Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
How to Lie with Statistics
“F.B.I. Treating Attack in San Bernardino as Terrorism” reads the New York Times headline, implying that perhaps the Times demurs. The sense of the paper being dragged, reluctantly, from a preferred narrative is accentuated midway through the article by a curious graphical island, appearing in splendid isolation from the actual text:
Clicking through to the source reveals the contention that there have been more deaths in the United States from “right wing attacks” than “jihadist attacks” since 2002. The take home point: Even if the San Bernardino shooting is the work of ISIS-inspired terrorists, as the FBI now acknowledges, what’s the big deal? The Second Amendment is the real problem.
In his 1954 classic How to Lie with Statistics, Darrell Huff devotes a chapter to manipulating impressions by resort to the Gee-Whiz Graph. The Times entry stands as a modern-day exemplar of the type. For starters, care to guess why 2002 was selected as the base year for comparison rather than 2000? What are the criteria used for categorizing murders as “jihadist” versus “right wing”? What murders are excluded from both categories and why? And what about the evident quiescence in violent death from both camps depicted from 2002-2008? What might possibly have changed in the United States beginning in 2009 to bring about a dramatic increase in successful domestic terror attacks? The authors offer no comment.
And that speaks volumes.
Both numbers seem rather negligible to me. How ’bout we stop blowing extreme outliers out of proportion?
The world is doing fine.
That is not be an Obama-approved graph. It doesn’t show the Fort Hood killings as workplace violence, and it doesn’t show the Boston bombing as “we shouldn’t rush to judgment.”
Compared to the number of people killed at the hands of ObamaCare, it doesn’t amount to much.
Or the number of people killed other ways…
I wonder if similar graphs could be made for Israel.
Why would extremist attacks of both kinds increase under Obama?
I wonder if similar graphs could be made for Israel.
*****
They look very different, because Americans are massively more violent than Israelis.
Lower risk of retaliation by US authorities. Higher probability of favourable treatment by federal law enforcement. Much higher probability of sympathetic media coverage. Maybe get an invitation for beers at the White House.
Not a graph, but a table:
http://www.jewishvirtuallibrary.org/jsource/History/casualtiestotal.html
Israel’s homicide rate is a little trickier to count–are they murders or acts of war or terrorism? According to wikipedia here, it goes from 2.4 down to 1.7 per hundred thousand in 2012 by counting nine killings differently. US homicide rate is now about 4/100000 and peaked at[ EDIT 10/100000] in the early 90s
https://en.m.wikipedia.org/wiki/List_of_countries_by_intentional_homicide_rate
Assuming the question is not merely rhetorical, Obama assumed office as the anti-Bush. The War on Terror was downgraded to Overseas Contingency Operations; police were demonized; Americans, we’ve learned, have racism in our DNA.
For one specific, surely you’ve heard that neighbors of the San Bernardino jihadists noticed suspicious coming-and-goings of multiple military-aged Middle Eastern men, but declined to contact authorities for fears of being deemed racist.
For another, consider that Ahmed Mohamed brought a clock repackaged to look like an IED to school, disregarded his teacher’s admonition to keep it to himself due to its alarming appearance, and when subsequently school officials contacted the police as per Homeland Security protocol, became a poster child for so-called Islamophobia, winning an official White House visit and, potentially, a $15 million litigation windfall.
In summary, tone at the top matters, as the President well knows.
This is the week the Left completely jumped the shark and went truly insane.
That is helpful. Thank you.
The biggest problem with the statistics in this graph is that they are comparing two things that are not remotely similar, other than that they both resulted in dead bodies.
Certainly more people have been murdered than have been killed by terrorists. But murder is a known factor. There is no chance in the future that the murder rate will somehow become 100 times greater.
Terrorism, on the other hand, has the potential to scale in tragedy by a huge amount, as it did on 9/11. That puts the risk of terror in an entirely different category. Current terror attacks should be considered an early warning of much bigger things to come, not a steady state condition or a manageable, acceptable risk.
In addition, the directed nature of terror has much greater implications for breaking down civil society and human rights in a way that murder does not. Ask yourself this: Even though an order of magnitude more people have died from murders, why is it that terror attacks caused wholesale changes to how we fly and caused the creation of an entire new federal bureaucracy?
The greatest danger we face as a society and a civilization comes from this simple fact: The march of technology is rapidly increasing the destructive capability available to non-state actors, faster than our ability to defend against them. At the same time, advances in communications are giving such groups the ability to coordinate on a massive scale, even from across the planet.
