An airman at the end of a transfer ceremony for two sergeants who died in Afghanistan. (Win McNamee/Getty Images)

Ronald E. Neumann was ambassador to Afghanistan from 2005 to 2007.

In theory, U.S. strategy in Afghanistan has been to train an Afghan army that can fight al-Qaeda, the Taliban, and now the Islamic State — and then largely to withdraw. After 16 years, it’s not surprising that many people think that strategy has failed. In fact, it hasn’t really been tried.

The Bush administration didn’t think such a force was needed. The Obama administration not only didn’t execute its own strategy, it also only grudgingly provided the resources necessary to avoid outright failure. The Trump administration has not increased those inadequate resources. In short, the strategy of trying to build an Afghan security structure that could successfully combat the Taliban only began six years ago and then was undercut after four years — scarcely enough time to build a whole new army.

When I left Afghanistan in 2007, our target of a 216,000-strong Afghan army and police force was unmet. This security force included no serious air force, artillery, logistics or medical services, since it was designed on the assumption that the war was largely over. Our advisory presence was small in the army and almost nonexistent in the police. Iraq was soaking up all available resources, and my warnings that the fighting was about to get worse were ignored. Only in fall 2009 did the Obama administration decide to build to the current goal of 352,000 security personnel, including essential supporting forces. That effort did not begin to receive funding and equipment until a year later.

These expanded goals were only really pursued for four years and were heavily undercut by policy shifts. Virtually every aspect of the training was rushed and under-resourced. The development of essential support functions, from logistics to artillery to air, was delayed for a year so that all available training facilities could be devoted to getting infantry into the battle. U.S. and NATO training teams never reached much over 50 percent of required personnel, and even that low level took several years to achieve. Advisory teams were delegated to the National Guard and Reserves rather than the regular U.S. Army, a clear indication that this was a lower priority effort. Proper advanced training for these teams took several years to set up.

Then, the decision to end U.S. involvement in active combat by 2014 converted a conditions-based strategy to one driven by Washington timelines. Withdrawal of advisers consistently outpaced the readiness of the Afghans to take over. By 2014, many of our forces were devoted to getting packed up and out of the country rather than to the fighting. But worse was yet to come.

Critically, Afghan security forces trained with U.S. and NATO forces, and thus learned to fight as we do — with air support. But from January 2014 to November 2016, that air support was withdrawn. The Obama administration declared that “we are no longer at war with the Taliban.” This nonsensical phrase, which I heard from senior officials at the National Security Council, left the Taliban free, except in the most extreme circumstances, to reinforce, maneuver and mass for attacks.

Only at the end of 2016 was the administration sufficiently shocked by the failure of its strategy to lift the prohibition on air support and end further withdrawals. But this left in place a greatly reduced U.S. and NATO force sufficient only to prevent immediate defeat. Since the Trump administration has largely maintained this inadequate force, it is puzzling that the White House wonders why it has not achieved a different result.

Our commanders may perhaps be faulted for not arguing harder against the political mistakes of Washington. I believe that they tried but found that the Obama White House resented every effort to speak truth to power. The Afghans can be seriously criticized for allowing politics to intrude much too far into senior military appointments (something that is being reversed by President Ashraf Ghani’s appointment of battle-tested commanders). Nonetheless, our advisory presence still does not cover every Afghan army corps, nor their subordinate brigades. Much of the rush to failure has been Washington-driven.

Even a great effort to correct the mistakes of the past would not produce results on the battlefield for a year or more. Whether this can or should be done deserves serious debate. But that discussion should be based on a clear understanding of what has — and what has not — already been tried in Afghanistan.