Local V. National Control

by Joshua Foust on 3/27/2010 · 8 comments

Context: Gameplanning a Solution In Medias Res
Discussing the Unpersuasive Ways Forward

One of the key assumptions behind the Small Wars Tribal Engagement Workshop (from now on I’ll call it TEW 2010) is that there must be some sort of middle ground between national governance objectives and empowering local communities. Several people suggested connecting the two objectives at the district level—think of a more refined version of the district development teams Tom Johnson has advocated (only rather than atomizing PRTs, this would entail empowering current district-level Afghan power elites). Most of the participants though this could best be achieved by holding district elections—a key failure of the international community and Karzai administration, since they should have happened in 2005.

But holding district elections is not a simple matter of holding them. For one, we don’t know how many districts there are, or where their boundaries lie: depending on which official source one uses, official district areas can vary by hundreds of square miles, and there are also dozens of unofficial districts nation-wide, including the newly-created unofficial district of Marjah, Helmand (which did not exist as a separate political entity until January). A pretty fundamental first step in holding district elections is deciding how many elections need to be held, and where people will be able to vote.

But even more than a practical consideration of whether they are practical to hold is the question—again, generally left unasked—of whether districts are even the appropriate structures to work through. Constitutionally, the provincial governors are not responsible for their provinces per se, but for keeping the Minister of the Interior and Hamid Karzai happy. This creates a pretty obvious perverse incentive, whereby a governor legitimately representing the needs of his people—say, Tamim Nuristani complaining about civilian casualties from errant U.S. air strikes—can be summarily fired for embarrassing the wrong person in Kabul. The current structure of the constitution, in other words, creates a range of headaches. We’re learning years after its drafting that it wasn’t designed for functioning so much as the ease for VIPs to interact with their counterparts.

That being said, the constitution requires district elections. Not holding them is violating the Afghan constitution. It is one of many ways in which the constitution complicates our objectives for the country. Some TEW 2010 participants looked at these complications and suggested we scrap the constitution, dissolve the government, and start over. And such a solution seems very appealing, until again we consider just what the hell we’re trying to do.

Forgive the slightly disjointed nature of this discussion—this is not a simple problem to contemplate. The objective of both the U.S., EU, NATO, and ISAF is to empower the Afghan national government, creating a stable society that will no longer play haven to terrorists. We cannot reasonably achieve that goal by arbitrarily dissolving the government we created because we find it convenient to do so. Now, if we are to modify our objectives—for example, by reducing our pretensions merely to allowing us the space to punitively attack elements we feel pose threats to Western interests—that is possible… but until that happens, we have to keep our national and international objectives in mind (there’s also the pesky problem of what Afghans want, which all the data on hand indicate is a functioning government).

So if we are stuck with a government we can easily or justifiably dissolve, we must work through it. Again, holding district elections is a critical component here, but I part ways with the larger group in seeing empowering districts as the key to connecting the much-beloved local defense forces to the national government. Despite the insistence of everyone that LDI/CDI/AP3 groups are not militias, and neither are arbakai, I remain deeply suspicious of these groups. Despite the current practices that limit what these local defense groups can do, there remains a deep distrust of any non-government armed group.

In 2007, at least, strong majorities of Afghans express at least some confidence in the National Army (84%) and National Police (76%), with opinions marginally lower in Kandahar. In sharp contrast, only one in four (26%) express such confidence in the local militia in their area (which includes private security forces and local warlords). Opinions were consistent across ethnic groups in that survey, though more recently non-Pashtun groups have expressed stronger skepticism of the idea.

I question that it’s even a good idea to use district officials to “connect” local security forces to the government, since I question the widespread adoption of these local security forces. That being said, we live in a universe of path dependency, and these groups are being created and expanded regardless of whether or not they’re a good idea in the long run. So in that sense, district administrations could, conceivably, be set up as a check on individual local forces. I have no idea how to do that, however—it would require having both ANSF and Coalition forces in reserve as a QRF of sorts, essentially defending communities from both the Taliban and their own security forces. I don’t see that as a sustainable solution. (Though we do have successful examples of the ANSF, including the ANP, doing their jobs admirably, and we should be looking to replicate those successes instead of re-inventing the wheel.)

