A Situational Model for Distinguishing Terrorist and Non-Terrorist Aerial Hijackings, 1948–2007

Abstract

Despite the centrality of situational variables to crime theories, they remain uncommon in criminology. Based on the hypotheses drawn from the literature on situational determinants of crime, we examine whether aerial hijackings perpetrated by terrorists are situationally distinct from other aerial hijackings. We define terrorist hijackings as those that include threatened or actual use of illegal force or violence to attain a political, economic, religious or social goal through fear, coercion, or intimidation. Other aerial hijackings include those perpetrated for transportation or extortion purposes. Using a newly updated dataset, we examined 1,019 aerial hijackings that occurred around the world from 1948 to 2007, out of which we classified 122 as terrorism. Results provide strong support for the argument that situational factors measuring organizational resources distinguish terrorist from non-terrorist aerial hijackings, and partial support for the argument that situational factors measuring publicity distinguish these events.

Keywords:

• aerial hijacking
• situational factors
• terrorism

Acknowledgments

Support for this research was provided by the Department of Homeland Security through the National Consortium for the Study of Terrorism and Responses to Terrorism (START), Grant No. N00140510629. Any opinions, findings, and conclusions or recommendations in this document are those of the authors and do not necessarily reflect the views of the Department of Homeland Security.

Notes

1. Note that in this definition of terrorism, the “economic goal” refers to a large-scale societal economic change, such as Marxism, and excludes incidents that are undertaken to personally enrich the perpetrator, and thus, it explicitly excludes extortion hijackings. Extortionists hijack in order to take money or valuables from national governments, airlines or passengers for personal gain, not for terrorist purposes.

2. Research assistants independently identified terrorist hijackings and obtained 0.91 inter-rater reliability. All disagreements were discussed and reconciled.

3. Coding the summer months for both the northern and southern hemispheres as “1” and the fall, winter, and spring as “0” produced no substantive changes in the results. Further, changing the variable to include the popular travel month of December produced no substantive change in the results. An additional model which included the summer variable and a dichotomous variable capturing whether the hijacking occurred in the month of December showed that all variables, including the original summer variable, had the same effects as reported here.

4. We also experimented with adding specific weapon types to the models (e.g. firearms, explosives, knives), but none of these variations produced significantly different results. In fact, the magnitude of the effects of each individual type of weapon, such as explosives, on the likelihood of the hijacking being conducted for terrorism purposes remained similar to the magnitude of the operationalization of the single weapon type and the combination weapon type in our original model. All of the odds ratios ranged between 2.8 and 3.7. Further, the differences between the coefficients (e.g. firearms versus knives; knives versus explosives; explosives versus firearms) were not significantly different from one another using the test described by Paternoster, Brame, Mazerolle, and Piquero (1998).

5. This decision may introduce error into our model, however no other equally comprehensive, open-source data exists for the early years. Fortunately, the composition of the top quartile countries changes infrequently—about once per decade (LaFree, Dugan, & Fahey, 2007). Thus, we expect the resulting error to be relatively small.

6. These measures also included stationing law enforcement personnel at all passenger checkpoints and an agreement between the US and Cuba to prosecute individuals who hijacked planes to those countries (Dugan et al., 2005).

1The predicted probabilities were calculated by holding all dummy variables at their mode, and continuous variables at their mean.

*The pseudo-r ² in this analysis was 0.2767.

7. Lebanon experienced 21 hijackings, 12 of which we classified as terrorist hijackings; all hijackings in Lebanon, regardless of type, departed from Beirut International Airport. We reran the analyses excluding the Lebanese hijackings and no substantive differences were observed in the capital city variable or the model in general. This supports our contention that although larger airports are often located in capital cities, particularly in smaller nations, this fact does not explain the observed connection between the capital city origination of the flight and whether the hijacking was perpetrated for terrorist purposes.

8. Although, we also expected that the number of casualties would be strongly correlated with whether the hijackers were armed (and whether they had multiple types of arms), in fact the connection between being armed and casualties was insignificant (r = −0.04) and the connection between casualties and multiple weapon types was significant, but relatively small (r = 0.17). We also found no evidence for a significant statistical interaction between multiple weapon types and casualties.