Assault Deaths within the United States

by Kieran Healy on July 22, 2012

The chart in “America is a Violent Country” has been getting a lot of circulation. Time to follow up with some more data. As several commentators at CT noted, the death rate from assault in the U.S. is not uniform within the country. Unfortunately, state-level and county-level mortality data are not easily available for the time period covered by the previous post—though they do exist, going back to the 1940s. What I have to hand is a decade’s worth of US mortality data courtesy of CDC WONDER covering 1999 to 2009. I extracted the assault deaths according to the same criteria the OECD uses (for the time period in question, ICD-10 codes X85-Y09 and Y87.1). The estimates are adjusted to the 2000 U.S. population, which isn’t identical to the standard OECD adjustment. But the basic comparability should be OK, for our purposes.

First, it’s well-known that there are strong regional differences in the assault death rate in the U.S. by state and region. Here’s what the patterns look like by state from 1999 to 2009.

Assault death rates by State

Trends in the Death Rate from Assault, 1999–2009, by State. Click for a larger PNG or PDF.

This figure excludes the District of Columbia, which has a much higher death rate but is also a city. Also missing are a few states with small populations and low absolute numbers of assault deaths—Wyoming, North Dakota, Vermont—such that the CDC can’t generate reliable age-adjusted estimates for them. If you want a “small-multiple” view with each state shown separately from high to low, here you go.

The legend for the figure above arranges the states from high to low, reading top to bottom and left to right. Although it’s clear that geographical region isn’t everything, those tendencies are immediately apparent. Let’s look at them using the official census regions.

Assault death rates by State

Trends in the Death Rate from Assault, 1999–2009, by Region. Click for a larger PNG or PDF.

As is well known, the South is more violent than the rest of the country, by some distance. Given the earlier post, the natural thing to do is to put these regional trends into the cross-national comparison and see—for the decade we have, anyway—how these large U.S. regions would fare if they were OECD countries. Again, bear in mind that the age-adjustment is not quite comparable.

Assault death rates by State

US Regions in Comparative OECD Perspective. Click for a larger PNG or PDF.

Despite their large differences, all of the U.S. regions have higher average rates of death from assault than any of the 24 OECD countries we looked at previously. The placid Northeast comes relatively close to the upper end of the most violent countries in our OECD group.

Finally, there’s the question of racial and ethic incidence of these deaths within the United States. Here are the decade’s trends broken out by the race of the victim, rather than by state or region.

Assault death rates by Race

Assault Death Rates by Race. Click for a larger PNG or PDF.

The story here is depressing. Blacks die from assault at more than three times the U.S. average, and between ten and twenty times OECD rates. In the 2000s the average rate of death from assault in the U.S. was about 5.7 per 100,000 but for whites it was 3.6 and for blacks it was over 20. Even 3.6 per 100,000 is still well above the OECD-24 average, which–if we exclude the U.S.–was about 1.1 deaths per 100,000 during the 2000s, with a maximum value of 2.9. An average value of 20 is just astronomical. And this is after a long period of decline in the death rate from assault.



Witt 07.22.12 at 2:44 am

This is fascinating, if tremendously sad. Thank you for putting it together.

The starkness of the South is pretty remarkable. I was going to speculate that it had something to do with being less urbanized, but I stopped to check and was reminded that 3 of the top 10 largest cities (Dallas, San Antonio, and Houston) are in the South.

One question on the last chart — does the CDC not break out by ethnicity as well as race? There are no figures for Hispanic/Latino on the chart.


Martin 07.22.12 at 3:15 am

Any world figures for comparison? eg where do say Guangdong or New South Wales or Gauteng or Metropolitan Tokyo or Bihar or whatever fit in relatively? Everyone in Hong Kong ‘knows’ Shenzhen is ‘really unsafe’ — how does it compare to states of the US?


JP Stormcrow 07.22.12 at 3:23 am

Yes, The % urban of the 4 census regions in 2010 are as follows:
West 89.8
Northeast 85.0.
Midwest 75.9
South 75.8

The percentage for the South is strongly influenced by Florida (> 91%) and Texas (> 84%). All the other states there are below the national average.


dilbert dogbert 07.22.12 at 3:41 am

Here is The Guardian:
Crappy graphics as it is colored for total deaths not deaths per 100,000. Click on a state and you will get deaths per 100k.


Craig 07.22.12 at 3:54 am

In terms of murder rates, the US South (and Maryland) is comparable to late sixteenth/early seventeenth century Western Europe, which was roughly 10/100,000 with minor geographical variations.


Matthew 07.22.12 at 3:54 am

Nice use of GGPlot to tell a story. Even better would be to code up an interactive so that someone could enter their own race and state (and age?) and see what their odds of being killed in an assault would have been over time, with appropriate comparisons included.


Tom T. 07.22.12 at 4:04 am

Re 1 and Hispanics: Yes, the CDC put the homicide rate for Hispanics at 7.2. See my link at #52 of Kieran’s last post. Kieran’s link above analyzes the data slightly differently and would likely yield somewhat different numbers, but the tool does allow searching by Hispanic origin.

I’m not sure the breakout by state really adds much that’s meaningful, since the variance is pretty much a proxy for race, where the differences are so wide as to overwhelm any other distinctions.


Dan Nexon 07.22.12 at 4:07 am

Given tremendous within region variation by state, I’m not sure what to conclude about aggregates like “the South.” Seems to me that neither regions, nor even “states,” are doing important explanatory work here.


Down and Out of Sài Gòn 07.22.12 at 7:12 am

Martin@2: for homicides, Wikipedia has a breakdown by states/provinces for Canada, Australia, United States and Mexico. The last is of interest because it is part of the OECD, but its homicide toll (15 per 100,000) is far higher than the US (with Chihuahua at 111 per 100,000!) However, I would guess that the major factor is inter-Cartel violence.

