Resident Scholar Frederick W. Kagan

President Obama has said many times that America must succeed in Afghanistan. He is right, and he deserves our full support in that effort.

Afghanistan is in many respects harder to understand than Iraq was. Even with a good strategy and sufficient resources, success will almost certainly come much more slowly. But as a great man said two years ago, hard is not hopeless.

The keys to finding the right approach lie in nine fundamental principles.

1. UNDERSTAND WHY WE'RE THERE
Afghanistan is not now a sanctuary for al-Qaeda, but it would likely become one again if we abandoned it. Mullah Omar, the head of the Taliban government we removed in 2001, is alive and well in Pakistan. He maintains contacts with Osama bin Laden, Ayman al-Zawahiri, and the other key al-Qaeda leaders, who are also based in Pakistan (although in a different area). Mullah Omar supports Taliban fighters in southern Afghanistan from his Pakistani havens, while al-Qaeda and its affiliates support insurgents in eastern Afghanistan. Allowing Afghanistan to fail would mean allowing these determined enemies of the United States to regain the freedom they had before 9/11.

Pakistan itself is another reason Afghanistan is vitally important to America. It's a country with 170 million people, nuclear weapons, and numerous terrorist groups. As long as Afghanistan is unstable, Pakistan will be unable to bring order to its own tribal areas, where many terrorist sanctuaries persist. It will also be distracted from addressing the more fundamental problems of Islamic radicalism that threaten its very survival as a state. Further, Afghan instability makes the U.S. dependent on Pakistan logistically--there is no way to replace completely the land route from Karachi with another route through Central Asia. This dependence in turn reduces our ability to influence Islamabad on other matters of great importance, such as stabilizing civilian rule in Pakistan and stopping support for terrorist groups like the one that attacked Bombay.

2. KNOW WHAT WE HAVE TO ACHIEVE
Success in Afghanistan does not require creating a paradise in one of the poorest countries on earth, but we cannot define victory down. Preventing Afghanistan from becoming a safe haven for terrorists again, helping Pakistan fight its own terrorist problems, and liberating ourselves from dependence on Pakistan will require building an Afghan state with a representative government.

Afghanistan has a longer tradition of such political organization than Iraq has. It has been independent since 1747, and had a functioning constitutional and parliamentary monarchy in the middle of the 20th century. Centrifugal forces in Afghanistan have always been powerful, making the prospects for a strong centralized government in Kabul poor, but the country is neither ungovernable nor artificial. It cannot be stable at this point in history, however, without a representative system. Its multiethnic makeup and decades of internal war mean that any attempt to impose a strongman or to break the country up into effectively independent, warlord-ruled fiefdoms will lead to perpetual violence.

3. UNDERSTAND OUR ENEMIES AND FRIENDS
There is no such thing as "the Taliban" today. Many different groups with different leaders and aims call themselves "Taliban," and many more are called "Taliban" by their enemies. In addition to Mullah Omar's Taliban based in Pakistan and indigenous Taliban forces in Afghanistan, there is an indigenous Pakistani Taliban controlled by Baitullah Mehsud (this group is thought to have been responsible for assassinating Benazir Bhutto). Both are linked with al-Qaeda, and both are dangerous and determined. In other areas, however, "Taliban" groups are primarily disaffected tribesmen who find it more convenient to get help from the Taliban than from other sources.

In general terms, any group that calls itself "Taliban" is identifying itself as against the government in Kabul, the U.S., and U.S. allies. Our job is to understand which groups are truly dangerous, which are irreconcilable with our goals for Afghanistan--and which can be fractured or persuaded to rejoin the Afghan polity. We can't fight them all, and we can't negotiate with them all. Dropping the term "Taliban" and referring to specific groups instead would be a good way to start understanding who is really causing problems.

Recognizing the limitations of the current government is a good next step. That government is ineffective and deeply corrupt. Provincial governors and district leaders were not elected, but appointed by Pres. Hamid Karzai, often with an eye toward marginalizing potential rivals and consolidating his power. Karzai's popularity is dwindling, and the postponement of Afghanistan's presidential elections from May to August allows his opponents to paint him as illegitimate. It is possible that even if Karzai wins the August election, many Afghans will continue to view him as illegitimate.

The U.S. cannot, however, turn away from the central government and seek solutions only at the local level. For one thing, important local leaders are Karzai's appointees. For another, building local solutions that do not connect with the central government is the path toward renewed warlordism and instability. The key, therefore, is to develop local solutions that are connected to the central government but not necessarily completely controlled by it.

