Comparison of the Kerry - Fingerhut - Connaly vote in the 2004 OH election

[Time for change]

As many of you are aware, one of the main findings (perhaps THE main finding) which led the DNC report to conclude that their analysis "strongly suggests" no widespread fraud in the 2004 Ohio Presidential election, was the strong correlation by precinct between Kerry's percent of the vote and the percent of the vote for the Democratic Senatorial candidate, Eric Fingerhut. Here is the Executive summary, where that conclusion is stated on page 11: http://a9.g.akamai.net/7/9/8082/v001/www.democrats.org/pdfs/ohvrireport/section02.pdf. And here is Section VI of the report, where the relevant discussion can be found on pages 14 through 16: http://a9.g.akamai.net/7/9/8082/v001/www.democrats.org/pdfs/ohvrireport/section06.pdf. The correlation coefficient for the correlation between the Kerry vote and the Fingerhut vote by county (not provided in the report) was 0.91, with r squared = 0.83.

I believe that there are numerous reasons to doubt the DNC report conclusion, including those which I discussed in my letter to Howard Dean, which I posted recently on another thread: http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x380878.

Some have wondered why the DNC report looked at the correlation between Kerry's vote and Fingerhut's vote (as well as some other things) but NOT the correlation between Kerry's vote and that of Ellen Connally, Democratic candidate for Chief Justice of the Ohio Supreme Court. Being very curious about that, and having read about various findings suggesting a discordance between Kerry's vote and Connally's vote, I decided to look at the relationship between the Kerry and Connally vote statewide.

It turns out that the correlation between Kerry's and Connally's vote by county, using simple linear regression analysis, is much weaker than the correlation between the Kerry and Fingerhut vote. Specifically, the Kerry-Connally correlation coefficient is 0.71, with r squared = 0.51. Here are the counties that stand out in this regard, showing a Connally advantage over Kerry of 9% or more, despite Kerry running ahead of Connally by 4.5% state-wide.

Mercer --- 16.3%
Shelby --- 16.0%
Butler ---- 13.7%
Auglaize -- 13.2%
Clermont - 11.7%
Darke ---- 11.4%
Warren --- 11.1%
Putnam --- 10.6%
Brown ---- 10.6%
Van Wert - 09.7%

The disparity between the Kerry/Connally vote correlation, compared to the Kerry/Fingerhut vote correlation is substantial. The question is, what is the explanation for this?

There could be either a fraud explanation or a non-fraud explanation. A fraud explanation would involve some reason why it would be easier to perpetrate fraud on the Kerry and Fingerhut vote simultaneously, while leaving the Connally vote alone. A non-fraud explanation would involve some reason why Connally would poll better than Kerry in the counties listed above, despite running behind him statewide, and why the state-wide Kerry-Connally correlation is so much weaker than the Kerry-Fingerhut correlation.

The reason I am posting this thread is to encourage exploration of the reasons for this anomaly, in the hope of gaining some understanding of what happenned in the 2004 election.

[roseBudd]

1. Clermont, Warren & Butler are all in SW OH and...

Warren of the infamous no press can watch the tally lockdown..

Recount volunters in Clermont observed white stickers on Optiscan ballots covering Kerry's bubble.

As far as name recognition, I saw not one Connaly sign or commercial and had never heard of her.

A Warren County DUer told me the R candiate, incumbent Moyer had yrad signs in lots of repuke yards.

Fingerhut's name has name recognition.

The county Connaly should have maybe outpolled Kerry is Cuyahoga.

[Time for change]

8. But Connally outpolled Fingerhut state-wide

She polled 45% of the state-wide vote, Fingerhut only 36%. Why would that be? Whatever the mechanism was that caused Kerry to lose votes, why would the same be done to the Fingerhut vote but not the Connally vote?

[Bill Bored]

11. Ask minvis to look at ballot order rotations

in the Senate and Judicial races.

