From the Good Ideas File

by Joshua Foust on 10/11/2008 · 3 comments

In the L.A. Times, we learn:

Confronting the prospect of failure after seven years in Afghanistan, the U.S. military is crafting a new strategy that is likely to expand the power and reach of that country’s tribal militias while relying less on the increasingly troubled central government.

Under that approach, U.S. forces would scale back combat operations to focus more on training Afghan government forces and tribal militias. The plan is controversial because it could extend the influence of warlords while undermining the government of President Hamid Karzai in Kabul, the capital.

That is certainly understating the problem. The Tribal Liasion Office, a European-funded NGO whose explicit purpose is to encourage “systematic and institutionalized engagement with tribal structures” thinks it’s a good idea… in three provinces. In a paper about the use of tribal militias as security forces during the 2004-5 elections, Masood Karokhail and Susanne Schmeidl warn that the arbakai—tribal militias meant to enforce decisions by a jirga—are only relevant in the Loya Paktya area: Paktya, Paktika, and Khost. More importantly, they emphasize that arbakai are “a very temporary body that is only established for solving specific problems and only for the length of time required to do so.” They continue:

Despite the fact that each Arbakee has a clear leader (mir), the accountability goes back… to its own community. Furthermore, Arbakee only function in the very limited realm of the tribe they represent. Their fighters are volunteers from within the community they represent. Their fighters are volunteers from within the community and are paid for by the community. This emphasizes again that their loyalty is with their communities…

There are other issues with their use, which are too legion to go into here. Suffice it to say, the British had very limited luck in getting tribal forces to fight on their behalf, especially in a jihad context. And the terms under which the British did so required such crimes against humanity as collective punishment metted out on civilian populations—something the U.S. explicitly forbids by law.

Elsewhere, Conrad Schetter, Rainer Glassner, and Masood Karokhail have warned that in areas without the relatively strong tribal pattern of Loya Paktya the funding of tribal militias can lead to outright warlordism. They also note the problem in their fundamental nature:

…the tasks of the >>arbakee< < depend strongly on tribal norms and values, which in many cases are diametrically opposed to Western norms and values… For example, the strictly obeyed exclusion of women from the public sphere in Paktia contradicts the idea of gender-equal communal participation. Also, the continuing legitimacy of blood feuds undermines attempts to introduce modern conflict-resolution mechanisms [such as the DIAG]. Finally, one has to underline that >>arbakee< < do not constitute neutral forces, but are time and again involved in tribal rivalries.

This is another way of saying that arming tribes can blowback in a major way. Complicating matters further still is the fact that the Taliban are not tribal, and they never were. Outside of a few areas, tribal rivalries don’t drive conflict—communal rivalries might (where tribal identity is incidental), or general grievances against the government or one group favored by the government. All of these create the space under which the Taliban and other militant groups can move in and sow discord, yet none could be realistically solved by trying to co-opt the tribal militias.

In short, the idea of arming the tribes fails the intimate knowledge test. Much like the previous discussion on legitimacy, solutions for Afghanistan will fail—quite miserably, I might add—if they are not tailored to interact with the specific social circumstance of Afghanistan. Afghanistan is not Iraq, and you cannot treat it as if it is.

Tribal and community engagement can work, however. It requires getting off the FOBs, getting out of the armored Humvees and MRAPs, taking off the sunglasses and body armor, and treating the locals as if they were people. In fact, in the areas where that has been done, it has been hugely successful—at least until that Brigade is cycled out and the new one throws all the lessons learned away so the officers can get good OERs by changing everything. Our problems in Afghanistan are only partially tactical; they are mostly systemic. Until those systemic problems are addressed—including the overemphasis on force protection, the general Army way of doing things—we’re just trying to solve problems that don’t even exist.

Subscribe to receive updates from Registan

This post was written by...

– author of 1848 posts on 17_PersonNotFound.

