Doesn’t “vested” mean that once they leave, they can continue in the insurance program (obviously they have to pay for it, but presumably the Feds get a good group rate so its better then they’re likely to find elsewhere)? Santorum was in the Senate for more then five years, so I’d think he’d qualify.
What it means is that (assuming it’s a 95% MOE as usual) there’s a 95% chance that Smith’s support is between 38% and 46% at the time of the poll, and there’s a 95% chance that Jones’s support is between 34% and 42% at the time of the poll, and there’s a 95% chance that Black’s support is between 16% and 24% at the time of the poll.
The chances that all of these things are true at once is somewhat less. But you can’t just take the product of .95*.95*.95 because these are dependent events.
And the ‘at the time of the poll’ caveat is rather important. Even if the poll’s taken on Monday and the election’s on Tuesday, the MOE doesn’t say anything about how much the support of any candidate can rise or fall over time. A poll is an estimate of their support at a specific moment in time.
I’ve got a kid at my elbow wanting to play, so I’ll talk about the MOE of a difference later.
Is there any reason to believe the last day’s polling is more important than the first day’s? Couldn’t that just indicate that Ron Paul’s supporters are more zealous in making sure their votes count, so they are more likely to vote early?
Well, the last day is three days closer to the day of the caucus, so there’s less time for further changes in voter opinion. Also, to the extent it can be done on the basis of small daily samples, the polls capture the direction of change of voter support over time.
No, this isn’t voting, this is polling. The pollster chose four random samples of households, and called them up on four consecutive evenings. People didn’t get to choose which sample they were in - they didn’t even get to choose whether they were in any of the samples or not, for that matter.
Now if an election was held over four days, and you had access to the daily totals, you’d be absolutely right: it wouldn’t matter which day a vote was cast on, and everybody would want to get their supporters to the polls on the earliest day possible, to make sure they didn’t forget to vote, or change their mind.
I don’t know if it is going to be by 14 votes or 4, but this is a bigger win for Santorum. On the one hand, he’s the not Romney. On the other hand, he’s also really not Ron Paul.
Did any Doper call the Santorum thing? I’m actually surprised he didn’t do better…I’d say the economy and the war really threw it with RP in the race.
edit: Santorum also wins a shit ton of free media for the next few weeks.
Nate Silver at fivethirtyeight.com called it, sort of – against his own model, sort of. He analyzed Iowa’s historical response to momentum, calculated momentum numbers for the candidates, and declared that Santorum could squeak out a win (this Iowa-specific model also had Romney in third place, but it was a pretty savvy analysis all the same).
Oh, and Santorum didn’t win by four votes, he was four votes ahead with one precinct remaining. The NYT has Romney winning by eight votes.
I should confess now: I took a shot, but it missed and he just got a few small scratches on his face, one on his upper lip. I’m sure it didn’t change anything…
Amid all the fuss over Santorum on abortion and personal sexual matters, the press is now digging into his record on things like earmarks. He’s now raising much more money but the New Hampshire primary is only a few days away.