Local governments--possibly at the level of individual villages--will have to play a role in selecting individuals to help maintain security once it has been established. Afghan villages often have representative bodies, or at least local elders who can identify needs and priorities while balancing tribal concerns. Local and provincial governments connected to Kabul will have to provide weapons and compensation to local security forces and will therefore acquire a certain limited control over them.

Similar approaches are likely to be required on the economic front--local groups and leaders, in some cases supported initially with funding from the U.S. Commander's Emergency Response Program, can get economic projects going, but they will have to connect those projects to central-government representatives for long-term funding and integration into regional and national economic systems. The bottom line is that we must work hard to develop local solutions to local problems, but always with the goal of integrating those solutions into a loose but real central support-and-control system.

4. COMMIT TO THE EFFORT
The consistent unwillingness of the U.S. government to commit to the success of its endeavors in Afghanistan (and Iraq) over the long term is a serious obstacle to progress. The Pakistani leadership appears convinced that America will abandon its efforts in South Asia sooner rather than later, and this conviction fuels Pakistan's determination to retain support for (and therefore control of) Afghan Taliban groups based in its territory. It also contributes to instability within Pakistan, because Pakistani leaders are tentative about committing to the fight against their internal foes as long as they are unsure of our determination to do our part.

At the local level within Afghanistan, people who are not convinced that coalition forces will stay to support them if they oppose the terrorists are unlikely to risk retaliation by committing to us. When U.S. forces moved into insurgent strongholds in Iraq in 2007, the first thing they were asked was: "Are you going to stay this time?" When the answer was yes (and we proved it by really staying and living among them), the floodgates of local opposition to the insurgents opened. The people of Afghanistan need the same reassurance. Until it is widely believe that the U.S. will remain in the fight until the insurgency is defeated, doubt about our commitment will continue to fuel the insurgency. If we are going to fight this war, as our interests require, we must make it clear that we will do what it takes to win.

Our history is very much against us in this effort. Islamists point to our retreat following the Marine-barracks bombing in Lebanon in 1983, the "Blackhawk Down" incident in 1993, our abandonment of Afghanistan following the defeat of the Soviet Union in 1989, and our abandonment of Shiite and Kurdish Iraqis to Saddam Hussein's retribution in 1991 and 1992. At the end of 2006, our enemies in Iraq were already declaring victory, convinced that the pattern would repeat itself. The question they are now asking is: Was the surge an aberration in U.S. policy or a new pattern?

Our friends have the same question. We are asking them to put their lives on the line in support of shared goals, and they need to know we will stand by them. More rides on the outcome of our effort in Afghanistan than the particular interests we have there. American security would benefit greatly if we changed the global perception that the U.S. does not have the stomach to finish what it starts.

5. LEARN AND ADAPT THE RIGHT LESSONS
We cannot dismiss our extensive and painful experiences in Iraq, but we must recognize the differences between that country and Afghanistan.

Perhaps the most important lesson of Iraq that is transportable to Afghanistan is this: It is impossible to conduct effective counterterrorism operations (i.e., targeting terrorist networks with precise attacks on key leadership nodes) in a fragile state without conducting effective counterinsurgency operations (i.e., protecting the population and using economic and political programs to build support for the government and resistance to insurgents and terrorists). We will never have a better scenario in which to test the limitations of the counterterrorism model than we had in Iraq in 2006. U.S. Special Forces teams had complete freedom to act against al-Qaeda in Iraq, supported by around 150,000 regular U.S. troops, Iraqi military and police forces of several hundred thousand, and liberal airpower. We killed scores of key terrorist leaders, including the head of al-Qaeda in Iraq, Abu Musab al Zarqawi, in June 2006. But terrorist strength, violence, and control only increased over the course of that year. It was not until units already on the ground applied a new approach--a counterinsurgency approach--and received reinforcements that we were able to defeat al-Qaeda in Iraq (even without killing its new leader).

In Afghanistan, we have nothing like the freedom of movement we had in Iraq in 2006, and nothing like the force levels. We have, furthermore, been targeting leadership nodes within terrorist networks in Afghanistan and Pakistan for seven years now, yet the groups are not defeated. Absent a counterinsurgency and nation-building strategy that leads the population to reject the terrorists, killing bad guys will not defeat well-organized and determined terrorist networks.

Enthusiasm has been growing for some time over the idea of generating "awakenings" in Afghanistan similar to the Anbar Awakening that helped turn the tide in Iraq in 2007. Conceptually, this enthusiasm is sound. As noted above, success will require developing local solutions that are integrated in some way with the central government--the most abstract rendering of the "awakening" phenomenon in Iraq.

