Data

Case Selection

Two primary criteria were used to select only those cases that fit the definitional requirements of a terrorist organization:

  1. Eliminate any group indicated to have targeted only property or military targets, with no indicated associated civilian injuries or fatalities.
  2. Eliminate any group that did not display sustained organizational capabilities, i.e. those groups with only one attacks or with only a single series of coordinated attacks within several days of one another and with no subsequent evidence of activity or communication.

This rubric worked effectively in most cases but suffered from certain complications inherent in the MIPT data.  One of the most common complications in selection was inconsistencies within the MIPT reporting for a single group.  In many cases, the MIPT incident statistics would indicate no civilian casualties as a result of the group’s activities.  In each of these cases the group’s descriptive profile was carefully considered, and if it suggested attacks in addition to those listed in the incident statistics the group would be included as a terrorist organization, unless the description explicitly indicated that there were no civilian casualties.

Another common complication was provided by the proliferation of name changes and breakaway groups listed in the database.  Again, for reasons of feasibility and consistency of interpretation, each listed group that met the requirements for terrorism was included as an individual group, regardless of any links to another group.  The exception was the very few cases where a clearly and consistently defined ‘armed wing’ and the general organization were both listed and made mutual reference to one another; in these cases only the ‘armed wing’ was included to avoided double-counting the group (see the case of Resistenza Corsa and Accolta Nazinuale Corsa for example).  In general, splinter groups, aliases, and name changes, so long as terrorist attacks could be ascribed to the name, were all included. For consistency, hijackings and kidnappings were held to the same standards of civilian injury as other incidents.  In the context of the construction of this database, both types of incident are only considered terrorist actions if civilians are injured or killed at some point in the incident.

Groups that fit the selection criteria for inclusion as terrorist organizations were then coded for lifespan, level of engagement in negotiations over the group’s fundamental aims or strategies, and the extent, if at all, to which the group achieved its strategic aims.  Other variables, for example those that could directly examine the presence trajectories of decline discussed in this book, would be very interesting to consider, but unfortunately are not consistently accessible in the information provided by the MIPT. In the very occasional cases where meaningful values of the variables in question could not be obtained from the MIPT data the variables were coded as missing.

This presentation of raw data can only be well understood in conjunction with the explanations and elaborations given in the Appendix (“Statistical Analysis of Terrorist Campaigns”) of How Terrorism Ends, pp. 207-222.

Please note: Unfortunately, the MIPT Terrorism Knowledge Database is no longer accessible to the public.  Following the completion of this study (which accessed the database directly online over a period of several years), the MIPT Terrorism Knowledge Database was transferred to the National Consortium for the Study of Terrorism and Responses to Terrorism (START) at the University of Maryland, athttp://www.start.umd.edu/.  It is no longer directly accessible on-line, and this is something over which we have no control. (The previous site was www.tkb.org.)  Lack of direct access to the MIPT database will prevent those who are using our data from studying the MIPT organization profiles and incident descriptions that we mainly relied upon (buttressed and double-checked with extensive searches of media accounts, interviews, secondary sources and other resources).  We regret the inconvenience.