Category Archives: Game Theory

Focal Point in the Libyan Civil War?

The current developments between the various parties in Libya, especially following Hafter’s recent interview and the Vienna conference on the 16th of May, have been interesting to say the least. In the variety of analysis on the solutions in Libya, including those editorials endorsing certain figures, there is a lack of understanding where each party stands in bargaining for a position of power.

To simplify the conflict I will do away with the smaller tribal conflicts in the south and the pervasive new crime wave across Libya. The reason is that compared to the other features I will mention they won’t determine the future of Libya, in other words their weight on any future state is negligible. I also don’t think it is important to analyze arms control when one wants to understand why some parties logically don’t give up. Yes, some groups receive weapons through the porous borders and banana republicans in the middle east, the weapons aren’t more useful or effective than brigade size. In the end it’s just bullets and Ak-47s. The technologically threshold has been maximized for almost all belligerents.

There are two sets of conflicts. This is reached by crude simplification of individual conflicts, and the web of allegiances can be bundled together to determine the aggregate conflict (Fajr vs. GNA vs. Dignity).

On the lowest level there are militias that protect regional interests. These interests are the cities they are from, usually they were the militia that removed Gadaffi loyalists in 2011 from said city. The have local support and political support from similar groups that view them as fellows in the fight against Gadaffi. Some groups are multi-city troops. For instance Hafter’s 2014 insurgency against Islamist militias was formed from anti-islamists militias and military units reformed by the General National Congress (GNC) and National Transitional Council (NTC). In the need to keep this analysis broadly about the general conflict, suffusive to say most of the Operation Dignity’s political support comes from Eastern Libya, due to figures from there supporting Hafter and Hafter’s reciprocal support for some of their military figures. In the west, Misratan militias have almost full support in Misrata and contain unique centralized leadership tied to the municipal government of Misrata. Again, due to the war in 2011 the Misrata faction receives political and regional support. This for Misrata extends weakly from Misrata all the way to Zawiyah and the Nafusa mountains. There is a inverse relationship between the distance from Misrata and the acceptance of Misrata based security forces. With the GNA, it seems Misrata may enjoy extended support if it becomes the “LNA” for the Government of National Accord (GNA). There are similar groups in Tripoli, Nafusa Mountains, Zintan, Zawiyah and Sabratha.

The groups mentioned above have alliances amongst each other. In the case of the non-LNA party they have allegiances formed due to cooperation in the revolution and subsequent cooperation to lobby (more accurately threaten) public officials of the first GNC to give out contracts to militias and ban Gadaffi-era officials from military posts amongst other positions. The dignity operation is the formed from former Gadaffi brigades who revolted at sometime in the war, a minority consisting of “new” army trained by the NTC, Zintani city militia and tribal militias From the Northeast.

In the Zintani and GNC government militias conflict, it is a proxy for the general Fajr (Misrata, Tripoli militia alliance)-Dignity conflict. For the last few months there has been a cease fire with territorial control centred around the Gheryan. The point of consideration here is that Gheryan is geographically a focal point. The road to Gheryah from Azziyah leads to the mountain pass to the city and then to much of the Nafusa mountains. Thus behind Geryhan is an expectation that one’s enemy will plan primary checkpoints around the city depending on their geographical starting point.  Much like a river is a common retreat point for armies. It is an equilibrium solution if both parties decide that eventually they prefer to concede (for a limited time), assuming they know each other’s expectation. This same focal point ceasefire is seen going west from Gheryan to the Tunisian border and down to Sebha. Both parties here have not confused their intentions. All parties have it in their best interest to avoid incidents like the Tripoli Airport conflict.

This is not observed in the more political and aggregate conflict. There is no focal point between Hafter and GNC (now GNA). The main reason is information. There have been no clichés, incentives or any informational clues that can solve the tacit bargain for seat of power in Libya. One important consideration is that in any war in which there is tacit limited conflict the consideration of focal points can help parties find some resolution since the war is not zero-sum. Misrata and LNA don’t face each other in conflict at all. There is no way to understand what plan they have as a contingency when they fail. Or what is is the “second” best result. There is clearly here confusion in the conflict. There is no river, no mountain pass, no Nash equilibrium in this limited war.

Moreover, my main argument is that in fact the confusion that leads to the institutional lack of focal point is that inside theses parties, in any of their parts, there is a lack of information on leadership, structure, and goals. Consider Hafter, he was decried by the newly elected parliament before its official formation in Tobruk. Then he was accepted, since then he has yet to appear in front of parliament to report on advancements in their fight against the Shura Council in Benghazi, a militant jihadist group. This lack of communication between a military general and the government makes the door to political reconciliation difficult to open. There is no focal point politically if the goals of Hafter are not known. He may be a rogue general, it is nearly impossible to make an educated guess.

