Lane Kenworthy, The Good Society
February 2019

Pundits routinely opine that we live in a violent era, and lots of Americans fear violence.1 About half say they worry “a great deal” about crime and violence, and a third report that they’re afraid to walk alone at night somewhere within a mile of their residence, as we see in figures 1 and 2.

This is understandable. Fear of violence is stoked by policy makers, aspiring or already in office, who aim to sound tough on crime.2 Campaigning for president in 2016, Donald Trump warned that “crime is out of control and rapidly getting worse.” But there’s a more basic reason. Few of us have any clue about actual rates of violence, and we tend to estimate the probability of something occurring based on the ease with which we can recall examples of it. Given our near-constant access to the internet, social media, television, and radio, we are bombarded with examples of violence — terrorist attacks, mass shootings, police killings of unarmed African Americans, gang violence in urban ghettos, sexual assaults on college campuses, brutal civil wars, dictatorships committing genocide. This makes it likely we’ll perceive the level of violence to be much greater than it actually is.

Figure 1. Worry a great deal about crime and violence
Share of US adults. Question: “How much do you personally worry about crime and violence?” Response options: a great deal, a fair amount, only a little, not at all, no opinion. The graph shows the share responding “a great deal,” with no opinion responses excluded. The line is a loess curve. Data source: Gallup, “Crime,”

Figure 2. Afraid to walk alone at night near residence
Share of US adults. Question: “Is there any area near where you live — that is, within a mile — where you would be afraid to walk alone at night?” The lines are loess curves. Data sources: Gallup, “Crime,”; General Social Survey,, series fear.

In fact, we live in probably the least violent period in human history. For roughly the past 5,000 years, violence around the world has been decreasing, not increasing. In the United States, violence did rise in the 1960s and 1970s, but it began falling in the 1990s and is now back near pre-1960s levels. How, and why, has this happened?


In The Better Angels of Our Nature, Steven Pinker documents a steady downward trajectory of human violence over the past 5,000 years.3 Pinker begins with fact that, contrary to long-held assumption, humans aren’t biologically programmed to be violent. We are susceptible to five motives for violence: practical (means to an end), dominance, revenge, sadism, and ideology. But other motives and practices — empathy, self-control, the moral sense, and reason — orient us away from violence.

The level of violence among humans thus depends on the degree to which institutions and policies reduce the benefits and increase the costs of violence and the degree to which they allow our anti-violence motives to flourish. Three institutions that do this have become increasingly prevalent: government, commerce, and information, knowledge, and science. Government helps because a centralized authority can prevent and punish violence more effectively and on a larger scale than other actors. Commerce reduces violence because exchange of goods and services makes other people more valuable alive than dead. Literacy, mobility, schooling, and mass media increase our awareness and understanding of other people, which encourages us to empathize with them. Application of knowledge and rationality allows people “to recognize the futility of cycles of violence, to ramp down the privileging of their own interests over others’, and to reframe violence as a problem to be solved rather than a contest to be won.”4 The growth of government, commerce, and information, knowledge, and science account for much of the decline in violence over time.

Pinker identifies a number of key advances. The first, beginning around 5,000 years ago, was the transition from “the anarchy of hunting, gathering, and horticultural societies … to the first agricultural civilizations with cities and governments,” which brought “a reduction in the chronic raiding and feuding that characterized life in a state of nature and a more or less fivefold decrease in rates of violent death.”5

A second major advance occurred in western Europe beginning in the late middle ages. It consisted of a tenfold to fiftyfold reduction in homicide due to “the consolidation of a patchwork of feudal territories into large kingdoms with centralized authority and an infrastructure for commerce.”6 Figure 3 shows the steady drop in homicide rates in five European nations.

Figure 3. Decline in homicides in western Europe after the middle ages
Homicides per 100,000 population. The vertical axis is on a log scale. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 3.3, using data from Manuel Eisner, “Long-Term Historical Trends in Violent Crime,” Crime and Justice: A Review of Research, 2003.