[Time for change]

24. What would be the significance of that?

And what is minvis and how do I contact him?

[minvis]

41. I'm here!

You can PM me if you'd like. I have the ballot order information for Butler and Clermont Counties.

[roseBudd]

12. They screwed up when they failed to fix the tally on the lowest race

on the ballot. Supreme Court was last. There is no way Bush voters accidentally voted for Connaly over Moyer having never heard of her in SW OH.

[papau]

2. The game was Senate and Pres - Chief Justice was not tampered with

"There could be either a fraud explanation or a non-fraud explanation." AGREED

A fraud explanation would involve some reason why it would be easier to perpetrate fraud on the Kerry and Fingerhut vote simultaneously, while leaving the Connally vote alone. - NO NOT "EASIER" - JUST MORE IMPORTANT


The "non-fraud explanation" requiring a reason why Connally would poll better than Kerry in the counties listed above, despite running behind him statewide, and why the state-wide Kerry-Connally correlation is so much weaker than the Kerry-Fingerhut correlation, is really lacking - and thereby becomes proof of the fraud..

[Time for change]

7. I agree that "more important" would also be a reason

But I question how important that Senate race was considered. I don't know how close the race was the day before the election, but Fingerhut lost by 28%. So I question whether they had a good reason to steal votes for that race. Do you know if that was considered a close race?

[roseBudd]

13. It is impossible for a significant amount of votes to be cast for a

candidate no one in SW Ohio had ever heard of. So underfunded I saw not one yard sign or commercial and I live in one of the urban University precincts in Cincinnati.

Yet a Warren County DUer verified that there were numerous Moyer for SC signs right next to Bush/Cheney signs in her county. In addition both papers endorsed the incumbent Moyer.

[Time for change]

14. Ok, but what is your explanation then

What does giving extra votes to Connally have to do with taking them away from Kerry? And why did Connally do so much better then Fingerhut, especially in southwestern Ohio?

[roseBudd]

16. Tabulation fraud ...

They didn't give extra votes to Connaly, that reflects the Ds who voted a straignt dem ticket having their Kerry votes converted to the Bush column.

Bush did not get 72% in Warren in 2004 and Kerry did not do no better than Gore in 2000 in Warren. Warren County tabulated the ballots with no oversight after declaring a phony Homeland Security threat. In addition Warren a sparsely populated county finished their tally AFTER Cuyahoga.

Warren County ended up having the highest percentage of Bush votes amongst all of Ohio counties that had a total vote turnout greater than 22,500.

Warren 92,251 votes - Bush: 72.1% Kerry: 27.53%

There were only four counties wiht a higher percentage vote for Bush, and they all had 22,279 voters or less:


Mercer 20,058 votes - Bush: 74.89% Kerry: 24.55%

Putnam 18,631 votes - Bush: 76.20% Kerry: 23.34%

Holmes 10,976 votes - Bush: 75.61% Kerry: 23.89%

Auglaize 22,279 votes - Bush: 73.78% Kerry: 25.71%

So what happened that night in Warren County when all of America was waiting for the counting to finish in Ohio to decide the election? Why was it the only county in all of Ohio where media was not allowed to witness the counting?

http://www.enquirer.com/editions/2004/11/05/loc_warrenvote05.html

http://www.enquirer.com/editions/2004/11/10/loc_warrenvote10.html

[Time for change]

18. I don't doubt that there was massive fraud in Warren County

But why was Kerry's state-wide vote so much more strongly correlated with Connaly's vote than with Fingerhut's? And, why did Connally do so much better than Fingerhut, especially in southwestern Ohio?

[roseBudd]

26. Because there was fraud in many counties. Warren was the icing

which is why it was so important for their tallying to run late enough to come in after Cuyahoga.

[newswolf56]

3. If any of the Ohio newspapers are still investigating...

...the election, you should get the names of the reporter(s) assigned to the project(s) and forward this material to them. It has been many many years since I studied sociological statistics, but I believe your analysis is absolutely valid.

