Skip to main content
ABC News
Tracking 125,000 Incidents Of Global Terrorism

Soon after the terrorist attacks in Paris earlier this month, when three assailants massacred people in the office of the French satirical newspaper Charlie Hebdo and a kosher supermarket, a researcher at the University of Maryland at College Park carefully logged the details of the violence (location, names of the perpetrators, weapons used, number of fatalities). The attacks became new entries in the Global Terrorism Database, a comprehensive clearinghouse of terrorist acts around the world.

The database, housed at the university’s National Consortium for the Study of Terrorism and Responses to Terrorism, includes information on 58,000 bombings, 15,000 assassinations and 6,000 kidnappings going back to 1970 — anything that falls under the database’s definition of terrorism: “The threatened or actual use of illegal force and violence by a non-state actor to attain a political, economic, religious, or social goal through fear, coercion, or intimidation.”1 Each case’s entry in the database includes somewhere between 40 and 120 variables.

It’s an ever-growing chronicle of terrorist activity that offers a high-level look at the distribution, frequency and scope of terrorist attacks across the globe.

As an authoritative source on both the history and dispersion of terrorist attacks, the GTD can confirm or challenge the veracity of claims about just how serious — and widespread — terrorism activity is today. The GTD allows researchers to compare domestic terrorism to transnational attacks and to follow terrorist organizations over time. It can map which group has been active the longest (Basque Fatherland and Freedom), which has the greatest geographic reach (Hezbollah), or which has caused the most casualties (the Islamic State in Iraq and Syria).

Its data shows that the vast majority of terrorist attacks don’t make international headlines. Instead, they’re concentrated in a small number of countries where thousands of people die every year in these incidents. Half of terrorist attacks claim no lives, but terrorism fatalities nevertheless number in the tens of thousands globally.

The GTD is now a computerized database that has more than 125,000 separate incidents going back to 1970, but it started with handwritten index cards. Between 1970 and 1997, a risk-advisory firm called Pinkerton Global Intelligence Services tracked incidents of terrorism across the globe so it could advise its clients about the safety of international travel. Every day, Pinkerton researchers, mostly ex-Air Force personnel, sat in front of the radio and listened for newswire reports of acts of terrorism. They recorded basic details about an attack, such as place, target, perpetrator group and number of people hurt or killed.

By the time Pinkerton stopped data collection, the company had amassed a record of 61,165 terrorist attacks over a 28-year period, all logged on those index cards. In 2001, Gary LaFree, a professor at the University of Maryland and now the terrorism center’s director, stumbled upon the trove. After the school obtained the index cards in 2001, researchers spent years digitizing 58 shoeboxes full of data.

The database still is primarily updated through media reporting on terrorist incidents, although the technology is different — the loggers aren’t listening to newswire reports on the radio anymore. “The huge jump in technology since this data collection started makes our jobs easier and harder at the same time,” said Michael Jensen, the data collection manager at the center. “On the one hand, nobody in the 1970s could jump onto Twitter to follow up an ISIS claim. But there’s a tremendous volume of material to deal with.”

Every day, Jensen and his team download approximately 1.2 million news stories from media outlets around the world — everything from obituaries to sports news to politics. Whittling down this avalanche of information is no easy task. At first, researchers rely on a sophisticated Boolean search to narrow the pool of articles to everything involving political violence. Then they run an algorithm to delete thousands of duplicate articles. Finally, they unleash a machine-learning model — like the algorithm that online stores like Amazon use to track purchases and suggest similar products — which determines how likely each remaining story is to include content about terrorism.

That leaves about 550 articles per day for Jensen and his 10-person team to sort through manually and code. They work with a month’s worth of material at a time, so that they can watch a news story evolve over days and weeks — like the unfolding coverage of the Paris attacks — and ensure that the details are accurate.2 “It’s a little like Groundhog Day, the same thing happening over and over and over again,” Jensen explained. “The Paris-style attacks are rare. What’s far more common is a roadside bomb somewhere in Afghanistan or Iraq. So it’s sobering reading, but it’s also extremely repetitive.”