Witt@1: less urbanisation may be a factor. The highest for Oz is the Northern Territory, and the highest for Canada is Nunavut – 18 per 100,000. Both are sparsely inhabited territories. But even the NT at its worse is less that many US States, and almost all of the Mexican ones.

(I have no idea why Nunavut is so high. But its population at 31,906 is so low that it probably more susceptible to statistical fluctuations.)


Danny 07.22.12 at 7:54 am

Anybody know what would be the resulting assault death rate among non-Hispanic Whites? Presumably somewhat lower than 3.6%. I imagine that if we would look at the geographic distribution among non-Hispanic whites we would see that some areas of the US have a European-range assault death rate.


Joey Fishkin 07.22.12 at 8:09 am

These are some extremely striking numbers. Thank you for posting them.

The one positive thing (I guess) one could say for African-Americans anywhere in the death rate data is that their suicide rate appears to be considerably lower than that of the white population or the population as a whole.

But the homicide data tell an amazingly horrific story. According to this report by the Children’s Defense Fund, the total firearm death rate for black males age 15-19 in 2006 was 67.65 per 100,000 (compared to a rate for white males age 15-19 of 14.40 per 100,000). That figure includes homicides, suicides, and accidents, but among black males 15-19, whose suicide rate is low, about 90% of those gun deaths are homicides.

It really makes me despair for my country. (I live in Texas.)


aepxc 07.22.12 at 8:35 am

Given the data by region and ethnicity, it would be interesting to see assault violence by median income and/or income inequality. Quickly playing around on Gapminder produces this:


Data Tutashkhia 07.22.12 at 8:40 am

Relevant characteristics are, of course, income/social status, population density, and so on. Surely you don’t believe that people kill and get killed because they are black or Am. Indian.


Scott Martens 07.22.12 at 9:12 am

@8 I have no idea why Nunavut is so high

90% 0f Nunavut’s population is Inuit. Natives on reserves in Canada have violent crime rates as high or worse than Nunavut, for the same basic reasons: No work, little money, overcrowding, substances, incompatible cultural traditions… If there was a US breakdown for reservations I suspect it would look similar. It’s not a statistical fluctuation but a long standing social problem. Note that the top jurisdictions for homicide in Canada most years are exactly the ones with the highest proportion Native populations: Yukon, NWT, Saskatchewan, Manitoba. The only variance is when, by some fluke, there are no murders in one of the territories.


Peter Erwin 07.22.12 at 10:52 am

Down and Out of Sài Gòn @ 8:
I would guess that the major factor is inter-Cartel violence

Given the numbers at this page (or this article from the El Paso Times), it appears that the homicide rate in Mexico has consistently been about twice that of the US all through the 1980s and 1990s, so cartel-related murders are only a minor factor (though they’re important in specific local cases like Chihuahua during the last 6 years, of course).


derrida derider 07.22.12 at 11:08 am

Scott Martens @11, the Northern Territory is an outlier for violent crime rates in Australia too mainly because it has a big indigenous population (Aboriginal and Islander). And for the same reasons as with the Inuit – “no work, little money, overcrowding, substances, incompatible cultural traditions…”.

Neither Canada, the US nor Australia has a lot to be proud of in its treatment of its original inhabitants.


AcademicLurker 07.22.12 at 11:12 am

I’m shocked at how high Maryland ranks. Do you suppose Baltimore accounts for all of it, or is there something else going on?


bjk 07.22.12 at 11:20 am

PG County too. And the mean streets of Bethesda Chevy Chase.


Christiaan 07.22.12 at 11:25 am

For the state-by-state chart the color coding is really badly chosen, it is impossible to distinguish different colors. The fact that they seem to be ordered rather arbitrarily (at least not in any recognizable way like the order at the start or the end) makes it even more impossible.
For instance, there seem to be an interesting “green” state that saw a large increase, which one is that? The other with a large increase (though earlier) was clearly Louisiana. Anyone has an idea, and why?


John 07.22.12 at 11:33 am

What would be very interesting would be to compare areas’ assault death numbers with the the relative strictness (if any) of local gun laws. My guess would be that their would be some correlations (stronger gun laws=fewer assault deaths). Of course the problem is that guns leak across these artificial borders here in the States.


Kieran 07.22.12 at 11:34 am

As I said in the post, you can also look at the states separately, ordered from high to low rates. The state you’re noticing is Delaware.


soru 07.22.12 at 11:53 am

How respected is this theory?

Research at Pittsburgh University found that adolescents arrested for crime in the city had lead levels four times higher than their law-abiding contemporaries, and a study of 3,000 possible causes of criminality in 1,000 young people by Fordham University, New York, found that high lead levels were the best predictor of delinquent and violent behaviour.

Seems plausible enough to me, and the evidence behind it seems fairly solid to a non-expert.

Of course, that doesn’t rule out the ‘some cranks who spent more effort on being plausible’ scenario…


AcademicLurker 07.22.12 at 12:14 pm


Wow. That’s pretty striking. So all the other states have been exporting their violent criminals to Delaware?


Matt McIrvin 07.22.12 at 1:28 pm

Louisiana’s rise (from an already-sky-high but declining base) was after Hurricane Katrina devastated New Orleans and environs. There’s probably some causal effect there.

I’m not sure what the heck is going on in Delaware, but it has a relatively small population and the northern part of the state is in the Philadelphia/Camden/Wilmington metro area, which has a lot of problems.


Marc 07.22.12 at 1:32 pm

@19: Delaware is a small state that is close to a couple of major metropolitan areas. It wouldn’t take a large absolute number of spillover crimes to make the rates high.


AcademicLurker 07.22.12 at 2:25 pm

Re: Delaware, I’m less puzzled by the absolute numbers than by the trend. They are showing a sharp upturn when almost every one else started to trend downward.