Another dimension you might explore -- forgive me if this has already been done, but I am a relative latecomer to this matter -- is the percent of correlation between voter-registrations and results. (I assume Ohio like most states requires registration by party.) Theoretically, the election returns from a precinct that is 60 percent Democratic, 40 percent Republican should parallel the registration ratio. This would provide additional context for the study you have undertaken. If otherwise true, it would also make the deviation you have noted all the more glaring. As you imply, follow-up polling could quickly eliminate (or confirm) any causes within the electorate. (Indeed a newspaper is advantageously positioned to undertake just such polling.) If no cause for the deviation can be found within the electorate, then fraud (or malfunction) is clearly indicated. It would probably require access to a renegade programmer to confirm just how such fraud/malfunction might happen.

A further index to voter behavior would be to attempt pattern analysis of individual precincts' inclinations to split tickets. From this material you could develop a generalization that said, for example, "in 20 years of election results, Y percentage of Precinct X's voters always cast straight party-line ballots. But this year only Z percentage did." Based on my own experience analyzing election data, any genuine deviation from the norm will be immediately apparent. If the deviation's cause were socioeconomic, that would probably be obvious from comparative census data; if otherwise inexplicable, the deviation becomes yet another indicative unknown. Again, additional foundation material upon which to establish the ultimate validity of the evidence.

(It would also be informative to DU readers to post -- if indeed any newspaper investigations are still ongoing -- current links to the relevant stories.)

That said, thank you for your efforts on this critical matter.

[LightningFlash]

4. Very good answer...

...Specifically, this must be done in correlation to Issue 1. If you study the actual flaws in this side by side, that is where you get the pattern.

A fully reproduceable pattern is required to prove whether widespread fraud happened. And you can find and get that pattern, by doing a poll in regards to those counties as well as Issue 1 on the ballot.

[liam_laddie]

5. party affiliation

Only by voting in primaries (asking for either a Dem or rep ballot)
is party affiliation established in Ohio. Around 30%, maybe 35% maximum, of registered voters bother to vote in primaries, so
most voters are listed as "ind" - thus no way to compare voting
results with registration numbers.

[roseBudd]

6. Ohio does not require registration by party, voting in a primary does

cause identification based on ballot requested. In SW Ohio counties Connaly was so underfunded as to be unknown. Both papers endorsed the incumbent AND Warren County flat out lies about a Homeland Security threat that they planned over a week in advance yet failed to tell reporters they would not be allowed in the BoE. Reporters were finally allowed 2 floors below the vote tally.

For Warren County info Google Erica Solvig, Cincinnati Enquirer, Pat South and Frank Young. Erica Solvig broke the story.

[Time for change]

9. Thanks for all the ideas Newswolf

With regard to the registration issue, some others have indicated that that information isn't available in Ohio. But still, I understand that new voter registration was off the charts in Ohio prior to the election. I think it would be interesting to learn anything we can about that -- e.g., whether this occurred in Democratic or Republican areas.

And I agree that past ticket splitting practices would also be important to look into.

I do volunteer work for the Ohio Project. I will contact them to see if they know where I can get any of this information.

[roseBudd]

17. New voter registration in OH favored the dems and the repugs knew it

That is why Blackwell fought so hard to disqualify as many provisional ballots as possible. That's why Franklin County held back 68 voting machines on election day. That's why the GOP sent registered letters to newly registered urban minority voters and then used that list for the voter challengers.

Check out the links section on http://bushcheated04.com

In addition Nathan Sproul's organization was operating in Ohio.

[Time for change]

19. I also heard that new voter registration in OH favored the Dems

But how do we prove that if party affiliation isn't part of the record?

[roseBudd]

20. They knew it based on where the new voters lived and because

it was in African American precincts. The GOP also bought marketing studies that identified cars, magazines subscribed to etc. It's why they sent my hubby letters even though we live in a very dem precinct. He drives a BMW and reads Esquire & GQ so they thought he might swing republican. What they sent was push polling at its worst.

