Two Terror Graphs

AFAIK, the only potentially misleading element is the region and time period chosen. Clearly the dominant contribution there comes from Northern Ireland, which as you might recall was a pretty big deal at the time. Restricting it to western europe omits things like Chechnya or September 11 or any of the four recent attacks in Turkey. And of course if you expanded further to include Asian countries the numbers would be dramatic.

Ultimately, whether this is a misleading graph depends on what question is being asked. If the question is “Is Islamic terrorism currently a worse problem for western Europe than the IRA was?” then I would say it is not. If the question is “Is Islamic terrorism currently a significant problem for the world?”, I would suggest it is.

Yes, there is something materially wrong with the first graph. Firstly, it lists attacks committed by muslim terrorists, inspired by islam as being “other”, so long as they did not have some other motive beyond just islam (ie Palestinian terrorism). Secondly, it is misleading to show islamic terrorism vs all other forms of terrorism, rather than each specific subset of terrorism/terrorists.

It is important to note that while terrorism levels might be reduced, the problem of terrorism has actually increased. Terrorism isn’t about killing people, it’s about spreading fear. Information is shared to a much greater extent today. People’s knowledge of a terrorist attack is going to be much greater today than it was 40 years ago. This means that per terrorist attack, more fear is created.

1) Pan Am Flight 103 is not classified as Islam inspired. If it was not Islam inspired then the term is meaningless. It should be “carried out by Islamic people.”

2) The blue and yellow colors dominating the second graph tell the main story. They cover the Troubles and the Baskian movement attacks. Both were geographically contained in a single country. As can be seen in the second graph, both are over. I think economists call this “time preference” or something of the sort.

So mainly the question is, what are these graphs supposed to prove? Maybe a world wide view including Algeria (over 100,000 victims), Syria, Lebanon and Iraq, would give a better picture.

Also, the graph is misleading because it excludes the 2015 Paris attacks, which might make red 2000 look more like blue 1970

Also, as above, Munich should be red.

Probability to be killed or injured by terrorists is very small. It is very small today, it was very small in 1970s. In this sense these two graphs are meaningless.

But if the goal of these two graphs to say that islamic terrorism is no big deal, then I would agree what Brian said.
Basque country and Nortern-Ireland were isolated events, local, with specific goals without universal ambitions. They did not challenge western values, societal norms and rules. But islamic terrorism does. For instance, after Jyllands-Posten and Charlie Hebdo, freedom of expression in Europe has de facto diminished. Not by law, but because of fear. Also it means that one religion has been able to achieve special treatment compared to other religions.

So , chance to be killed by terrorist is small and was small. But islamic terrosim is more dangerous because it attacks (and at least partly succesfully) values and societal norms what Basque and NE did not.

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]

@Brian,
You said:

“Those two conflicts were resolved after centuries in the 1990s because Europe was finally rich enough and professionally governed enough to buy off the losers. That is a great achievement but it does not retroactively make the resistance freedom fighters into terrorists.

I’m not sure that “freedom fighter” and “terrorist” are mutually exclusive terms. Even if one were to grant the the IRA had legitimate reasons to wage war against a foreign colonial power, certain acts by the IRA can still be classified as terrorism.

1) Western Europe isn’t representative of the entire world

2) The past isn’t a prologue

3) Instances of terrorism are highly volatile and inherently unpredictable

4) Terrorism can spread rapidly from current global hot zones

5) The west has a stated enemy who has declared an intention to increase such attacks

6) Threats =/= risk events, and these graphs don’t illustrate what is behind the curtain, so to say (many threats are mitigated by law enforcement / national intelligence prior to a risk event)

It’s like showing a graph of the 3 decades prior to the financial crisis – it makes for a nice history lesson that has zero relevance for the future.

It’s all good, until it’s not.

Seems about accurate. The Troubles in the UK alone killed a few thousand people, which is a lot more than Muslims have killed subsequently, and the magnitude difference looks about right.

It may be misleading in that the Muslim population of Europe was much smaller in 1975 than it is today. If the implication is that one should average Muslim terrorism over the whole time period, it’s misleading. If the implication is that recent Muslim terrorism fueled by recent immigration has been much less severe than past nationalist terrorism, it is not.

But to my mind the most interesting feature of this graph is not to do with Muslims, but rather how the death rate due to nationalist terrorism collapses following the fall of the Soviet Union. Having lived through that time in the UK, my impression was that the IRA was pretty fearsome in the early 90s, maybe more so than it had been previously, and that the Good Friday Agreement was the real break. Looking at this graph, though, the IRA had already been operating at much reduced capacity for almost a decade at that point. That ETA did the same, at the same time, makes this a bit too much to be a coincidence.

Between the fall of the USSR and the introduction of a lot of Muslim immigrants, terrorism in Europe is clearly running on fumes.

Exhibit A would be Vladimir Ilyich Sanchez Ramirez (AKA, Carlos the Jackal);

Ramirez allegedly took hijackings, bombings and killings in mercenary style, with links for years to causes like the Popular Front for the Liberation of Palestine and in far-left European terror groups.

Shadowy alliances that thrived during the Cold War kept him beyond the reach of Western secret services. But after the fall of Communism, French operatives nabbed him in 1994 in Khartoum, Sudan, and flew him to Paris.