Marahall 07.22.12 at 2:43 pm

@22: urbanization is spreading. I notice that Oregon is right down there next to Massachusetts, despite having plenty of heavily armed poor people. Our largest city is Portland, only middling. Could it be this has to do with the way we maintain our cities, eg esp policing policies, esp vis a vis Blacks? Be interesting to see some urban/suburban/rural breakdown.


Alex 07.22.12 at 4:14 pm

The West seems to be an outlier – rising when the rest was falling, and then falling especially fast.


Christiaan 07.22.12 at 4:51 pm


The rise in Louisiana ended about 2 years before Katrina, so that cannot be the explanation (unless you believe in Deja Vue).


Alex K. 07.22.12 at 5:26 pm

I’m skeptical of a causal link between lack of urbanization high assault crime. New Hampshire and Maine have among the lowest rates of urbanization, and both states seem to be dead last in the rate of assault deaths category. (Maine is also almost last in terms of GDP per hour worked.)

I think cultural explanations are necessary here. I’m thinking of David Hackett Fischer’s “Albion’s Seed”, where he divides colonial US into four different “folkways”:

The first was an exodus of Puritans from the east of England to Massachusetts during a period of eleven years from 1629 to 1640. The second was the migration of a small Royalist elite and large numbers of indentured servants from the south of England to Virginia (ca. 1642-75). The third was a movement from the North Midlands of England and Wales to the Delaware Valley (ca. 1675-1725). The fourth was a flow of English-speaking people from the borders of North Britain and northern Ireland to the Appalachian backcountry mostly during the half-century from 1718 to 1775.

You’re never going to explain the difference between, say, West Virginia and New Hampshire or Maine, if you’ve picked up bad habits of thinking in terms of race rather than in terms of culture, or if you’re too focused on “objective” measures like lack of urbanization rather than focusing on culture.


Cahokia 07.22.12 at 6:13 pm

I’ll withhold quoting rap lyrics, but here is a scene:

Furious Styles:I want y’all to look at that sign. See what it says?
“Cash for your home.” You know what that is?

T,C:- It’s a billboard. – Billboard.

F:What are y’all, Amos and Andy?
Are you Steppin’ and he’s Fetchit?
I’m talking about the message. What it stands for.
It’s called “gentrification.”
It’s what happens when property value of a certain area is brought down.
– You listening? – Yeah.
They bring the property value down. They can buy the land cheaper.
Then they move the people out, raise the value and sell it at a profit.
What we need to do is keep everything in our neighbourhood, everything, black.
Black-owned with black money.
Just like the Jews, the Italians, the Mexicans and the Koreans do.
Ain’t nobody from outside bringing down the property value.

O:It’s these folk!
Shooting each other and selling that crack rock and s==t.

F:How you think crack gets into the country?
We don’t own any planes. We don’t own no ships.
We are not the people who are flying and floating that s==t in here.
Every time you turn on a TV, that’s what you see.
Black people selling the rock, pushing the rock. Yeah, I know.
It wasn’t a problem when it was here.
Wasn’t a problem until it was in Iowa…
…and on Wall Street where there’s hardly any black people.
If you want to talk about guns…
…why is it that there’s a gun shop on every corner here?
– Why? – I’ll tell you why.
Just like there’s a liquor store on every corner in the black community.
Why? They want us to kill ourselves.
You go out to Beverly Hills, you don’t see that s==t.
They want to us to kill ourselves.
The best way to destroy a people is to take away their ability to reproduce.
Who is it that’s dying out here on these streets every night? Y’all.
– Yeah. – Young brothers like yourselves.

G:What am I supposed to do? Fool roll up, try to smoke me?
I’m gonna shoot the m======f=====r if he don’t kill me first.

F:You doing exactly what they want you to do.
You have to think, young brother, about your future.

C:Damn, man. You know, Furious is deep.
– He used to be a preacher? –

T: He ain’t never been no preacher.
Just reads a lot.

Boyz n the Hood (1991)

That is a window into some thoughts of the past, stemming from the worst rates. Looking back, I don’t see benevolent gentrifiers swooping down on mid-size cities affected by the violence.


bianca steele 07.22.12 at 6:32 pm

Within region, eyeballing (and anecdotal observation) suggests warmer places are easier to be homeless or to live in inadequate housing–or to live isolated from other people. IOW to be poor, or to be poor and simultaneously anti-social.


Metatone 07.22.12 at 7:28 pm

Thinking of another way to grope blindly at explanations – anyone have theories on the rise and fall? 1960 to the present day?

In 1960 things were worse than in other countries in the chart by a bit, but then it just rockets. Things have been improving since 1978 or so… but why?


AcademicLurker 07.22.12 at 7:59 pm


The standard explanation I usually hear is that the males of Baby Boom generation were in their prime troublemaking years from roughly 1960 through the late 70s.

I’m not sure how convincing that explanation is, although it probably accounts for at least some of the trend.


Demografik 07.22.12 at 8:02 pm

Age matters, especially with dramatic shifts in generational pop (boom, boom)
As well as income, education, changes in laws/ reporting…
Charts are interesting, causal stories need serious research.

Perhaps someone who actually knows about these things could be invited to guest-blog? Maybe even a panel discussion?

Thanks for the charts!


aepxc 07.22.12 at 8:06 pm

@30 Metatone:

People feel more frustrations/perceive fewer social barriers to violence during periods of social change than they do during periods of stabilisation, maybe? A la some sort of variation on Strain Theory?

USA in the 1960s-1980s saw a lot of things change or fall away – both in social mores and in the economy – that had been previously taken for granted.