The internal communication issues exist in the alliance between the Libyan Revolutionaries Operations Room (LROR) and Misrata. Besides a few skirmishes in eastern Tripoli that I remember being reported only faintly (only because it was common in Tripoli to attack groups who arrest your militia members for crimes), the alliance is weak. They don’t form a proper operations room or military. They don’t share a point of command in any government. Thus, they don’t share a plan or strategy. The formation of an operations room by the new UN-supported GNA from Misrata brigades and their closest allies is trying to fill this gap by having the Presidential Council become the point of decision making for all militias so that they can eventually organize themselves into a structured command without relying on Gadaffi-era members.  Regardless, they are independent operations with self-involved goals. This makes it unknown to the parliament, the UN, and Hafter what is their internal logic and expectations from the conflict.

The solution is clearly a third party mediator. Sadly the UN has been involved aggressively throughout the process. But when comments by Hafter reveal distaste for UN intervention and LROR threatens to fight a UN-endorsed government one begins to see that mediation is a point of contention amongst the parties.

So, the focal point is non-existent in the Libyan war. The Vienna talks might reveal a new focal point if it pushes the parliament in Tobruk to accept the GNA, causing concessions between groups as a communication channel opens. However, if Hafter remains difficult with the media (and public officials who govern him from Tobruk) and the Tripoli based militias lack internal control then the rising confusion leads to no general political focal point for Libya. Simple symbolic signals are important too. LROR and Misrata brigades don’t have the same uniforms, not between both groups or within each one separately. This further makes Libyans and their opponents skeptical of their current efforts anywhere.

It may seem the details I have discussed here are unnecessary for the conflict resolution. Certainly this thesis of confusion in the game of war is never discussed by any articles I have read on Libya. Articles focus on ideology of groups, which is important when discussing the ISIL, Ansari Alshari, and Shura Council of Benghzi. However, there is little to no use in islamist or liberal labels to explain the rest of the war. To put it bluntly, the most anti-Hafter city is Misrata due to the fear of rogue generals. The experience of Misrata in 2011 is dreadful and the city arguably will need another decade to recover, politically and economically. Same in Benghazi, where militias who use the word “revolutionary,” to excuse their assassinations and murder recked, havoc and killed a beloved ambassador. This experiential perspective of two unique cities in the conflict started the issues we face today. To solve it, some communication must happen so that at least both sides get systemic access to each others true expectations. This seem frightfully simple, and supposing this is the solution I won’t be surprised by the hindsight laden article in any publication noting that dialogue was a solution all along. Dialogue may seem simple in hindsight, but dialogue is hard to do and many individuals are preoccupied with ISIL sleeper cells and new governments. Hardly do they have time to revisit the structure of their command completely and commit to rationalizing the expectations of their enemy.

There is more to discuss about the conflict. In fact, as I noted in one of my previous posts, I find the conflict ripe for the type game theory analysis that deals with information asymmetry and reputation, likely with  multiple equilibria. The question becomes, in which “solution” is Libya a democracy and can one predict it. There is no focal point in Libya and if there is no focal point I can’t adequately predict a resolution.

Sources on militias:

More updated source:

Note: All sources on militias completed by foreign or local sources is bound to be incomplete, like I noted the structures intra-militias isn’t always well defined due to lack of proper leadership and anarchic banditry.

PS I highly recommend Chapter 3 and 4 from Thomas Schelling’s great book, The Strategy of Conflict. There are a lot of great Libyan writers and journalists. Mohamed Eljarh is great. Also, this site  run by Jason Pack, a History PhD at Cambridge University, is wonderful.



Escalation of Commitment in Libya

I should disclose that a lot of topics in economics dealing with political economy and growth theory that I am motivated by come from interest in the affairs of Libya.

Nonetheless, I currently acquired a copy of Thomas Schelling’s The Strategy of Conflict . My interest in the game theoretic nature of the Libyan civil war has occupied my day dreams and detached thoughts for weeks now. In particular, there are two elements of the “game” in Libya I find particularly interesting. First, it is escalation of commitment brought by a sunk cost fallacy and tribalism. The tribalism is essentially a sunk cost fallacy but I think there is a dimension to Arab tribalism that is almost a mapping of what someone thinks the hive mind behaves like.

For instance, in a discussion with a relative, they were oblivious about what was happening in Benghazi or in Sirte or in Derna. Areas where substantial and unusual islamist radicals have flourished. The individual was reasoning because his local area’s leaders have said it doesn’t exist then no one thinks it exists. This is different from the presumption I had that people tied to groups because they were attached to their identity then followed their deeds, but in Libya it is like the group is a short cut to normal thinking. A secondary dimension is added.

Secondly, the proxy war in Libya can exist even when it doesn’t through rumours. While the interests of various Arabic parties and Turkey are murky it doesn’t stop the discussion between militia leaders from divulging into lumping a proxy onto another militia. At times with negligence. Tribalism plus false narrative can certainly fuel the “game”.

The two governments in Libya almost ask to be modelled into a paper. Leaving me to wonder if a game with sunk cost fallacy or escalation of commitment exists and whether simple “narratives” can be reconstructed in a game. I’ll certainly try and find out while the weather is cold this weekend.