WHEN RIOTING suddenly broke out in Kabul in May, sparked by a fatal traffic accident involving the U.S. military, most in the city were taken by surprise. Less shocking, alas, was the response of the Afghan National Police, or ANP, to the unrest. Rather than dispersing the mobs and restoring order, Kabul's cops were reported fleeing their posts and, in some cases, joining the looters. "The reaction of our police was really shameful," acknowledged Jawed Ludin, chief of staff to President Hamid Karzai.
Unfortunately, the sorry performance of the ANP was not an isolated event, but a reflection of a much bigger problem. Nearly five years since the ouster of the Taliban and more than three since the fall of Saddam, the Bush administration has repeatedly stumbled in its efforts to create effective foreign police forces. In marked contrast to the army-building efforts in Iraq and Afghanistan, which have begun to yield encouraging results, the indigenous police in both countries appear stuck in a transition to nowhere, slaughtered by insurgents and infiltrated by militias and warlords.
Admittedly, there are good reasons why police are harder to recruit and train than an army. Militaries are structured hierarchically and tend to operate in large formations, allowing their development to be managed top-down rather than bottom-up, and with less manpower. Even more important, an army by its very nature operates at a distance from society, and can consequently be better insulated from its problems--whether ethnic rivalries, patronage networks, or corruption. Police are harder to wall off from these forces, operating as they do in close proximity to the population.
But that is also why police are so important--especially in counterinsurgency, where the need to gather intelligence and win public trust demands a security force that can stay close to the people. Indeed, as a superb U.S. War College study by James Corum recently argued, "in counterinsurgency, organizing and training the indigenous police often attains a higher priority than training the indigenous army."
Police are likewise crucial for democracy. Far more than soldiers or parliamentarians, they are the representatives of state power with whom ordinary citizens have regular contact. Rule of law, civil liberties, human rights--all presuppose the existence of a certain kind of police.
Why, then, has police assistance in Iraq and Afghanistan gone so poorly? As always when bureaucracies fail, there's an enormous temptation to blame insufficient resources and inadequate planning. And certainly, given the Bush administration's often lackadaisical attitude toward postwar reconstruction, less deserving scapegoats can be found. Yet a closer look at efforts in Afghanistan reveals another, more troubling dynamic at play--one that suggests that, absent sweeping reforms, police-building will continue to be a weak spot in the global war on terror long after George Bush leaves the White House.
The story of the Afghan National Police begins in late 2001, when Hamid Karzai's interim administration came into existence, inheriting tens of thousands of poorly trained, poorly disciplined, and poorly equipped constabularies. Although in theory answerable to the interior ministry in Kabul, these forces were "national" in name only, a balkanized rabble whose loyalties tended toward local powerbrokers.
Faced with this mess, the Bush administration first tried to hand it off to someone else. In early 2002, responsibility for the ANP was given to Germany, under a plan for Afghan reconstruction in which different countries took charge of different problems. It was thought that assigning ownership of a particular issue to a particular government would bolster accountability for solving it. In practice, however, this stab at hardheaded multilateralism proved a disappointment, as nations interpreted their mandates in wildly divergent ways.
While the United States, responsible for the Afghan National Army, understood its task to mean building the new military, the Germans insisted they were only coordinating police reform. As a result, although Berlin set up a police academy in Kabul, it made no systematic effort to develop the professionalized, countrywide force so desperately needed--a gap the Afghans soon turned to Washington to fill.
But building foreign police, it turns out, is something that the American government is expressly designed not to be able to do--the legacy of a 1974 congressional ban that abolished USAID's Office of Public Safety, previously charged with these missions. Although exceptions to the act have since crept onto the statute books, their cumulative effect has been to make police assistance into a second-tier, ad hoc responsibility of several different agencies and actors scattered throughout the executive branch.
Worse yet, the infrastructure that does exist for police assistance consists of more bureaucracy than capa city. Because America doesn't have a national police force of its own from which to draw for deployments abroad, Washington has come to depend on contractors like DynCorp, which in turn hire retired state and local cops and dispatch them to post-conflict zones.
In Afghanistan, police reform fell to the State Department's Bureau for International Narcotics and Law Enforcement Affairs (INL)--despite the fact that the bureau's core mission is counter-narcotics, and that it had almost no personnel for the job of building foreign police forces.
INL's plan amounted to little more than sending Afghan police, as quickly as possible, through a handful of regional training centers run by DynCorp. Although this approach allowed Washington to congratulate itself for having "reformed" a large number of ANP in short order, it scarcely affected their behavior or capabilities at the operational level, where it actually mattered.
"The police would get trained, but then they would go back into the system with nothing to support them, and they'd tend to fall back into their old bad habits," recalls one Afghan policy insider--a process another official compares to making batch after batch of ice cubes, only to keep dumping them into a vat of boiling water.
The shortcomings of INL's plan were especially glaring to U.S. soldiers dispersed throughout the country, who had to live day to day with a weak, corrupt ANP. Early last year, when I visited a Provincial Reconstruction Team in Ghazni, its commander confessed he was spending half of his time on the police, even though he had no mandate to do so. The local ANP were simply too corrupt and inept to safely ignore, he explained, and no one else was volunteering to fix them.
