The Atlantic Home

President Obama has made it clear that any strategy he commits to in Afghanistan must stabilize the country while accounting for our exit. But a very significant hurdle stands in the way: the notorious weakness of Afghanistan's police and military. Of the troop-level plans Obama has reportedly considered, even the smallest emphasizes training and assistance for Afghan forces. After all, for us to leave, Afghan institutions must be able to replace the 100,000 foreign troops currently providing security. This makes building a massive, national Afghan military one of our top priorities in the region. Critics of this plan say the Afghan military is hopelessly disorganized, ill-equipped and corrupt. Supporters say it's crucial to our success. But there may be another way.

Bolstering the Afghan military carries significant risks. Given how illegitimate Afghan President Hamid Karzai's government is perceived to be by Afghans, a Karzai-led army would be poorly received and perhaps worsen anti-government sentiment. If a national Afghanistan army has a fraction of the national government's corruption, it could inspire disastrous backlash. Under Karzai's corrupt governance, the application of a national security force would wax and wane with political whims. With no personal stake in security outside Kabul, would Karzai really risk his resources and military strength to counter every threat or pacify every skirmish?

Afghanistan has not been a stable, unified state with a strong centralized government in three decades. The cultural and political institutions for a single national force may simply no longer exist. But Afghanistan, owing in part to necessity and in part to the tumultuous processes that have shaped the country, retains functional, if weak, security infrastructure at the provincial level. In the post-Soviet power vacuum and throughout periods of civil war, warlords arose to lead local militias. Many of them still remain in place--they were among our strongest allies in routing the Taliban's hold on the government--and have settled into more stationary roles somewhere between warlord and governor. Local rule has become the Afghan way. Local leaders who operate their own provincial forces, after all, stake their very lives on the security of their realm. By working with these leaders to establish and train local militias and police, rather than troubled and mistrusted national forces, the U.S. could find its route to Afghan stability and exit.

In parts of Afghanistan, strong provincial leadership has already developed security separate from national leadership. In the relatively peaceful and prosperous northern region of Mazar-E-Sharif, Governor Atta Mohammad Noor, himself a former warlord who fought against the Soviets and Taliban, commands authority rivaling that of President Karzai. Unlike Karzai, Noor is popular among his constituents and his province enjoys remarkable stability. The local military officials are loyal to him before Karzai, if they are loyal to Karzai at all. By promoting local governance and directing our military training and assistance to forces loyal to that governance, the U.S. could promote other strong provincial leaders like Noor.

Like Noor, many of these are likely to be former or current warlords. Warlords, despite their scary name, can be our strongest allies. They tend to be non-ideological and fervently anti-Taliban. Their fates are tied to the local populaces they govern. They're corrupt and tax heavily, but they provide real security and are trusted. Their ambitions are not for anti-Western war or fundamentalism, but sovereignty, security, and domination. None of these men is Thomas Jefferson, but in a country of many evil and exploitative forces, they are the best that Afghan civilians or American forces are likely to get.

Just as important, local security forces would better suit the region they protect, with more religious militias in the devout south and east but conventional police in the secular north. As General Stanley McChrystal, the top commander in Afghanistan, wrote in his much-discussed report calling for more troops, "Focusing on force or resource requirements misses the point entirely." He insisted that Afghans' "needs, identities and grievances vary from province to province and from valley to valley." A national security force would struggle to overcome the inevitable Goldilocks problem: Either it would be too secular for the south and east or too religious for the north but never just right. After all, the Taliban's initial support came in part from Afghans who desperately wanted religious rule. Though we may find the idea of supporting Islamic militias discomforting, forcing secular rule would risk another Taliban-like uprising. Better, perhaps, to establish local Islamic governance that is religious enough to satisfy the populace it serves but moderate enough to resist the Taliban.

The U.S. is already enacting a micro variant of this strategy by hiring and arming locals to provide security. The informal militiamen must come from within 50 km of their deployment site, which in addition to providing local jobs (Afghanistan's unemployment rate is a catastrophic 40%) also deters insurgents, who would be less likely to attack a familiar neighbor than a foreign invader. The principles that make this so effective would also apply to a larger, standing provincial force.

This does not preclude a national government with its own separate, standing force in the style of the national guard. Karzai's government could function much like a miniature European Union, setting economic and social policies while facilitating interactions between the provincial leaders. An economically centralized Afghanistan would in fact be crucial in this case so that provincial leaders remain dependent on Karzai for funding.

It may be tempting to point to Iraq as a model for putting stock in national security forces. After all, the strong roles of Iraqi military and police were crucial to stabilizing the country and phasing out American control over the past two years. But modern Iraq has never lacked the traditions or institutions for national security. If anything, Iraq under Saddam Hussein had one of the world's strictest and most oppressive regimes since the fall of the Soviet Union. Saddam's Iraq was in many ways a polar opposite from the chaos of frontier Afghanistan. Any rebuilt security in Iraq has been a matter of replacing one national security system with another. In Afghanistan, there is none to be replaced.

Of the many problems likely holding up President Obama's decision on Afghanistan, the public contradiction between two of his top officials is likely high on the list. General McChrystal famously warned of "mission failure" without an additional 40,000 troops. More recently, U.S. Ambassador to Afghanistan Karl Eikenberry, a general who previously held McChrystal's command, cautioned in two leaked cables against bolstering the notoriously corrupt Karzai. Their requests are not mutually exclusive. Working with provincial leaders to establish local security forces could meet McChrystal's security priorities while getting around Eikenberry's concerns about Karzai. Most importantly, it would meet Obama's goals of stability in Afghanistan with a foreseeable exit strategy.