Wednesday, February 29, 2012

Police development: you're doing it wrong

I can't remember who tweeted a link to this excellent report on ANP development (speak up and be recognized!), but if you're interested in police development it is a must read. A product of the Peace Research Institute - Frankfort, the authors (Cornelius Friesendorf and Jorg Krempel) titled the work Militarized versus Civilian Policing: Problems of Reforming the Afghan National Police, looked at best practices in building a civilian police force and the German and US involvement in Afghanistan with the ANP.

The paper dives into a lot of the nitty-gritty of the ANP development program dating back to 2002 and it's worth a read. (I wish it had been translated when I wrote this paper that touched on some of these issues this past spring.) But I want to focus in on some of the conclusions and attempt to draw them into bigger thoughts beyond Afghanistan. They cite numerous issues of donors not effectively supporting the ANP development, recruiting problems, training problems, corruption problems, and so on and so forth. All important.

But what's important in this paper is that they look very closely at the intersection of police and military responsibilities and point how horribly wrong we've been delineating responsibilities between the ANA and ANP. Due to a number of factors - lack of a democratic policing tradition in the years running up to 2001, the use of military forces to train the nascent ANP, and operational needs - ISAF is building a paramilitary force not a police force. While the first of these factors can be overcome in my opinion (Kosovo provides a pretty example of making an adequate police force essentially from scratch), I think the latter two are catastrophic to police development programs if not done correctly. And we're not doing it correctly.

In brief, if a military force trains a police force, that police force is much more likely to look like a military force than a police force. Just as in Iraq, the ANP are formed into military structures (unless that's changed recently), they carry military-grade weapons (including RPGs on occasion), and they have little interest in the tedium of enforcing the law. Simply put, if you train your police with the military, you end up with police who can't police and are an under-trained and under-equipped military force. What you end up with is a force that exacerbates conflict instead of quelling it that primarily acts as a strong-arm of local factions or even national entities (such as in Iraq where the INP became the MoI's private Army). Historical studies have shown that military training of police forces nearly invariably result in this scenario and we need to stop doing it. Of course, that would require building capabilities and/or capacities in our civilian agencies such as ICITAP, OPDAT, USAID, or even INL. Something that seems unlikely given Congress's fetishization of military power over civilian power. Alas.

However, this pales in comparison to the operational necessities that drive police development away from policing and towards military operations. When conducting police development in countries embroiled in an intra-state war, priority is given to the operational needs to defeat the state's enemies, not to stemming crime. The U.S. and other donors have been building the ANP into a counterinsurgency force in their own right to "thicken the lines" as the old Iraq saw went. Community police simply wouldn't be able to defend themselves against any quasi-military insurgency. As they would be among the first targets of the insurgency, police being among the most visible realizations of a state's power, they would simply be slaughtered. This would then require precious military resources to be pulled from finding bad guys and instead guard the police. It's a lose-lose situation for any operational commander. So we've taken the middle path: forget policing and build a quasi-military police force. A force that, again, stinks at policing and stinks at military operations.

This is an offhand synopsis and analysis of the paper I linked to, but it distills to the fundamental flaws of our attempts to build a police force in Afghanistan. Because of the myriad problems underlying our efforts and our militarization of the ANP (because of the two factors discussed here), we're failing Afghanistan by not fielding an effective police force now and virtually ensuring that the ANP cannot transition into a true police force. Friesendorf and Krempel put it much more eloquently, but that's the basis of their conclusion. Nice, huh?

So what do we do about it? Our boys from Frankfurt call for reforms to the program and to ensure that those in ISAF who are assisting in driving the ANP strategy are police experts and not military officers (they don't say it, but an MP is not a policeman, so no, not even an MP should be doing this). Concur entirely. I don't know what else we can do for the ANP at this point. But the lesson we need to draw from Afghanistan and Iraq is that in conflict countries maybe we shouldn't do police development at all until violence reaches levels where they can operate as police and not as soldiers. Police development in the early stages of the war should focus on the ministries responsible and a cadre of officers who are provided with long and intensive training who don't actually do any policing until a violence threshold is met. At that point, DDR programs could help staff the police force around this framework of cadre who can then instill a proper police culture throughout the force and train their policemen to be policemen. Instead of throwing billions of dollars at sub-par police who also are sub-par counterinsurgents as well as conflict drivers, maybe we shouldn't do any of that until the environment is safe enough for an actual community police force to operate.

