Lies, damn lies and counterinsurgency benchmarks

Whatever strategy President Obama chooses for Afghanistan, you can be sure that "benchmarks" or "metrics" will be a big part of the prime-time news conference. "Going forward, we will not blindly stay the course," Obama said in March, when he first reassessed the war. "Instead, we will set clear metrics to measure progress and hold ourselves accountable."

So, how do we measure success in Afghanistan? If Obama opts for a narrow counterterrorism approach, the ultimate benchmark is simple: no terrorist attacks against the American homeland. But if he goes with the full McChrystal -- a long-term, fully resourced counterinsurgency, with lots of new troops -- the indicators of success become murkier.

Acknowledging that "using metrics in Afghanistan is more art than science," Brookings Institution scholars Jason Campbell, Michael E. O'Hanlon and Jeremy Shapiro map out the key indicators for a counterinsurgency war in the latest issue of Policy Review. In Iraq, they note, the most critical measures focused on violence and civilian deaths; in Afghanistan, "the most important metrics are those that gauge progress in the capacity and viability of the government."

The authors highlight three indicators, relating to security forces, the opium trade and public opinion.

The growth and quality of the Afghan army and police forces are paramount, they write, but the "capability milestone" scale normally used to evaluate local forces is of "questionable merit" because it is measured by the very trainers responsible for the units. A better metric, the authors suggest, is the number of units "that have appropriate mentoring teams embedded within them, and NATO units partnered with them" -- though those are prerequisites for, not indications of, improvement.

The opium trade, meanwhile, is a measure of security (because it funds the insurgency) and of government effectiveness (because it is linked to pervasive corruption). Though the acreage devoted to poppy cultivation and the tonnage of opium produced fell last year, "future trends are not particularly encouraging."

Finally, Afghan public opinion is relevant; "after all, it is the civilians that are the focal point of counterinsurgency missions," the authors write. They cite polls showing that Afghans worry far more about quality-of-life issues such as unemployment, health care and roads than about security. And even though public support for the Karzai government has declined, overwhelming majorities prefer to have the current government, rather than the Taliban, running Afghanistan.

The authors lament the lack of quality data on Afghanistan and caution that any wartime benchmarks should be viewed with skepticism. There is "a very human tendency," they note, to highlight those things that are easily measured and to assume they matter most. They also warn against "our own incentives in using and abusing quantitative measures." Military commanders may feel compelled to emphasize the positive to boost troop morale, they write, while politicians are often tempted to stress favorable indicators to maintain public support for the war.

"Some amount of message control is necessary and inevitable in any war effort," the authors conclude, "but in accepting this we must be careful not to spin ourselves."

-- Carlos Lozada