But this July was worse than last July!

Benjamin Disraeli is reputed to have said: “There are three kinds of lies: lies, damned lies, and statistics.” A prescient man, that Ben.
General Petraeus’ execution of the troop surge in Iraq appears to be reducing U.S. deaths from hostile action. This angers the Left because it doesn’t fit their “all is lost” narrative. So what’s a poor Lefty to do? Why, skip past the damned lies and head right for the statistics, of course! Take a look at some of the left’s statistical sleight-of-hand:

I [sic] afraid you’re reading the data incorrectly.
In fact, the chart you link to shows more Americans have died every month this year than in the same month in ’06.
What looks like a big drop in September is because September 07 is only through the first 9 days . . . adding the 9 more who died today brings September ’07’s total to 42 in the first 10 days of the month.
American deaths are clearly rising rather than falling as you conclude . . .

Oh dear. He’s right. Almost every month this year was worse than the same month last year:

  Jan Feb Mar Apr May Jun Jul Aug
2006 42 46 26 65 57 57 38 58
2007 78 71 71 96 120 93 67 56

 
But wait a moment. Two can play that game. I like the surge, and I want to “acc … sennnnnn … chu-ate the positive”. I’ll look back all the way to 2003 when we invaded Iraq, and then I’ll compare the months I want (in red below) with this year’s deaths (in blue) to minimize the deaths in 2007:

  Jan Feb Mar Apr May Jun Jul Aug
2003 0 0 58 50 8 18 28 16
2004 39 12 35 126 62 37 44 55
2005 54 42 31 46 67 69 45 75
2006 42 46 26 65 57 57 38 58
2007 78 71 71 96 120 93 67 56

 
Much better. This statistical stuff is fun!
But come to think of it, my lefty counterparts can just turn the tables on me like so (their favorite months in green and this year’s deaths in yellow):

  Jan Feb Mar Apr May Jun Jul Aug
2003 0 0 58 50 8 18 28 16
2004 39 12 35 126 62 37 44 55
2005 54 42 31 46 67 69 45 75
2006 42 46 26 65 57 57 38 58
2007 78 71 71 96 120 93 67 56

 
Are you beginning to see the problems with this little game? I’ve studied statistics enough to spot B.S. even when it’s offered on a silver platter.

The Left’s fancy figures don’t really tell us anything worthwhile. When we compare August ’07 to August ’06, and July ’07 to July ’06, things look bad at first. But that’s not a meaningful way to look at the data. We only compare data from the same month in successive years if we’re trying to spot seasonal patterns. Think of a hotel comparing this year’s spring break occupancy with last year’s spring break and you’ll see what I’m driving at. I don’t think the seasonal approach applies here, unless our jihadist enemy migrates to cooler climes in the summer.
Our lefty pals also choose to completely ignore September through December of 2006. Now maybe I just don’t get The New Democratic Math, but I don’t understand why combat deaths in the last third of 2006 don’t fit into the picture they’re painting here. I thought we were all supposed to rend our clothes and gnash our teeth over the uselessness of every death in Chimpy Bushitlerburton’s Neocon Killbot Adventure©. Silly me. I guess we can ignore the troops who died from September to December whenever liberals have some politically advantageous statistical cherry-picking to do.
Let’s get serious. The troop surge reached its peak in mid-June of this year, and from May through the end of August the number of American troops killed by hostile action has been dropping steadily.
Iraq casualties
A halfway competent person looking at American combat death statistics would identify the start of a given strategy and track the trends in deaths by hostile action from that date forward. When a significant shift in strategy takes place, the halfway competent person would flag that point in time and track any changes in the trends. If combat deaths are high and growing during Strategy A, but drop after switching to Strategy B, then something worth examining has happened.
Of course a shift in strategy isn’t the only variable in play here. As any statistician knows, correlation isn’t always proof of causation. War is horribly messy and complex and often irrational. But we need to look for what else might have changed in Iraq from around June of ’07 onward. Something (or maybe several somethings) caused American deaths to drop, and it happened right about the time that the troop surge reached its peak. Was it mere coincidence? I don’t have the kind of superhuman faith required to believe that. If we can’t find some other good explanation, then we have to conclude that the troop surge likely had something to do with the drop in combat deaths.
Unless we’re so committed to the “all is lost” narrative that we prove Disraeli right.

Comments are closed.