Unlikely, but maybe in the next war.


  1. A-stan is the Wild West. So you need your lawmen to be like Wyatt Earp,
    when they've evolved only then can they learn from top cops like Bill Bratton and his COMPSTAT bs.

    Until then, it's the Wild West.

  2. Ah, but the problem BachaBozo is that while true, it leads to ineffective police. This has proven true in every case study I've ever written or seen. See my last (full) para on what we should really do to account for that. Let the military sort it out until the police can do it.

  3. Wyatt Earp was ineffective? OK, maybe the OK corrall isn't how you deal with Afghan protestors.

    How about we let China train ANP and we train the ANA, so long as ANA don't shoot us.

  4. Afghanistan is not the Wild West, and our relentless application of false analogies because we're too lazy to understand it on its own terms is the root of most our problems there. Not to suggest that doing so is easy, but at the policy/strategic and institution levels, we have consistently prioritized action over analysis. Ready, fire...(not sure we ever really get to 'aim').

    To wit - Jason, while (as you know) I largely agree with your critique of militarized police training, it strikes me that a doing a better job of building a Western-style police service (whether a single national one, or multiple regional/municipal ones) in a context like Afghanistan is still likely to fail. We need to start with a vision for a political system and associated security institutions that matches the political, economic and social context of the country in question. I don't see a lot evidence that ICITAP, OPDAT, and INL seriously consider alternative models that might be a better fit. USAID seems somewhat more open-minded, but still defaults to fairly conventional models.

    Seriously - is there any case we can point to where SSR models (or for that matter state-building) has been genuinely adapted to the challenges of a conflict-ridden country?

  5. I'm sure another problem is the fact that most police officers or deputies sub-contracted thru DoS are small town cops from powdunk county USA. Who think of themselves as modern day Wyatt Earps who own hundreds of firearms and love Rick Santorum.

  6. Not true at all. Most of these police officers come from the largest police departments in the country.

  7. What's the breakdown then? 30% LAPD, 30% NYPD, 30% CPD & others? Let's face it we ourselves face the same phenomena of our cops militarizing and gearing up like they were SOCOM. Community policing is still very theoretical, ask Bill Bratton, many cops don't buy into it.

  8. MK--on SSR, I don't know of any example where we can say: "in this case, we did a careful assessment of the context and came up with this plan for SSR." In fact, all the cases I'm familiar with have a similar lesson which is that "we should have done a better job assessing the context and tailoring our SSR program."

  9. It seems almost like a Maslow-type hierarchy (I'm sure there are other hierarchies and orders of operation out there; his is just the first that popped into mind): first one must stopped militarized conflict before focusing on criminality and associated problems of law-and-order. That seems intuitively plausible and relatively uncontroversial on its face - for those who work on SSR/DDR, does that reflect the conventional wisdom inside the Beltway? Also, I can't help but note that Jason Fritz is a former military officer writing about policing within the context of a nation-state undergoing a military insurgency - what are the backgrounds, generally, of those who work on SSR/DDR, and is there any general patterns that transpire which may arise from differing backgrounds (in other words, people with law enforcement experience and backgrounds tend to favor A/B/C, and people with military experience and backgrounds tend to favor X/Y/Z)?

    Thanks for indulging my curiosity,

  10. ADTS - This is the one part of Jason's post I don't entirely agree with. Military and criminal depredations against civilians are often difficult to distinguish from each other in conflict environments, and insurgents/terrorists are often detained, prosecuted and punished under domestic legal regimens (even if they are enacted under national security or emergency powers legislation) rather than as combatants under the Geneva Conventions.

    And from a political perspective, the 'predictable and tolerable' conditions required to obtain and sustain the cooperation of the civilian population doesn't find a meaningful distinction between criminal and military acts of violence. Whether in the form of a mob boss running a protection racket, or a highly formalized law enforcement and justice system, there's got to be something to keep criminals in check.

