Think Again

Challenge common wisdom

First Person

Look through expert eyes

Policy Riffs

Dig into tough policy issues

Culture Shock

Weigh in on social trends

Writer’s Corner

Cross the academic divide

Home » Think Again

A crime puzzle

Submitted by on May 2, 2011 – 12:05 am13 Comments

The economy is down. Shouldn’t violent crime be up?

Illustration by Chicken Billy via Flickr.

By Claude Fischer

Fischer is professor of sociology at the University of California, Berkeley. He was the founding editor of Contexts, the American Sociological Association’s magazine of sociology for the general reader, and edited the magazine through 2004.

Violent crime went down in America again last year. According to preliminary statistics from the FBI covering the first half of 2010. The number of violent crimes dropped by about 6 percent from the same period a year before. Given population growth, that means that the rate of violent crime dropped even more. And property crime dropped too.

This is a puzzle because crime is more common among the poor; the percentage of poor Americans has been trending up since about 2000. The economy tanked in the last couple of years. One would have expected a rise, not a fall, in violent crime.

But this head-scratcher is just part of a larger puzzle – understanding long-term trends in America’s criminal violence.

Murder History

The most reliable measure of violent crime is the homicide rate. Americans kill one another at a much higher rate – double, quadruple, or more – than do residents of comparable western European nations. This gap persists despite a roughly 40 percent drop in our homicide rate in the last 15 years or so. Americans have been notably more violent than western Europeans since about the mid-to-late 19th century.

This graph shows the American homicide rate over the last century-plus. (A few notes about construction of the chart are at the end of the post.)

Homicides 1900-2010

The puzzle compounds. We see a roughly cyclical pattern: a high plateau in the 1920s and early 30s; a rapid drop of more than half to a low point in the late 1950s; then, a sharp rise, more than doubling, by 1980 and 1990. That’s followed by what will probably be a drop of about half by 2010. These are huge swings.

We can put this story into yet greater perspective with the graph below. The line in that graph represents my rough estimate of fluctuations in the U.S. rate of homicide over many more generations, drawing on the historical literature (see some references at the end of this post). While the details are informed guesses, the general trend is well established.

The overall story is that homicide rates declined substantially (as did rates of interpersonal violence of all sorts). The drop in violent crime in the U.S. after about 1850 was not as fast or as consistent as it was in Western Europe. That is when the striking violence gap between the U.S. and Europe opened up. The graph also shows that progress was hardly uniform, as there were many upswings of violence. Spurts often coincide with wars and the aftermaths of war – notably having many demobilized soldiers, trained and armed fighters, roaming the land. (See this paper for one analysis of the war effect.) Another short-term influence is bloody competition among armed criminals – for example, over alcohol distribution during Prohibition and over crack cocaine during the 1980s.

Scholars have offered several explanations for the centuries’-long decline of violence in the West. Here are three common ones:

Government: Over many generations, political authorities gained greater policing power and legitimacy. This allowed them to suppress criminal attacks, intergroup battles, and personal feuds. Also, court systems provided a peaceful way to resolve conflicts. And mandatory schooling swept dangerous boys off the streets.

Economics: Greater and more broadly distributed wealth reduced people’s motivation for crime and raised the costs of getting into trouble. (Barroom brawling seems less attractive if it will cost you a steady and well-paying job.) This explanation may seem to have foundered in the last several years, but in the long run – if not the short run – goes the argument, the affluent society is the pacific society.

Culture: Over the centuries, westerners increasingly came to feel that violence was uncouth and distasteful. Historians refer to the “civilizing process,” a phrase German sociologist Norbert Elias used to describe how the royal courts of Europe suppressed bloody feuds among lords. The repression of violence spread to the bourgeois who, in turn, taught it to the working classes – or forced it on them through, for example, schooling. Over time, hitting, knifing, and shooting came to seem (to most people) as vulgar as smelling from body odor or defecating in the castle hallway.

Back to the Present

How might any of this explain the latest – the post-1990s – downswing in homicide and in criminal violence more generally? The rates are now at roughly the level of the least violent era in American history, the late 1950s.

Researchers point to some similar factors, although they disagree about their relative importance. Some stress government authority, namely that longer criminal sentences and the prison-building boom kept many more bad actors off the streets longer. Others point to the economic boom of the 1990s, when unemployment, even in poor communities, sunk to low levels. And others argue – although it is difficult to confirm with hard data – that a cultural shift occurred, that increasing revulsion toward violence eventually spread into even the most violent communities and corners of the United States. (One piece of evidence is that a similar, though much less volatile, pattern occurred in Canada, which is more culturally than legally similar to the United States.)

Recently, scholars have added yet another explanation: Immigration – although not in the way that some people might expect. Cities and neighborhoods that have received the largest influx of immigrants (including Mexican immigrants) have had – despite popular stereotypes to the contrary – the largest drops in criminal violence. (See, for example here and here. Thus, increased immigration may explain part of the crime drop since 1990.

In a wider view, perhaps the more puzzling part of the story is the rapid upswing in violence from around 1960 to 1990 (see the first graph above). Two generations of scholars have yet (it appears to me) to satisfactorily explain why that happened. Some of the upswing in crime can be attributed to the baby boom: Put a lot more 15-to-25-year-old males into a society and you will get an upsurge of violence. Some of it has to do with what happened in the black ghettos of the North: The population grew rapidly just when the well-paying blue-collar jobs for men were disappearing. Some of it involved the growing drug trade. And perhaps some of the upswing reflected a short-term cultural shift – maybe the baby boomers’ rejection of authority – that encouraged violence.