Richard Skinner 07.22.12 at 8:18 pm

Where is Vermont?


mojrim 07.22.12 at 9:04 pm

Alex K@28

That theory (I think it’s a good one) shows up as well in James Webb’s book “Born Fighting” which chronicles the scots-irish tradition in the US, as well as it’s effect on the cultures of the other cultural groups. Walter Mead brings it up in his essay “The Jacksonian Tradition” though I find his love of it rather nauseating. The reality is that one of our major cultural traditions is violent by genetics and early childhood instruction, while the other OECD countries managed to weed those people out or ship them to North America in the 18th and 19th centuries.

Of course there are other correlates (e.g. age-demographics, economics, cultural collision) which have followed the same curves in the other OECD countries, but from a lower baseline. As was pointed out in the original thread, firearm restrictions in Australia reduced suicide rates but not homicide rates; the same is true for the UK. Clearly, this is a cultural problem.


Bill Murray 07.22.12 at 9:30 pm

AcademicLurker @ 30

i think soru at 20 has a good explanation too.


clifford carr 07.22.12 at 9:33 pm

Is there a graph of different states charting deaths when compared to strength of gun laws?


uffy 07.22.12 at 9:36 pm

Maybe I’m misinterpreting the views of the majority of commenters here, but it seems odd that a graph of falling violence would elicit calls for stricter laws. Have not gun laws been rather widely loosened during this time of decreased assault deaths?

This is certainly not to say that guns are decreasing crime, especially since ownership rates have been decreasing as well.

Shouldn’t we be focusing on mental health initiatives, environmental protections (lead exposure has a striking relation to juvenile crime), social services generally, access to birth control, and related?


Witt 07.22.12 at 9:51 pm

Alex K at 28: You’re never going to explain the difference between, say, West Virginia and New Hampshire or Maine, if you’ve picked up bad habits of thinking in terms of race rather than in terms of culture

I would be helpful for those of us who have not read Fischer’s book if you could connect the dots between the paragraph you quote and this statement. How do those four groups play out in terms of culture? How do 200-300 year old Mid-Atlantic settlement patterns influence trends in the contemporary US, especially in the South and West?


Bruce Wilder 07.22.12 at 9:58 pm

I think David Hackett Fischer’s folkways strains more than a bit on a tribalism born of geographical origins. Culture diverged by latitude for many decades, from the 18th century well into the 20th, with strong economic underpinnings.

The cultural imperatives and lack of empathy and ethics required by an economy centered on plantation slavery poisoned the culture of the Deep South. The violence against the aborginal population was an overlay across the whole country, but a deeper overlay for the Deep South, which had a much larger population of settled Indians to uproot and/or slaughter. The violence of the Wild West, as is well-known, was a consequence of a diaspora of Confederate guerilla fighters from the Civil War, and the migration to the northern urban industrial areas from the South during the 1930s and continuing into the 1950s, gave rise to significant outbreaks of violence and fear of violence.

The crime wave of the late 1960s remains a mystery, but that lead poisoning was a significant contributor seems difficult to doubt.

The cultural incomprehension continues. Most of the country simply cannot comprehend the legal environment surrounding guns and violence in states like Texas and Florida, and the cultural “thinking” that gives rise to it.

The rise of the plutocracy, though, will require that we rationalize, culturally, a great deal of violent repression by security forces in the years ahead.


Witt 07.22.12 at 9:59 pm

Whoops, that should be “It” in my first sentence, not “I.”


Peter Erwin 07.22.12 at 10:17 pm

soru @ 20:

I confess to being a bit skeptical about that explanation… but the apparent correlation between lead levels and crime (roughly twenty years later) over about a century and across multiple countries (e.g., this article) is pretty intriguing.

Academic Lurker @ 31:
The standard explanation I usually hear is that the males of Baby Boom generation were in their prime troublemaking years from roughly 1960 through the late 70s.

The discussions I’ve seen suggest that demographic shifts only seem to account for some of the variation; in particular, the continuation of the decline since ~ 2000 is apparently puzzling given that the Echo Boomers (Baby Boomers’ kids) were entering “prime troublemaking years” then, and should have produced an upturn in crime.


Chris Johnson 07.22.12 at 11:05 pm

I like Fischer’s book very much, and I think it has good explanatory power in some ways. But his four folkways — Puritans from East Anglia, Cavaliers from southern England, Quakers from the Midlands, and Scotch-Irish borderers — going respectively to New England, the Chesapeake, the Delaware valley, and the American backcountry, are about English immigration only. Surely that’s been diluted out by now, although I suppose the area where it’s been the least diluted is the South, where the traditioinally violent Scotch-Irish folkway may persist.

It’s a good book — thick, but readable.


Chris Johnson 07.22.12 at 11:45 pm

I suppose I should have written British, not “English immigration only.”


Watson Ladd 07.23.12 at 12:12 am

Bruce, by that logic Spain should be the most violent country in Europe, followed shortly thereafter by France. Also wild west violence was often fueled by resource conflicts, as well as the economic questions of policies favoring mining or ranching. The violence ended long before the current A/C driven mass settlement, so the culture of violence would seem to be swamped by the immigrants to the area during the 1950’s.


heckblazer 07.23.12 at 12:32 am

AcademicLurker @21

Given that most major American corporations are headquartered in Delaware, yes, other states are sending all of their criminals there :).


Muscat 07.23.12 at 12:49 am

What’s going on with AZ, NV, and (possibly) ID?

OH, PA, & NJ also have interestingly similar patterns.


gordon 07.23.12 at 12:57 am

Chris Johnson (at 39) – Wikipedia has a nice little map of ancestry by county here:

I can’t quickly find estimates of total immigration to the US by country of origin except for the most recent years given in the linked article. I have a memory that German and Irish immigration both exceeded English/Scottish, but I can’t remember where I get this from.