    Although it is only one example from a very particular context with a relatively low level of violence, Northern Ireland offers a case where the police and military had primacy at different points in the conflict.

    Likewise, the working relationship between UN police and military forces in Haiti in 2006/07 holds lessons for lower intensity environments. Haiti also provides a clear example that civilian police-led training does not guarantee that police and justice reform will succeed.

    However, as Jason points out, asking police to stand up to threats that exceed their capacity and distort their role is just as counter-productive as asking the military to deal with more run-of-the-mill, inter-personal crime issues.

  11. ADTS--on the background question. This varies tremendously.

    I don't know if there's been any work done on the issue and on how biases can affect program design and implementation. In the work I've done on this, the practitioners--by that I mean those charged with implementing the SSR programs in the field, for example the various donors that contribute to SSR in DRC-- I interviewed fell into several categories: government bureaucrats, diplomats, military officers, police officers, lawyers, corrections officers (not a lot of those ever), people who had worked in UN peacekeeping a decade or more, and people who work for NGOs and relief organizations.

    As to whether they were or perceived themselves qualified for their jobs, well that varied too.

  12. MK - Lil is right that donors have rarely assessed programs before beginning SSR. They have a tendency to create security forces in their own likeness. I also agree with you that civilian police-led training does not guarantee success, Haiti not being close the only example of that. But I can't think of a single case where military-led police training has succeeded in the long term. If I'm missing one, let me know.

    I also find N. Ireland to be a difficult example for the U.S. to follow for the obvious reasons. For the UK it was a justice issue, unlike ISAF in Afghanistan or the US in Iraq. I'm not saying that we should delay police training in every intervention that involves violence, just in environments where violence exceeds certain parameters (as yet undefined) that by their nature preclude effective employment of police. Iraq and Afghanistan have both been one-offs (two-offs?) for the US in trying to build a police force under such intense and pervasive violence - and the military-led police training efforts there have been, on the whole, failures. (I understand DOS has played a role in both programs, but without spending too much time here, I think their role only supports my position on this.)They may be anomalies, and they may not be. But I think we really ought to consider it.

  13. MK, Lil and Jason:

    This is directly primarily at MK and his comment, but thanks to each of you for your responses. I suspect my prior comment betrays a relatively superficial and unsophisticated understanding of SSR, the challenges it presents, and the context in which it occurs. I further suspect that there are fairly complex conceptual and definitional issues regarding what is military and what is criminal, what is legal, etc., and that all harken back to the classic Weberian question raised of, "What is legitimate? What does 'legitimacy' mean?" Tilly's provocative analogy of the state as a racketeer comes to mind (perhaps unsurprisingly), and I suspect I would perhaps profit by reading Woods's article in the "Annual Review of Political Science" on social dynamics during civil war, along perhaps with Race, Elliott and (the usual suspect) Kalyvas.

    Enough name- and citation-dropping: I still am curious about the enterprise of SSR and SSR consulting, the backgrounds of those who undertake it, and the potential biases those backgrounds impart. I recall Ray Kelly (NYPD) worked in Haiti and (perhaps less illustriously) Bernard Kerik (also NYPD, currently c/o Federal Bureau of Prisons, I think) worked in Iraq. In my personal life, I encountered a few North Carolina ex-state law enforcement personnel en route to Pakistan (2009). The case of Northern Ireland (perhaps embarrassingly) didn't even come to mind, even though it probably is paramount for the question of who assumes primacy in a unified COIN campaign (i.e., military or police?) - thanks for raising it. I'd be curious to note any variance in mindsets, approaches, strategies undertaken, etc., that could be explained due to differing background. The comparison of the Marine Corps versus Army in terms of institutional legacies and differing viewpoints about how to prosecute the war in Vietnam is perhaps an analogous interservice dispute (see, e.g., the sketch offered by Sheehan, in "Bright Shining Lie"); it'd be interesting to see something similar in which, say, a career law enforcement officer offered a dramatically different suggestion than a career military officer, and if that could be traced back to their different career paths and professional affiliations.

