consistent with the age-old devaluing of the lives of soldiers, received a fraction of the news coverage. The next rows down reveal that in 2015 an American was more than 350 times as likely to be killed in a police-blotter homicide as in a terrorist attack, 800 times as likely to be killed in a car crash, and 3,000 times as likely to die in an accident of any kind. (Among the categories of accident that typically kill more than 44 people in a given year are “Lightning,” “Contact with hot tap water,” “Contact with hornets, wasps, and bees,” “Bitten or struck by mammals other than dogs,” “Drowning and submersion while in or falling into bathtub,” and “Ignition or melting of clothing and apparel other than nightwear.”)3
In Western Europe, the relative danger of terrorism was higher than in the United States. In part this is because 2015 was an annus horribilis for terrorism in that region, with attacks in the Brussels Airport, several Paris nightclubs, and a public celebration in Nice. (In 2014, just 5 people were killed.) But the relatively higher terrorism risk is also a sign of how much safer Europe is in every other way. Western Europeans are less murderous than Americans (with about a quarter their homicide rate) and also less car-crazy, so fewer die on the road.4 Even with these factors tipping the scale toward terrorism, a Western European in 2015 was more than 20 times as likely to die in one of their (relatively rare) homicides as in a terrorist attack, more than 100 times as likely to die in a car crash, and more than 700 times as likely to be crushed, poisoned, burned, asphyxiated, or otherwise killed in an accident.
The third column shows that for all the recent anguish about terrorism in the West, we have it easy compared with other parts of the world. Though the United States and Western Europe contain about a tenth of the world’s population, in 2015 they suffered one-half of one percent of the terrorist deaths. That’s not because terrorism is a major cause of death elsewhere. It’s because terrorism, as it is now defined, is largely a phenomenon of war, and wars no longer take place in the United States or Western Europe. In the years since the attacks of September 11, 2001, violence that used to be called “insurgency” or “guerrilla warfare” is now often classified as “terrorism.”5 (The Global Terrorism Database, incredibly, does not classify any deaths in Vietnam in the last five years of the war there as “terrorism.”)6 A majority of the world’s terrorist deaths take place in zones of civil war (including 8,831 in Iraq, 6,208 in Afghanistan, 5,288 in Nigeria, 3,916 in Syria, 1,606 in Pakistan, and 689 in Libya), and many of these are double-counted as war deaths, because “terrorism” during a civil war is simply a war crime—a deliberate attack on civilians—committed by a group other than the government. (Excluding these six civil war zones, the terrorism death count for 2015 was 11,884.) Yet even with the double counting of terrorism and war during the 21st century’s worst year for war deaths, a global citizen was 11 times as likely to have died in a homicide as in a terrorist attack, more than 30 times as likely to have died in a car crash, and more than 125 times as likely to have died in an accident of any kind.
Has terrorism, whatever its toll, increased over time? The historical trends are elusive. Because “terrorism” is an elastic category, the trend lines look different depending on whether a dataset includes civil war crimes, multiple murders (which include robberies or mafia hits in which several victims are shot), or suicidal rampages in which the killer ranted about some political grievance beforehand. (The Global Terrorism Database, for example, includes the 1999 Columbine school massacre but not the 2012 Sandy Hook school massacre.) Also, mass killings are media-driven spectacles, in which coverage inspires copycats, so they can yo-yo up and down as one event inspires another until the novelty wears off for a while.7 In the United States, the number of “active shooter incidents” (public rampage killings with guns) has wobbled with an upward trend since 2000, though the number of “mass murders” (four or more deaths in an incident) shows no systematic change (if anything, it shows a slight decline) from 1976 to 2011.8 The per capita death rate from “terrorism incidents” is shown in figure 13-1, together with the messy trends for Western Europe and the world.
Figure 13-1: Terrorism deaths, 1970–2015
Sources: “Global Terrorism Data
base,” National Consortium for the Study of Terrorism and Responses to Terrorism 2016, https://www.start.umd.edu/gtd/. The rate for the world excludes deaths in Afghanistan after 2001, Iraq after 2003, Pakistan after 2004, Nigeria after 2009, Syria after 2011, and Libya after 2014. Population estimates for the world and Western Europe are from the European Union’s 2015 Revision of World Population Prospects (https://esa.un.org/unpd/wpp/); estimates for the United States are from US Census Bureau 2017. The vertical arrow points to 2007, the last year plotted in figs. 6–9, 6–10, and 6–11 in Pinker 2011.
The death rate for American terrorism for the year 2001, which includes the 3,000 deaths from the 9/11 attacks, dominates the graph. Elsewhere we see a bump for the Oklahoma City bombing in 1995 (165 deaths) and barely perceptible wrinkles in other years.9 Excluding 9/11 and Oklahoma, about twice as many Americans have been killed since 1990 by right-wing extremists as by Islamist terror groups.10 The line for Western Europe shows that the rise in 2015 came after a decade of relative quiescence, and is not even the worst that Western Europe has seen: the rate of killing was higher in the 1970s and 1980s, when Marxist and secessionist groups (including the Irish Republican Army and the Basque ETA movement) carried out regular bombings and shootings. The line for the world as a whole (excluding recent deaths in major war zones, which we examined in the chapter on war) contains a spiky plateau for the 1980s and 1990s, a fall after the end of the Cold War, and a recent rise to a level that still falls below that of the earlier decades. So the historical trends, like the current numbers, belie the fear that we are living in newly dangerous times, particularly in the West.