NH: Zombie Narrative

It just won't die! Here's but one example from Joe Conason in Salon:

"What happened during the contest's last few days was that the undecided broke for Clinton, and the question is why."

But the undecideds did not break for Clinton. Again, the exit poll:


"Day-of" deciders narrowly went for Clinton (and by a margin well within the margin of error). Voters who decided "last three days," "last week," and "last month" all went for Obama. Clinton's only category where she leads Obama beyond the margin of error is the category of voters who decided more than a month ago.

The readily available empirical data simply does not support a "late Hillary surge" theory in New Hampshire. The data undermines such a theory. The simplest theory remains that the tracking polls did not properly model likely voters.

But simple theories are simply no fun. Simple theories do not lend themselves to the kind of self-important speculation that passes for "keen political insight." Here's another illustrative example on MSNBC immediately following the declaration of HRC winning New Hampshire. Pat Buchanan (a mean and nasty man, but he's been around elections for quite a while) speculates of a possible "hidden vote," but the rest of the bobbleheads prefer the less-plausible "crying and/or backlash" theory:



The zombie is unleashed!

Yes, Chris Matthews is a buffoon who acts like a little boy at the tree-house "He-Man Woman Haters Club." It's funny to see Rachel Maddow stick it to him. But the empirical data neither supports a "Clinton cried her way to victory" theory nor a "backlash against the media" theory. The Zombie Narrative continues to walk the earth, because the Zombie Narrative fits right in with the self-important pundits and bloggers who think that average voters care what pundits and bloggers obsess over.

That, and "likely voter models" is a phrase without much zing.