The humanitarian revolution during the enlightenment period in the 17th and 18th centuries brought a third major advance. It featured “the first organized movements to abolish slavery, dueling, judicial torture, superstitious killing, sadistic punishment, and cruelty to animals, together with the first stirrings of systematic pacifism.”7 Figure 4 shows the decline in the use of judicial torture in major countries beginning around 1700, figure 5 the spread of slavery abolition, and figure 6 the reduction in executions in the United States.

Figure 4. Decline in use of judicial torture in major countries
Number of major countries using judicial torture. The countries, in the order in which they ended judicial torture, are England, Scotland, Prussia, Denmark, Sweden, Bohemia, Austria, Italy, France, United States, Switzerland, Netherlands, Spain, Portugal, Russia. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 4.2.

Figure 5. Decline in legal slavery
Number of countries with slavery abolished. The vertical axis is inverted. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 4.6.

Figure 6. Decline in executions in the US
Number of executions per 100,000 population. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 4.4.

World Wars I and II were notable exceptions to the downward trend in violence, but they were, in all likelihood, genuine exceptions rather than a shift in direction. Since the end of the second world war, violence has declined along a number of dimensions. Figure 7 shows that deaths produced by wars involving the “great powers” have dropped sharply, and are now at their lowest point in recorded history. Figure 8 shows a similarly dramatic decline in deaths from all wars between and within countries since the 1950s. Figure 9 shows the decrease in deaths from genocide. In addition, the rights revolutions beginning in the 1950s brought a “growing revulsion against aggression on smaller scales, including violence against ethnic minorities, women, children, homosexuals, and animals.”8

Figure 7. Decline in deaths from wars involving the great powers
Number of battle deaths. “m” = million. The vertical axis is on a log scale. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 5.15.

Figure 8. Decline in deaths from interstate wars, international civil wars, and civil wars
Number of battle deaths. “k” = thousand. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 6.4.

Figure 9. Decline in deaths from genocide
Deaths per 100,000 population. Data source: Steven Pinker, The Better Angels of Our Nature, 2011, figure 6.7, using data from Rudolph Rummel (light line) and the Political Instability Task Force (dark line).

The only noteworthy exception to the downward trend in violence since the middle of the twentieth century is terrorist killings. The number has jumped sharply since 2010, as figure 10 shows. The level, it’s worth emphasizing, remains very low.

Figure 10. Deaths from terrorism
Deaths per 100,000 population. Data source: Global Terrorism Database.


In the US, the long-run trend in violence has mirrored the worldwide pattern. Figure 11 shows estimates of the homicide rate going back to 1700 (with reliable statistics beginning in the 1930s). The decline is marked.

Figure 11. Homicides
Per 100,000 population. Data sources: 1700-1900 from Claude S. Fischer, “A Crime Puzzle,” Made in America, 2010; 1900-1932 from Douglas Lee Eckberg, “Estimates of Early Twentieth-Century U.S. Homicide Rates: an Econometric Forecasting Approach,” Demography, 1995, table 4; 1933-1950 from M. Duane Smith and Margaret A. Zahn, eds., Homicide: Sourcebook of Social Research, Sage, 1998, figure 2.1, using FBI data; 1950ff from Bureau of Justice Statistics, using FBI Uniform Crime Reports data.

The long downward trend in violence was interrupted in the 1960s and 1970s. As figure 12 shows, all four types of violent crime — homicide, rape, aggravated assault, and robbery — increased sharply beginning in the 1960s. Then, in the early 1990s, they began to fall. The homicide rate dropped all the way back to its 1960 level, and robbery isn’t far behind. Rape and aggravated assault have declined less rapidly but still a good bit. What caused this unexpected rise and fall of violent crime in the United States?

Figure 12. Violent crime
Per 100,000 population. Each series is rescaled to vary from lowest value to highest value. Data source: Bureau of Justice Statistics, using FBI Uniform Crime Reports data.