Ohio had MoveOn, ACT, ACORN, Young Voter Alliance, VoteMob, Kerry campaign, etc. People came from all over to descend on Ohio. Unless you never came out of your house you got registered.

[roseBudd]

21. new registrations in OH & FL...

http://www.truthout.org/docs_04/printer_092704K.shtml

Big Increase of New Voters in Swing States
By Ford Fessenden
The New York Times

Sunday 26 September 2004

COLUMBUS, Ohio - A sweeping voter registration campaign in heavily Democratic areas has added tens of thousands of new voters to the rolls in the swing states of Ohio and Florida, a surge that has far exceeded the efforts of Republicans in both states, a review of registration data shows.

The analysis by The New York Times of county-by-county data shows that in Democratic areas of Ohio - primarily low-income and minority neighborhoods - new registrations since January have risen 250 percent over the same period in 2000. In comparison, new registrations have increased just 25 percent in Republican areas. A similar pattern is apparent in Florida: in the strongest Democratic areas, the pace of new registration is 60 percent higher than in 2000, while it has risen just 12 percent in the heaviest Republican areas.



snip

(in Franklin County; very long lines in AA precincts ) Nevertheless, an examination of county registration records shows that the groups have added thousands of new Democrats to the rolls and have far outnumbered new registrations in Republican areas. In a 300-square-block area east of the courthouse in downtown Columbus that voted nine to one against Mr. Bush in 2000, for instance, 3,000 new voters have registered this year. That is three times as many as in each of the last two presidential election years. The number of registered voters in the area is up 18 percent since January.

By comparison, in a prosperous area north of downtown with a similar number of voters who are overwhelmingly Republican, just 1,100 new voters have been added this year, increasing registration rolls by 7 percent. These numbers are similar across Ohio. The Times examined registration from Jan. 1 to July 31 in a sample of counties that included seven of the state's nine largest, along with some smaller rural and suburban counties. Voters do not give a party affiliation when they register in Ohio, but The Times looked at the voting history of ZIP codes to gauge the political inclinations of the new voters.

[Time for change]

27. Thank you Rosebud

Do you know where we can get specific county data on that?

[Bill Bored]

10. Also, look at Bush vs. Moyer

Moyer only got more votes than Bush in 4 counties. Connally got more votes than Kerry in 12 counties (although some say it's 19?????).

Anyway, if you compare Bush and Kerry's totals in all 16 counties in which either judicial candidate got more votes than his or her Presidential counterpart, bush got 150,000 more votes than Kerry.

What do you make of that TFC? I've been citing it since Nov. but no one has really picked up on it, despite the fact that the Kerry/Connally discrepancy has been widely discussed and misstated.

[Time for change]

15. I think you're on to something Bill!

I went and looked at the numbers myself, using this site:
http://election.sos.state.oh.us/Results.aspx . I didn't come up with exactly what you state, but it is very close, and I really do think it suggests something -- though I don't know what yet.

Here's what I found:
There are 16 counties where Kerry's total vote was less than Connally's and 4 counties where Bush's vote was less than Moyer's. In those 20 counties, Kerry lost to Bush by a net of 172,833. So that's very close to what you state.

But there's lots more. First, the whole difference is in the 16 counties where Kerry's total was less than Connally's. Kerry lost every one of those counties, by a total of 240,535. The four counties where Moyer's vote was more than Bush's were all won by Kerry. So let's concentrate on the other 16.

I then looked at how Bush did against Gore in 2000. In those 16 counties Bush also beat Gore by quite a bit -- but by a lot less than what he beat Kerry by. In those 16 counties Bush beat Gore by 171,270. So, the net difference in 2004 was a Bush gain of 69,265.

The overall association of those 16 counties with poor Kerry performance compared to Gore is quite striking. There is a very strong relationship of those 16 counties to relatively poor performance for Kerry compared to Gore's performance in 2000 (and remember that Kerry did substantially better than Gore state-wide in Ohio). I sorted counties by relatively poor performance of Kerry compared to Gore. Among the top 8, 7 of them were counties where Connally got more votes than Kerry. Among the next 32, there were 9 where Connally got more votes than Kerry. And the last 48 counties contained none where Connally got more votes than Kerry.

So, what we have is two anomalies (Kerry doing much worse than Gore, and Connally getting more votes than Kerry) that are very strongly associated with each other. Either one by itself wouldn't seem terribly striking to me. But why should they be so strongly associated with each other? I can't think of a non-fraud reason for that. Well, I can't specifically think of a fraud-related reason either. But maybe someone else can.

What do you think?

[Bill Bored]

22. My figures were based on the initial unofficial tallies.

I haven't looked at Ohio in a while.

They were correct before the provisionals were counted, although it's hard to imagine that Kerry lost even MORE counties to Connally but gained on Bush statewide since then! I'm still working from the old spreadsheet which had Bush winning the state by 136,483 votes. There were only 12 counties in which Connally got more votes than Kerry at that time.

When you say he did worse than Gore, do you mean in terms of margin percentage or just numbers of votes? There was higher turnout in 04 for both candidates.

I think there are plenty of things that could be investigated in OH, but without a proper recount, and given the fact that the punch cards don't even have the candidates' names printed on them, or the precincts in which they were supposedly cast, it seems very hard to prove anything. I think a recount would help though. The only way to get around that would be to switch ballots between precincts. I was hoping that a recount could detect that, but apparently, it really can't.

That said, a full hand count of Butler, Clermont and Warren would certainly seem like a good idea, no matter what! But don't hold your breath.

[Time for change]

23. What I meant by counties where Gore did better was by % of the vote

I agree that a recount needs to be done.

But in the absence of that, given the these anomalies, especially the very strong association between Connally getting more votes than Kerry and Kerry's relatively poor performance compared to Gore, what else can we look at now?

[Bill Bored]

28. Past presidential races with OH Chief Justice candidates?

I don't think this happened in 2000, but perhaps earlier.

We should also look at Early Voting in FL and elsewhere and I have some ideas about how to do that. One idea is to find out if Election Day WPEs were higher in Early Voting states than non-Early Voting states. There were 30 of the former and 20 of the latter.

Since the 1250 precincts in Mitofsky's report excluded most of the Early/Absentee voting ones, it may be possible to see if there were larger exit poll discrepancies on Election Day in Early Voting states than in "regular" states that still have some semblance of an electoral process (i.e., NO Early Voting).

If we can show that Election Day voting didn't match the exit polls in states with Early Voting more than in states without it, this might require some additional explanation.

Remember, Red Shift was based on projections that included Early/Absentee voting AND Election Day voting. By sticking to the 1250 precincts, we remove most of the Early/Absentee stuff and focus on what happened on Election Day. If Early Voting is as big a scam as I suspect it is, we should see larger exit poll discrepancies on Election Day in Early Voting states because Election Day is when the fix would have been in. The cover story is of course that the Dems had "an aggressive Early Voting campaign." But that would not necessarily explain why Election Day voting would not match exit polls in these states, would it?

Just a hypothesis of course.

[Time for change]

32. I don't quite understand several points that you're making

1. Are you saying that, whereas the exit poll vote percents include early voting and election day voting, the WPEs based on the 1250 precincts include almost no early voting?

2. What reason do you have to believe that election day voting in early voting states would be more fraudulent than election day voting in non-early voting states?

3. You make a number of references to looking at "election poll discrepancies". Do you mean by that: a) WPE or overall difference between election poll result and official count? b) looking just at election day votes, or all the votes?

[Bill Bored]

33. Q and A

1. Are you saying that, whereas the exit poll vote percents include early voting and election day voting, the WPEs based on the 1250 precincts include almost no early voting?

Yes; 210 precincts that had >15% Early/Absentee voting were excluded from the 1250. 1460 - 210 = 1250

2. What reason do you have to believe that election day voting in early voting states would be more fraudulent than election day voting in non-early voting states?

If one knew that their candidate were losing based on the Early Voting numbers, they might be inclined to change the E-Day results and they'd have plenty of opportunity to do so in advance. They are defined as separate elections in e-votin' systems (at least by Diebold). This is less likely in states without early voting where they'd just have to guess the outcome and hope they got it right, or change the results only AFTER the election when it's easier to detect. Early Voting gives the perpetrator an opportunity to see the handwriting on the wall and act on it.

3. You make a number of references to looking at "election poll discrepancies". Do you mean by that:

a) WPE or overall difference between election poll result and official count?

Our notion of "Red Shift" is based on early exit poll releases to the media which I believe included the results of early/absentee voting (all 1460 precincts). It therefore does not distinguish between election day and pre-election day votng.

b) looking just at election day votes, or all the votes?

I'm suggesting just looking at election day results since they are what the WPE calculations are based on. We already know that the mean WPE of the 210 Early/Absentee precincts was lower than that of the 1250 (~ -5% vs. -6.7%). But of course, we must be cautious since we don't have raw WPE data, only the public domain stuff.

[Time for change]

34. Sounds good to me

Do you know what the 30 states were? I have all numbers on the red shift (and I assume that you do also), so if we know what the 30 states are we should be able to figure it out pretty quickly -- although if you're saying that you want to look at pre-election vs. election day results separately for individual states, and compare each of these with exit poll data for pre-election vs. election day, that data I think might be quite difficult to obtain.

With regard to your last paragraph, I think I have a slight disagreement with you: I believe that the WPE for the 1250 precincts was -6.5. The overall red shift for the country (which included all precincts I assume) was -5.5. So there is a difference of 1.0 there.

But I don't think that that necessarily means that the difference was due to the fact that the 210 early/absentee precincts were very different. The -5.5% is overall difference. The -6.5 is the value of each precinct averaged together. So it is possible (though perhaps unlikely) to have that 6.5 vs. 5.5 difference even if they referred to the exact same precincts. This could happen if there were a lot of very small precincts that had very large red shifts. That would pull the WPE towards greater negativity, but wouldn't have as much effect on the overall red shift because of their small size. I don't know how likely that is, however.

Well, anyhow, I think that's right.

[Bill Bored]

36. Have to check the Jan report again

but I think that when I added up the 1250, which are presented as quintiles of varying size on page 36, I weighted the quintiles by the number of precincts in each and got a weighted average of -6.77, if mem serves me. There are a few other tables that include all 1250 that could be checked in the same way. Page 39 has one "Size of Place." Page 40 has some with 1249. But I've only checked the one on pg 36. They should all come out the same though.

The -6.5 is Mitofsky's number, which included all 1460 precincts. If you then calculate the average WPE of only the 210, it comes out to about -5.

I agree that Red Shift may be different because of varying precinct sizes, but also because the early exit polls on which the Red Shift is based were weighted demographically and the precinct data were not. We know this because the earliest exit poll release had overweighted women which they corrected later on even before adjusting to the vote count.

See if my -6.77 is right.

[roseBudd]

25. The one day I went to the Warren County BoE w/Andy Stephenson...

you could tell they did not like us being there. Andy did discover signatures that did not match and missing signatures.

IMHO the key to Warren County is Frank Young Emergency Services Director. I would like to see hin deposed. He is the originator of the so-called level 10 Homeland Security alert. See the Cincinnati Enquirer articles for info.

As far as local reporters who might be Woodward & Bernstein material I'd go with Howard Wilkinson [email protected]

The Enquirer can't let the Toledo Blade win all the journalism awards this year...