But we must be very careful about trying to apply Iraq "lessons" of greater specificity. For one thing, what happened in Iraq was not a single phenomenon. The Sunni-Arab rejection of al-Qaeda and turn to the coalition consisted of myriad local developments rather than being a coordinated movement. The coalition response to and support of those local developments was coordinated--we coined the term "Sons of Iraq" and treated SOIs as though they were a coherent group for certain funding and bureaucratic purposes--but each group remained independent. The SOIs never developed a corporate identity, and the local movements transformed their local political contexts rather than evolving into a country-wide movement.

The same will be true in Afghanistan. Local groups in Konar will not identify with local groups in Helmand, nor should they. There is no "Sons of Afghanistan" program that can be centrally defined and directed during its formation. As in Iraq, we must allow and encourage local movements to grow organically--in accordance with local conditions and traditions, but moderated by Afghan and coalition forces that understand the local area. It should go without saying that any effort to develop local security forces in areas that have not been cleared of insurgents will fail, either exposing the locals to vicious retribution or helping the insurgents co-opt new fighters.

6. CONSIDER THE HUMAN TERRAIN
Pashtuns are not Arabs. They have different traditions, different tribal structures, different ways of resolving differences. One of the most important (and least remarked-upon) differences is that Iraqis fight in their cities and villages while Pashtuns, on the whole, do not.

Saddam Hussein planned his defenses against U.S. attack with the intention of drawing us into urban fights he thought we would fear. Indigenous Iraqi insurgents dug into villages and cities and blended into the population. So did the external terror groups.

Coalition forces fought their way through Iraqi cities and villages, sometimes doing fearful damage to the cities and local populations. We devastated Fallujah and Ramadi, for example. But local grievances did not focus on the collateral damage. Considering the scale of the destruction, Iraqi complaints about it were very mild. In 2007, victorious coalition troops who had fought their way through insurgent and terrorist sanctuaries in Baghdad were more popular at the end of the fight than at the beginning. Iraqis generally recognize that their wars are fought in their cities, horrible though that is, so they have a fairly high tolerance for collateral damage and even for the presence of foreign forces in their urban areas and villages. They are generally more interested in who is going to win.

Pashtuns don't work that way. The Soviets invaded Afghanistan at the end of 1979 and quickly occupied all of the major urban areas. The insurgents, for the most part, did not contest that occupation. They focused instead on cutting off communications between the cities, on ambushing Soviet troops moving outside urban areas and villages, and on attacking isolated Soviet outposts. The Soviets did not know how to respond--they had no context for thinking about a rural insurgency. They had fought the Second World War city by city, and had suppressed rebellions in their Eastern European satellites by fighting through their capitals. They tried to subdue the Pashtuns with ferocious and indiscriminate bombing of Afghan villages, generating 5 million refugees and strengthening the resistance rather than breaking it.

Today's situation is similar. The major urban centers are not insurgent sanctuaries, and most insurgent attacks occur not only beyond the city limits but outside of the villages as well. American troops accustomed to setting up positions within Iraqi cities and towns may find that the same procedures in Afghanistan incense the population rather than reassure it. That does not mean the problem lies with our overall "footprint" in Afghanistan, but rather that we should rethink where to put our feet. We must also remember that Afghan tolerance for attacks within villages and cities is much lower than Iraqi tolerance, which is why complaints about collateral damage in Afghanistan are much louder than Iraqi complaints were, even though the damage is milder.

Understanding this principle is vital, because if we misinterpret the nature of the "footprint" problem we might come to the erroneous conclusion that success requires fewer forces rather than more--or, as some senior leaders are increasingly suggesting, that our presence is the problem. In fact, to solve the problems in Afghanistan we must have a deep understanding of local dynamics in many different areas. In the current security environment, only American and allied military forces can understand those dynamics, and they can do so only by living among the people in a way that is mutually acceptable to our forces and the Afghans. Pulling back to bases may reduce local resentment of us, but it will also deprive us of any ability to interact with Afghans and their leaders at the level necessary for success. As General Petraeus is fond of saying, you can't kill your way out of an insurgency. Neither can you defeat one long-distance. Success in Iraq required finding the right way to deploy American forces among the Iraqi population. Success in Afghanistan will require finding the right way for Afghanistan, which will almost certainly be different from the right way in Iraq.

7. UNDERSTAND WHAT WE MUST DO, CAN DO, AND CAN'T DO
The Afghan National Army consists of perhaps 70,000 troops (on paper). This number will rise gradually to 134,000--itself an arbitrary sum, based on assumptions about what the fifth-poorest country in the world can afford to pay for an army that is certainly too small to establish and maintain security. The Afghan National Police are ineffective when not actively part of the problem. Afghanistan is significantly larger than Iraq, its terrain is far more daunting, and its population is greater. The Iraqi Security Forces that defeated the insurgency (with our help) in 2007 and 2008 numbered over 500,000 by the end. There is simply no way that Afghan Security Forces can defeat the insurgents on their own, with or without large numbers of coalition advisers.

