Editor’s note: We’ve partnered with the National Consortium for the Study of Terrorism and Responses to Terrorism (START) to publish a series of infographics based on data from their Global Terrorism Database and related START projects. Each week we’ll release a new set of graphics that depict trends in global terrorism activity. Sign up for the War on the Rocks newsletter to make sure you don’t miss any of them!
These graphics were designed by Michael Jensen.
Experts have suggested that there are marked differences in the behaviors of leftist and religious groups, namely that religiously motivated groups tend to target civilians at a greater rate and kill more people per attack than do leftist organizations. This week’s graphics examine the three most active leftist and three most active religious groups in the last 10 years to see if these assessments are supported by the data. The first graphic looks at a breakdown of target types by leftist and religious groups using these six groups as representative samples. Based on target selection, there isn’t much difference between leftist and religious groups. They target civilians roughly the same amount. Leftist groups do tend to target businesses more than religious groups, who, in turn, tend to focus more on security targets.
However, what is supported by the data is the idea that leftist groups tend to be less deadly than religious groups. The next graphic shows the number of attacks and the average lethality of attacks for each of the six groups. As the chart shows, each of the religiously-motivated organizations killed more people on average than did the leftist groups. In the case of AQI/ISIL, the group averaged nearly 7 deaths per attack, while the most lethal leftist group—CPI Maoist—killed on average 1.24 per attack.
The final graphic shows the individual breakdown of target types for each of the six groups. When we look at each group individually, we see that AQI/ISIL and NPA seem to be the archetypes for the hypothesized behaviors of religious and leftist groups. However, the story is a bit more complicated for the rest of the groups. All of them include a substantial number of civilians among their targets, with CPI-Maoist targeting civilians more than any other target type. Each group also appears to prioritize security forces as targets, which is not surprising.
Ultimately, what these graphics show is that neat conceptualizations of the targeting practices of terrorist groups are difficult to develop based purely on ideological motivation. Efforts to create them will very often be frustrated by groups that stubbornly refuse to fit the explanatory mold.
**Note: Beginning with 2012 data collection, START made several important changes to the GTD collection methodology, improving the efficiency and comprehensiveness of the process. In general, comparisons of aggregate statistics over time and between locations should be interpreted with caution due to these methodological improvements, as well as the considerable variation in the availability of source materials.
Michael Jensen is the data collection manager for the Global Terrorism Database at START.
Photo credit: Johan Althén
Michael Jensen 
Ryan Evans 
This is a great question, but the answer given here suffers from serious selection bias because of the short time period covered by the source data. Leftist groups were much more widespread, more active, and more deadly in the 20th century, while ostensibly religious groups have figured more prominently so far in the conflict zones of the 21st century. Why that’s the case merits bookshelves of its own, but my point here is that data from just the past 10 years compare one set of groups on the wane with another set arguably in their prime. The result you get can be explained by those historical phases without any reference to the content of the leftist or religious ideology.
Excellent thought and worth a look at the data. You are certainly correct that leftist groups were more numerous and active in the 20th century. However, if we extend the analysis back to 1970 and throw in a few leftist groups popular in the 20th century, like the Red Army Faction or Red Brigades, we don’t see the numbers change much. Civilians were targeted in 22% of the attacks by FARC, RAF, Red Brigades, CPI-Maoist, NPA, from 1970-2013. The average lethality doesn’t change much either, with 1.95 killed per attack.
Thank you for checking that out. I remain unconvinced, however, and it’s the sharp distinction between terrorism and other forms of political violence that I suspect as the culprit. To take one extreme case, the Khmer Rouge were arguably leftist and killed millions in the 1970s, but they don’t show up in data on terrorism because they were so successful that they didn’t need to use that tactic. In other words, the choice of tactics is endogenous to success, which, as Kalyvas shows, is related to lethality in complicated ways.
Interesting point. Great exchange
This is an important point, but one that changes the scope of the analysis in fundamental ways. You (and Kalyvas) are right. If we extend this analysis to include state actors (leftist, religious, fascist, etc.), with their access to the tools of state and a greater ability to kill than non-state terrorist groups, we could capture the deaths of millions. The point here, however, is to look at a common argument about the behaviors of non-state actors. In particular, the belief that leftist groups, who seek to portray themselves as vanguards of the people and who want to win over the population, will go out of their way to avoid targeting and killing civilians. This argument says nothing about how those groups will behave once they’ve taken control of the state apparatus. The results of this quick look at the data are mixed and certainly don’t support the argument that leftist groups (at least the ones looked at here) are kind to civilians.