No doubt somebody will conclude that putting all those people in gaol hasn’t helped much, but then no doubt somebody else will reply that without so much incarceration the statistics would look even worse!

Here’s a link to some other US statistics, just for comparison with violent deaths:

From the second link: “…Fullbrook’s book reminds us that there’s a rational order in the world – that countries can learn from one another’s experience in tackling social problems and challenges, and that by striving to match what already works elsewhere, they can make their own countries better”.

What an optimist that writer is!


Matt McIrvin 07.23.12 at 1:26 am

I suspect at least part of the general decline in violent crime in the US is simply that we brute-forced it by locking up alarming numbers of people for alarming lengths of time. We warehoused the violence away in prisons and ended up with an amazing world record in fraction of the population incarcerated, which is a solution in one way, but not a very satisfying one.

Of course this doesn’t explain the parallel declines in less-violent countries that experienced no such prison expansion. Also, the last I looked into this, it seemed like the US prison explosion started earlier than I’d thought.


faustusnotes 07.23.12 at 1:37 am

I’m with Watson on thinking Bruce’s explanation in genocide and slavery can’t be the whole story. Well, slavery could, I suppose. But Australia has a history of genocide and just can’t mix it up with the US on murder patterns. Britain has a history of colonialism and bad, bad things but has a very low murder rate – but anyone who has lived in London knows that it’s a bad, dangerous place and the problem is not immigration.

So once again, it’s very simple: cohort effects and guns.


Omega Centauri 07.23.12 at 3:35 am

Its only wikipedia level knowledge, but it is claimed honour culture develops when a group of people is isolated with respect to law enforcement, and have to defend their own interests, by threat of violence. The settlers for much of the south came from the pastoral parts of the UK where those conditions prevailed. Also in the early years of the US west, one couldn’t count of the law to enforce ones property rights, so an “honour culture” is a natural way to cope with the situation. Once established these attitudes can persist for a very long time.


heckblazer 07.23.12 at 4:02 am

Omega Centauri @ 49

That would also apply to every slave plantation in the American South through 1865. If you include Jim Crow that pushes things up to circa 1965.


drs 07.23.12 at 5:25 am

Pinker puts a lot of weight on Hobbes’ Leviathan, or rule of law, in pushing down violence levels, and says violence levels are typically high in failed states, or in areas within states where law enforcement doesn’t extend, by neglect or fear or distrust by the locals. “Duh”, in a sense, but that applies to not just drug gangs and other criminal activity but to US blacks in general, given police racism and black distrust. Point is, a lot of that violence isn’t just people being criminally mean to each other, it’s also private attempts at justice or at least revenge and deterrence in the absence of law enforcement. Feuding.

And IIRC, he says that in the South, the higher homicide rates aren’t from higher rates of profit-oriented violent crime, muggings or home invasions and such, but from higher rates of “he insulted me” or “she cheated on me” violence. Honor or impulse killing. Which could connect to both the Scotch-Irish initial settlers and the later effects of slavery and then Jim Crow and lynchings and all. Basically, higher acceptance of private violence for some reasons.

IIRC, that even got tested, with fake resumes, identical but for “I was in prison for stealing to feed my family” vs. “I was in prison for killing someone because of [some honor reason]”, and Southern employers turned out to be friendlier to the latter reason.


Katherine 07.23.12 at 8:51 am

The point about honour killings being the possible reason for higher murder is an interesting one. Here in the UK, the phrase “honour killings” has come to be attached to a subset of murder carried out by immigrant parents from South Asia, outraged by the Westernised behaviour of their offspring. That does of course come with the usual Daily Hate pile on which closely resembles Islamophobia and/or racism.

It would be interesting to see the statistics though – does a subgroup of the UK population, who might fulfil the criteria outlined above – areas within states where law enforcement doesn’t extend, by neglect or fear or distrust by the locals – also have a statisticall significant higher murder rate? Ie is there a way of measuring this hypothesis by comparison?


Bruce Wilder 07.23.12 at 9:09 am

Watson Ladd: “Bruce, by that logic . . . ”

Please do not impute stupidity to me, Watson; at the very least, it is impolite. If you do not understand what I have written, own that, and ask an appropriate question. Or, simply refrain from commenting, altogether.


Data Tutashkhia 07.23.12 at 9:27 am

I get the impression that in the US context ‘honor killings’ translates into gang violence, hits and retaliations.


bianca steele 07.23.12 at 12:02 pm

Katherine makes good point–besides “feuding” I would add “dueling.” (Which in Europe would hardly have been tolerated among ordinary folk, I suppose, even at times when aristocrats were expected to defend their honor with arms.)


Barry 07.23.12 at 12:49 pm

“Katherine makes good point—besides “feuding” I would add “dueling.” (Which in Europe would hardly have been tolerated among ordinary folk, I suppose, even at times when aristocrats were expected to defend their honor with arms.)”

It might well have been, with the proviso that it was informal, used fists/clubs/knives rather than swords or guns. The local police/elites could have an attitude that a certain amount of ‘letting off steam’ or ‘dealing with scum’ would be tolerated.


chris 07.23.12 at 12:50 pm

Surely you don’t believe that people kill and get killed because they are black or Am. Indian.

No, they get killed because they are perceived as threatening, and they are perceived as threatening because they are black (not sure if that applies to Native Americans too).

Do these statistics include people killed by police? The police are often (usually?) not charged in such cases, but I would still consider it a violent death, even in cases where it could be argued that the killer is not morally to blame (e.g. if the deceased actually *was* a threat to the police, as opposed to merely being perceived as one).


rf 07.23.12 at 1:15 pm

‘Honor or impulse killing. Which could connect to both the Scotch-Irish initial settlers’

I remember hearing a good bit about that a few years ago, (can’t find many links now), how the Scot’s-Irish culture of honour and ‘herder society mentality’ contributed to the high crime rates in the Deep South.
Here’s one link anyway, might be nonsense, I wouldn’t know.

