Violent Societies

by Kieran Healy on March 24, 2005

While thinking about the deterrent effect of the death penalty I wondered about cross-national variation in rates of violent death. Comparative data on homicide rates undoubtedly exist, but I don’t have them to hand. I do have OECD data on rates of death due to assault, though, so here’s a nice picture of this trend for eighteen capitalist democracies from 1960 to 2002.

Death rate graphic

You can get this as a higher-resolution PDF file (with appropriately rearranged layout), too. Countries are arrayed from top to bottom based on their mean death rate over the period, with the lowest at the top left. As is immediately clear, the U.S. rate is both much higher and much more variable than any of the other countries. It rises rapidly through the 1960s and ‘70s. It bounces around between the late 70s and early 1990s, falling then rising sharply again, before beginning a sharp and sustained decline that brings it back to about 1965 levels, though still more about three times as high as other countries.

Variations in other countries are harder to pick out because their differences look small when compared to the U.S. But many countries experienced small but steady increases until the 1980s. From there, rates generally leveled off or declined, though there are exceptions. Italy experienced a surge in violent deaths in the late 1980s. Dutch rates seem to have gradually climbed across the whole period, while Japan’s have consistently declined. Though it doesn’t change much after 1970 or so, Finland’s rate bounces around its mean value much more than other countries.

I’m a sucker for data visualization. Maybe I’ll post more of this sort of thing.

{ 31 comments }

1

Cryptic Ned 03.24.05 at 5:38 pm

Why is Belgium steadily increasing?

2

Matt Weiner 03.24.05 at 6:07 pm

One design nitpick–since the data are almost always at the bottom of the chart, I associate them with the country listed below, not the country listed above (which is the right one).

3

dipnut 03.24.05 at 7:56 pm

Isn’t “death due to assault” “homicide”?

I guess you can have homicides that don’t involve assault, like poisonings and, uh, electrocutions or something. I just wonder what’s being measured here, and whether it’s measured the same way for every country.

Whatever it is, it appears to validate my proud ideology of American exceptionalism.

4

rd 03.24.05 at 8:08 pm

There have been fascinating attempts, particularly in the work of Eric Monkkonen, to extend such comparisons back through history using what data exists. Basically the pattern is a sharp drop in homicide rates throughout much of Europe during the seventeenth and eighteen centuries, the Middle Ages having been more violent than even the worst urban areas today. As a result, America has been an outlier in violence from the very beginning of its history. Throughout the nineteenth and twentieth centuries it has had homocide rates several multiples higher than comparable developed European countries. As for causes of this difference, he speculates that the
lack of anything like the kind of coherent state authority developed in Europe during the early modern period gave early America both a higher incidence of and a higher tolerance of violence,
a pattern that has proved pretty lasting.

5

Mill 03.24.05 at 8:39 pm

What I want to know is what sequence of events happened in Italy to cause such a clean jag. Can it really be just a fluke?

6

dm 03.24.05 at 8:50 pm

It’s too bad you can’t get the same statistics for the regions of the United States. Homicide rates in the US vary a great deal from region to region — New England basically looks like a European country, with the upper Midwest not far behind (at Finland levels). Statistics for the US as a whole are skewed by the South, where homicide rates are comparatively high.

7

Brett Bellmore 03.24.05 at 9:12 pm

If you extend the period of the analysis to cover the entire 20th century, rather than just the latter 40% of it, you will note that European rates of violent death are actually much more variable, and on average much, much higher, than in the US. It’s just that we maintain a fairly moderate level of violent death, while they alternate low levels with binges of genocidal horror.

Interestingly, if you combine homicide and sucide into one “violent death” number, the US falls right in the middle of the pack, and the spread between the best and worst of the above nations is much narrower. For some reason, nations with extremely low murder rates tend to have very high suicide rates, and visa versa.

8

Sebastian Holsclaw 03.24.05 at 9:22 pm

My understanding was that much of the drop in the murder rate in the last 15 years in the US was due to medical advances rather than a decline in people trying to kill each other. No similar drop is seen in the other countries but I have no idea what that means.

9

luci phyrr 03.24.05 at 9:26 pm

Assuming an even distribution of killers across countries, this demonstrates the higher productivity of US murderers. American killers should give seminars.

10

carla 03.24.05 at 9:27 pm

I’ve been thinking a lot this week about the erosion of ethics and morality in the US.

I wonder how much the murder rate in the US has to do with economic factors?

11

carla 03.24.05 at 9:27 pm

Oops..forgot to close the html tag. Sorry.

12

Iceman 03.24.05 at 10:04 pm

“My understanding was that much of the drop in the murder rate in the last 15 years in the US was due to medical advances rather than a decline in people trying to kill each other. No similar drop is seen in the other countries but I have no idea what that means.”