    Thanks once more,

  14. That's probably the problem there, NYPD commishes shouldn't be out there teaching others.
    Because they'll know how to spout "community policing" bs, but not really understand it.

    Maybe the LAPD, more professional than NYPD, would be the right resource, but it's theyre too
    far away from DC, to actually be considered.

    but also you'd
    want to look at the "career" police officers' record, whether he spent most of his time in investigative
    units, specialize squads, or behind a desk polishing his resume.

    If someone's going to understand
    community based policing it'll be that 25 yr beat cop, whose worked the same beat or that shift sgt.
    whose worked patrol all his career.

  15. Kind of along the lines of BachaBozo, do people who study, teach and research (e.g., professors) criminal justice for a living work on, consult on or perform SSR, or is it for the most part the preserve of political scientists (or - gasp - others)?


  16. ADTS and BachaBozo--great discussion.

    I don't have good answers on your questions though I would say that I encountered two big categories when I was involved in a comparative assessment of international efforts to support SSR in conflict environments a few years back: the practitioners and the M&E/political scientist/lessons learned (not all the same I know) people.

    Just to give you some examples of practitioners: Mark Kroeker (Police Commissioner for UNMIL after a 32 year LAPD career, rising to Deputy and then Chief in Portland, Oregon, so no, the West Coast is not too far away to be considered!); the senior French gendarmerie officer who was police commissioner for UNOCI (and who told me he had too many senior officers on staff who hadn't been beat cops in too long and therefore weren't much use helping to train beat cops (so agree completely on that front BachaBozo) or the German general who was head of EUSEC Congo when they were working on separating chain of payment from chain of command for the FARDC). I met a civilian advisor to the SRSG in Liberia who had previously advised the SRSG in Sierra Leone. I met attorneys who had worked in/run the rule of law offices in several UN missions.I think the head of the DDR Unit in Cote d'Ivoire had previously been the head of DDRRR in DRC (not sure what he did before that).

    Then, there are the think tank people and as I said the M&E people who work for governments or the assessment/best practices units at international organizations like OECD, the UN, etc.

    Now on to some people who might have some actual numbers on the real spread (rather than my impressions based on my little think tank perch). You could just give them a call:

    1)Mark Downes who runs the International Security Sector Advisory Team at DCAF (
    2) Nicole Ball at the Center for International Policy ( because I think she and a couple others have been working the M&E/best practices side the longest.
    3) Whoever runs the SSR shop at OECD-DAC (the former head of that office is at a UN mission) but I can't seem to figure out who that is. Nicole would know, as would Mark.
    4) you could try the SSR Unit at UNDPKO (they have a roster too of people to send out) but they're always a bit overrun with requests for info so I wouldn't hold my breath on a response.

  17. Lil:

    Many thanks for your thoughtful, and extremely thorough and comprehensive, response.


  18. City cops are basically your modern day flâneur, after they
    take care of business, they usually philosophize about life, living,
    society in general, while remaining as detached observers.

    This stance, simultaneously part of and apart from, combines
    sociological, anthropological, literary and historical notions of
    the relationship between the individual and the greater populace.

    SWAT and detectives, and other specialist coppers are one thing,
    but you want that "gentleman stroller of city streets" if you want a person
    who understands American policing, someone who can transfer that
    knowledge abroad.

  19. You might be interested to look at the Australian Federal Police's International Deployment Group: .

    Not sure that it has a specific or sufficient training (training of the host country's police, that is) element, but it's a start.

  20. Patrick Henry: the AFP IDG is interesting to look at and does have a training mission but it is not really a model for the US. Mainly because we don't have a uniformed federal police from which to draw such a group or any US law enforcement agency with a mandate, authority, or funding to do what they do. Also their training doctrine is written for post-conflict environments, not for during conflict. So we don't get much from them there either.

    I like that the Australian government has a force that does just that and that AUS law enforcement folks perceive service in the IDG to be good for their careers. But other than that, we can't get a whole lot out of them for our own purposes. But thanks for bringing it up!