Breaking the insurgency will have to be a real team effort. Coalition units must partner with Afghan army units to clear critical areas, and then work with local leaders to develop local security solutions that smaller numbers of residual U.S. and Afghan troops can support while other areas are cleared.

It is better, in general, for Afghans to take the lead in moving into or through Afghan towns, but this is not always as desirable as we might think. In many regions, Afghan villagers are highly localized. Iraqis were accustomed to traveling across their country, maintained active links with and made frequent visits to relatives in various regions, and were willing to see the Iraqi army as their army even when its units were drawn from other parts of the country. Many rural Afghans are not nearly as mobile, particularly after decades of fighting in which the insurgents worked studiously to disrupt communications. In some areas, any outside forces--even Afghan forces--are seen simply as outsiders.

We can observe this phenomenon clearly in Pakistan today, as Pakistani soldiers (largely Punjabis) move into Pashtun areas and are attacked as foreigners. It is not remotely in our interest to generate a similar situation in Afghanistan. We must also remember an important lesson from our efforts to transition security responsibilities prematurely in Iraq in 2005 and 2006: It does not matter much if the local population resents us; it does matter if they resent and mistrust their own security forces. Some counterinsurgency operations are better conducted by outside forces simply because the resentment they generate will leave with them rather than stick to the indigenous government.

8. HAVE A GOOD PLAN
Adding more troops to a failing strategy rarely works. Current military and political leaders recognize this, which is why reviews are underway in CENTCOM, the Joint Staff, and the White House to develop a new strategy for Afghanistan. At the end of the day, however, the detailed campaign plan for implementing a new strategy has to come from the commander in the theater. That commander, Gen. David McKiernan, suffers from a number of significant handicaps that Generals Petraeus and Odierno did not face in Iraq in 2007.

Developing a detailed campaign plan requires a large military staff. Coordinating the use of force with political, economic, and social projects also requires a large staff, on both the military side and the civilian side. In Iraq in 2007, General Petraeus had a large staff (Multinational Force–Iraq). He had a terrific civilian partner in Amb. Ryan Crocker, who headed the largest U.S. embassy in the world and had the power to coordinate most of the non-military efforts in Iraq. Petraeus also had the support of Lt. Gen. Raymond Odierno and the large and excellent staff of the III Corps. Odierno and his staff did most of the work developing the military plans to defeat the insurgents, working through five division-level (two-star) commands and as many as 22 combat brigades. Every part of that command structure was necessary to understanding the problem and developing plans to respond to it that were truly integrated at every level.

General McKiernan has no such resources. His staff is too small and is a hodgepodge of U.S. and allied officers whose main function, when the staff was formed, was the coordination of an allied reconstruction effort. The much larger number of allies in Afghanistan, and the fact that NATO took control of the operation in 2006, places an enormous burden on McKiernan and his staff that Petraeus did not face. There is no corps headquarters in Afghanistan, moreover--no equivalent to Odierno's III Corps and the staff that actually developed the war plan in Iraq. There are five subordinate headquarters (regional commands), but some have few troops and only one has the resources that the five division staffs in Iraq provided. Current plans may put as many as six U.S. brigades on the ground by the end of this year. The U.S. mission in Afghanistan has nothing like the authority that Ambassador Crocker had; on the contrary, the proliferation of allies and international aid efforts has frustrated attempts to unite them in a coherent civil-military campaign plan.

The situation in Afghanistan requires a significant augmentation of McKiernan's staff: the addition of a corps headquarters under him and at least one division headquarters in the south. It also requires a body that can coordinate international efforts and mesh them with military planning, either through the U.S. mission in Afghanistan or through the U.N.'s special envoy. Without such an increase in headquarters and planning capabilities, even the best work by our commanders can mitigate only a portion of the problems. The solutions that emerge will likely be suboptimal.

Before it departed, the Bush administration decided to send reinforcements to Afghanistan, and the new administration has supported that decision. Rightly so--Afghanistan needs more U.S. troops. But until a thorough and detailed joint campaign plan has been developed in the theater--with buy-in from the overall military commander, our allies, and the civilian organizations that will have to help execute it--it will not be possible to know exactly how many troops are needed, what exactly they should be doing, or what resources they will require. Developing such a plan and evaluating the resource requirements should be an urgent priority--more urgent even than getting more troops into the theater.

