As Pinkerton Was to McClellan, part 3

Powerline talks about the data-massaging issue here. Most important is the link to this report by Larry Kudlow.

Two things to notice:
1. even a relatively honest pollster like Ed Goeas has to make a lot of value judgements about the electorate to produce a poll number.
2. even a professional liar like Mark Mellman needs to make his poll move toward what the data really seems to be telling him as the election draws near (but only then).

One thing the pollsters can't hope to account for is the shenanigans that their customers, the news show biz, will engage in on election day.

Four years ago the news show biz called the election in Florida before the polls had closed in that state. By doing so they drove potential voters away from the polls and turned what would have been a 10,000-ish vote win for W in to a litigatable "tie".

Some suggest that they don't dare do something like that again. I ask, "why wouldn't they?" What consequences did they actually suffer for their interference in the democratic process? (Hint: sack of air)

UPDATE: (2-Nov-04 2:00 PM) Well it looks like the shenanigans from the news show biz this time come in the form of rigged exit polls. They apparently mean to use these as their basis for proclaiming Kerry in the lead all day long. It remains to be seen whether they will start calling states before the real polls close.

UPDATE: (2-Nov-04 4:13 PM) Now it looks like the exit polls in question are not so much "rigged" as "rumored". Not much difference there in the end.