(Also interestingly, perhaps, crime rates in Northern Ireland and Scotland seem to be higher than any other regions of the UK)


bianca steele 07.23.12 at 1:20 pm

Now that I think of it, though, the more traditional explanation for honor culture in the South is the (still widespread) belief that it descends from a true transferral of the English aristocratic tradition in the form of the so-called “Cavalier” legacy. This is in part behind the respect Southern culture has historically received among, say, European elites, compared to general “vulgar” Americans who may even be rich. This is very hard to reconcile with what people are saying about the Scots-Irish heritage, though.


Matt McIrvin 07.23.12 at 1:42 pm

The ethnic discrepancy could be aspirational: everyone wanted to be identified with the aristocratic English gentleman planters of Tidewater Virginia, not with the scrappy lower-class Scots-Irish.


SamChevre 07.23.12 at 1:43 pm

bianca steele @ 62

The “Cavalier” culture is Tidewater/plantation culture (think Washington and Jefferson); the Scots-Irish culture is Appalachia (think Davy Crockett and Andrew Jackson).

They are very different cultures, and historically do not get along very well with each other.


drs 07.23.12 at 2:19 pm

Who colonized the Deep South, MS and AL?

For that matter, Louisiana is the lead state, so maybe Cajun culture should get some scrutiny.

Maryland could be Tidewater, or could just be Baltimore.

South Carolina? It’s high, and the traditional high violence state. Edgefield is long famous for violence, and an article says Scotch-Irish. Also:

“It is an extraordinary historical detail: as late as 1910 white landlords would give each of their black sharecroppers a few lashes of the whip in a little ritual of degradation–a flesh-memory of slavery”


sanbikinoraion 07.23.12 at 2:52 pm

Does it take a genius to work out why violence levels in Northern Ireland are high…?


rf 07.23.12 at 3:05 pm

“Does it take a genius to work out why violence levels in Northern Ireland are high…?”

According to the link homicide rates in Northern Ireland are lower than Scotland. (2.3% in Scotland 1.92% Northern Ireland per 100,000 – 1.5% England and Wales – between 2003-2007)


Marc 07.23.12 at 3:22 pm

I’ve always thought that the legacy of slavery is a substantially better explanation than ideas related to particular European ethnic groups.

Slavery didn’t just impact blacks. Physical labor was associated with slaves, and devalued for all. Education was strongly discouraged – in fact, it was made illegal for slaves in the period prior to the Civil War. In the rest of the US public education had a broad constituency. In the South education itself was a low priority, especially for minorities. Prior to desegregation the black schools had minimal resources. Post desegregation whites built up a shadow private system (“christian academies”) and defunded the overwhelmingly black public schools.

The above two give you a large population of people with minimal opportunities for advancement, and with all of the negative environmental factors of extreme poverty and poor nutrition.

The South has also always had white and black, rich and poor in close proximity to each other. This always causes crime and unrest, which was governed by extremely harsh sanctions on black vs. white crime, real or imagined. The legal system virtually ignored black vs. black crime. You therefore get large prison populations (from the first effect) and weak legal protection for minorities (from the second.) Both amplify the presence of minorities in the criminal justice system, furthering the cycle of violence by diminishing job prospects. Add everything up and you get pretty much the crime patterns that you see. They also act as reinforcers of racist attitudes, thus generating a self-sustaining cycle.


Wonks Anonymous 07.23.12 at 4:10 pm

Fischer argues that immigrants to different regions tended to assimilate into the dominant culture. So New England retains many of its old Puritan characteristic, even if the inhabitants aren’t Congregationalists from East Anglia any more.

It seems wildly implausible to me to blame violence in the West on Confederate soldiers, frontiers have been violent places for a long time and all the world over (Canada is unusual in that the mounties arrived before many settlers as described in Keeley’s “War Before Civilization”). And while the tidewater “Cavaliers” of the plantations certainly had a culture of dueling, really serious violence comes from feuding where each attack calls for retaliation. An economy of pastoralism also gives rise to an incentive to frighten others away from stealing your (very mobile) assets through the threat of retaliation. Mountainous areas are both hard for states to “climb” (in James Scott’s terms) and more amenable to pastoralism than intensive agriculture.

Marc, the studies I’ve heard of certainly indicate that black on white crime is more severely punished (legal briefs on the death penalty often point out that bias), but to say crime against blacks is “virtually ignored” seems an understatement. My guess is that there are more people in prison for killing blacks than whites, even while blacks are less adequately protected and vulnerable (I recall Kleiman having some findings on that but can’t recall them now off the top of my head).

If Canada, the U.S & Australia all have little to be proud of regarding their treatment of indigenous populations, which nation may be proudest? What little I’ve heard of Finland’s treatment of the Saami doesn’t sound too bad. Among the Anglosphere settler nations, Canada seems to come off looking best, but that could just be my ignorance talking.


Bruce Wilder 07.23.12 at 5:28 pm

“Who colonized the Deep South, MS and AL?”

A diaspora of low-land South Carolina planters’ (and overseers’) sons led the way.

Hill country was a different story.

“It seems wildly implausible to me to blame violence in the West on Confederate soldiers . . .”

Of course, we have names and biographies, despite the alleged “implausibility”.


bianca steele 07.23.12 at 5:47 pm

Changes things a bit if we try to think of John Crowe Ransom and Allen Tate as “aspirational,” though, doesn’t it? Fascinating.


Marc 07.23.12 at 5:53 pm

@69: The case of capital punishment is especially clear; the odds of a death sentence are much higher if the victim is white. See

For racial trends in being crime victims see


William Timberman 07.23.12 at 6:17 pm

bianca steel @ 62

This is very hard to reconcile with what people are saying about the Scots-Irish heritage, though.