The terror attacks to date are better looked at as warnings or precursors of something much deadlier in the future, not as a statistic to be compared against personal conflicts or individual acts that result in murder.
Extremely well put, Dan.
I’d be curious to know what these folks define as “right wing,” as obviously the intent is to equivocate.
Ask yourself this: Even though an order of magnitude more people have died from murders, why is it that terror attacks caused wholesale changes to how we fly and caused the creation of an entire new federal bureaucracy?
*****
Because it’s a sexier problem. Spending money on anti-terror makes you look like a tough guy, and bexause terrorism is sporadic, politicians can get away with spending a lot of money on it without any of their allocation decisions being called into question. Terrorism comes with a Romance Cost.
Current terror attacks should be considered an early warning of much bigger things to come, not a steady state condition or a manageable, acceptable risk.
*****
Why? We have a fourteen year track record and counting since Sept 11th. If there are sleeper cells in this country capable of mounting bigger attacks, what are they waiting for? They should either be regularly succeeding in killing tens of thousands at a stroke, or our agencies should be capturing terrorists, on a regular basis, just before they kill tens of thousands of people. Instead, the pattern from 1993 to now looks very much like a manageable, acceptable risk with one big anomalous attack.
The greatest danger we face as a society and a civilization comes from this simple fact: The march of technology is rapidly increasing the destructive capability available to non-state actors, faster than our ability to defend against them. At the same time, advances in communications are giving such groups the ability to coordinate on a massive scale, even from across the planet.
*****
Again, if this is true, why are they taking it so easy on us?
I’m talking about general technological advancement and its tendency to put more and more power into the hands of smaller and smaller groups. This is just a simple fact. Not that long ago, the technology of the time was an effective protection from global terrorism. It was much harder to organize and to have the capability to launch attacks at capital targets. Today, terrorism has gone global and is fully embracing the capabilities of modern technology. Technology is enabling smarter IED’s, and making it easier to build them – by far. And I think that’s one reason why IED’s have become an increasingly potent weapon. A terrorist today can build a drone, load it with a large explosive, and fly it into a crowd or a building. That’s the kind of stuff only the U.S. government could do 30 years ago.
As another example, the internet group Anonymous has very sophisticated hacking capabilities. There are a number of private hacker groups that can wield an awful lot of digital power if they really want to.
The point is that if we don’t figure out a way to get a handle on this, the attacks will continue to grow increasingly destructive over time. How quickly that will happen is an open question.
I suspect there are other things going on as well, and perhaps even more importantly. The world has changed a lot since Obama took office. I just scrolled through the source to which the graph links, and the methodology they’re using is unclear. To say the least. So I’m not going to appeal to their data, because I can make no sense of how they’re collecting it.
But let’s go with the assumption that political violence of a non-Islamist stripe has risen since 2009 (not at all clear from their data as presented, but let’s assume it for the sake of argument). If political science were a science, I’d want to run these experiments: Let’s go back in time and keep every other variable constant, but let’s see what happens with a) an improved economy; and b) no Internet. Keep everything else the same, just change those variables (not both at once, of course). Do we see the same patterns in political violence?
This is actually a really good point, and I’ve been struggling to articulate the effect of the Internet on this kind of political violence.
I think maybe I’ll write a post…
It isn’t that simple is it? Islamic terrorists are willing to die in their attacks, which opens avenues of killing that wouldn’t normally merit prophylactic measures. Airplane hijacking and flying into buildings being the foremost example.
Instugator raised a good point at Kevin Creighton’s post:
https://en.wikipedia.org/wiki/John_Allen_Muhammad
The graph numbers here match up with the numbers cited in the list by Kevin C. The 16 killed by John Allen Mohammed in 2008-2009 are missing.
It may be that they are buying the dodgy case that he was motivated by mental illness and anger at the U.S.A., and not by Islamic zeal, but, if so, why were his murders not included in the “others” list? Did they leave him out because it was just a local crime story in five states plus D.C.? Or did they leave him out so we would not notice that they didn’t count him as Islamic?
No, this is the week that they went public as anti-American and anti-Christian ideologues. They feel free to go public because they think they have the votes to win without needing to hide their dark side any more.
I note three things:
1) The graph above and the counting box below are inconsistently labeled. The graph says “non-Islamic extremism,” while the tally box says “Right wing.” Yet the numbers are the same. Are they claiming there has not been a single death from left-wing violence since 2001?
