Here’s how the report summarizes the situation in straight prose: “Some individual islands of security exist in the sea of instability or insecurity.” The authors muster only two islands: the town of Mazur-i-Sharif in the north and “small contiguous areas” near the Ring Road in the south. The level of security, they add, is “significantly related to the presence of well-led and non-corrupt” units of Afghan soldiers or police.

The problem is that “well-led and non-corrupt” Afghan security forces are, as yet, rare commodities. The Afghan army and national police force are making “slow progress” toward its manpower targets because of “high attrition and low retention.” Between 60 percent and 70 percent of uniformed police are “hired and deployed with no formal training.” By this August, NATO troops will be mentoring Afghan police in 45 of the 80 most important districts. Yet the report notes that even well-trained police units “have regressed” after a mentoring team is reassigned elsewhere…

But here is the most gulp-worthy sign of all. In how many of Afghanistan’s 121 key districts do majorities of the people say they “support” the Afghan government? Zero! In only 29 districts do they express “sympathy with” the government. By contrast, eight districts support—and another 40 are sympathetic with—the insurgents. (In the other 44 districts, people see themselves as “neutral.”)…

Still, the numbers are bad; the trends, even those tilting slightly upward, are worrisome. The next report is due in October. Two months after that, the U.S. military leaders will conduct their “assessment of progress toward meeting our strategic objectives.” At that point, President Obama will decide whether to keep it up, escalate, or withdraw. If the numbers haven’t markedly improved by then, he will be in a very tough spot.