America Must Give the South to the Taliban

In spite of the commitments made at Tuesday's conference on the future of Afghanistan in Kabul, the current US counter-insurgency strategy (Coin) is likely to fail. The Taliban cannot be sufficiently weakened in Pashtun Afghanistan to coerce it to the negotiating table. America cannot win over sufficient numbers of the Afghan Pashtun on whom Coin depends. President Hamid Karzai's deeply corrupt government shows no signs of improvement. The Afghanistan army cannot stand up to the Taliban for many years, if ever. Pakistan's military continues to support its Afghan Taliban proxies. And the long-term Coin strategy and the far shorter US political timeline are incompatible.

President Barack Obama has promised to review the administration's Afghanistan policy in December. After this review the US should stop talking about exit strategies, and accept that the Taliban will inevitably control most of the Pashtun south. Instead Washington should move to ensure that north and west Afghanistan do not fall too, using for many years to come US air power and special forces - some 40,000-50,000 troops - along with the Afghan army and the help of like-minded nations. Such a de facto partition would be a profoundly disappointing outcome to America's 10 years in Afghanistan. But, regrettably, it is now the best that can realistically and responsibly be achieved.

This week media reports suggested another approach gaining favour: negotiation. But as CIA director Leon Panetta said recently about Taliban behaviour, why would they negotiate in good faith, if they think they are winning? Some instead think the US should withdraw all of its military forces over the next year. But that would be a major strategic defeat for the US and its partners, with negative global repercussions for many years to come.

Equally wrong-headed are those arguing the US should stay the course, no matter how long it takes. The CIA now thinks there are barely 50-100 al-Qaeda fighters left in Afghanistan, facing 100,000 US troops. The original Afghan objective was to destroy al-Qaeda, not fight the Taliban. That has largely been accomplished.

Even if the Afghan Taliban invited al-Qaeda to join them in greater numbers, the estimated 300 or so al-Qaeda fighters in Pakistan moving across the border would not substantially increase the threat. Is it worth an indefinite ground war, and thousands more US and allied casualties, to try to prevent that happening? The US can attack al-Qaeda on both sides of the border in any case.

Others worry the Taliban would not adhere to the rough boundaries of such a de facto partition, and would seek to reconquer the entire country. But US and allied military might and growing Afghan army capabilities could stop that from happening. Indeed, without such a long-term US military presence, a renewed civil war is probable. With such a commitment, it is unlikely. Small islands of non-Pashtuns in the south and east would be an unfortunate but unavoidable consequence, as would the comprehensive violation of women's rights in Taliban territory. But the US could still assist those Pashtun tribal forces that wish to resist the Taliban.

Wider threats to the region should be taken seriously. An irredentist "Pashtunistan", and perhaps the fracturing of Pakistan, could happen. Ironically, the Pakistan military is making such a development more likely through its support for the Afghan Taliban. But why should the US be more concerned about the territorial integrity of Pakistan than the country's General Ashfaq Kayani and his colleagues? Indeed, the spectre of de facto partition in Afghanistan might even produce the change of heart in the Pakistani military's attitude to the Afghan Taliban that successive US administration have failed to achieve.

Henry Kissinger has observed that: "For other nations, utopia is a blessed past never to be recovered; for Americans it is just beyond the horizon." With its many flaws, de facto partition is hardly a utopian outcome in Afghanistan. The overriding virtue of this concept is only that it is better than all available alternatives.

The writer was US ambassador to India, and a deputy national security adviser under George W. Bush