Three months ago, I wrote that the war in Iraq was wrecking the U.S. Army, and since then the evidence has only mounted, steeply. Faced with repeated failures to meet its recruitment targets, the Army has had to lower its standards dramatically. First it relaxed restrictions against high-school drop-outs. Then it started letting in more applicants who score in the lowest third on the armed forces aptitude test—a group, known as Category IV recruits, who have been kept to exceedingly small numbers, as a matter of firm policy, for the past 20 years. (There is also a Category V—those who score in the lowest 10th percentile. They have always been ineligible for service in the armed forces and, presumably, always will be.)
The bad news is twofold. First, the number of Category IV recruits is starting to skyrocket. Second, a new study compellingly demonstrates that, in all realms of military activity, intelligence does matter. Smarter soldiers and units perform their tasks better; dumber ones do theirs worse.
Until just last year, the Army had no trouble attracting recruits and therefore no need to dip into the dregs. As late as 2004, fully 92 percent of new Army recruits had graduated high school and just 0.6 percent scored Category IV on the military aptitude test.
Then came the spiraling casualties in Iraq, the diminishing popularity of the war itself, and the subsequent crisis in recruitment.
In response to the tightening trends, on Sept. 20, 2005, the Defense Department released DoD Instruction 1145.01, which allows 4 percent of each year's recruits to be Category IV applicants—up from the 2 percent limit that had been in place since the mid-1980s. Even so, in October, the Army had such a hard time filling its slots that the floodgates had to be opened; 12 percent of that month's active-duty recruits were Category IV. November was another disastrous month; Army officials won't even say how many Cat IV applicants they took in, except to acknowledge that the percentage was in "double digits."
(These officials insist that they will stay within the 4 percent limit for the entire fiscal year, which runs from October 2005 through September 2006. But given the extremely high percentage of Cat IVs recruited in the fiscal year's first two months, this pledge may be impossible to keep. For the math on this point, click
Some may wonder: So what? Can't someone who scores low on an aptitude test, even very low, go on to become a fine, competent soldier, especially after going through boot camp and training? No question. Some college drop-outs also end up doing very well in business and other professions. But in general, in the military no less than in the civilian world, the norm turns out to be otherwise.
In a RAND Corp. report commissioned by the office of the secretary of defense and published in 2005, military analyst Jennifer Kavanagh * reviewed a spate of recent statistical studies on the various factors that determine military performance—experience, training, aptitude, and so forth—and concluded that aptitude is key. A force "made up of personnel with high AFQT [armed forces aptitude test] scores," Kavanagh writes, "contributes to a more effective and accurate team performance."
The evidence is overwhelming. Take tank gunners. You wouldn't think intelligence would have much effect on the ability to shoot straight, but apparently it does. Replacing a gunner who'd scored Category IV on the aptitude test (ranking in the 10-30 percentile) with one who'd scored Category IIIA (50-64 percentile) improved the chances of hitting targets by 34 percent. (For more on the meaning of the test scores, click here.)
In another study cited by the RAND report, 84 three-man teams from the Army's active-duty signal battalions were given the task of making a communications system operational. Teams consisting of Category IIIA personnel had a 67 percent chance of succeeding. Those consisting of Category IIIB (who'd ranked in the 31-49 percentile on the aptitude test) had a 47 percent chance. Those with Category IV personnel had only a 29 percent chance.