There’s also the larger conceptual question of whether districts are appropriate interfaces between the national government, as represented by provincial governors, and communities, as represented by… well, we don’t know yet. For starters, there are so many local/regional governance and representative bodies—ASOP, NSP, CDCs, and so on—that I don’t know what has responsibility for what, nor do I know how these communities will respond to having marginally functioning systems of local administration swept away for a U.S.-administered and nationally-run elective and administrative system (this is another downside the “districts first” crowd didn’t discuss to any detail).

To reference my previous post about the military’s crippling experientialism, while you don’t see much informed discussion of how government normally works in Afghanistan in pop histories, the media, or soldier and marine AARs, you do find it in academic literature and detailed histories. More amazingly, none of this is a secret—it is available, mostly for free, on the Internets, and it was out there many years ago.

José Oberson, for example, wrote a really interesting article (pdf) exploring the incorporation of social fracture into models governance:

Whether the imperfect success of religious politics the failure of Pashtun national politics are due to a balanced product of either their conceptual weakness or strength, respectively, is an open question. But the contribution of disjointing forces that result in an impression of deep-seated propensity for political and societal fragmentation may even account more for political events than other factors…

In Afghanistan, [a leader overcoming person-based politics by using religion] has been successfully activated in order to overcome tribal constraints and kin-based political economy with considerable success in the beginning, as waging jihads against non-Muslim invaders or the initial triumph of the Taliban movement have shown. Another historical example is a form of internal colonialism imposed on non-Pashtun communities gives evidence for this strategy.

Oberson, however, makes sure to caution that even the religious example is at best spotty and not-permanent (and I want to emphasize this again, since many TEW 2010 participants got it into their heads that we need to coopt or manage the influence of mullahs, which for a crowd of mostly-white Christians didn’t strike many as a guaranteed epic fail). For lack of a better term, Afghanistan’s politics are more or less permanently personal, and creating systemic or material solutions to personal problems will not be very effective. A second opinion here comes from Olivier Roy:

Th[e] Afghan identity is based on a common political culture which could be summarized as follows. ‘Real’ political life is played out at the local level and primary loyalty lies with a ‘solidarity group’, whatever its sociological basis…

What these local and ethnic networks and groups need is a distant but benevolent and legitimate state, regarded as a broker or an ally helping to establish a favourable local balance of power and influence. They also expect the state to deal with general services, education, health, transportation etc. The state is seen as a means of enhancing local status and power, and must therefore be effective, without being disruptive – and a state that bases itself on ideology, whether Communism, Islamic radicalism or ethnicity is a disruptive state.

This Afghan political culture has of course been shaken by over 20 years of war. In 1978 the new communist state was perceived as an enemy of the people, based on an alien ideology and working for an alien country. The revolt therefore took on an anti-state dimension.

The key point here, for the purposes of this specific discussion, is to avoid the state being seen as disruptive. And using district officials as a buffer between national representatives at the provincial level and local security forces is deeply disruptive. It might be more efficient in a theoretical sense, but it violates how these communities tend to entreaty the government for help, especially if we are creating local security groups and then neutering them with district officials before they can get to national administrative institutions. The other concern is government being seen as fair and representative that can serve as a broker—while the current structure of the provincial governments is not that, neither is the mere establishment of district elections.

So, in a way, the big question of what to do about districts can be answered simply: hold elections, but leave it at that. If the goal is to build up a massive and widespread local security apparatus, then it does not need district-level intermediators to manage their relationship with the government: that should be the job of the provincial officials, who then also command the ANSF elements responsible for the area. The elected district bodies can and should be responsible for representing individual villages and family concerns to the provincial governor, and for dispersing development money (mirroring the CDC method of handing out developmental assistance). In one way, this would be cribbing from the British idea of administering through political agents (another discussion, perhaps?), which does pose its own problems. But it seems like accomplishing the goal of connecting local initiatives to the national system of administration doesn’t need any help, it just needs barriers removed. We don’t have to be disruptive and reorder the entire district system to be effective at reducing space for insurgency while empowering local Afghan control, just incremental.