One prominent hypothesis points to the economy. Crime is influenced by changes in the economy and in people’s (real and perceived) economic well-being. When jobs are plentiful and wages are rising, there is less incentive for people to commit robberies or other types of theft that can sometimes escalate to violence.9 However, the condition of the economy is mainly helpful for explaining short-run (year-to-year) fluctuations in violent crime. The US economy has been more difficult to navigate for many poor, working-class, and even middle-class Americans since the late 1970s,10 but that can’t account for the large rise in violent crime in the 1960s and 1970s, nor for the steady drop since the early 1990s.11

In the 1980s and 1990s it was common for conservatives to attribute rising crime to the breakdown of the traditional male breadwinner family, the weakening of community organizations, and the turn away from religion. However, since the early 1990s those institutions haven’t strengthened — indeed, most measures suggest they’ve continued to weaken12 — while violent crime has declined significantly.

Most violent crime is committed by males aged 15 to 30. The baby boom from 1945 to 1964 produced a surge in the number of 15-to-30-year-old males beginning in the early 1960s. This demographic bulge surely contributed to the rise in violent crime in the 1960s and 1970s. Yet the baby boom’s peak 15-to-30-year-old population year was 1980, so if demography is a strong driver of trends in violent crime, we would expect a decline commencing around that time. Instead, the crime decline began nearly a decade and a half later.13

A fourth theory, the most influential one in guiding our policy choices, focuses on incarceration. Locking up more criminals and keeping them in jail or prison longer ought to reduce crime by getting perpetrators off the street (incapacitation) and by deterring potential perpetrators (deterrence). From 1973 to around 2000, we steadily ramped up aggressiveness in putting criminals in prison and keeping them there for long periods of time. The “war on drugs” began in the 1970s under the Nixon administration. Mandatory minimum sentences, “three strikes” laws, and “truth in sentencing” laws, which forced judges to assign long prison terms irrespective of case circumstances, were added in the 1980s and 1990s. The likelihood of being arrested for committing a crime didn’t change much, but the likelihood of being convicted if arrested, the likelihood of being sentenced to prison or jail given a conviction, and the typical length of sentences all increased sharply.14

While incarceration surely helps to reduce violent crime, there are limits to its effectiveness. As one recent summary puts it15:

“Deterrence theory is underpinned by a rationalistic view of crime. In this view, an individual considering commission of a crime weighs the benefits of offending against the costs of punishment. Much offending, however, departs from the strict decision calculus of the rationalistic model…. Offenders must have some knowledge of criminal penalties to be deterred from committing a crime, but in practice often do not. Furthermore, suddenly induced rages, feelings of threat and paranoia, a desire for revenge and retaliation, and self-perceptions of brilliance in the grandiose phase of manic-depressive illness all can limit a potential offender’s ability to exercise self-control. Also playing a role are personality traits and the pervasive influence of drugs and alcohol: in one study, 32 percent of state prison inmates reported being high on drugs at the time of their crime…. The influence of crime-involved peers who downplay the long-term consequences of punishment is relevant as well. Taken together, these factors mean that, even if they knew the penalties that could be imposed under the law, a significant fraction of offenders still might not be able to make the calculation to avoid crime.”

The timing of the trend in violent crime is broadly consistent with the incarceration hypothesis. As figure 13 shows, we began locking more people up in the early-mid 1970s, and violent crime began falling in the early-mid 1990s.16

Figure 13. Incarceration and violent crime
Per 100,000 population. Incarceration in federal prison, state prison, or local jail. Data source: Bureau of Justice Statistics, “Total Correctional Population.” Violent crime: homicide, rape, aggravated assault, and robbery. Data source: Bureau of Justice Statistics, using FBI Uniform Crime Reports data.