Whatever the reason, the downward trend of violent crime in the U.S. seems consistent with our longer history, although still high by first-world standards. It’s the upsurge of violent crime starting in the early 1960s that is now ending that remains the larger puzzle.

* * *

Suggested Readings:

Adler, J. S. 2001. “`Halting the Slaughter of the Innocents’,” Social Science History.

Blumstein, A. and J. Wallman (eds.). 2000. The Crime Drop in America.

Eckberg, D. L.  1995. “Estimates of Early 20th-Century U.S. Homicide Rates,” Demography.

Fischer, C. S. 1980. “The Spread of Crime from City to Countryside,” Rural Sociology.

Gurr, T. R.  (ed.). 1989. Violence in America, Vol. 1: The History of Crime.

Johnson, E., and E. Monkkonen (eds.). 1996.  The Civilization of Crime.

Lane, R. 1997. Murder in America: A History.

Monkkonen, E. H. 2001. Murder in New York City.

Rosenfeld, R.. 2002. “The Crime Decline in Context.” Contexts.

Zimring, F. E. 2007. The Great American Crime Decline.

Notes on the Chart: The basic sources are listed on the chart. Eckberg (1995) corrected the raw federal data for missing states in the early part of the century. I excluded the victims of 9/11. The 2010 point is a rough estimate from the first half of the year.

An earlier version of this article appeared on Fischer’s blog, Made in America.


  • [...] This post also appeared on the Berkeley Blog on June 16, 2010 and on The Public Intellectual on May 2, [...]

  • David Greenberg says:

    In a paper that I published in 2001 in the Journal of Quantitative Criminology, I showed that homicide rates over a period of about half a century were consistent with a random walk (a process in which change scores are white noise). Robert O’Brien found the same thing to be true for a period of about 100 years. Random walks can look like trends for significant stretches of time, but the appearance can be illusory.

    • So, what are you suggesting? That the fluctuations in the 20th century were just stochastic, or that this applies to larger frames of time. Surely, you concede the major point of the post, that there really is such a thing as a modernization curve?

  • [...] Link: [...]

  • Hoyt Drayer says:

    The latest drop in American violent crime has been shown by the authors of “Freakonomics” (Steven Levitt and Stephen Dubner) to be only partly the result of the above-mentioned factors. Each of those factors only account for a small percentage of the drop in violent crime. The most significant factor that the authors found was Roe vs. Wade. Starting about 20 years after abortion was made legal in the United States is when the crime rate started to drop quickly, thus suggesting that the people most likely to commit violent crimes weren’t being born in the first place anymore…namely…unwanted children. The authors of “Freakonomics” also found the opposite case in Romania, when abortion was banned in 1966 creating a 10% gain in birth rates and a spike in violent crime in the 1980s (when the unwanted babies born after the ban became adults), inspiring the post-revolution government to legalize abortion again in 1989. I am quite surprised this article makes no mention of this correlation.

  • Sadly missing from this article is any reference to the upheavals of racism, the civil rights movement, Jim Crow laws, and the Great Migration. To suggest a blame on the violence between 1960 and 1990 to ” what happened in the black ghettos of the North” is to grossly neglect the violent white mobs which were responsible for initiating most of the deadly and destructive riots in those areas. To summarize “The population grew rapidly just when the well-paying blue-collar jobs for men were disappearing.” fails to observe the systematic racism in real estate laws prohibiting blacks from buying and renting outside of specified neighborhoods to get to the few jobs available to them. Even in the North racism persisted and many jobs were still not available to men or women.

    I suggest adding to the reading list:

    Family Properties- Race, RealEstate and the Exploitation of Urban America by Beryl Satter

    The New Jim Crow, Mass Incarceration in an Age of Colorblindness by Michelle Alexander

    The Warmth of Other Suns, The Epic Story of America’s Great Migration by Isabel Wilkerson

    • admin says:

      Hello Melanie, you might be interested in the other articles in this package on policing and surveillance, including “Something smells like a pig, you say?” and “Jamaicans targeted for deportation.”

  • Mary Stevens says:

    Great art for the piece! In reference to the above comment, by Hoyt Drayer, if we added in those deaths caused by legalized abortion or any legalized killing maybe the violent crime rate really isn’t that low for the western world.

  • [...] is a graph of American homicide rates, the earlier results should be taken with a grain of salt of course, but the trend is clear. N.B. [...]

  • oboe says:

    Would love, love, love to see a graph comparing western European murder rates to US since the 1700s. Had no idea the US murder rate was so high going back pre-1900.

  • [...] on A crime puzzle for a report by Claude Fischer in The Public Intellectual, which includes the originals of the [...]

  • Tom C. says:

    I wonder what effect the emergence of modern emergency medicine (or for that matter, the availability of modern medical treatment in general) has had on these trends? Shooting or stabbing victims would have been much more likely to have become murder statistics decades past and prior. I wonder if that would flatten out the long term trend somewhat. To what extent are we less murderous vs. less successfully murderous? Other and more varied violent crime rate statistics would be interesting.

    • Bia says:

      That’s an interesting point. It seems to me that many reports fail to take every possible factor into account. But then again that may be impossible. However I think an inclusion of statistics regarding shootings and stabbings that don’t lead to death is important too, and not overly prohibitive.

Leave a comment!

Add your comment below, or trackback from your own site. You can also subscribe to these comments via RSS.

Be nice. Keep it clean. Stay on topic. No spam.

You can use these tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

This is a Gravatar-enabled weblog. To get your own globally-recognized-avatar, please register at Gravatar.