In a blog entry discussed by Greg Pollowitz, Karen Tumulty of Time finds that generals have historically made only average presidents–though, like a true journalist, she compensates for the lack of a story by ladling on the adjectives and adverbs.
Of the 10 former generals who served long enough as president to be rated in a 2000 poll of historians, “just one,” she says, was listed as “great.” Considering that of the 39 presidents rated, only 3 achieved this rank, 1 out of 10 is exactly what you would expect from the subset of generals.
In fact, if you look at the entire poll, you’ll see that the 39 rated presidents fall into its six categories (great, near great, above average, average, below average, and failure) as follows: 3/8/7/8/9/3. Since generals make up roughly one-quarter of the sample, you can get the expected distribution for generals by dividing each of these figures by 4 and rounding off: 1/2/2/2/2/1. The actual distribution for generals is 1/2/0/2/3/2.
So the generals come off a little worse than expected, but not a lot–especially when you look at how closely the ratings are spaced, and how arbitrary the division into groups is (Benjamin Harrison, for example, barely misses being average).
Here’s another way to look at it: Take the numerical ratings of the 10 generals and average them. The composite rating comes out to 2.87, which would fall between No. 21 (George H. W. Bush) and No. 22 (Rutherford Hayes) on the list of 39 presidents. In other words, generals come off very slightly below average–much too slightly to base any generalizations on. (There’s also the fact that three of the survey’s four presidential “failures” served just before or just after the Civil War, an unusually stressful era that tends to exaggerate a president’s rating.)
For comparison, let’s split the presidents up a different way–alphabetically. The first group will contain those whose last names begin with A through J (21 presidents), and the second will contain those whose last names begin with K through Z (18). The first group is distributed among the six categories (best to worst, remember) 0/3/3/6/6/3, while the latter group is 3/5/4/2/3/1. As you can see, the second group is much more densely clustered among the higher categories. The first group’s average rating is 2.74, while the second group’s average rating is 3.38. This is a far greater discrepancy than Ms. Tumulty found with the generals.
So what do we take from this? It would seem that waiting until the end of the alphabet for your name to be called in grade school gives you lots of time to think, which gets a boy in the habit of mulling things over at length, instead of being rash and impulsive. I’m sure my colleagues, Mr. Williamson and Mr. VerBruggen, would agree with that. Or maybe it’s just a meaningless statistical aberration–like the much smaller difference Ms. Tumulty found with her survey of generals.