applying some political science to American Idol

The third season of American Idol was surprising, but not for the reasons you might think.

In an opinion piece for Slate, Matt Feeney introduces some new analyses on the apparent oddities of this season's American Idol.

Late Sunday, Kate Aurthur of The New York Times complained that American Idol had been Hijacked by Its Viewers. It "used to exude a sense of pop-cultural justice," but has now "lost its innocence." She was referring to an episode a couple of weeks ago during which capable La Toya London was voted off while the inconsistent but very pretty Jasmine Trias moved into the final three. Aurthur blames the audience (and to a lesser extent, the phone-voting technology) for the fact that "the votes have been so capricious," and have pushed the show far from its stated purpose — to find the best unknown singer and make him or her a star.

But Feeney thinks otherwise.

But the complainers have drawn the wrong conclusions from this year's weirdness. By Idol's own standards, it was far from a bad year. It was the opposite, and that was the problem. The contestants for the third Idol were the strongest ever, something that the judges, when they weren't complaining about the feckless voters, repeatedly acknowledged. There were six different contestants who could, with the requisite throaty loudness, hold a tune and inflict a little vibrato on it.

... Such a large, tightly bunched field invites chaos. It is one of the few remotely scientific principles of political science that voting procedures involving three or more candidates or preferences are unpredictable and sometimes perverse in their outcomes, especially when, as in weighted-voting arrangements, voters can choose not just a single favorite, but can express the order and strength of several preferences. (American Idol voting registers strength of preference in several indirect ways: the decision to get up and vote at all; the willingness to persist through busy signals and other telecom snafus; and, given the option of unlimited voting, the motivation to do all these things repeatedly.)

He also dismisses the idea that producers had fixed results.

There was also race, which came up most conspicuously on a selection show at the end of April when the three lowest vote getters were all black women with excellent voices. (If there are any suspicions that Fox manipulates the votes, that image of three black women huddling together awaiting elimination should banish them. Network executives must have been cringing.) But this showed that the race question cuts in several directions. Some observers immediately inferred that these three were splitting the black vote — Idol voting is for, not against contestants—thus implying that black voters were voting their race.