Why Are Killing Rampages Increasing? A Guest Post


Peter Turchin is a professor of ecology and mathematics at the University of Connecticut and author of War and Peace and War: The Life Cycles of Imperial Nations. Much of his work concerns a new field known as cliodynamics, which attempts to discover general principles that explain the functioning and dynamics of historical societies. He has agreed to be a guest blogger this week; this is his first of two posts.

Why Are Killing Rampages Increasing?

By Peter Turchin

A Guest Post

So far in 2008, The New York Times reported at least six shooting rampages. Just two weeks ago a man shot or stabbed 10 people in northwest Washington, six of whom died. After his arrest, he told the police that God told him what to do and told him to “kill evil.”

Are these episodes of senseless mass murder increasing? A systematic search of The New York Times from 1946 to the present suggests that the increase is real and very dramatic:

MassacreNumber per year, plotted by decade.

Over the last half-century the incidence of massacres — shooting rampages, killing sprees, etc. — increased roughly 10-fold. These numbers exclude crime-related (along the lines of Reservoir Dogs) and family-related (“Man Shoots Wife, Kids, Self”) multiple murders.

Why is this happening? Note that I am not concerned here with why killing rampages occur, but why they are on an increase.

The rise in massacres began during the 1960’s and shows no signs of abating. It’s a long-term trend, and I think it is telling us something about fundamental ways in which our society is changing.

I am interested in this trend because my research, in general, focuses on investigating long-term dynamics of historical societies (I call this cliodynamics, Clio being the muse of history). One empirical result from this program is that societies tend to experience recurrent waves of violence and political instability; these waves are themselves the results of long-term social and economic trends (more on this in my book War and Peace and War). In other words, when we want to explain trends evolving on a slow time scale, we look to mechanisms operating on the same scale.

A study by The New York Times in 2000 found that the majority of massacres happen in two situations: at the workplace (about one-third of the total) and at school (one-fifth). Here’s my hypothesis for the increase of work-related rampages (I will deal with school rampages in my next blog):

We know that during the 1970’s something changed in the American economy, and in a very fundamental way. Between 1930 and late 1970’s, real wages grew essentially monotonically (overall, they grew by a factor of 3.5). Since then the wage stagnated (white-collar workers) or even declined (blue-collar workers). These are official statistics (Bureau of Labor Statistics); the actual situation must be worse, because the real rate of inflation is probably underestimated by creative folk at government statistical agencies.

In any case, the costs of big items that define the middle-class way of life — houses, college education, medical insurance — have increased faster than the official inflation rate.

The implications are obvious, and it is surprising that they are rarely brought up in the context of massacres. As their economic prospects deteriorate, many breadwinners find themselves under unendurable pressure to maintain the socially expected level of consumption. Under these conditions, people — whose psychological problems would be borderline in the gentler economic climate of the 1950’s — today “go postal.” So the harsher the economic conditions, the greater the numbers of those whose latent psychological problems develop into full-blown psychosis.

The New York Times from 2000 provides some indirect support for this hypothesis: 57 percent of rampage killers in their database were unemployed when they went on rampage. Remember the 1993 movie Falling Down? The resemblance between the fictional character, brilliantly played by Michael Douglas, and real rampage killers, described in the press reports that I read, is uncanny.

To close, I submit that there is a plausible connection between worsening economic conditions (for most Americans, except the rich) and an increase in workplace rampages. Substantial and growing proportions of massacres, however, occur at schools and universities. I will address this issue in my next blog.

William B

I submit that there is a plausible connection between worsening economic conditions (for most Americans, except the rich) and an increase in workplace rampages

I would be interested in hearing by what metric you claim that economic conditions have been worsening for "everyone except the rich." Certainly not income per capita.

Imad Qureshi

To David #8

"So, your grand unifying theory is that people under a lot of stress may become violently mentally unstable and that not having enough money is stressful? "

Professor also said

"As their economic prospects deteriorate, many breadwinners find themselves under unendurable pressure to maintain the socially expected level of consumption."

I think it's the expected level of consumption by society that causes a lot of stress and some people turn violent.


The NY Times also reports fewer mass-fatality bus crashes, because there is less need for one- or two-paragraph filler material (page design programs have advanced dramatically over the past 20 years).

So does this mean there are fewer traffic crashes? No, it means that the NYTimes reporting has changed over the years.

The dataset is flawed.

Dan Q. Public


I'd have to disagree. Yes, government employment is weighted toward veterans (and rightly so), but over the course of time on Prof. Turchin's graph the number of veterans (defined as those who have served honorably in the armed forces) is small compared to the number of combat veterans. (I have served in the US Army and thusly receive preference in applying to government jobs, but I have never seen combat.)

'Going postal,' I would posit, has less to do somehow with PTSD than it does with the specific job of working in a post office: intensely repetitive and non-stimulating, while simultaneously having to deal with the outbursts and tantrums of the angry, non-understanding public. Veterans get preference in a wide range of government jobs, but you never hear of someone grabbing a gun and 'going Forest Service.'