A similar sense of frustration gnawed at the U.S. military leadership in Kabul, who contrasted the lackluster performance of the ANP with that of the increasingly capable Afghan army. The latter, they noted, was being overseen by a large, U.S.-led office of military cooperation, along with hundreds of American soldiers embedded inside the force. These tactical trainers represented an especially important innovation: Living alongside Afghan troops and accompanying them on operations, they provided constant reinforcement and mentoring, as well as serving as liaisons with coalition forces and a check against abuses.
Given the success of this model, the military began arguing in mid-2004 for a new approach to the Afghan police, one that would allow the U.S. military to oversee their training, as it does that of Afghan soldiers. Not only would this allow the Pentagon's vast resources to be funneled toward supporting the ANP, providing the personnel that the State Department lacked, it would also facilitate an integrated civil-military strategy for Afghan istan's security forces.
Although the proposal won approval from Zalmay Khalilzad, then-U.S. ambassador to Afghan istan, it was seen at the State Department as nothing less than a military coup, sparking massive resistance. The stage was thus set for what one U.S. official would describe as "the most frustrating, bureaucratic, counterproductive interagency battle I've ever known."
The argument, which persists to this day, boils down to a nasty collision of ideologies and institutional cultures. INL, in brief, insists that police assistance must remain civilian-led and that the Pentagon's involvement threatens to "militarize" the program; rather than building an Afghan police force focused on rule of law and human rights, it warns, the U.S. military will turn Afghan cops into auxiliaries for counterinsurgency. As one Foggy Bottom employee griped to me last summer, "The Defense Department fundamentally doesn't understand rule of law."
The military--along with much of the Afghan national security leadership--responds by pointing out that, like it or not, Afghanistan is a country at war. In the south and the east, in particular, Taliban and other insurgents have been murdering police as representatives of the national government. Regardless of whether officials in Kabul or Washington wish to think of the ANP as combatants, the enemy is treating them as such.
Publicly, both sides claim that they have now reached an amicable compromise: a composite training command, responsible for both the army and police, run by a two-star U.S. general but with an INL representative who retains oversight of the ANP.
The reality on the ground is far darker, however: a shotgun wedding between the military and INL, characterized by pervasive distrust and recrimination at the staff level, and recurring skirmishes over issues like which contractors to hire, what tactics the Afghan police can be taught, and whether key individuals should work out of the U.S. embassy or the military compound. "INL is constantly trying to split stupid hairs," complains one officer. "Teaching the police how to react to an ambush: Is that offensive or defensive? They say it's offensive and shouldn't be taught."
Unsurprisingly, the biggest losers in this unhappy marriage are the Afghan police. Although some reforms have lurched forward over the past two years, such as a series of personnel changes in the ANP's upper ranks, the most important question--how to get large numbers of U.S. personnel embedded with police at the operational level--remains unanswered. In part, that's because INL has held the line against using soldiers to train police. It's also because any effort dominated by interagency sturm und drang is likely to remain more focused in Kabul and Washington than in the field.
Whatever the excuse, the result is that the Afghan police--despite fighting bravely in numerous engagements--all too often have found themselves isolated, outmanned, and outgunned against a revived insurgency. The failure of the international community to deliver effective police also prompted President Karzai to suggest the formal creation of village militias--a controversial proposal that speaks volumes about the disillusionment and disappointment of our Afghan allies, whose public credibility is being chipped away by their inability to secure their country.
The problem here isn't that the American officials involved are ill-intentioned or egomaniacal. On the contrary, one of the most striking things about the civil-military tension over the ANP is its persistence despite successive staff turnover.
Rather, the difficulty lies in the fundamental misalignment of capa city and responsibility for police assistance inside the U.S. government, and the extent to which the institutions of American foreign policy simply aren't organized for this purpose. Instead of confronting the need for painful bureaucratic reforms in Washington, however, U.S. officials have shifted the burden almost entirely to Kabul: Over there--and only there--are people expected to disregard their institutional identities, disentangle their respective mandates, and then jerry-rig some sort of mechanism to accomplish the mission. It's a rare constellation of personalities who can make this work; most of the time, it's a recipe for gridlock.
This arrangement still might make sense if we were convinced police assistance in Afghanistan were an anomaly, a onetime requirement that won't recur. But that's hard to swallow, given the string of interventions over the past decade--Iraq, Kosovo, East Timor, Bosnia, Haiti--all of which have required some sort of ambitious police-building. And given the nature of the war on terror, especially as the Bush administration has defined it, with a dual emphasis on security and liberty, there's every reason to believe foreign assistance to indigenous police is going to become more, not less, important in the years ahead.
On the positive side, the Bush administration is spending more money to help the Afghan police than ever before, but new squad cars and refurbished police stations aren't going to fix the institutional disconnect in Washington or Kabul.
Here, then, is the paradox: Police assistance will continue to be a critical American mission for the foreseeable future, while the U.S. government will continue to be organized in such a way as to be bad at it.
Perhaps those Afghans have good reason to riot.
Vance Serckuk is a research fellow at AEI.