Now, though, the University of Maryland researchers aren’t content with the what of terrorism — they want to tackle the why. It’s the same question that has prompted journalists’ detailed explorations of the Paris attackers’ paths to radicalization. How can a young man like Paris assailant Cherif Kouachi go (in the words of The New York Times) from “easily spooked amateur to hardened killer”? Kouachi’s story includes some possible explanations — affiliation with a radical Islamist preacher, a stint in prison with other jihadists, a trip to Yemen to train with al-Qaida — but offers few systematic lessons in how radicalization occurs.

A team of three full-time staff members at the terrorism center are trying to move beyond anecdote. They’ve amassed a data set of more than 1,500 people radicalized to violent and non-violent extremism in the United States since World War II and put them into three categories: Islamist, Far Right, and Far Left. The database — which hasn’t been released publicly — has detailed information about the terrorists’ lives and backgrounds, including criminal records, social networks and histories of abuse. The researchers believe it’s among the first of its kind.

“Radicalization suffers from a lack of data,” said Patrick James, a researcher at the center. “And that’s a problem — it’s hard to form an effective response to something if you don’t have an empirical basis for studying it. You can’t distinguish between people who are likely to commit violence and people who aren’t. That’s the real golden goose we’re trying to get.”

The preliminary findings of the study already have yielded some basic demographic patterns about which extremists in the U.S. are most likely to resort to violence. Compared to violent domestic terrorists on the Far Left and Far Right, Islamists stand out. They’re more likely to be young (between 18 and 28 years old), unmarried and unassimilated into American society. They are also more likely to be actively recruited to an extremist group.

But in other important ways, Islamist extremists in the U.S. as a whole — violent and nonviolent — are not so different from other extremists. People in the three groups were equally likely to have become radicalized while serving time in prison — complicating the narrative that Muslim prisoners are unusually likely to commit to extremism from behind bars — and to be composed of individuals who have psychological issues, are loners, or have recently experienced “a loss of social standing.”

“Social networks are incredibly important to radicalization, but that’s not unique to Islamists at all,” James said. “There’s almost always a facilitator — a personal relationship with a friend or family member who’s already made that leap.”

When the project on domestic radicalization launched, James and his fellow researchers decided to borrow their strategy from the Global Terrorism Database, combing through news articles for traces of information about terrorists’ lives. Their task is more difficult, though, than simply tracking terrorist attacks. They need to get inside terrorists’ heads — to understand their motivations, influences and goals. News articles often provide cursory information about traits thought to be linked to terrorism, like having a history of childhood abuse. But a first-person blog post or an article that links to a homemade video or mentions a particular jihadi forum can also be a goldmine.

“The good thing about studying terrorism is that terrorists tend to want to be known, to spread their ideology, so what they believe isn’t secret,” James said. “Oftentimes that makes it easier for us, because they’re more than willing to put out propaganda that gives us a kind of line to their thoughts.”

But there are still many gaps in the data. One potential workaround is to study the deradicalization of former extremists. That way, policymakers can understand not just what incites terrorists to violence, but how to intervene. The University of Maryland’s terrorism center has just begun a project that will combine in-depth qualitative interviews with former terrorists with a larger quantitative data set.

“It’s a necessary complement, when you think about it,” James said. “We want to understand how people develop radical beliefs and when they’re likely to commit radical acts of violence, but you also need to know — well, how do you stop it?”


  1. This differs somewhat from other definitions that do include state actors (for example, if the Assad regime launches a chemical attack against its own people in Syria, that won’t go in the database) and definitions that don’t include attacks against military personnel. The center has stuck with this definition to achieve consistency, since this is how the database’s originators defined terrorism starting in 1970.

  2. A number of the articles the researchers read each month are updates to cases from previous months (kidnapping outcomes, perpetrator claims, revised casualty figures). When information changes, they update the entry accordingly.

Amelia Thomson-DeVeaux is a senior editor and senior reporter for FiveThirtyEight.