2) Let’s say it only purports to show Islamic terrorism and “right-wing” violence. Since this comes through the New York Times, through the New America Foundation, I know WHY they excluded left-wing violence. But I wonder what this would have looked like if they’d included those numbers. Come to think of it, aside from the murderer in Charleston (I won’t say his name), I can’t think of any large number of deaths from sociopolitically-identifiable attackers. I never heard the Charleston attacker’s views on income tax levels, government regulation, and religious belief. But if we’re going to call hatred against blacks to be “right wing,” then I suppose hatred against white policemen must be left wing, and any deaths from anti-police riots should be chalked up to left-wing political violence. The attacks that came out of the Occupy Wall Street movement were certainly left-wing, although I don’t remember if any deaths resulted.
3) In debating the moral responsibility for these attacks, because surely the purpose behind the graph is to say that right-wingers are worse people and more dangerous than jihadists, the number of attacks or deaths by them should be normalized by the population from which they’re drawn. In particular, I suggest that we divide by the total number of American adherents of “right-wing” politics in one case and the total number of Moslems in the other. The extremists in each group may be a minority, but we’ve got to compare them on similar terms. “Right-wing” describes about 30-40% of the country. “Right-wing extremist” describes a very tiny number. “Moslem” describes about 5% of the population. (I’m having trouble loading the actual number, but this is what I remember.) “Jihadist” describes something smaller. So let’s say the jihadists are drawn from the parent population of Moslems, and right-wing extremists are drawn from the parent population of right-wingers. That means that when we want to ask how likely is a member of one of these groups to go on a violent rampage, we divide the number of right-wing attacks by about 100 million, and we divide the Moslem attacks by about 15 million. If we stipulate a rough parity between the absolute number of attacks, then we see that there is a far greater risk posed by an individual Moslem than by an individual right-winger.
I only mean this as a response to my left-wing friends who say from these numbers that “the real threat is from the right wing, not Islamic terrorists.” Each individual Moslem believer is not a jihadist, of course.
Tim, you remind me of something I neglected to note in my post. New America, the source for the Times’s graphic, categorizes killings as “Deadly Jihadist Attacks” versus “Deadly Right Wing Attacks.” However, the New York Times changes the labels to “Islamic extremists” versus “non-Islamic extremists.”
Apparently, there is no such thing as a left-wing extremist, and the only imaginable non-right-wing extremist is of the Islamic variety.
It wasn’t a rhetorical question—I just wanted to make sure I wasn’t missing something.
There were terrorist attacks (Islamic and otherwise, foiled and otherwise) before Obama, of course, 9/11 being only the most obvious one. I wonder if there are additional factors—as Claire says, there’s the internet and the economy to consider, plus the strange, inspiring effect that prior massacres seem to have on proto-mass-shooters of the non-Islamic variety; just as watching the videotapes of Dylan and Klebold provides a blueprint for Root to follow, might watching videotapes of Jihadis abroad have been enough to galvanize the Tsarnaevs, even without top-down confirmation of America’s flaws? If so, we would expect to see more copy-cats no matter who the president was.
Which is not to say that the president has handled the rhetorical problem well—only that if we’d had John McCain, and some unabashed patriotism instead, the result might easily have been more terrorism rather than less. Hard to say whether “America is a great country and you people suck” is more or less infuriating than “America, sadly, sucks and y’all are peaceful people, now please stop blowing us up…”
Why are all the non-Islamic attacks called “Right Wing” attacks? The facts don’t seem to support that contention.
Taking off from Claire’s and Kate’s comments, I wonder if the rise of attacks around 2009 has any connection not just to the internet in general, but to social media in particular. Facebook was started in 2004, but it took a few years to get big. Twitter began in 2006, and I think it took off a bit more quickly. Could it be the instant-sharing aspect of these kinds of media that contribute more than other parts of the internet?
Why are deaths the only thing counted? Is it because that allows the Times to ignore the 280 wounded in Boston?
Does anyone know what the picture would look like if all casualties were counted?
And isn’t the number of foiled plots relevant as well? I may be wrong but it seems to me that I hear about a lot more foiled Islamist plots than I do about plots by “other extremists.”
It seems that their definition of “right wing” is “violent, and non-Islamic.” Again, there is so much dishonest in those numbers, a person really doesn’t know where to start.