Joshua Foust is a Fellow at the American Security Project and the author of Afghanistan Journal: Selections from His research focuses primarily on Central and South Asia. Joshua is a correspondent for The Atlantic and a columnist for PBS Need to Know. Joshua appears regularly on the BBC World News, Aljazeera, and international public radio. Joshua's writing has appeared in the Columbia Journalism Review, Foreign Policy’s AfPak Channel, the New York Times, Reuters, and the Christian Science Monitor. Follow him on twitter: @joshuafoust

For information on reproducing this article, see our Terms of Use


Daniel Pineu October 12, 2008 at 1:36 pm


Overall, I agree with the point you are making, but I think the problem is not so much, or not necessarily, to do with using informal security governance structures like the arbakai, but (i) what do you use them for, and (ii) how do you go about doing that. In other words, instrumentalizing local institutions such as the arbakai (or the jirga system, or lashkars, or…) for counterinsurgency purposes according to the US’s strategic agenda in Afghanistan may be the real problem. Arbakai should not be seen as a useful proxy, nor as a useful substitution force for the ANA, ANP or ISAF forces. They are not.

Sorry, not sure if I’m making much sense. Perhaps I should explain where I’m coming from. My doctoral research is on U.S. police reform/assistance as a tool of U.S foreign policy/national security, and I have been focusing on Afghanistan as a case study. On of the things that gets me mad about the whole thing is how police assistance and police reform (or indeed the whole “security sector reform” effort) is seen primarily through the lens of “our” goals. The U.S. has the most decentralized police force in the world, with more than 13.000 police outfits (I’m quoting from memory, but can get you an exact figure). When it occupied Germany and Japan, it went to great lengths to ensure that police forces in those countries were similarly decentralized (ordering each city of over 5.000 people to have their own municipalized police force).

In Afghanistan, such a decentralized model would – so far as my interviewees go at least – be a good fit. But we have determined that the crucial problem in Afghanistan is state-frailty, state-failure or state-collapse, and the existence of “ungoverned spaces” whence evil comes. So, all security forces must be national. All security assistance should go to the formal (state) sector. And the crucial thrust of their training and deployment should be ancillary to NATO’s COIN strategy, and secondarily to bost the central government’s control of… well, whatever it controls at the moment. You know, police calling in close-air-support, and “holding” areas already “cleared” by soldiers.

Meanwhile, in real Afghanistan outside Kabul, Herat and Kandahar, people still mistruts the police. An Afghan who wrote cogently on this asked the question: “Can the people secure themselves?”. Another asked: Why not have a look at improving (not instrumentalizing) local structures of justice delivery and security governance? TLO and some NGO’s working with it, as well as a number of Afghans, have for the last 5 years been talking about it. The US has either ignored or downright marginalized this perspective, putting virtually all the eggs in the same basket: ANP+ANA= ASF, ASF+ISAF=COIN.

Now they start looking at this (too little too late), and what do they see? Militias. Proxies. Partners in COIN. That is a fundamental misunderstanding of arbakai. They are not a militia. They are not made to clear, control, and hold territory. They are made to dispense justice, act as watchmen, perfom basic police functions at the village level. If the U.S. was serious about improving security governance in Afghanistan, it would be worrying right now how to interface these structures better with the central government. It would be sitting down and learning from them – what do they do, what they don’t, what they need, how can they improve. Sure, they’ll probably be limited to Loya Paktia, but that’s a start.

Before the creation of a national police in Afghanistan, or the opening of the Police Academy in kabul in 1924, the police _function_ already existed in Afghanistan, in society everywhere. Why not learn from that, build on that so as to achieve a model of plural policing, with service delivery spread through the formal and informal sectors of security governance?

But, if on the other hand the program now is to _arm_ the arbakai, and train them in light infantry tactics, and use them as the next Concerned Local Citizens/Awakening Councils, then you are absolutely right – it will fail, and the fallout will be awful. This is not an issue for military engagement. And you’re quite right, it will require “treating the locals as if they were people”. People whose pressing needs – including security needs – have very little to do with the mainstream narratives of COIN being transmitted from the corridors of DC all the way down to the private contractors training young Afghan policemen.

Sorry fo the length and possible incoherence of the post, this is a subject that touches a raw nerve with me.

Joshua Foust October 13, 2008 at 9:01 am

Daniel, I think you’re making a lot of sense. Under the normal center-periphery relationship ordinary communities used to have, the arbakai system seemed to work pretty well.

That is a great point.

cizgi film izle October 14, 2008 at 8:10 am

Before the creation of a national police in Afghanistan, or the opening of the Police Academy in kabul in 1924, the police _function_ already existed in Afghanistan, in society everywhere. Why not learn from that, build on that so as to achieve a model of plural policing, with service delivery spread through the formal and informal sectors of security governance

Previous post:

Next post: