Cleveland paper>Rest assured, we checked out Election 2004 thoroughly

[Algorem]

http://www.cleveland.com/readers/index.ssf?/base/opinion/1150619659219900.xml&coll=2
Rest assured, we checked out Election 2004 thoroughly

Sunday, June 18, 2006
Ted Diadiun
Plain Dealer Columnist

Atop the June 15 issue of Rolling Stone magazine you will find the following, in white capital letters on a black background: "DID BUSH STEAL THE 2004 ELECTION? How 350,000 Votes Disappeared in Ohio."

Yes, he did, writes Robert F. Kennedy Jr. in a long, exhaustively footnoted piece that flows over 16 pages in the magazine.

The early exit polls that showed John Kerry winning the election; the stories of lost, delayed and denied voter registration cards; the long lines and other problems at many polling places; and the presence of Secretary of State Ken Blackwell as an overarching malevolent Republican presence all led Kennedy to conclude that a vast and wide-ranging series of conspiracies resulted in Kerry losing an election that the majority of Ohio voters wanted him to win.

So why, some readers have asked, hasn't The Plain Dealer written anything about this story? More importantly, why has Ohio's largest newspaper forced readers to depend on Robert F. Kennedy Jr. and Rolling Stone for the real story about what happened in the 2004 election?...

[TomClash]

1. I am puzzled by the one example he cites

"(Brad Coker, president of the Mason-Dixon polling firm) cited an example from the 2004 Florida exit poll that was partially based on the pollsters' expectation that 18- to 29-year-old voters, a group that leaned heavily toward Kerry, would account for 17 percent of the vote there.

The exit poll was weighted accordingly for that age group, but it turned out that the age group actually accounted for only 13 percent of the vote, which skewed the poll inaccurately toward Kerry."

I don't get it. If the exit poll was weighted at 17% how do they know that only 13% actually voted?

I was in Florida in 2004 and the voting was rife with errors.

This is the first time I've seen the argument that Republican votes were not counted. Where's the evidence of that?

[Algorem]

3. I remember in this clown's first column as ombullsman of Pain Dealer he

wrote that he's a Bush supporter and conservative and on and on.They call him "reader representative" when what he does is defend the paper against readers.He's a joke.

[OnTheOtherHand]

4. I think that may actually be a mistake

I totally endorse the premise that exit polls can be wrong, but either Coker or the writer seems confused about the age weightings. The final weighted results show 17% 18-29.* The exit poll interviewers actually try to keep tabs on refusals by rough age range, so they aren't likely to be so far off on Florida. Where does Coker get 13%? No clue. Dunno, I could be missing something -- I don't have that dataset on this computer.

The exit poll estimate put Bush ahead in Florida by 1.1 points, so I don't know why someone would choose that as the state to discuss anyway.

What kinds of errors did you see in Florida? I think it has gotten somewhat less attention than Ohio because it was less close, and because the initial quantitative arguments for fraud didn't pan out.

"This is the first time I've seen the argument that Republican votes were not counted."

I can't tell what you're looking at, so I won't try to reply.

* EDIT TO ADD: since I don't have the data here, I'm relying on the CNN.com tabulation.

[EFerrari]

12. Maybe because they knew the other 77% wouldn't be able

to vote? They went after the student vote with a vengeance.

Robocalls threatening to cut their aid if they voted "in the wrong place" -- alternate lines for students -- absentee ballots that never arrived or were sent to the wrong address -- registration campaigns whose aim was to tear up the registration forms.

We went to swing states to help students vote. They went there to block that vote.

[AAARedskin]

2. How about we look to the future for a change?

This is more wasted energy and time going on. How about we try to get some people to run for office that can actually inspire people to vote for them? This is exactly what the GOP did after Clinton kicked their asses in both national elections..All these accusations and conspiracy theories about stolen elections are really getting old, folks.

[Botany]

5. Not to me.

Sure, don't report, write or care about the biggest story in US history ....
focus on the future ..... I am sure they will not do it again.

"Those who do not study the past are certain to repeat the same mistakes
in the future."

[Gman]

7. That "it's time to move on" attitude only serves to

perpetuate the fraud. In other words it means "you didn't catch us so forget about it".

[mod mom]

17. Ah...we tried that after 2000, concentrated on GOTV and look what happened

We are up against a party of thugs who will win at any cost. You don't really think they will decide to play fairly this time-do you?

BTW, if you look back at Clinton's win over HW you will see that Clinton /Gore only received 43% of the vote. HW received 37% and Ross Perot's reform party (ie 3rd Party) picked up 19.7 million votes. (Source: Greg Palast's Armed Madhouse pg 304)

Anyone who call it "conspiracy theory" is ill informed or.....

[blm]

28. You believe Bush found 11million MORE voters to vote for him in 2004

when Rove was only originally targeting 4 million new fundies? And NO defections?

I worked in NC offices during the 2004 campaign and we sold many a Kerry yard sign to disaffected Republicans on a daily basis. I am quite certain that many other GOPs did the same - like Lee Iacocca and John McLaughlin who also turned against Bush.

[mod mom]

29. We had Republicans for Kerry clubs in Ohio, and I agree with you. No Way!

[OnTheOtherHand]

30. you're talking to a tombstone n/t

[Botany]

6. Cover your ass time

<There was no shortage of mistakes made in vote counting. There were voters who should have been registered but weren't, polling places with lines that were too long and without enough voting machines, and decisions from Blackwell that appeared to be partisan.

All these mistakes and misjudgments took votes from both candidates, but probably more from Kerry. But they didn't add up to nearly enough votes to swing Ohio from Bush to Kerry.

The mistakes were often a result of lack of foresight and bad judgment, but they were bipartisan in nature and not a result of Republican chicanery.>


Relationship between machines per registered voter, voter turnout, and Kerry vote share

More specifically, here are tables for all three counties, which specify the number of precincts, machines per registered voter, and voter turnout, according to the Kerry vote share (percent of voters voting for Kerry). These tables were computed using the data that Professor Walter Mebane used to produce his portion of the 2004 Ohio Election Report sponsored by the Democratic National Committee (DNC).

Lake County

Kerry %     Precincts    Machines/1000 reg. voters    Turnout
20-30       4                3.82                                 67.5%
30-40       32              3.51                                 68.4%
40-50       76              3.55                                 69.1%
50-60       88              3.40                                 66.1%
60-70       15              3.19                                 62.0%
70-80       2                3.10                                 50.0%


Mahoning County

Kerry %     Precincts    Machines/1000 reg. voters    Turnout
30-40       17               6.90                                 68.2%
40-50       51               6.48                                 66.1%
50-60       68               6.48                                 65.6%
60-70       75               6.41                                 62.0%
70-80       42               5.72                                 56.0%
80-90       38               5.01                                 47.6%
90-100     21               5.31                                 50.0%


Franklin County

Kerry %     Precincts    Machines/1000 reg. voters    Turnout
20-30       18              4.02                                 67.8%
30-40       130             3.92                                 66.4%
40-50       236             3.66                                 60.6%
50-60       154             3.32                                 57.1%
60-70       90              2.94                                 51.1%
70-80       57              3.10                                 48.6%
80-90       69              3.08                                 49.4%
90-100     32              3.03                                 47.5%


There are two important things to note from observation of these tables. First, there is a general and significant trend in all three counties for the number of machines per registered voters to be very high in the heaviest Bush precincts and then become progressively lower the higher the Kerry vote. Secondly, voter turnout demonstrated precisely the same relationship, being highest in the heaviest Bush precincts and then becoming progressively lower in the Kerry precincts.

[Febble]

8. Yes, but....

You are right of course. Turnout was higher where there were more machines per active voter, and there were more machines per active voter in more Republican precincts. No question.

But there is an important confound: the number of active voters per machine (in Franklin at any rate) correlates not only with turnout in 2004 but with turnout in 2000. In other words, it looks as though, given that not enough machines were purchased, turnout in 2000 was used to guesstimate the likely throughput.

Which is incompetent for a lot of reasons, not least because where turnount is low, the capacity for increase is likely to be higher. And it may have been deliberate, although I think the BoE in Franklin at least has an alibi.

But it also means that in trying to quantify the reduction in turnout due to the shortage of machines you have to consider the degree to which the shortage of machines was due to low historic turnout patterns. I did an analysis which you can download here:

http://uscountvotes.org/ucvAnalysis/OH/FranklinCountyReport_v2.pdf

in which I concluded that even after controlling for turnout in 2000 additional votes were lost due to the long lines that resulted from the inequitable distribution of voting machines, and that that the majority of these votes (two thirds) were likely to be Kerry votes.

But you can't make a straight inference from 2004 turnout alone.

Elizabeth Liddle

[Botany]

9. I live in Columbus, OH .... Franklin County

"Which is incompetent for a lot of reasons, not least because where turnount is low, the capacity for increase is likely to be higher. And it may have been deliberate, although I think the BoE in Franklin at least has an alibi."

No they don't have "an alibi." 62 machines were deliberately held from the voters in dem precincts in Franklin County.
Election workers started asking for new machines @ 7:00 AM and they never received them. The Franklin
County BOE's rooster of voting machines shows that those machines had beens deleted from official tally
sheets. "Rumor" has some of the machines on trucks driving around Franklin that day.

Lack of machines were a deliberate causative agent in suppressing the vote in certain areas.

[Febble]

10. Well, I don't live anywhere near

so I'm not in a position to judge intent. On first look at the data, it certainly looked deliberate. It was only when I started looking into the data in more detail that I realised that historical turnout must have been used in the allocation algorithm, whether maliciously or not.

But seeing as it clearly did (for whatever reason) that does need to be controlled for when you estimate the total number of lost votes. My estimates were that net votes were indeed lost (and this of course is supported by anecdotal evidence) and that the majority of these were for Kerry.

But what is absolutely clear, no matter what the reason, is that far fewer machines per active voter were allocated in Dem precincts, resulting in far fewer machines per actual voter, and explaining the long lines.

And what you say about what happened on the day is certainly disturbing.

[Febble]

15. Oh, and I should have made clear

what that of course turnout in 2000 tended to be lower in Democratic precincts (strong correlation).

So by rationing on the basis of past turnout, they were de facto rationing on the basis of partisanship.

[mod mom]

19. Not to mention that inner city precincts had less machines that they did

in the primary. There is also the issue of QUALITY of machines. Look at where the complaints came of machines breaking down. If you read Palast's book, you'll find the same was true for NM. Planned assault PERIOD.

[Febble]

23. As I said

I don't have evidence on which to judge intent, and Mebane's and OTOH's analyses are more thorough than mine anyway.

The point is that there is no doubt that Democratic precincts had fewer machines per voter.

Let's agree on that.

[Innocent Smith]

13. Very good post

The first thing I thought when I saw the post above yours was what was the turnout in the previous election and were they using that data to decide how to allocate machines. If, for example, you had a location that had close to 100% turnout in the past then they would do fine with the number of machines, but if for some reason another location that in the past had low turnout this time turned out in large nunbers then they would have number-of-machine problems. If fact the lower the turnout you had in the past the more likey you would be to have problems.

The simple solution would be to just allocate machines based upon registered voters, hmmm... but then the higher turnout areas would have longer waits relative to other areas. The best solution would be to just buy a whole lot more voting machines to have available to allocate.. I can't imagine that costing that much relative to the benefit you'd get from it.

[Febble]

14. Well, a number of states

(I googled it when I was trying to figure out Franklin) base their allocation on "active voters" which seems to mean, usually, voters who have voted at least once in two electoral cycles. If they'd done that in Franklin, they'd have done better, because what matters is the size of the pool of potential voters, not the predicted turnout, and certainly not the turnout predicted from a low turnout year.

There are so many things wrong with DREs, but their tendency to create bottlenecks, especially if they are unfairly rationed, is not the least of them.

One of the great things about pencil and paper is that all you need is pencil and paper.

[Innocent Smith]

16. Makes sense

It makes sense to average it out from the turnout of previous elections instead of using just the one election before as a yardstick. I would guess that mid-day things are pretty quite and there are small bottlenecks in the morning before work and much larger bottlenecks in the evening after work.

[OnTheOtherHand]

21. the trouble is

first of all, precinct boundaries change constantly, so averaging across previous elections is sort of a non-starter.

Beyond that, based on Mebane's results (and my secondary analysis of his data), the allocation was pretty reasonable using active voters as of April, but inequitable using active voters in November. In other words, they needed to allow for new registrants, and they didn't. (Of course in principle it isn't hard to allow for new registrants!)

About the most charitable interpretation I can put on the data is that they had a deer-in-headlights breakdown: faced with the need to take machines out of precincts that were already short on machines, in order to put them into precincts that were even shorter on machines, they gambled that Blackwell would get them off the hook by allowing some sort of quasi-absentee voting for people stuck in lines. "Shockingly," he didn't.

There are some less charitable interpretations that I can't rule out, although I can say that even some Republican precincts were short on machines.

EDIT TO ADD: by the way, apparently the basic reason they didn't buy more machines in 2004 was that the Danaher machines wouldn't comply with Ohio's then-new VVPAT requirement, and they didn't want to buy machines they could only use once.

[mod mom]

20. This ignores NEW REGISTRATION INCREASES AMONG DEMS. It was

intentional to disenfranchise new voters!

[Febble]

25. Whatever

I didn't commit one way or the other - I don't live in Ohio. All I can tell you is that if it was intentional, they set it up to be plausible, because machine allocation correlated with past turnout - and that matters when you are trying to quantify the effect on turnout.

But it was certainly incompetent, and maybe worse. You know better than I whether it was worse.

[mod mom]

18. Febble, they knew about the huge surges in registrations. The use of

"ACTIVE VOTERS" was a way to penalize the huge surge of DEM voters. Blackwell went down to FL after the 2000 theft to learn the ropes. This was a planned assault to insure a GOP victory.

[Febble]

24. Look, please

read my post. I said I did not have information on which to judge intent. All I said is that they had an alibi. Most crooks have an alibi.

I do not have information on which to judge intent. I just looked at the numbers, and confirmed, statistically, what you know from observation.

And whether it was incompetence or deliberate (and I DO NOT presume to say which) it seems to me it was a civil rights violation.

Go for it.

[mod mom]

27. I love this excuse:

"The mistakes were often a result of lack of foresight and bad judgment" but they endorse Blackwell regardless.

[EFerrari]

11. What is the email addy for this fatuous Quisling? Can't find it. n/t

[Algorem]

22. I looked at his "weblog" for first time,has link to e-mail form-

http://www.cleveland.com/weblogs/readers/?/weblogs/readers/email_author.html

http://www.cleveland.com/weblogs/readers/



i see he's posted a few things in last few days,but most recent before that was in February:what's the deal with that?

i think i don't trust them enough to let them have my name or number

[kster]

26. K&R...... nt