A response to Palast's "Kerry Won. Here are the Facts." Is it accurate?

[Fridays Child]

I email blasted the media in three or four states with the Palast article (http://gregpalast.com/printerfriendly.cfm?artid=392) and received this response from Frank Taylor, "Florida Today" Associate Editorial Page Editor:

Oops.
The final exit poll totals didn't disagree with the results. They were conducted in three waves at different times of day. The first wave, conducted in the morning, showed the Kerry leads in several key states. But exit polls correctly showed the race tightening later in the day, erasing that early lead. However, the early numbers were widely diseminated without sufficient explanation and are were wildly spun. Hint: If you have an exit poll but the polls are still open, it's an incomplete poll. The demographic breakdown on voters who show up in the morning when most people with jobs have to be at work is very different from the breakdown of those who show up after 5 when they get off work! Early polls tend to break inaccurately for Democrats, except in places where there is a large bloc of retirees who lean GOP. The Associated Press and others have already explained this.
Like Kerry said, the election is over.

[robbedvoter]

1. Dear, the last exit polls were put up AFTER the results

It was medja catching up to the script. Pay attention to stories from all over. Kerry may want out, my right to vote is not his to give away.

[Fridays Child]

2. Thank you. Agree, and kicking for more comments (hopefully).

[regnaD kciN]

13. What is hard to argue about, though...

...is that, for all states shown above except Illinois, and whether paper ballots or e-voting (with or without paper trails) were used, Kerry's actual results were lower than his exit poll results. Sometimes by a percent or two, sometimes by a lot more.

I'm not saying that the Republicans didn't steal Florida and Ohio. But the very fact that, in 8 out of 9 races, there was a drop in Kerry's numbers from the exit poll to the final result (and in 0 of the 9 did it go up) tells me that it isn't simply a matter of electronic voting machines without paper trails. Maybe all of the systems, electronic or print, were gamed for Bush. Maybe the polling methodology was flawed, and inadvertantly created a pro-Democratic bias of a few percent. (The fact that the turnout was evenly split between the two parties, while previous elections had shown a 3% advantage to Democrats, might explain some of the latter.) But it isn't a simple matter of "the vote matched the polls everywhere except in e-vote states without paper trails," no matter how much people may try to write it off as such.

[Awsi Dooger]

14. Very good point, and not merely this election

It happened in 2002 (my examples in post #9) and also in 2000. People tend to forget that in 2000 the GOP was pissed off that states they won by huge margins, primarily in the South, were called only after long delay because the exit polls indicated they were closer than they turned out to be. That happened again ths time, namely North and South Carolina.

Someone else posted tonight the same guy is running AP who headed defunct VNS. He needs to get his shit together and change the models.

[m berst]

3. misdirection

Whether or not exit polls are accurate is irrelevant. They were accurate when and only when there was a paper trail. They were wildly inaccurate where there wasn't a paper trail.

That doesn't prove anything, but so what? The burden is not on any of us to prove that the election was stolen, the burden is on elections officials to prove that it wasn't.

[Fridays Child]

4. Thanks. I'm writing him again and sending the above graphic.

[regnaD kciN]

15. Not quite true...

They were accurate when and only when there was a paper trail.

Looking at the graphic above, Florida seems to have had a change of around 4% and Ohio around 3%. But, by the same token, North Carolina shifted 5%, New Hampshire 6%, and Pennsylvania (which Kerry won along with NH) a whopping 9%. All those make the Florida and Ohio shifts look small indeed. Did all three of those use machines without paper trails? If so, which states did have paper trails?

For that matter, it appears that Maine and Wisconsin, which used paper ballots, had a Kerry drop of around 2%, which isn't that different from Ohio's 3%, as far as I'm concerned.

What's also significant to me is that none of the states, even those with paper ballots, showed what I would normally have expected: at least one case where the final result was one or two points higher for Kerry than the exit polls. That might point to a coordinated nationwide conspiracy, involving manipulation of paper ballots as well as eVotes (which moves it onto a somewhat higher level than Diebold and BBV, IMHO), or it might simply point to the formula used by the pollsters being outdated (as in, expecting turnout to break 38-35-27 instead of 37-37-26).