Could it be that medical advances have vastly reduced the death rate of GUNSHOT victims, while murders in Europe are much less likely to be gun slayings?

13

Sebastian Holsclaw 03.24.05 at 10:35 pm

“Could it be that medical advances have vastly reduced the death rate of GUNSHOT victims, while murders in Europe are much less likely to be gun slayings?”

It could have been, but my ER doctor friend says a lot of the advances are in stabbings and blunt trauma-to-the-head so I don’t think that explains it.

14

goesh 03.24.05 at 11:07 pm

Of course the gun folks are going to cite concealed weapons permits as a significant contributor in the decline in the US.

15

bad Jim 03.25.05 at 3:32 am

So, what’s up with Finland? Is it the saunas? The torment of their notoriously difficult, non-Indo-European language? The long dark winters?

I think that the spikes for Ireland, France (Algerians) and Italy (Sicilians) can be plausibly explained.

Someone, somewhere noted that when American cars were made safer, people started driving faster, implying that we maintained a constant rate of danger. Another might note that our cars continued to improve, and we have always driven as fast as we could.

Someone else wondered how many of our traffic fatalities are in fact suicides. I would not be surprised to find that American suicide rates, when automobile accidents are factored in, are as anomalously high as our homicide rates.

16

Rob 03.25.05 at 4:57 am

Kieran,

this might be helpful:

http://www.homeoffice.gov.uk/rds/pdfs/hosb601.pdf

I’m imagining the homicide spike in Italy is to be explained by the car-bombings and assorted shootings associated with the anti-mafia campaigns (or perhaps, if the time-scale’s not quite right, the violence accompanying the consolidation of the mafia that preceded them). I’d be a bit careful about international comparisons of crime rates – although homicide is probably the ‘cleanest’ – because of different legal systems and different police recording practices. Infamously, Derbyshire and Nottinghamshire, two more-or-less socially identical neighbouring counties in England, had totally different levels of recorded crime in the seventies and the eighties, which investigation found was almost totally attributable to different police recording practices.

17

Joshua W. Burton 03.25.05 at 10:04 am

The mischaracterization of this data as “violent death” rather than “assault death” is obscuring a very interesting point. If _all_ violent death (including suicide) is included, the US winds up right in the middle of the pack.

See, for example, .

There is, of course, a school of thought that excludes self-murder from the “violence” column of public-policy analysis, moving it over to “personal autonomy”. But with all major religions and most of the history of Western moral philosophy weighing in on the other side, it’s surely tendentious to treat this as a closed question, as these graphs implicitly do when headlined as “rates of violent death”.

18

Joshua W. Burton 03.25.05 at 10:05 am

Sorry, lost the URL.

See, for example, http://www.guncite.com/gun_control_gcgvintl.html.

19

Mike Maltz 03.25.05 at 10:15 am

Kieran:

For another view of homicide (in the US), you might take a look at http://tigger.uic.edu/~mikem//homicide.PDF; the relevant figures (which don’t come out too well in the printed version) can be found at http://www.uic.edu/~mikem/3%20homicide%20figures.pdf

20

Joshua W. Burton 03.25.05 at 10:18 am

It’s also worth noting that the US homicide rate starts falling sharply just 16 or 17 years after Roe v. Wade. I’ve seen, but can’t find on the web, graphs showing that states with early “soft” policies on induced abortion (first CT, then CA, NY, and a couple of others) each began to drop the appropriate number of years ahead of the national rate. The obvious hypothesis is that a significant number of future murderers are no longer being born.

There’s no use in taking this idea into the pro-life/pro-choice debate, of course: even if there’s something to it, you have to kill a hundred fetuses for every murder averted. So everybody winds up on the same side of that debate as before, at slightly higher stakes.

21

Kieran Healy 03.25.05 at 11:03 am

Interesting figures, Mike — especially the child homicide part of the surface. Infancy is much more dangerous than later childhood.

22

Barry 03.25.05 at 12:39 pm

“If you extend the period of the analysis to cover the entire 20th century, rather than just the latter 40% of it, you will note that European rates of violent death are actually much more variable, and on average much, much higher, than in the US. It’s just that we maintain a fairly moderate level of violent death, while they alternate low levels with binges of genocidal horror.”

Posted by Brett Bellmore · March 24th, 2005 at 9:12 pm

Of course, doing the same for the US might increase figures a bit.

23

bellatrys 03.25.05 at 2:17 pm

The US of course has never ever in our history committed a genocide or internal extreme violence…

[/irony]

We just got it out of our system – perhaps – a hundred-fifty years ago. Until eliminationism and witchhunts come back in fashion.