Yet there are several empirical problems for the incarceration hypothesis. The lag between the rise in incarceration and the decline in crime was two decades. That’s longer than we might expect if incarceration has a large impact, though this may owe partly to the fact that much of the rise in incarceration was for drug crimes, rather than for violent crimes.17 Other rich countries didn’t follow the US lead in ramping up incarceration in the 1970s, 1980s, and 1990s, yet they too experienced a drop in violent crime beginning in the mid-1990s, as figure 14 shows.18 In the US, states that have reduced incarceration since 2000 have enjoyed the same continued decline in violent crime as states that have maintained or increased incarceration.19 According to the most thorough assessment to date, a 2014 report prepared for the National Academy of Sciences, “On balance, panel data studies support the conclusion that the growth in incarceration rates reduced crime, but the magnitude of the crime reduction remains highly uncertain and the evidence suggests it was unlikely to have been large.”20

Figure 14. Homicides
Homicides per 100,000 population. Data source: OECD. The non-US average decreased from 1.5 in 1991 to 0.8 in 2010.

A fifth explanation points to two innovations in policing strategy. One is commonly known as “broken windows” or “zero tolerance” policing. The hypothesis is that aggressive deterrence and punishment of small offenses, such as vandalism and public drinking, creates an atmosphere of order and lawfulness and thereby reduces violent crime.21 The other is “place-based” or “hot-zone” policing, whereby police departments use computers to identify areas in a city where a large number of crimes are reported and quickly blanket the area with police.22

The principal evidence supporting a large impact of policing comes from New York City, which adopted these two strategies in 1993 under new police commissioner William Bratton. New York’s murder and violent crime rates fell sharply over the next two decades.23 The problem for this explanation is that many other large cities that didn’t adopt these strategies, or did so only later, experienced drops in violent crime similar in timing and magnitude to that of New York, as figure 15 shows.24

Figure 15. Homicides in New York and other large cities
Homicides per 100,000 population. The other cities are Atlanta, Boston, Chicago, Cleveland, Dallas, Denver, Detroit, Houston, Los Angeles, Miami, Minneapolis, Philadelphia, Phoenix, San Diego, San Francisco, San Jose, Seattle, and Washington DC. Data source: Bureau of Justice Statistics.

Recently, some analysts have begun looking into the role of exposure to lead during childhood. Such exposure, via car emissions or old paint, is associated with lower cognitive ability, weaker aggression and impulse control, increased likelihood of attention deficit disorder, and greater likelihood of being arrested for violent crime in adulthood.25 Use of leaded gasoline in cars rose sharply between the late 1930s and early 1970s. We then began to phase it out. As figure 16 suggests, a plot of the violent crime rate matches very closely the rise and decline of lead in gas 23 years earlier. The same is true for a number of other rich nations. In addition, states varied in the rate at which lead exposure rose and fell, and state variation in trends in violent crime has tended to match these patterns. So too for six large cities.26

Figure 16. Lead in gasoline and violent crime
Each series is rescaled to equal 1 in 1960. Lead in gasoline: tons per 1,000 population. Data source: Kevin Drum, “Lead: America’s Real Criminal Element,” Mother Jones, 2013, using data from Rick Nevin. Violent crime: per 100,000 population. Data source: Bureau of Justice Statistics, using FBI Uniform Crime Reports data.

What, then, are the chief causes of the rise and decline in violent crime in the United States? Social scientists are nowhere close to consensus.27


Intentional violence isn’t the only threat to our safety, or even the main one. In the United States, around 15,000 people are murdered each year. Vehicle accidents kill more than 35,000.

Figure 17 shows that the share of Americans killed in vehicle accidents has been falling since the early 1970s. Indeed, the death rate per mile driven has been decreasing since the early 1900s, but until the 1970s that was offset by the fact that more and more Americans were driving more and more miles.

Figure 17. Vehicle accident deaths
Per 100,000 population. Includes vehicle occupants, pedestrians, and other non-occupants. Data source: Wikipedia, “Motor Vehicle Fatality Rate in U.S. by Year,” using data from the National Highway Traffic Safety Administration (NHTSA).

Figure 18 suggests that the US story is more nuanced. Vehicle accident deaths have fallen sharply in all affluent democratic nations over the past three decades, but the rate of decline has been much slower in the US than in other countries. Reducing our vehicle accident death rate to that of the next-highest country, South Korea, would save 10,000 lives per year.

Figure 18. Vehicle accident deaths
Per 100,000 population. Includes vehicle occupants, pedestrians, and other non-occupants. Data source: IRTAD database, via the OECD.

Could we do that? Yes. Government policies and technological advances — crash testing, padded dashboards, seat belts, air bags, helmets for motorcyclists, speed limits, reflective road lane markers, guard rails, lower legal blood alcohol levels, driver education courses, and more — have been the key to pulling down the death rate from car accidents over the past century. We can do more in both areas.28 According to a report by the RAND Corporation, “up to 10,000 lives could be saved if currently available advanced driver assistance [ADAS], such as automatic emergency braking, lane departure warning, and blind spot detection systems, were deployed fleet-wide. Passive alcohol ignition interlock systems, such as the Driver Alcohol Detection System for Safety (DADSS), might save more than 7,000 lives annually if all cars were equipped.”29 A requirement that smartphones automatically disable texting by drivers would also help. Redesigning city streets to create more car-free zones, to reduce average vehicle speed (speed cameras, more turn lanes), and to replace four-way intersections with roundabouts will further reduce fatalities, especially among pedestrians. We also can reduce driving by those who do so least safely. Ride-sharing services and automated driving technology will help to decrease driving by the elderly. Nearly one-quarter of vehicle accident deaths involve drivers aged 16-20; changes in permit requirements can reduce driving among this group.


Nearly every type of violence is less common today than at any previous point in human history, due largely to the expansion of government, commerce, and information, knowledge, and science. There have been hiccups — World Wars I and II, the rise in violent crime in the US in the 1960s and 1970s, the recent rise in terrorist killings — but so far these have proved to be exceptions to the long-run decline in violence rather than reversals of the trend.

The rise in violent crime in the 1960s and 1970s was one of the most significant developments in the US in the past century, with consequences not just for crime victims but for the large number of young Americans who, due to our “lock ’em up” policy reaction, became much more likely to serve a lengthy spell behind bars.30 There is little agreement among social scientists about the key determinants of the violent crime explosion and its equally pronounced decline beginning in the early-mid 1990s.

Accidents are an underappreciated threat to safety. In the United States, vehicle accidents kill more than twice as many people as homicides. That rate has been falling in recent decades, but recently the decline has stalled and it’s now higher than in any other rich democratic nation. Policy changes and technological advances, the keys to past improvements in vehicle accident deaths, could sharply reduce them going forward.

