In a Washington Post op-ed on Friday, Charles J. Dunlap argued that reversing existing tactical limitations on international forces' airstrikes in Afghanistan would allow international forces to kill more insurgents, and thus help save Afghan and foreign military lives in Afghanistan. Dunlap's arguments are based on mistaken assumptions about Afghan attitudes toward the warring parties and the conflict. If military and policy leaders acted on those flawed assumptions as Dunlap suggests, by lifting airstrike restrictions, it would not only put more Afghan lives at risk, but would erode any progress that has been made in the last year. Instead of reversing tactical limitations, military and policy advisors should be considering whether they go far enough.

In July 2009, as part of the new counterinsurgency strategy to deny insurgents the Afghan population's support, then-International Security Assistance Force (ISAF) commander General Stanley McChrystal placed restrictions on tactics that risked high collateral damage, like airstrikes in populated areas, or those that provoked extreme offense.

Though it may be too soon to tell whether the changes will permanently reduce civilian harm or have the desired tactical effect of reducing support for insurgents, so far they seem to be working. The United Nations Assistance Mission in Afghanistan (UNAMA) reported a 29 percent decrease in civilian deaths caused by international and Afghan forces in the first half of 2010. A study by the National Bureau of Economic Research (NBER) (selectively cited by Dunlap to suggest that airstrikes are not that lethal) found that reduced civilian casualties in Afghanistan resulted in a decrease in insurgent attacks. The restrictions were re-affirmed when General David Petraeus took command in Afghanistan this year.

Despite this evidence, Dunlap argues that these restrictions should be reviewed. More bizarrely he argues that airstrikes can be used more frequently because Afghans actually do not care about civilian casualties -- a startling claim given the years of Afghan protests and recriminations on civilian casualties, and the warnings of military and humanitarian officials on the ground that have seen the visceral reaction of Afghans to civilian casualties. Ignoring this evidence, Dunlap instead points to the lack of Afghan public backlash against insurgent attacks. "Although insurgents caused almost 76 percent of the civilian deaths, according to a U.N. report published in August, Taliban strength is reportedly nevertheless increasing," he wrote. Thus, he concludes, civilian casualties are not as strategically important to maintaining Afghan support, and reducing the number of airstrikes was a mistake.

My organization, Open Society Foundations, recently conducted research across Afghanistan to gauge Afghan attitudes toward insurgent-caused civilian casualties, and the impact of the tactical restrictions on civilian attitudes. We found that, contrary to Dunlap's assertions, civilian casualties and perceived impunity for them are still extremely important to Afghan communities. The issue is both a central cause of disenchantment with the international coalition-backed Afghan government, and a boon to insurgent recruitment.

Yet despite his faulty conclusions, Dunlap raises an often overlooked question: why is there such a significant backlash against international forces over civilian casualties when insurgents are causing the vast majority of harm (76 percent in the first half of 2010)? Our research suggests that the issue is not that Afghans do not care about civilian harm, but that they attribute blame for overall civilian casualties and instability in Afghanistan differently than commentators like Dunlap would. These differences have contributed to significant policy missteps in the past.

First, insurgents suffer less backlash than international forces even though they cause a greater proportion of civilian harm because Afghans have lower expectations of insurgents. Afghans are more outraged when international forces cause collateral damage than when the Taliban do because they credit international forces with seemingly limitless precision technology. They expect them to have the capacity to avoid such harm, whereas they do not have a similar expectation of the low-tech suicide bomber.

Second, civilian casualties generate more resentment against international forces -- and relatively less against insurgents -- because Afghans often blame international forces for both the harm they cause and harm caused by insurgents. Given their promises of protection, high troop levels, and international forces' real and perceived capabilities, Afghans expect international forces and the Afghan government not only to refrain from harming civilians but to protect them from insurgent attacks. When they fail to do so, they are blamed for the deaths that result. Thus the more insurgents cause civilian casualties, it perversely causes more blowback for international forces.

This reaction to insurgent violence is a strong reason for the increased resentment in provinces like Helmand and Kandahar provinces; Dunlap notes this fact, but rather than exploring why more troops might anger civilians (they make communities a target for insurgent attacks and decrease rather than increase their overall security), he uses this fact to argue for the tried-and-failed policy of inaccurate airstrikes.

Finally, part of the reason there is not the same blowback for harm caused by insurgents is that they have managed the issue better for many years. Those who have analyzed Taliban propaganda and messaging have found that Taliban spokespeople and commanders have paid attention to the civilian casualty issue for years, using it as a recruitment tool and as a way to undermine support for the Afghan government. The Taliban have gone to great pains to defuse local anger when they cause civilian casualties, by only accepting blame for incidents that cause low civilian casualties, by issuing statements or propaganda blaming international forces for prompting an attack, by directly pointing the finger at ISAF for perpetrating attacks committed by the Taliban, or by justifying indiscriminant attacks by saying their fighters lack the capacity to be more precise. For most of the last nine years, ISAF has made little effort to respond at all to accusations of civilian casualties, aside from frequent denials issued in the wake of an incident.

However, international forces' knee-jerk denials of civilian casualties allegations that later proved to be true have weakened their credibility and thus their ability to counter Taliban propaganda when it counts. This general lack of accountability of foreign forces to affected communities has also been a useful tool for the Taliban. Unchecked by any competing accounts for years, Taliban have been able to spread real and exaggerated stories of incidents of civilian casualties for many years, ratcheting up local resentment against international forces.

The solution to this dilemma is not to lift restrictions on airstrikes. Airstrikes are only as effective as the intelligence that guides them and it is difficult to get that intelligence if families are furious over civilian casualties, and instead begin supporting insurgents in the area. This was one of the key strategic rationales for the restrictions in the first place, and many military officials have told me that they are beginning to pay off.

To address some of the negative perceptions of international forces, military planners should instead look at whether the tactical revisions went far enough. Afghan cooperation will be essential not only to prevent and identify insurgent threats, but also in order to promote longer-term stability in the region. The strategists who developed the current restrictions were correct in thinking that high civilian casualties make it impossible to build those relationships.

But the international community and international forces have struggled to build this trust because they have sent mixed messages. While air strikes, one source of Afghan discontent, have been limited, night raids, which are equally inflammatory, have increased dramatically under McChrystal and now Petraeus. Money and energy have been poured into addressing Afghan government accountability, but international forces rarely publicly investigate civilian casualty incidents or impose meaningful disciplinary procedures to improve conduct. Greater emphasis has been placed on encouraging the rule of law, but U.S.-supported base guards, CIA paramilitaries, and other local proxies continue to commit abuses against the population, extort money, and even support insurgent groups, with impunity.

The continuing high death toll in Afghanistan, of both foreigners and Afghans, is a harsh reality check on statements that the situation is actually getting better. But one thing we have learned in the last nine years is that we can't shoot our way out of this fight. More airstrikes may sound like a viable and even good option from the safe perch of Washington, D.C., but they're not helpful to the reality on the ground.

Erica Gaston is a human rights lawyer for the Open Society Foundations, specializing in civilian casualty issues. She is based in Kabul and New York.