refugee : indicators, uncertainty and triggers

Somewhere, somehow, somebody must have kicked you around some/Who knows maybe you were kidnapped tied up/Taken away and held for ransom/Honey, it don’t make no difference to me baby/Baby, everybody has to fight to be free, you see

This post is more a series of questions, as opposed to answers. It is not exhaustive.The questions relate to events in Northern Mali and as the situation is on-going definitive answers are a long way from being established. The underlying question is could the events of the past months have been predicted. The literature on warning intelligence is abundant. A key theme of the literature is how to predict over the horizon events and put in place measures to mitigate against the likely risks these events could create. For a sample see here, here, here and here.

The basis for warning intelligence is the generation of indicators:

Early warning indicators — precursor events that provide  initial signals into the nature, severity, and direction of emerging industry conditions — are the basis of an intelligence early warning system.  

In the case of Northern Mali were there indicators of the events that subsequently occurred? A review of open source material suggest that there were, at least, four indicators that could have been pertinent for analysts. Others with more local knowledge (and time) may have others.

  1. The consequences of the end of the Gaddafi regime.
  2. The death of Ibrahim Ag Bahanga
  3. AQIM kidnappings
  4. The emergence of new armed groupings

There have been numerous reports making the case that changes in Mali are a result of the end of the Gaddafi-regime – weapons, fighters etc moving back into the Northern Mali. Samples are here, here and here. The overall impact of this indicator may have have been limited by itself but when combined with the other indicators is perhaps more significant, providing additional men and material that would otherwise have been unavailable.

The death of Ibrahim Ag Bahanga could have been a second indicator for analysts;  the death of a historic leader of armed revolts in the region would inevitably lead to competition for influence among other leaders. His death removed a known individual whose decision making processes could be tracked across time and were predictable.

An third indicator could have been the AQIM kidnappings. Franklin Charles Graham has a view of the kidnappings here (pay-wall). This would have required recasting the analytic paradigm through which the kidnappings were portrayed, at least in the media from a money making vehicle for AQIM; a ‘terrorism financing’ optic to an analytic cadre that saw them as a tool to project power. Certainly, at a tactical level the kidnaps are widely alleged to have allowed the group to obtain money, but also to negotiate prisoner exchanges, and limit strikes and military intervention to protect the lives of the hostages.  At the strategic level, AQIM was able to project power by forcing negotiations and making governments seem weak. An alternative analysis would be to examine the geographical scope of the kidnappings and to concentrate on their expanding area of operations. For example, tracking travel advisories from foreign governments about where it was safe for tourists to travel could show the expansion of their area of influence. Arguably, prior to the fall of Kidal, Timbuktu and Gao, this area was already to a certain extent under the ‘influence’ of AQIM and their associates. The kidnapping operations had a ‘corrosive effect’ on the local populations, undermining state authority and official regulation, gradually shifting perceived power and influence to AQIM and their kidnap associates.

The emergence of new armed groupings would have suggested that the established group dynamics – AQIM, secular Tuareg nationalists (MNLA) – in the region were changing. The key question was and remains, how to interpret the emergence of MUJWA and Ansar Edin. For coverage on MUJWA see The Moor Next Doors posts (here, here and here). For Ansar al-Din See Lebovich’s post on al-Wasat (here). For an overview by GRIP in French see here. An analyst could have interpreted the emergence of these groupings in a number of ways – as a sign of dissension or break-down in relations between the AQIM commanders, particularly in the case of MUJWA. An alternative explanation would have been to see the groups as appealing to different constituencies, cutting across various divides in the region – Ansar al-Din leveraging support among the Tuareg and MUJA in local Arab community as well as amongst the West Africans. (For an overview of the role of ethnicity and race in recent events in Mali see the analysis by Hannah Armstrong on Current Intelligence here.)

These indicators do not, of course, explain everything, and a more detailed evidence based research project could disprove the influence of any one or all of the indicators. Alternative and/or complementary explanations could include issues raised at the recent BISA conference, where the panel, “Stability and Change in North Africa: known knowns, known unknowns and unknown unknowns” saw a discussion on Tunisia, Egypt and Morocco. Two issues of relevance to Northern Mali are perhaps the role of uncertainty produced by revolutionary processes and trigger factors.

Frederic Volpi in a paper on Tunisia, “Framing Stability and Change in North Africa: The Tunisian Kaleidoscope” commented on the role of structural factors versus the ‘uncertainty’ created by the revolutionary process. Revolutionary change should not be seen as a logical consequence of structural changes undergone by regimes; reconstruction of regime discourses to include ‘democracy’, ‘liberty’,  as well as limited political opening that legitimized some political actors or economic opportunity which benefited a select few –  these were not necessarily the cause of the revolution. The initial stages of the unrest were typical or standard forms of behavior, what social movement theory tends to call modes or repetoires of contention – action by protestors and then reaction by the regime – money and repression. At the beginning of the unrest there was not a demand for regime change but for change. The uncertainly that developed was when the state became unclear about how to respond. The shift to revolution came when uncertainly led to an end to negotiation for change. The flight of Ben Ali was not predictable. Laurence Whitehead, one of the discussants, suggested that one way of understanding the relationship between structural factors and abrupt change was through the concept of ‘triggers’. He indicated that a potential trigger was a succession crisis (the crisis was different in each of the North African states; but present in some form or another).

Reflecting on Northern Mali, is it possible that the overthrow of ATT could have been ‘triggered’ by a succession crisis and that this gave an opening to the various groups in the North to take advantage of the uncertainty generated by the ‘coup d’etat’ to press their demands for an alternative to the status quo, among them AQIM who had already been working at underming the official state in the North?

For the analyst, the challenge is double, identifying and tracking relevant indicators but also revising established analytic paradigms and interpretations. In Northern Mali, a  paradigm that saw AQIM as being primarily interested in criminal financing, less interested in establishing territorial control in urban areas where they could expose themselves,  and their local leverage constrained by the Algerian origins of the group, would perhaps have read the indicators wrong or discarded them. Warning intelligence needs both the right information but also a certain humility to revise an analytic position as evidence accumulates against it. It would be interesting to see gaming of the Northern Mali situation as used by Kristan Wheaton and his Mercyhurst students.

For a paper focused on warning intelligence in relation to insurgencies see  Cormac, Rory. “From the Periphery to the Centre: British Intelligence and the Warning of Irregular Threats” presented at the BISA conference. The paper can be accessed here by searching in the author field.