Once again, I'm not saying that the result, at least in Ohio and possibly nationwide, wasn't stolen. There are other ways to cheat aside from hacking touchscreen voting machines, and we certainly saw some of them in Ohio this year. But, throughout the past couple of days, I've heard a repeated mantra: that, in all states where paper ballots or a paper trail were provided, the results nearly matched the exit polling, while in those states without paper trails, there was a dramatic 5% shift to Bush between those two numbers. Based on what I see in the graph above, that simply isn't the case.

[DoYouEverWonder]

5. Here's a chart for just Florida that you can send too

and it has nothing to do with exit polls.

These are stats by county, what system they used, how many register REP & DEM, how many REP & DEM actually voted, and how many voted in 2000.

There are significant increases in counties that have a majority of registered DEMS and touch screen machines, of REPs voting in much, much higher percentages all of a sudden and with sharp increases compared to 2000.

However, in counties with optical scanners and more Registered Republicans, we are not see these increases. What we are seeing are numbers that are in line with the percentage of DEMS in the county and the ratios for 2000 in that county.

Statistically, this dog don't hunt.

http://ustogether.org/Florida_Election.htm

Oops: Forgot Link. It is really way past my bedtime.

[Awsi Dooger]

6. Please repost with chart: it didn't show up the first time

I'm not a believer in BBV theft, but I want to look at any related studies. Thanks.

[DoYouEverWonder]

7. Sorry

It's there now.

[Awsi Dooger]

10. Thanks

That looks like an interesting one

[KharmaTrain]

8. So Whose Going To Replace These Secretaries Of State

Time to move ahead and prevent this from happening. The best way is to control the machines, and the person who does is the Secretary of State and other officials, and if you haven't noticed, almost every one is a Repugnican. Ironic isn't it??? It'd be a nice goal for Democrats to replace these and other Repugnican officials...building up a true national party rather than bits of this and that.

This arguement reminds me of the scene in Citizen Kane where they have to choose from two newspaper headlines the day after the election..."Kane Wins" or "Fraud At The Polls".

Put the shoe on the other foot, had Kerry won a majority and then * contested contested ballots in a state or two he was leading in, you'd be singing a completely different song.

[Awsi Dooger]

9. I am getting conflicting information on final exit polls

That chart posted in this thread was also presented in a Diebold thread. I was working GOTV on Tuesday and didn't see the earliest exit polls. A friend of mine who saw them disputes these are the final exit polls; he insists they were the first wave, including the ridiculous 60/40 in PA and 57/41 in NH. Those two numbers alone make me want to dismiss the entire sample, and therefore the argument.

Exit polls do come in waves. I have looked at them every cycle since '96 and exit polls are updated all day and night. The best ones are available long after the election, a combination of more than one source. That is more nationwide than state.

In 2002, the early exit polls did skew heavily toward Democrats. Strickland in CO was listed as 19 points ahead, or some outrageous fugure. He lost by 6. Bowles in NC was supposedly winning by 9. He lost by that amount. McBride in FL was only down by 2. He lost by something like 15. Kathleen Kennedy Townsend was winning in Maryland. She lost by 3 or 4.

I am not posting these numbers to dispute exit polls from 2004, but early exit polls in general. I posted a thread late Monday night in GD2004, mentioning those exact numbers from 2002 and warning the exit polls on Tuesday likely would be terrible and not to have too much faith in them, one way or another. I understand Skinner initially banned them. That's a good idea for the future.

[JohnnyCougar]

11. That doesn't explain how Kerry LOST votes in a CNN update

After adding more numbers to the exit poll, CNN TOOK AWAY votes for Kerry. You can't get negative votes later in the day.

[Awsi Dooger]

12. No, it doesn't

I didn't see that one

[TruthIsAll]

16. kick

.