24

Alan 03.25.05 at 3:15 pm

I’m opposed to the death penalty, and am as appalled as anyone at the glorification of violence in American society. Nonetheless, I’ve always thought there was something unfair in these comparisons of U.S. violent crime rates with those of other “advanced industrial” (mostly European) nations. Far more accurate and revealing would be comparisons of U.S. violence rates with those of, say, Latin American nations like Brazil, which are far more similar to the U.S. in terms of the history and sociology relevant to these things.

25

bellatrys 03.25.05 at 7:53 pm

Well, perhaps, alan – but we’re not going around saying “look how much better of we are than the third-world countries we’ve been exploiting for the past 100 years,” instead we in this country (particularly conservatives) have been holding ourselves up as models by comparison to “godless Old Europe.”

I doubt very much – though who knows, what with all the “But Saddam killed more Iraqis than we did!” rhetoric, the race to the bottom may get there yet – that George Will and Rich Lowry and Peggy Noonan are going to be going “Aren’t we great, just see how much less violence and corruption we have than Columbia and Mexico!” anytime soon.

26

Brett Bellmore 03.25.05 at 9:57 pm

“Of course, doing the same for the US might increase figures a bit.”

Only if you go back to the Civil war, Barry. And WE don’t seem to have one of those every few decades.

27

BadTux 03.26.05 at 1:16 am

We actually have data on the deterrant effect of the death penalty. When the death penalty was outlawed, murder rates… well, they kept increasing at the same rate as would be predicted considering that the Baby Boomers were coming of age (murder rates closely track the size of the age 17-35 male cohort). When the death penalty was re-instated, murder rates… well, they kept decreasing at the same rate as would be predicted considering that the Baby Boomers were aging past the age 17-35 range where most murderers come from and thus the size of that age 17-35 male cohort was decreasing relative to the U.S. population, and furthermore, THEY DECREASED AT THE SAME RATE IN THE STATES WHERE THE DEATH PENALTY REMAINED BANNED! I.e., when, for example, Texas re-instated the death penalty, it had the same effect as when Massachusetts did NOT re-instate the death penalty — murder rates declined in both states at the same rate (but Texas has a far higher BASE rate, as do most states with the death penalty — most death penalty states have far more murders per 100,000 population than in states where the death penalty is outlawed).

I will try to get a cite to you on this one, I read about this in _The Sociology of Crime_, a sociology textbook, and that book is packed away in a 10×10 storage locker right now along with about 10,000 other books (uncatalogued, alas) so it may be easier for me to just go check it out at the university library (sigh!). But one thing that was clear from looking at their nice pretty graphs of murder rates vs. cohort size, murder rates, etc. was that a condom has more deterrant value as the death penalty when it comes to murder. Specifically, use of a condom 18 years prior. Further data regarding the cause of murder showed why this was so. A very small percentage of murders (under 5%) were premeditated. Most murders were “heat of the moment” type murders that occurred during an argument, while intoxicated, during a barroom brawl, or otherwise in situations where the murderer acted impulsively and then, only AFTER he’d already killed the person, had that “oh shit” moment: “Oh shit, I could get the death penalty for this!”. Which is a few seconds too late to do any good for the murder rate :-(.

– Badtux the Sociology Penguin

28

Raimo 03.26.05 at 5:41 am

“So, what’s up with Finland? Is it the saunas? The torment of their notoriously difficult, non-Indo-European language? The long dark winters?”

The murder rate goes UP in summer, so maybe heat has something to do with it, Jim.

I don’t know why our rate is so high compared with say, Sweden, but I think I can explain the zig-zags: although the rate is high, the actual number of murders is quite small, about 130 per year. So you only need a couple of multiple murders to produce a spike in the chart.

And unfortunately, we do sometimes get such multiple murders. A typical case would be when a drunken husband shoots his family (and then himself).

And fyi, the language is no problem – I could speak it when I was three!

29

Barry 03.26.05 at 11:18 am

Brett:

Me: “Of course, doing the same for the US might increase figures a bit.”

Brett:
“Only if you go back to the Civil war, Barry. And WE don’t seem to have one of those every few decades.”

Brett, how many people did the US kill in the 20th century?

30

bellatrys 03.26.05 at 1:46 pm

Barry, those don’t count, because we’re the Good Guys™.

Also the ones that we outsourced don’t count either. If the corporate media doesn’t report it, Joe Smoe doesn’t know about it, and it didn’t happen.

Welcome to Oceania.

31

Barry 03.26.05 at 5:50 pm

Truer words were never spoken.

Comments on this entry are closed.