The truly misleading part of the graphs is that they’re basically the “seatbelts kill” argument applied to terrorism.

The data doesn’t tell us much of anything about the underlying response function a society might employ against terrorism. At the “don’t bother taking any precautions at all” extreme, we can expect lots of terrorism. At the “employ a complete totalitarian lockdown of society” we can expect no terrorism, but at an incredibly high cost.

In this sort of analysis there’s another big problem with the graphs, in that many of the “victims” of terrorism noted above (i.e. soldiers and police) were actually “resources” being employed to combat the terrorists political goals.

That is, they’re more properly considered a cost of preventing terrorist attacks than a cost of terrorist attacks. Which is a funny sounding distinction, but an important one.

I wonder what would be a good value of X here:

BC will bet $1000 even money that there will not be an Islam-inspired terrorist attack in Europe that kills X people before 2026.

The truly misleading part of the graphs is that they’re basically the “seatbelts kill” argument applied to terrorism.

Interesting. To me the graphs are the “car accidents kill” argument applied to terrorism.

Overlay the death rate from automobiles on this graph and you’ll see how minor the problem of terrorism is — the graphs showing that it’s even more minor today than yesterday. Yet we cope with the risk of car accidents all the time, suffering only minor inconveniences such as seatbelts, a couple thousand dollars per car in mandated safety, and harsher traffic laws that decrease that risk by perhaps half. The reaction to terrorism is not only an order of magnitude more disruptive: disruption is exactly the reaction the terrorists are looking for!

Look at 9/11. $25 billion of physical capital lost. About the same in human capital. Trillions of dollars lost in the reaction, and an outcome in the Middle East that is more conducive to exported terrorism, not less.

It was indeed sad that news media thought they had to move their anchors to Paris after the attacks there. We should all treat terrorism on the ground as a crime perpetrated by asses, not as a change in the way we have to live.

1. From 2001 onwards, preventative action against terrorism massively increased in Western Europe. So at best the graphs post 2001 tell you the probability of terrorism given a certain level of enhanced preventative action and vigilance. They don’t tell you the probability of terrorism if current precautions were to be abandoned–just as the statistical infrequency of deaths by lightning strikes doesn’t mean it’s safe to wander around in open fields during electric storms.

2. Prior to 9/11, even the nastiest terror groups in Western Europe sought the deaths of at most dozens in a single attack. 9/11 showed that there are terrorists who seek the deaths of thousands. And terrorists who seek the deaths of thousands would probably feel no compunction about going much further: they just don’t care about the lives of people unlike themselves. So the worst case scenario for terrorism is a great deal worse now than it was before 9/11, and that should enter into a rational calculation of our response to terrorism. The graphs fail to show this.

Bonus: Though the absolute chance of dying in a terrorist attack in Northern Ireland during the so called “Troubles” was small, it was high enough to make living there at that time pretty miserable. Contrary to what is sometimes assumed, terrorism doesn’t have to achieve anything close to the death rate of auto accidents to fundamentally transform our lives. That this is so may well reflect the irrationality of the masses, but that irrationality (if that’s what it is) isn’t going away any time soon. We have to reckon with it in determining anti-terrorism policy.

I would think I was conversing with a certified idiot. Martin Luther King was not exaggerating when he called the bombing “one of the most vicious and tragic crimes ever perpetrated against humanity.”

Nonetheless, however barbaric the killings during the Civil Rights movement were, I would still hope that society did not overreact and dramatically remake itself in response to a bunch of asses who killed a few people a year for a couple decades — and not particularly conspiratorially at that.

What’s the alternative? Metal detectors and bomb-sniffing dogs at every church? Prohibiting the free movement of white people?

Back in the 1970’s and 1980’s there was a wave of new religious movements (also known as cults). The Moonies, Scientology, Hare Krishna, Bhagwan Shree Rajneesh, to name four.

I went off to University in 1978. Parents then didn’t just worry about drugs and sex, they worried that their 18 year old “children” would drop-out and join a cult. Now-a-days they might worry that their “child” is going to get mixed up in Islam and become radicalized and move on to fire-and-sword Islam.

You could make a case that fire-and-sword Islam is super-retro revivalism, and not new. I don’t accept that. Picking the date of the Golden Age that you pine for is something that each generation does for itself, in reaction to their own time. They chose anew.

I see fire-and-sword Islam as a new religious movement or cult. The graph should be comparing the terrorism it inspires with that perpetrated by other cults, Moonies, Mormons, etc. You can make quite a good case that Marxism is a cult and belongs on the same graph.

The IRA and Basque separatists are at the nasty end of guerrilla war. They belong on a different (and much bloodier graph) along with civil wars, wars between states, and revolutions.

Brian: “2. The first graph classifies Lockerbie and Munich as non-Moslem inspired”

Matthew Moore: “Also, as above, Munich should be red.”

The leader of the Palestinian terrorists of Munich was a Christian, using “Issa” (“Jesus”) as war name; the code name of the attack was “Operation Iqrit and Biram” (two Arab Christian villages destroyed in 1948). Some people seems to forget that the Palestian terrorism in the 1970s was disproportionally Christian (or at least from Communists born in Christian families).