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 Registan.net. 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


TJM March 27, 2010 at 2:01 pm

district administrations could, conceivably, be set up as a check on individual local forces. I have no idea how to do that, however—it would require having both ANSF and Coalition forces in reserve as a QRF of sorts, essentially defending communities from both the Taliban and their own security forces.

Why not make the local body (whether district, province, or community) responsible for compensating any ANP force that is manned entirely (or almost entirely) of men from the community that they police? If the central government funnels money to that more local body, which then pays the ANP, then this creates a situation where the ANP are enforcing the laws of the state, in return for money from the state, distributed through a local administrative body that gains legitimacy as a conduit through which money is channeled.

The elected district bodies can and should be responsible for representing individual villages and family concerns to the provincial governor, and for dispersing development money (mirroring the CDC method of handing out developmental assistance).

Definitely, I think the elected local bodies should be responsible for making financial decisions related to government funds, but I think the CDC method is a poor example as it is presently implemented. The CDC seems to be a program divorced from any long-term strategy. ISAF is the largest organization with the largest budget and influence in Afghanistan. NSP may be second, if not close to it. Why isn’t NSP synchronized with ISAF? Were there NGOs working for NSP following the Marines into Marjah to establish a CDC that would do projects in conjunction with CERP projects and establish some kind of local administrative body for the Marines to interface with? If so, great. If not, why?

Joshua March 29, 2010 at 8:55 am

TJM, in answer to your first question: they can’t afford it. For the foreseeable future, the entire ANSF budget will be shouldered by the international community, primarily the U.S. In a manner of speaking, there already are those communally-responsible structures, called arbakai. But they exist only in a small handful of provinces (and there is universal academic consensus that it would be a bad idea of force such an institution on communities that don’t already have it).

TJM March 29, 2010 at 12:06 pm

Regarding whether they can afford it, what I was getting at was to have the central government give the money to the local paying agent and have him/them pay the ANP, rather than using the current payment channels. If the local district office, CDC, or whatever is the place that people get their wages from, then this can reinforce its legitimacy, so long as there is no perception of officials skimming money from the account.

Sailani March 28, 2010 at 3:16 pm

Dear God no more elections please! The Arbakai can answer to the local CoP, who in turn is answerable to the provincial CoP and through him to the Provincial Governor AND the ANP command chain up to the minister in Kabul. Does not require us to create yet another layer of local government, and integrates security forces into a security force chain of command.

Not quite sure where the LDI project is currently. Depending whom you speak to at ISAF it’s either been killed off until Karzai approves the overall concept and launches it with Afghan ownership, or it is bubbling along in the care of SF ODAs across the Southern, Southeastern and Eastern regions. I don’t share Josh’s concern that these are major problems for the future, essentially because they will be very small in scale and capability, and will be answerable to the local ANSF command structures (all the members are also known, badged, and their weapons registered, so their ability to go rogue, absent a total meltdown of the Afghan state, is limited).

Joshua March 29, 2010 at 8:56 am

One caveat: district elections are mandated by the Constitution, and should have happened in 2005. Every year they don’t happen is another tangible proof that both Karzai and the International Community intend to discard pieces of Afghan law they find problematic or annoying.

Unless we reform the constitution. Which, as I said in the post, poses its own headaches.

Also, for the 31 provinces without arbakai, how would you structure the cooperation between local initiatives and provincial/national administrations?

BruceR March 28, 2010 at 4:35 pm

Oh, good, their weapons will be *registered.* That makes me feel a whole lot better.

(You know, the thing we saw with all those cheap AK knockoffs coming out of Pakistan a year ago that the real ANP were swapping their real guns for for personal profit was how, even though the gun itself might be better used as a club, they could always get those serial numbers exactly right… it was remarkable…)

Madhu March 28, 2010 at 5:53 pm

Interesting post.

A side question: What kind of census has been conducted in Afghanistan to date?

Christian March 28, 2010 at 6:53 pm


There was one started in 1979 but canceled for obvious reasons. For security reasons the planned 2008 census was postponed to… 2010. I’m guessing that it’s not going to happen this year either.

Previous post:

Next post: