Instead of prolonging the war in Afghanistan, US and NATO must tackle the root of the problem – in Pakistan.
NATO leaders committed to continuing support for Afghan security forces until 2020 at the summit in Warsaw over the weekend. In addition, last week President Barack Obama confirmed that US troop levels would remain more or less constant until the end of his presidency.
It is an important reversal of his planned drawdown to 5,500, which sends a message to the Taliban that the United States is not about to walk away. This is welcome relief for President Ashraf Ghani, who is encircled by profound security, economic and political threats.
Despite these chronic pressures Ghani continues to push for an ambitious reform agenda, and deserves to be given this breathing space by the Afghan government’s international military allies.
That said, US and NATO support should not be unconditional, and should be directed at getting Afghan security forces to better protect their civilians.
They should do this not just for moral and legal reasons, but because a failure to protect civilians can have profound strategic implications, as I documented in a the recent The Strategic Costs of Civilian Harm report published by Open Society Foundation with former Pentagon adviser, Christopher Kolenda.
Afghan forces have performed better than many expected, and the transition from international forces, with their air supremacy and precision weapons, to more traditional ground engagements was always going to pose a greater threat to the civilian population.
But civilian harm from Afghan security forces has jumped sharply, up 70 percent in the first quarter of 2016 compared with 2015, according to United Nations data.
There is also a strain of thinking within the security forces that sees the Taliban as an existential threat whose defeat justifies any means necessary.
Everyone remembers the shock of Kunduz falling to the Taliban last September, which led to the devastating US air strikes on the Doctors Without Borders hospital. But those air strikes resulted in less than a tenth of the total civilian casualties in the Kunduz attack and counterattack.
Of the 896 civilian casualties recorded by the UN in Kunduz, the vast majority resulted from ground engagements between the Taliban and government forces.
Although the UN was unable to attribute responsibility for most of these deaths and injuries, we can assume that a large share were caused by the pro-government forces, since in 2015 they caused 30 percent of civilian casualties from ground engagements, compared with the Taliban’s share of 25 percent (the rest was unattributed).
In Kunduz, both sides were implicated in grave crimes including the mutilation and desecration of bodies of the opposing side, as well as the use of low-precision weapons in an urban environment. This contributed to making Kunduz a terrifying experience for civilians who fled the city in their thousands, also contributing to the surge in migration to Europe.
Afghan leaders take civilian protection seriously, as witnessed by numerous statements by the president and other senior leaders. But there is also a strain of thinking within the security forces that sees the Taliban as an existential threat whose defeat justifies any means necessary.
Recent history shows the folly of this thinking: the Afghan government would be facing a far less grave security situation today had the US and NATO not caused so much civilian harm prior to transition, as we heard from many senior US and Afghan officials and experts in our research.
This harm was caused in a number of ways, including entrenching abusive local commanders for short-term security gains, or being hoodwinked into targeting civilians accused of being Taliban by their rivals, as well as civilian casualties in air strikes and detention operations.
Despite the very real short-term threat from the Taliban, the Afghan security forces need to recognise that restraint is a strategic necessary in the mid to long-term. In practical terms this means that the Afghan National Security Council should publish its national policy on the protection of civilians, which was promised months ago.
Second, the Afghan government needs to put more into accurate reporting of civilian harm, and analysing mitigation efforts.
The Afghans have a civilian casualty tracking unit, but it is currently hugely underestimating the harm caused by government forces – they estimate responsibility for 200 civilian casualties in 2015 compared with the UN’s estimate of 1,200.
This is a result of poor reporting channels, low literacy rates, and a culture that doesn’t encourage transparency. International mentors could boost the capacity of the tracking unit and result in real improvements. Government investigations into major civilian harm incidents should be professionalised, instead of ad hoc and often highly political delegations being handpicked after each emergency.
It is important for NATO leaders to keep faith with Afghanistan to the extent that their support helps maintain the gains of the past 15 years. But that needs to be tempered with unflinching attention to the protection of civilians. The Afghan National Defence and Security Forces still have a great deal of national support. This can only be maintained if they are seen as their people’s protectors, and don’t sink to the level of the Taliban.
Rachel Reid is the advocacy manager for Middle East, North Africa, and Southwest Asia at Open Society Foundations.
The views expressed in this article are the author’s own and do not necessarily reflect Al Jazeera’s editorial policy.