  1. For examples of pundit statements, see Steven Pinker and Andrew Mack, “The World Is Not Falling Apart,” Slate, 2014. 
  2. Bruce Western, Punishment and Inequality in America, Russell Sage Foundation, 2006, ch. 3. 
  3. Steven Pinker, The Better Angels of Our Nature: Why Violence Has Declined, Viking, 2011. 
  4. Pinker, The Better Angels of Our Nature, Preface. 
  5. Pinker, The Better Angels of Our Nature, Preface. 
  6. Pinker, The Better Angels of Our Nature, Preface. 
  7. Pinker, The Better Angels of Our Nature, Preface. 
  8. Pinker, The Better Angels of Our Nature, Preface. 
  9. Richard B. Freeman, “Why Do So Many Young American Men Commit Crimes and What Might We Do About It?,” Journal of Economic Perspectives, 1996. 
  10. Lane Kenworthy, “A Decent and Rising Income Floor,” The Good Society; Kenworthy, “Shared Prosperity,” The Good Society. 
  11. Franklin E. Zimring, The Great American Crime Decline, Oxford University Press, 2007, ch. 3. 
  12. Lane Kenworthy, “Family,” The Good Society; Kenworthy, “Civic Engagement,” The Good Society; Kenworthy, “Religion,” The Good Society. 
  13. Zimring, The Great American Crime Decline, ch. 3. 
  14. Jeremy Travis, Bruce Western, and Steve Redburn, eds., The Growth of Incarceration in the United States, National Academies Press, 2014, ch. 2. 
  15. Travis, Western, and Redburn, eds., The Growth of Incarceration in the United States, pp. 133-34. 
  16. Incarceration data that include federal, state, and local inmates begin in 1972. Data that don’t include local jails are available farther back in time; they too suggest the rise in incarceration began in the early-mid 1970s. See Travis, Western, and Redburn, eds., The Growth of Incarceration in the United States, figure 2.1. 
  17. Zimring, The Great American Crime Decline, ch. 3; Travis, Western, and Redburn, eds., The Growth of Incarceration in the United States, ch. 2; “America Is Waking Up to the Cost of Mass Incarceration,” The Economist, 2013. 
  18. Zimring, The Great American Crime Decline, ch. 6; Eric P. Baumer and Kevin T. Wolff, “Evaluating Contemporary Crime Drop(s) in America, New York City, and Many Other Places,” Justice Quarterly, 2014. 
  19. Bruce Western, “Crime and Punishment,” Boston Review, 2012. 
  20. Travis, Western, and Redburn, eds., The Growth of Incarceration in the United States, p. 155. 
  21. George L. Kelling and James Q. Wilson, “Broken Windows,” The Atlantic, 1982. 
  22. David Weisburd, “Place-Based Policing,” Police Foundation, 2008. 
  23. Franklin E. Zimring, “How New York City Beat Crime,” Scientific American, 2011. 
  24. See also Richard Rosenfeld, Robert Fornango, and Andres F. Rengifo, “The Impact of Order-Maintenance Policing on New York City Homicide and Robbery Rates, 1988-2001,” Criminology, 2007; Baumer and Wolff, “Evaluating Contemporary Crime Drop(s).” 
  25. Christopher Muller, Robert J. Sampson, and Alix S. Winter, “Environmental Inequality: The Social Causes and Consequences of Lead Exposure,” Annual Review of Sociology, 2018. 
  26. Jessica Wolpaw Reyes, “Environmental Policy as Social Policy? The Impact of Childhood Lead Exposure on Crime,” Working Paper 13097, National Bureau of Economic Research, 2007; John Paul Wright, Kim N Dietrich, M. Douglas Ris, Richard W. Hornung, Stephanie D. Wessel, Bruce P. Lanphear, Mona Ho, and Mary N. Rae, “Association of Prenatal and Childhood Blood Lead Concentrations with Criminal Arrests in Early Adulthood,” PLOS: Medicine, 2008; Howard W. Mielke and Sammy Zahran, “The Urban Rise and Fall of Air Lead (Pb) and the Latent Surge and Retreat of Societal Violence,” Environment International, 2012; Kevin Drum, “Lead: America’s Real Criminal Element,” Mother Jones, 2013; Rick Nevin, Lucifer Curves: The Legacy of Lead Poisoning, 2016; A. Aizer and J. Currie, “Lead and Juvenile Delinquency: New Evidence from Linked Birth, School, and Crime Records,” Working Paper 23392, National Bureau of Economic Research, 2017. 
  27. Baumer and Wolff, “Evaluating Contemporary Crime Drop(s).” 
  28. Steven Pinker, Enlightenment Now, Viking, 2018, ch. 12. 
  29. Liisa Ecola, Steven W. Popper, Richard Silberglitt, Laura Fraade-Blanar, “The Road to Zero: A Vision for Achieving Zero Roadway Deaths by 2050,” RAND Corporation, 2018, p. 8. 
  30. Jeff Manza and Christopher Uggen, Locked Out: Felon Disenfranchisement and American Democracy, Oxford University Press, 2006; Western, Punishment and Inequality in America; Devah Pager, Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration, University of Chicago Press, 2007; Sara Wakefield and Christopher Wildeman, Children of the Prison Boom: Mass Incarceration and the Future of American Inequality, Oxford University Press, 2013; Jeffrey D. Morenoff and David J. Harding, “Incarceration, Prisoner Reentry, and Communities,” Annual Review of Sociology, 2014.