Having grown up embedded in that culture, (a father who was half Appalachian hillbilly, half Pennsylvania Dutch, but was raised on the hillbilly side), I can tell you from personal experience that inviting someone outside (away from the womenfolk) was as common to the culture as breathing. Insults, even subtle ones, required a man to offer to inflict a beating on the offender. If he didn’t, he wasn’t a man. On later reflection, I suppose it was the freeholder’s equivalent to dueling, as no weapons were usually involved, although one didn’t shy away from using weapons if the offender brandished them first.


heckblazer 07.23.12 at 7:16 pm

Marc @ 68
You’ve managed summarize my thoughts better than I could. I would add that a strong honor culture is a contributing factor in urban black violence. “Dissing” someone (short for “disrespecting”) is a good way to start a fight. I make no claims as to the origins of that, it could be an echo of cavalier culture, it could just be that when you’re poor honor is the only thing you have of value, or something else.


Bruce 07.23.12 at 7:49 pm

Excellent graphs!
Looking at the one that overlays individual US regions with the other countries highlights that it’s somewhat hard to reconcile the strong trend in the overall US OECD number with the regional numbers. It’s complicated because of course the four regions don’t have the same populations, and total populations in region/country may have changed, but in the overlap region, the only one that shows a decline from start to end is the midwest, and that decline is proportionally smaller than the OECD US decline. Any sense of what’s going on? if you add all the CDC numbers together do they come out at all comparable to the trends in the OECD?


Bruce 07.23.12 at 8:09 pm

(As a footnote, let me apologize if I confused the midwest with other regions – I’m colourblind enough that I may have mistaken shades; but the overall point I think remains – none of the four regions appear to show as proportionally large a decrease as the total.)


Alex K. 07.23.12 at 10:28 pm

“The “Cavalier” culture is Tidewater/plantation culture (think Washington and Jefferson); the Scots-Irish culture is Appalachia (think Davy Crockett and Andrew Jackson).

They are very different cultures, and historically do not get along very well with each other”

Quite so. Violence also played different roles in those cultures, even though honor played a central role in both.

On the Tidewater Cavalier culture side, the role of violence was the enforcement of a strict hierarchical order. The violence flowed from the superior against the subordinate or the equal — and the occasional violence from the inferior was rare.

Here is Fischer:

Violence was thought to be the legitimate instrument of masters against servants, husbands against wives, parents against children, and gentlemen against ordinary folk. But violent acts by servants against masters, or common folk against gentle folk was followed by savage punishment.

This was not the case in the Appalachian culture. There the expectation was one of equal esteem, regardless of one’s rank or wealth. Visitors who thought of themselves as being of high-rank were complaining about a lack of deference in the backcountry:

The Anglican missionary Charles Woodmason filled his journal with angry accounts of “ill treatment” by “insolent” and “impudent” settlers who stubbornly refused to display the deference which he thought his due. He complained that these people were “the most audacious of any set of mortals I ever met with.”

Violence was still linked to defending honor, but in a different way:

In the backcountry, honor had very little to do with gentility.[…] Backcountry ideas of honor were understood more in terms of valor and virility. To behave dishonorably was to commit an “unmanly act.” Folk punishments in the backcountry were designed to inflict humiliation by depriving an offender of his manhood — sometimes in a direct and literal sense.

Violence often took the form of family or clan feuds, and a powerful hostility against strangers (where a stranger was anyone not in the extended family or in the neighborhood) was ubiquitous. There was also a strong preference for avoiding legal institutions and settling disputes in a non-legal framework:

The mother of President Jackson prepared her son for this world with some very strong advice.”Andrew,” said she, “never tell a lie, nor take what is not your own, nor sue anybody for slander, assault and battery. Always settle them cases yourself.”

Sometimes, the lack of respect for legal institutions manifested itself against magistrates:

In 1767, when several gentlemen justices in South Carolina tried to bring some “banditti” to trial, the magistrates themselves were seized and tried before a kangaroo court by the intended defendants. One justice was dragged eighty miles at a horse’s tail.

This is just a sample of the evidence David Hackett Fischer offers supporting the case for a tradition of violence in the Appalachian culture. But it seems to me that the above is sufficient to show that you simply can not treat Puritan New England and the backcountry in remotely similar ways, when you analyze the role of violence in those cultures. Even the Tidewater Cavalier culture deserves a separate treatment.


Shelley 07.23.12 at 10:31 pm

Hard to think of a more embittering public issue than knowing that because of the power of the NRA, it will only be a few months before we are grieving again.

Is it real grief if you keep letting the loss repeat?


Nathanael 07.24.12 at 12:28 am

@22: The lead theory has MASSIVE amounts of evidence behind it.

Nowadays the main lead exposure is from firing guns (a lot of the lead in the bullets evaporates). Lead bullets need to be banned in favor of iron bullets. This would probably cause the NRA to become sane within a generation!


Watson Ladd 07.24.12 at 2:02 am

bianca, I’m curious about this claim of respect for Southern culture in Europe. Most European immigrants to the US ended up fighting for the Union in the Civil War and regarded slavery with abhorrence. Certainly within Europe love for the Confederacy was mostly from economic considerations and in the case of England hampering a rival.

Wonks Anonymous., Fischer argues that immigrants assimilate. In the case of the South there was a period of military occupation aimed at destroying all vestiges of slavery. Later in the century northerners moved south with the industrialization of the region, and ultimately more federal force was used to end discrimination. True, there is a cultural difference today. But the New South

Bruce, why does the US bear the legacy of slavery and colonization but not Manchester, France, or Spain despite even darker pasts? What are we to make of life in a New York City where the culture of assimilation and dominance is one that rejects assimilation, or a Boston where Irish Catholicism is as strong a presence as the faiths of the old families? To which historical pattern of settlement is the violence on the South Side of Chicago linkable?


