After arm-twisting in the eleventh hour, Afghan President Hamid Karzai has accepted a runoff election with his rival, Abdullah Abdullah. This is a face-saving move for both Karzai and the international community, but a runoff is unlikely to address the reality that many Afghans see Kabul as part of the problem. What happens the day after the election will be more important. It will be the last opportunity for the United States and the new Afghan government to define a political strategy to win the trust of the people and to buy precious time in the face of growing Western discontent.

The Taliban will soon start their winter campaign of political consolidation and intimidation. It is urgent that the United States and its NATO allies match that with a political surge of their own. Progress can be made on two fronts: by launching a massive Afghan civilian surge through the regional capitals and municipalities, and by announcing a two-year security transition plan whereby indigenous security forces will take over security tasks by 2012, starting with symbolic and stable areas like Kabul and the West and moving to more complex ones.

The August elections marked the ugly but necessary death of the 2001 Bonn process; when international and Afghan leaders first met after the ousting of the Taliban to draw Afghanistan's new political contours. The spirit of Bonn was based on short-term political expediency, mistakenly included warlords and closed the door to elements of the Taliban, and never summoned enough investment from the international community. Unsurprisingly, the fragmentation and privatization of state power has only accelerated since 2001, thanks to a mix of greed, opium money, and a lack of meaningful international oversight.

The power brokers in Kabul have sold governorships and police and judicial appointments, sometimes for hundreds of thousands of dollars, and used them to further their own private interests in the provinces. According to a group of independent Afghan electoral observers, the Aug. 20 fraud was orchestrated by the very local clients who owed favor or positions to the main candidates, especially those who have been in government. The result of years of neglect and impunity has been a governance vacuum in which Afghans across the country have been left stranded between local power brokers and warlords, the Taliban, and the international community. What little most Afghans see of local government -- mostly through their experiences of province and district governors and police chiefs -- convinces them that it is a presence to be feared, avoided, or bribed. The massive electoral fraud has only reinforced popular disillusionment with Kabul and the West.

U.S. President Barack Obama is right to think political strategy before resources. For too long, the West has thrown troops and money at Afghanistan, without any clearly articulated objectives for the mission. While NATO has fought already-lost battles against the Taliban, or against the opium business, no one has taken charge of building effective and legitimate institutions of Afghan government. The one international organization that should do just that -- the U.N. Assistance Mission in Afghanistan -- has seen its operations significantly shrunk over the last two years and has emerged from the elections divided and discredited. Meanwhile, the European Union, which often boasts of its soft power and crisis management tools, is still struggling to deliver the 400 police trainers it committed to deliver years ago.

But so far the debate in Washington has focused too heavily on the assessment put forward by the NATO International Security Assistance Force commander, U.S. Gen. Stanley McChrystal. The real story is what the report only indirectly alludes to and what has been seldom debated until the electoral crisis: the Afghan political "context" in which Afghans will be given reasons to bet on their government rather than sit on the fence or support the Taliban. If McChrystal's counterinsurgency proposition proposes an ambitious if not artificial approach to win the trust of local communities, in the end the one institution that should and can durably win Afghan hearts and minds is an Afghan government.

An Afghan political surge must take place simultaneously at top and bottom levels, combining formal and informal institutions. First, a greater share of international aid money needs to be channeled to the government's coffers by strengthening the system of trust funds co-administered by U.N. agencies. Under the current system, 80 percent remains in international hands, where each dollar spent comes with a $2.50 overhead. A recent internal survey by the Ministry of Rural Development shows, against received wisdom, that aid money spent by Afghan communities in Helmand is more cost-efficient and less wasteful -- even taking into account Afghanistan's well-known corruption problems.

The government should send administrators from basic services ministries such as rural development, agriculture, and Justice to major cities and municipalities. Kabul has failed to do so, consumed by its politics of survival, but also for lack of serious incentives from the international community, which too often has substituted itself for an Afghan presence on the ground, and for lack of capacity. To address the latter, European governments should commit to fund a new Afghan civilian academy, which, if created this winter, should be able in two years time to send its first graduates to local institutions. District Development Assemblies, which are community-based consultative bodies for development projects, are present in almost all 365 districts in Afghanistan and should be beefed up to something more like mini-administrations with programs rather than just projects. Informal institutions like community-based jirgas and shuras, which resolve most local disputes, should be given legal status, helping to make the judiciary more responsive to the Afghan people's demands.

On the security front, training of the Afghan National Security Forces has to balance the need for numbers with an imperative for professionalization. Today, less than half of the 95,000-strong Afghan National Army (ANA) is fit for combat operations. But provided the NATO training mission receives sufficient support -- and this will mean a generational commitment from the United States and its partners -- this number can reach 100,000 combat-ready troops in the next five years, according to senior NATO officials.

NATO and the Afghan government should devise a 2012 security transition plan. Within two years, the ANA could start assuming security responsibility in the most stable provinces in the north, center, and west of the country. This will send a strong message and buy much-needed lenience for Western governments in the court of public opinion.

Simultaneously, NATO's Provincial Reconstruction Teams must begin shifting their resources to Afghan institutions, and especially to the Afghan provincial and district governments. Within fewer than five years, Afghan civilian and security institutions should be able to play a frontal role in sustaining security and development, while NATO troops will be freed to focus more on a combat mentoring and training role, and the United States will maintain bases and quick-intervention  forces.

There will be setbacks and delays, and U.S. soldiers are likely to pay a heavy price in the next few years. And there will be hard choices to be made on where NATO and Afghan forces should focus their efforts. Today, under what can be called the "Helmand syndrome," more than 50 percent of NATO combat forces fight for territories with less than 15 percent of the population. In contrast, the most dynamic economic and population centers in the north and the west are grossly neglected.

If the new Afghan government is to earn public support, and NATO is to buy a way out of Afghanistan, a civilian surge will be vital. This will come at a cost, and it carries some real risks. But pouring more lives and more money into Afghanistan, with no clear Afghan strategy and no end in sight, will only hasten a return to chaos.