Great Moments in Political Reporting
An AP article, out all over the country this weekend, announces a Bush lead in the electoral college after a purported "analysis of state polls." The geniuses at AP awarded the solid states (25 for Bush, 14 plus DC for Kerry) and "leaning" states (3 for Kerry, 7 for Bush), and decided that Bush leads 217-193 in the electoral college count, leaving out the other 11 states, deemed "toss-up."

Here's the dumbest part (my emphasis), in the paragraph about those states in this last category:

TOSSUPS -- Bush and Kerry are running even in 11 states with a combined 128 electoral votes. Florida, Ohio, Iowa, Nevada, New Hampshire, New Mexico, Wisconsin, Michigan and West Virginia are the toughest battlegrounds. Two other tossups, Pennsylvania and Oregon, could soon move to Kerry's column.

So what does it mean that PA and OR "could soon move to Kerry's column" if it doesn't mean that they are "leaning"? What evidence exists that they could soon move one way, but not the other, except the kind of evidence that should convince you to count it as leaning? And if PA and OR are counted, as they should have been, whaddya know, it gives Kerry the lead and changes the entire headline and point of the piece from claiming a Bush advantage to claiming a Kerry advantage.

The race is virtually tied, and that's the only responsible horse-race reporting we should be getting. But if either has the lead right now, it's Kerry. The AP piece presents no methodology, no evidence for its claim, and completely ignores the fact that its conclusion runs contrary to every major polling firm, all of which give leads, however slim, to Kerry in most of those 11 tossup states.

Crazy as it sounds, some people's votes are swayed by a desire to go with the winner. I know, it's crazy, but I know some of those people. And so when the press touts a leading candidate in a close race, they should be especially careful, and they have a particular responsibility to realize the impact they have.