Developing a coherent plan for the entire country requires the involvement of our many allies. That involvement, in turn, requires coming to a common understanding of the situation, the tasks to be performed, and the challenges we face. When Afghanistan became a NATO mission, the presumption was that it was primarily a nation-building exercise. Many allied countries committed troops without intending to participate in counterinsurgency efforts. Although it is natural to complain about the national caveats that restrict some (but by no means all) allied troops from leaving their bases or fighting, we must recognize that many of our allies never signed up for this kind of war. They have therefore been reluctant to admit that we now face a full-fledged insurgency. The Obama administration and its newly appointed envoy, Amb. Richard Holbrooke, have a real opportunity for constructive diplomatic engagement here. It should be their priority to help our allies accept the reality in Afghanistan, at the same time making it clear that we do not expect them to engage in combat operations they never intended to undertake. As in Iraq, we should accept whatever contributions they are willing and able to make, but avoid allowing tensions over those contributions to distort the overall understanding of the fight.

9. PRIORITIZE EFFORTS
While the situation in Afghanistan is indeed deteriorating, it would be wrong to rush forces out of Iraq this year in response. Most important, as detailed above, we have not yet established the conditions in Afghanistan that would allow a surge to be decisive. Also, the theater cannot absorb too many reinforcements too quickly. The surge in Iraq brought U.S. troop levels up to something over 160,000 soldiers--about the same number we had had there at the end of 2005. By contrast, coalition force levels in Afghanistan are already at their highest levels. The logistical base that supports them is very sparse. In Iraq there was enough reserve logistical and infrastructure capacity to integrate five additional brigades and two battalions in the space of six months. Because similar resources are lacking, it would be much harder to accomplish such a feat in Afghanistan at this point.

It would also be wrong from the standpoint of U.S. global interests and grand strategy. The dramatic improvement in the situation in Iraq has already increased our options and flexibility--forces are moving from Iraq to Afghanistan this year without imposing unacceptable risks on our position in Iraq. General Odierno has identified 2009 as a critical year for Iraq, starting with the successful Iraqi provincial elections that just occurred and ending with the election of a new central government.

Maintaining American presence in Iraq in support of this effort is essential. Every estimate suggests that, if we maintain such a presence this year, the requirement for continued U.S. forces in Iraq after 2009 will drop dramatically. We can surge troops into Afghanistan, in other words, in 2010 without compromising success in Iraq, and after we have developed the command and logistical structures--and, above all, the plan--to support them in Afghanistan. Therefore, sound grand strategy means using 2009 to set the conditions for decisive operations in Afghanistan while ensuring that Iraq remains stable enough to permit dramatic force reductions.

The key problem with this approach is that Afghanistan must elect a new president this year, and many areas of the country are not secure enough for a legitimate election. Unfortunately, there is not much we can do to address this problem through troop redeployments. Two additional combat brigades are already on the way and will arrive in time to make a difference. Redirecting other combat brigades now meant for deployment to Iraq requires a good six months of advanced warning--among other things, the troops have to train for an entirely different climate, culture, and situation. Any additional brigades would therefore be arriving shortly before the elections. Considering that it takes a unit anywhere from 30 to 60 days on the ground to get deployed and gain enough situational awareness to develop reasonable plans and methods, it is already too late to get more troops to Afghanistan (at least in any prepared and orderly fashion) in time to make much of a difference to the elections.

The theater commander might be able to mitigate the problem to some extent by committing the theater reserve to help; our European allies might be able to help a little with a mini-surge of their own. But rushing out of Iraq now is far more likely to ensure that we are distracted by problems in Mesopotamia in 2010 than to turn the tide in South Asia.

PROLEGOMENON TO A PLAN FOR WINNING IN AFGHANISTAN
This essay does not provide a plan or a strategy for success in Afghanistan. It provides, rather, a set of guidelines for thinking about how to develop one, and for evaluating plans articulated by the administration, its generals, and outsiders. Ultimately, a plan for winning in Afghanistan has to be developed in Afghanistan, just as the plan for winning in Iraq was developed in Iraq. It is a truism that any plan must involve not only the U.S. and allied militaries, but all relevant civilian and international agencies, and must deeply involve the Afghans themselves at every level. Our military and civilian leaders understand that truism. We have failed to date in accomplishing the objective not because we haven't known that we must, but because it is very hard to do.

But hard is not hopeless in Afghanistan any more than it was in Iraq. The stakes are high, as they always are when America puts its brave young men and women in harm's way. President Obama has an opportunity in the difficult challenge he faces. So far, he appears determined to try to do the right thing. He deserves the active support and encouragement of every American in that attempt.

Frederick W. Kagan is a resident scholar at AEI.