Facts and statistics are pointless to look at in a vacuum. Numbers are facts. Truth can be found in trends. Someone needs to tell that to The New York Times, and maybe mention it to prospective Columbia doctoral students

MANY Americans and Iraqis feel that 2007 was the year the war in Iraq turned around: the “surge” strategy has pacified large sections of the country; previously hostile factions like those of the cleric Moktada al-Sadr and the sheiks in Anbar Province have dropped their opposition or even sided with American and government forces; and the number of insurgent attacks has dropped steadily. Still, numbers don’t lie: for those in uniform, 2007 was the deadliest year since the invasion.

First, there is a tell. We know we’re dealing with a liberal anti-war advocate because she starts her thought with the word “feel”. That needs to be edited.

Second, the aggregate number, no matter what the subject, is secondary to where the numbers are going. A true look at the data shows amazing improvement since May, 2007, when 126 American fatalities were reported. Following a monthly linear trend down, in December, 2007, 23 American soldiers lost their lives.

New York City averaged 41 murders a month in 2007. Los Angeles averaged 28. If the trend continues (and seven months of data is instructive that it will), there will be twenty-one states more dangerous to Americans than Iraq in 2008. That comparison sounds like a more compelling, truth telling factoid to me.

Yes, 2007 was a record year for American fatalities in Iraq, but both the Times and someone pursuing an advanced degree at an elite American university know that that’s not the story.

UPDATE Proof that The New York Times understands the concept of statistical trends, at least when blaming Republicans for economic troubles

On Friday, the Labor Department reported that employment in December had buckled as well. Over all, a meager 18,000 jobs were created. Even worse, hiring in the private sector contracted by 13,000 jobs, a harbinger of recession. The figures are subject to change, but job growth has been slowing since June, making a big upward correction unlikely. The unemployment rate, which is not subject to revision, also jumped in December, rising to 5 percent. As recently as last March, unemployment was only 4.4 percent. Such a big swing in such a short time also suggests a recession.

Take note of this instructive subject lesson on the distinctions between convenient truths and inconvenient truths.