Even in the age of instant communication, it takes three months or more for developments in Iraq to have any impact on the U.S. political debate. The war is like a distant star whose light we only see well after the fact.

So Democrats still warn that we'll never be able to police a sectarian civil war, even after violence has significantly declined in Iraq -- because we have successfully policed a sectarian civil war. Some critics of the war have seamlessly passed from lamenting the unstoppability of the Iraqi civil war to warning that the rise of Sunni security volunteers could be a harbinger of ... a civil war.

The outdated anti-war sound bite of the moment is that the surge has failed because the Iraqi government hasn't met 18 benchmarks set out for it by Congress last year. It is routinely asserted that only a handful of the benchmarks have been met. In Newsweek in March, columnist Fareed Zakaria darkly noted that a few newly passed laws "add up to only three or four of the 18 benchmarks."

The benchmarks are much cited, but apparently little read. Of the 18, seven have to do with supporting the surge and the effort to establish security in Baghdad: things like providing three brigades to support operations in the city; establishing joint security stations with U.S. forces in neighborhoods; and reducing sectarian violence and eliminating militia control of local security.

By any standard, almost all these security benchmarks have been met. They were formulated at a time when the Iraqi government's will to secure Baghdad was in question. Forget three brigades -- as Fred Kagan of the American Enterprise Institute points out, soon enough the Iraqis will have three divisions in and around Baghdad. The neutralization of militias has been more problematic, but now Iraqi Prime Minister Nouri al-Maliki has declared himself against the most dangerous Shia militia, Moqtada al-Sadr's Mahdi Army.

The highest-profile benchmarks are the seven legislative ones. Four of the key ones have been passed: a law undoing the excesses of de-Baathification; a provision granting amnesty to former insurgents; legislation allowing the formation of semiautonomous regions; and measures setting out provincial powers and a date for provincial elections. Another important one, a hydrocarbons law, is stalled, but the passage of a budget sharing oil revenues around the country serves some of the same function.

The balance of the other benchmarks has to do with the performance of the Iraqi government and protecting minority rights. They are harder to evaluate. Of course, all the grading is somewhat subjective, but roughly 12 of the 18 benchmarks have been met (and there's been movement on the others), which makes a much less seductive anti-war talking point. As the reality on the benchmarks slowly sinks in, opponents of the war will surely move on to something else -- probably the war's cost.

Needless to say, if a benchmark has been met, it doesn't necessarily mean the underlying law is wise or will be effective. The war's critics argue that, in its fine print, the new de-Baathification law may exclude as many Sunnis from government as the original, offending law. They're right. Which is why it was always foolish to try to judge the progress of a nascent, violence-plagued democracy by a crude checklist.

Already, there has been a shifting of goal posts. Zakaria warned that some of the new laws passed only "after months of intense wrangling." Horrors! What was so remarkable about the Feb. 13 passage of a package including a budget, provincial powers law and amnesty provision wasn't the intensity of the wrangling but the cross-ethnic and -sectarian logrolling that produced a grand compromise unlocking the stuck wheels of the Iraqi parliament.

Logrolling, alas, is not one of the benchmarks. The last time Gen. David Petraeus came to Washington, he heralded tentative but widely discounted security gains. Now he brings news of tentative but widely discounted political progress. We'll know he's had an impact when the benchmarks fade away from anti-war discourse.