I did a quick google to test the population growth idea. There is a chart here for the same time period showing population growth in the US.


The increase in population does not account for the upswing in massacres. The point as to reporting is valid for comparing say 1945 to 2005, but within the last twenty years, the reporting should be consistent.

Another factor that could have changed is the receptiveness of NYT readers to these reports. Has the media always focused on violencce and mayhem?

John F

"...the actual situation must be worse, because the real rate of inflation is probably underestimated by creative folk at government statistical agencies."

I'm sorry, but this flawed assertion, while I don't disagree with the premise, took away from your credibility, for me. It draws into question your critical thinking ability. Not because either statement by itself is incorrect but because you can't definitely assert condition A because "probably" condition B.

It makes me wonder what other logical leaps-of-faith you've taken in your analysis, particularly in respect to questions as raised by #2.


I don't think the right approach is to look at population changes but rather changes in population with regards to combat exposure. For instance the phrase "going postal" comes from the large number of postal employees who would go on "rampages", this in fact had nothing to do with being a postal employee so much as the people hired by the post office. It used to be (and may still be) that the USPS was required to hire wartime veterans how passed the civil service test over civilians who passed. Returning veterans with mental or psychological issues weren't very employable elsewhere so they ended up with the post office; when they "snapped" the causation appeared to be the post office when in fact it wasn't.

The first thing I thought when looking at the graph was that the spikes looked to be following the Gulf War, and as soilders are returning from Iraq, leading me to wonder if it is a combat issue and not a civilian issue. I wonder what percentage of those people had seen combat experience.


Michael C

This post was an embarrassment to the Freakonomics blog, it's regular writers, and the NYTimes. The most serious flaws in Prof. Turchin's analysis have already been mentioned above. I will just say that it is ludicrous to assert that economic conditions have been steadily worsening for 40 years for all but the rich in this country. It is equally preposterous to argue that the standard measures of inflation have been underestimating inflation during that same period. You don't even define "killing rampage" or bother to search for a more reliable data set. You ignore trends in population, population density, gun ownership, treatment of the mentally ill (the hospitals hypothesis mentioned earlier seems very plausible), or any number of demographic trends. The unsupported assertion that life is more stressful today than it was in the "gentler" in the 1950s is almost offensive. Finally, what can you really conclude from the fact that those who go on rampages are often unemployed? Could it be that people who are on the brink of slaughtering innocent people aren't generally very good at holding down a job?

Also, regarding the term cliodynamics (I cringe on your behalf), which you define as the search for the "laws of history," what is it that you believe historians do?(see Prof. Turchins website for even more evidence of his pretentiousness and intellectual laziness).

I wonder if Prof. Turchins has the courage to either admit his silliness or to attempt to defend his risible hypothesis either here in his next post.

Finally, as this is my first post, I wanted to say that I read this blog everyday and find it fascinating, informative, and entertaining. Let's have more of Peter Leeson and less of Turchin.


erik de koster, brussels

I guess your data cannot be used for this question: you need a more secure data source, and I also would suggest an international comparison. For data, data from law enforcement offices would be probably a lot more reliable (I confess I do not know whther these data would be centralized in the USA at the FBI or whther they would be stored at state level or even county level); for comparison, Canada, western europe and Japan immediately come to my mind. Without serious internal and external controls this observation is, I am afraid, essentially worthless, it is just interesting enough to initiate some serious work.

Johnny Mudshark

@21 - It still all boils down to the stunningly obvious that as people are exposed to more societal pressures -- a stagnent economy with more diverse players = higher competition for prime breadwinner spots -- these pressures drive the marginally stable over the line into instability. And some of the newly unstable become violent.


Intriguing theory... would like to see more of the data.

21/Petr - Thank you for expanding my usable vocabulary by forcing me to look up the English usage of "epsilon," previously but a Greek letter to me.


First, I think that the likelihood that the NYT just didn't cover small town massacres is epsilon. Regardless, if that view is valid ( that NYT simply missed some killings) it's still a statistically significant sampling and a clear pattern.

Secondly, Professor Turchin elides the other great societal changes that occurred during the period under scrutiny: the wholesale inclusion of an entirely different gender into the workforce. I don't think the two strands, stagnant wage pressures and the sudden inclusion of female co-workers, are un-linked. There's also the very real growth of the black middle class.

As a young child in the 70's, my earliest recollection of father getting angry was often, and loudest, at 'women drivers'... further upheaval occurred when mother wanted to 'get a job' (meaning, a corporate job, with corporate responsibilities, not teacher, nurse or secretary...) Yes, anecdotal as these are, I could sense, very young, the pattern emerging out of the narrative. "Breadwinners" were no longer SOLE breadwinners. Prior to WWII a man could work for a corporation his entire career and never face competition from a women, or risk be demoted into capitulation to a women (which is how they saw it...). Maybe there's a case to be made that some faced sudden and greater exposure to different pheromones...

That many have largely adjusted to this reality doesn't mask the stress this may have placed on some and, coupled with the aforementioned wage stresses may have pushed some over the brink.

I think just the general stress of coming to work changed over that period of time as people adjusted to a more diverse workforce and stagnant wages. The generalized anxiety over women in the workforce and (sometimes forced) integration of white and black might simply have been too much for some. I don't think this derives from particularly sexist and/or racist undercurrents: change is just harder for some people. I think the tendency to romanticize prior experience (rose-colored glasses) may have been contributive as well...



If there was merit to this theory, then you would also see increases in murderous rampages in places where economic hopelessness has really increased like the Gaza Strip and the West Bank. Hasn't happened there has it? Oh, maybe it has.

Laurence Frank

I would like to suggest an entirely different explanation, one which might also help explain the rise in depression and some other mental disorders. I believe that modern life is so utterly divorced and distant from the biological and environmental circumstances under which we evolved that many mental disturbances are a reaction against an existence that is biologically meaningless. Our evolutionary ancestry, going back long before there were even apes and hominids, was rooted in the vast complexity that is Nature. Each individual was responsible for feeding itself, avoiding or defending itself from predators/enemies, finding shelter from weather extremes, coping with the complexities of its social group, navigating through a complex natural landscape with all its daily,seasonal and geographic changes, etc., etc. By contrast, modern life is grossly oversimplified, lacking in the daily variety and challenges of 'life in nature' to which we are adapted. This simplified and unchallenging artificial existence is in itself a chronic stressor: at a very basic biological level, our lives no longer make sense. The process of domestication in animals is in large part a matter of losing reactivity to circumstances that are stressful in the real world: overcrowding, close contact with predators (i.e.humans)which are normally avoided, monotonous diet, boredom,confinement, etc. Although modern man is pretty well domesticated, i.e. nonreactive to those stressors, they still take a constant toll which often manifests in some sorts of mental illness like depression or anxiety. Under additional stress such as job loss, this chronic but subconscious organic dissatisfaction with existence occasionally erupts as murderous rampage.



So the premise is that the "unendurable pressure" of the daily life is causing those massacres? Okay but I would think that such an "unendurable pressure" would have also caused things like: alcoholism, domestic violence, suicide first and only after these options for venting the pressure have been exhausted, only then people would turn to killing rampages to express their anger.

So are alcoholism, domestic violence and suicide on the rise following the same correlation that is discussed here?

Also the author noticed a correlation but what about causality?


In workplace violence, I would be a cold Coca-Cola that there are two factors at work....

1) In the increasingly global economy, with downsizing and repurposing, employees can no longer "depend" on working for company (and having the company provide for them) over a lifetime. This issue--some "postals" are related to losing jobs--along with the heightened competition (having to do more with less), and receiving less for your work due to "competitive issues," are all stressors. And when you can't get ahead...you take heads.

2) It may be part and parcel of a growing distrust of the legal system. Consider, if you kill a child tomorrow, you MIGHT go to trial in a year. And if you are found guilty and sentenced to death, it will likely be AT LEAST a decade before they will execute you...long enough for the grandparents of the child to die, and for people to forget the heat of the moment. That inspires a vigilante approach to justice (and rightfully so--a child killer shouldn't last a week, let alone 10 years or more). And if a person PERCEIVES injust treatment toward themselves, and realizes that it would take years and millions to fight the corporation, they just take a do-it-yourself approach and blow away the person/persons that they perceive as part of the injustice.



I think that your data needs some controls. Have you controlled for:

* Population size? The US population grew from 131.7 million in 1940 to 281.4 million in 2000.

* Region? The NY Times is a national paper that primarily focuses on one of the largest regions in the US. Do your figures focus strictly on the NYC region, or do they include state and national reports too?

* News velocity? In 1946, the time for news to spread from one town to a state/region, and then to become national news could be days or weeks. In 2008 the time can be measured in minutes or seconds.

Personally, I'd put my money on the development of the "If it bleeds, it leads" mentality in news departments.

Kirill Pankratov

This is very intriguing - a relatively steady, long-term increase of "rampage killings", even in times when overall murder rate goes down. Some changes can be clearly identified with overall state of the society, others are fairly hard to relate to in the broader sense.

The jump in killings in mid-60's is likely related to overall turmoil of that time. Interestingly, in early 80's, during severe recessions of 1980 and 82, the number actually decreased a bit, while it went up again in late 80's. And how to explain another jump in mid 90's - the time of relative peace and prosperity? Clearly there are lonegr-term trends at work, not just ups and downs of an economic cycle and the stock market.

There is also an interesting related book "Going Postal" by Mark Ames, which draws a very interesting parallel between today's rampage killings and slave rebellions in the 18-19th centuries:




Well it certainly makes sense at the common-sense level. Most social actions end up having roots in economic triggers.

And while the cliche tells us that money cannot make us happy - it is perhaps equally true that money can make us UN-happy, both too little and too much can cause stress.

Of course too little in this context means little-relative-to-others and too much has the same connotation.

So money (or economic well being) is probably not a great satisfier, but a lack of it relative to your peers can certainly be a great DIS-satisfier.

I look forward to the next installment




I assume that your numbers are corrected for population increase during that time?