In this article and the next, we shall evaluate the likelihood of the primary scenarios for foreign military intervention, which we started to detail in “Scenarios for the Future of Libya – Scenario 2: The Joint Arab Force Takes a Side (1).” We shall focus on preliminary methodological work allowing for better describing the intervention cases for likelihood estimates. In the last article we discussed the likelihood of Scenario 1, where the Libyan actors negotiate a peace settlement—a scenario for which the probability we assessed was less than 20%, or highly unlikely.

As detailed previously, we shall use the methodology developed by The Red (Team) Analysis Society, building upon Heuer (“Assessing Probability of a Scenario”, in Psychology of Intelligence Analysis, pp.156-157) and the capability given by indicators. This methodology allows us to obtain an estimated likelihood, which is considered not only as good enough for the purpose of anticipation through scenarios but also as remaining usable by analysts.

Note: In the following article, we shall use the acronym COR for the Council of Representatives (nationalists), GNC for the General National Congress (Islamists), and GNA for the UN-backed Government of National Accord (unity government).

Organizing the Scenarios & Indicators

In order to mathematically deduce the likelihood of this scenario and its sub-scenarios, we organized the sub-scenarios in such a way as to correctly account for scenarios not detailed in our articles previously because they were not necessary in terms of narrative and understanding of the future of Libya – they were implicit.

For this scenario, we also had to add a supplementary step to account for intervention in support of the three separate governments, as well as the order in which intervention could occur­ as an intervention taking place for one of the governments could affect the likelihood of subsequent interventions occurring (see graphs below). With that in mind, we developed the graphs in such a way as to easily estimate the scenario likelihoods on different tiers and determine their overall likelihoods in various order of interventions.

In the first graph, external actors intervene (or not) first for the General National Congress, then on behalf of the COR depending on whether intervention has occurred (or not) in support of its rival, the General National Congress.

Click to access larger image

In the second graph, external actors intervene (or not) first for the GNA, then in support of the COR depending on the level of intervention for the GNA.

Click to access larger image

In the third graph, external actors intervene (or not) first on behalf of the Council of Representatives (COR), then for the GNA depending on whether intervention occurs in support of the Council of Representatives.

Click to access larger image

For tiers 2 and 3, we also had to add additional indicators that considered the potential intervention occurring in favor of the actors on tiers 1 and 2, because some indications may become more likely in the case of rivalry between competing Libyan authorities (GNC vs. COR). For example, the United Arab Emirates may be more willing to militarily intervene for the COR if Qatar has begun to intervene on the side of the GNC, possibly more so than a case where no external actors intervened in support of the GNC. Thus, we added this additional indicator to the pair of scenarios that follow an intervention in support of a particular actor on tier 1 or tier 2. In the tier 2 scenarios following the branch starting with “No Intervention in Support of [Actor]” on tier 1, we used the regular set of indicators for intervention, since no intervention occurred for the tier 1 actor.

Tier 3 accounts for the third actor receiving external intervention on its behalf according to the various branches of the “tree of scenarios for likelihoods”. As for Tier 3, we added supplementary indicators for scenarios that followed intervention of the tier 2 actor.

With the ability to estimate likelihoods (depending on the tier 1 actor) and thus calculate probabilities for three different orders of interventions, we are able to cover a broad range of scenarios. Having discussed the methodology of how we organized the various trees of intervention, we shall discuss the sets of indicators according to tiers and if necessary revise them, detail their evaluation and proceed with a first likelihood estimate in the next article.

Featured Photo: Norwegian F-16 Libya 2011 by Metziker, [CC BY-NC 2.0], via Flickr