President Hamid Karzai’s visit provides an opportunity to repair the badly strained relationship between Kabul and Washington. It also provides an equally important chance to redefine how we deal with Afghans in general. Just as the Obama administration has sought to restructure Washington’s relationship with Pakistan away from wheeling and dealing into a strategic partnership, so it needs to move away from the transactional relationship with the Afghans.

Rebuilding cooperation with Afghanistan’s leadership is an important start. Much of the counterinsurgency and development efforts are dependent on the quality of governance in Afghanistan and international partnership with Afghan government officials. Bypassing already-antagonized and underperforming Kabul only gets us so far. “Going local” by relying on local officials for actions that would be problematic for Kabul sometimes works. But often, the local politicians and officials are just as corrupt, incompetent, and parochially focused as their superiors in Kabul. Moreover, a host of policies – from water-use allocation among provinces for rural development to police reform to setting parameters for negotiations with the Taliban  — need to be handled at the central level.

Frustrated with poor governance in Afghanistan, the Obama administration adopted a tough-love approach to Kabul. But instead of inducing Kabul to improve governance and crack down on corruption, the public confrontations with U.S. officials only persuaded President Karzai that Washington was trying to undermine him. Thus, the presidential elections ended up deeply flawed and corrupt, police clean-up has not been forthcoming, and corrupt and predatory government officials have acted with impunity. Meanwhile, President Karzai has increasingly been blaiming foreigners for many of Afghanistan’s problems. At the same time, he has been actively courting new friends, such as Iran and Russia, to balance pressure from NATO.

There have been some indications of improvements in Kabul’s governance over the past few weeks. Washington should seize on them to reassure President Karzai that it is willing to work with him on friendly terms. Meanwhile, Washington needs to quietly lay down some redlines about behavior that would be absolutely intolerable to Washington, such as territorial concessions to the Taliban during the upcoming reconciliation jirga or egregious corruption and tribal discrimination.

Washington and NATO also need to move away from a transactional relationship in everyday dealing with Afghans to a steadfastly principled, if interest-based relationship. For example, economic aid and development should not be offered as a payoff for actions against the insurgency, but because economic growth is to the benefit of Afghanistan. Instead of rolling into a village and upon request building a well and then demanding intelligence on the Taliban, NATO soldiers and civilians should only provide economic aid for the aspect of projects the Afghans cannot accomplish themselves and only if there is a community consensus about the project. And we should not expect intelligence in exchange, at least not right away. Trust takes time to build, and the Taliban are skilled at intimidation. Treating aid as a principled and long-term approach, rather than an immediate buy-off is of course very hard in an insurgency setting where lives are at risk, political expediency runs high, distrust and frustration are rampant, and the withdrawal clock is ticking.

Structuring the relationship as a buyoff is bound not to work. All too often, the internationals have been following the old adage that one cannot buy an Afghan, but one can rent one. The problem has been precisely the core message of the proverb: Afghan loyalties as a result have been fickle, and compliance often minimal, with policies carried out as little as possible. Other bidders on the block often outbid us – be they problematic powerbrokers, the Taliban, or regional powers. Too frequently, Afghans see the farangis (foreigners) as marks out of whom to con money.

Thus, for example, we should not be paying for the standing up of militias. Either local antagonism toward the Taliban will generate self-defense (and with plenty of small arms to draw upon), or our money will only generate lukewarm efforts that are likely to deteriorate into discriminatory and predatory behavior toward fellow Afghans. Similarly, we should not shower a community with aid simply because it has now signed up with us, but because we believe that the purpose of Afghan government and international presence is to improve the lives of the local people. We should not structure reconciliation around payoffs, including preferentially offering jobs, to the Taliban men who demobilize just because they demobilized. As elsewhere in the world – such as Colombia or Nigeria — such payment-induced demobilization is not effective reconciliation, and it is likely to wither once the payoffs and jobs run out. We should not be paying villages to stand up local shuras; they should do so because they take ownership of their lives, even while we assist them.

Ultimately, breaking out of the unproductive buyoff system requires changing the time-horizons of the Afghans. As long as they see the future as bleak and the existing order about to worsen when the United States starts withdrawing its forces in July 2011, they will get sucked into short-term power and profit maximization. The more narrowly we define our goals in Afghanistan, to be, for example, the absence of al Qaeda, the harder it is to change Afghan time-horizons. And we’ll be ineffectively stuck with trying to rent the Afghans for a long time.