Kaveh 07.24.12 at 3:54 am

Wattson @47 & 80: why does the US bear the legacy of slavery and colonization but not Manchester, France, or Spain despite even darker pasts?

Much more violence in the the South and Latin America is exactly what I would expect from Bruce’s logic. I thought this is obvious, but maybe I misunderstood your question? Spain and former Spanish colonies (for example) aren’t some monolithic, tightly-connected social entity. The colonists are linked to Spain mainly by descent. Spanish colonists came from Spain, yes, but while there was a long period when Spain could claim political sovereignty, there was not close social contact between European imperial centers and their colonies–i.e. how many Spanish, French, or British in Europe ever actually met a Black slave or a native American, much less dealt with them on a daily basis? Violence lives on in places where the violence was actually inflicted, among descendants of people who were doing or receiving it, not people who were only distantly connected to them via political sovereignty and sharing the same language and ancestors (i.e. Europeans). Unlike Europeans, Spanish ‘creoles’ in the Americas were, if not directly involved in the violence of colonialism, then far, far closer to it. And Native Americans had to live in fear of it. So of course it is in the Americas where that violence lives on. Not that the colonial relationship as a whole didn’t have a profound effect on European attitudes, and shape their behavior when, for example, later generations traveled outside Europe (see e.g. Mary Louise Pratt, Imperial Eyes, on this, or Said’s Culture and Imperialism), but that is a different relationship from the direct and personal one that exists in the colonies where the violence actually took place.

(I guess you could consider current European hostility to (especially) non-European immigrants a legacy of colonialism more generally, since that history of colonialism is the outstanding factor shaping European attitudes towards Asians, Africans, and other ethnic others, i.e. the obvious connection between European colonialism and the rise of theories about race, and the way these theories were mobilized against Jews in Europe, by the Nazis). And I have heard it said that the brutality of WWI and WWII were partly a legacy of the violence of colonialism, because various Europeans acquired a belief in their own invincibility and a habit of using all kinds of indiscriminately violent tactics (e.g. chemical weapons) on their (non-European) enemies.)

To which historical pattern of settlement is the violence on the South Side of Chicago linkable?

The Great Migration? Aren’t most Blacks on the South Side are descendants of slaves, just like most other Blacks in the US, and thus inheritors of a culture that was profoundly shaped both by slavery and by interaction with the ‘Cavalier’ culture of the Deep South? Am I missing something here?


Kaveh 07.24.12 at 4:15 am

re Fischer’s argument and later immigration, apart from assimilation, I would expect that elective affinity would also be a factor. Anecdata: there was an interesting This American Life segment about Cicero (an incorporated town bordering Chicago’s SW side) that notoriously excluded Blacks from settling there since the mid-20th c. The city politics are notoriously, stubbornly corrupt. But in the last couple decades, Mexican immigrants settled there much more successfully, even to the point of success in city politics, because they had the skills (and expectations/approach) to handle politics based on backroom deals & personal ties. Combine elective affinity with chain migration and you could have a very strong tendency against cultural change.


heckblazer 07.24.12 at 5:20 am

Watson Ladd @ 80
They weren’t necessarily fighting on the side of the Union by choice. The NYC draft riots were sparked by the immigrant resentment of being drafted to fight a war they had no interest in.


Allen Hazen 07.24.12 at 7:43 am

A bit of “folklore” from New York City (where I grew up, and where my mother grew up: so, family memory goes back to before WW II): there is ALL THE DIFFERENCE IN THE WORLD between New York City blacks of U.S. ancestry and New York City blacks “from the islands” — West Indians. Caribbean blacks had much less social disfunction, were upwardly mobile. (When General Powell became the top U.S. military man, there were news stories about his background, and my response as a New Yorker to learning that his parents had immigrated from Jamaica (I think) was “That figures.”
——So. (1) There is a big “cultural” difference here. (2) I’ve always assumed, though with no particular knowledge, that it had something to do with post-slavery history: American blacks were (particularly but not exclusively) an oppresseed and stigmatized minority, but being black was “normal” in Jamaica (etc). (3) I would be interested if someone knows some statistics about comparative violence rates in the islands, among New Yorkers from the islands, and among black New Yorkers of southern U.S. origins.


Katherine 07.24.12 at 9:02 am

How much then does the culture of “honour”, however that is expressed in a particular geographical area, cross over with a toxic masculinity? It seems to me, just going on descriptions above you understand, that both the Cavalier “honour” and the Appalachian “honour” have some commonalities in the expectation that a man must act in a certain way, and be expected to react to insult (to him, or his -property- woman) in a certain way, and those ways involve a display of strength via violence.


bianca steele 07.24.12 at 1:15 pm

You see it in passing into the twentieth century–in Trilling’s and Arendt’s evident beliefs that the South had a unified culture in a way the North had, and that this was a good thing, and even might serve as a model for the US as a whole–in the way Beauvoir and Sartre wrote about William Faulkner, etc. It even infected Communist propaganda (in a way that seems to me to contradict Marx, but I’m not a Marxist so I’m not going to worry about the contradiction that much).

I don’t have notes, though, and obviously it would take me some time to bring forward substantiation for a real argument.


Wonks Anonymous 07.24.12 at 2:52 pm

Bruce Wilder, I am not saying that confederate soldiers were absent from the west or did not commit violence. I am saying frontiers are violent even in their absence. The U.S had a violent frontier in the colonial era as well.


Shatterface 07.24.12 at 7:38 pm

Just a hunch, but the regional distribution of firearms deaths doesn’t look like it correlates positively with the teaching of evolution or tolerance of gay marriage.

Comments on this entry are closed.