The Dangeorus Companions project codes external state support for each armed group targeting a state in the post-1945 period. Besides the dataset, it also includes information on each group including the year founded, objectives, ideational characteristics, states providing support, how support terminates and sources used for coding the information for each group. The groups coded in this dataset are the ones from the UCDP/PRIO Armed Conflict Dataset. The NAG dataset includes both the original UCDP/PRIOgroup codes and a new set of codes. In addition, the start dates for groups are determined not according to violence onset year (the criterion of the UCDP/PRIO dataset) but when there appears a declared name or leadership for a group. The idea is to capture the non-violence years in an effort to find out whether states that are external to an internal conflict contribute to the onset of violence.

The main novelties of the NAG dataset in comparison to existing datasets on external or third party state support torebel groups, such as Third Party Intervention, Nonstate Actor (NSA) dataset and UCDP External Support Data are as follows:

(a) Distinguish between active vs. passive support.
(b) Expand the time span to include all the years between 1945 and 2015.
(c) Specify the ethno-national and religious identity of each group.
(d) Specify the objectives of each group going beyond a dichotomy of whether a group has political and/or territorial claims.

The NAG dataset’s emphasis is on covert support, thus distinguishing it from third party intervention, which is not necessarily covert. The distinction between active and passive support is borrowed from Daniel Byman (2005). Byman defines passive support with three criteria:

(1) Theregime in question itself does not provide assistance but knowingly allows other actors in the country to aid a terrorist group,
(2) Theregime has the capacity to stop this assistance or has chosen not to develop this capacity, and
(3) Often passive support is given by political parties, wealthy merchants, or other actors in society that have no formal affiliation with the government.

The second criterion is not used in the assessment of whether a form of support is active or passive. Rather, the emphasis is on whether a government directly provides assistance to a group or not and within a country. In other words, when making the decisions about support, we looked for some evidence that the government or a political actor or organization that are formally affiliated with the government provide support. We also determined a precision level for the evidence that we used to code support. Furthermore, it is very difficult todetermine whether a regime or leadership knowingly or intentionally allows support. The best bet is to look for hints showing whether government creates direct and indirect channels to aid a NAG. The codebook provides more details for coding criteria and sources used for coding.