An interesting coincidence is that one of the Warren County officials is named Leslie Spaeth also an elector for the Bush-Cheney ticket in
the November 2, 2004 election. Some of you may remember Merrie Spaeth of Swift Boat Vets fame.

Warren County OH officials that locked down Vote counting are REPUBLICANS

Board Members

Susan Johnson, Director
Sharon Fisher, Deputy Director
Robert Hammock, Chairperson
Stella Hagemeyer, Board Member
Barbara Sizemore, Board Member
Leslie J. Spaeth, Board Member

http://www.co.warren.oh.us/bdelec



Stella Hagemeyer is also a Republican:

Barbara Sizemore's husband gave to Bush

SIZEMORE, DWIGHT
LEBANON, OH 45036
SELF-EMPLOYED/N/A BUSH, GEORGE W (R)
President
BUSH-CHENEY '04 INC $500 05/14/04

Warren County DU member lizzieforkerry told me that on the day they planned the lockdown no dems were present.

[Time for change]

31. I understand what you mean about Connally being an obscure candidate

And I also agree that there seems to be something fishy going on there, and that this may hold a key to an understanding of where or how the election was stolen.

But here is what is unexplained by all this: Despite her obscurity, she did considerably better than Fingerhut statewide as a percent of the vote. I don't see how fraudulently taking votes away from Kerry could have anything to do with this. Taking votes away from Kerry could cause Connally to do better than Kerry in some places, but it wouldn't cause her to do better than Fingerhut. And why would anyone want to take votes away from Fingerhut? That race didn't appear to be anywhere near close.

And furthermore, if votes were fraudulently taken away from Kerry, that could cause the correllation between his vote and Connally's vote to be reduced -- but it should also do the same thing to the relationship between Kerry's vote and Fingerhut's vote. And that apparently didn't happen, since that correlation was found to be so strong.

I'm harping away at these points in the hope that if someone can come up with an explanation for these apparently strange findings, that might provide us a clue as to how this happened.

[minvis]

40. What if votes were stolen from Kerry and Fingerhut?

It would make sense if votes were switched for both Kerry and Fingerhut. If you think about it, where would someone immediately look for anomolies in vote counts? I know I looked at the Senate and House races as well for abnormalities. Nothing was immediately apparent.

However, if you switch votes in all the top of the ballot races it makes it harder to detect fraud in the race that you want to influence, in this case, the presidential race. Obviously, there was no reason, other than covering your tracks, to decrease Fingerhut and Democratic House candidates vote total since the Democratic candidates for these offices were not going to win in this area. But, if Fingerhut and other Dem House candidates actually were closer percentage wise to their Republican opponents, this would be a key that Kerry did a lot better than the count indicated.

[Time for change]

42. That's what I was thinking for a while

But then I thought about this whole thing a great deal, and I changed my mind. One thing that caused me to change my mind about that was the realization that Kerry vote, according to the DNC report, was also very highly correlated with Hagan's vote in the 2002 governor's race. I don't think that Hagan's 2002 vote was manipulated in 2004 just so that it would correlate with Kerry's vote. Also, I did a comparison of my own, and found out that Kerry's 2004 vote was very highly correlated with Gore's 2000 vote by county -- even more highly correlated than with Fingerhut's 2004 vote.

Soooo, the upshot of this all, is that I now believe that the Ohio election was stolen in one of two ways, or more likely both: Either by electronically causing the disappearance of ballots in highly Democratic precincts or counties, which was apparently done in Cuyahoga County at least (which would not have changed the correlations), or by addition ballots in Bush precincts or counties (which would not have changed the correlations as long as the added ballots represented roughly the same proportions for all parts of the ticket as legally cast ballots in the rest of the precinct or county), which was apparently done at least in Miami and Warren Counties.

Also, I believe, for reasons that are too complex to discuss right now, that voter registration data (see posts 37-9 on this thread) could be of great help in coming close to proving this -- if I'm correct that voter registration was huge in Democratic precincts and counties compared to Republican ones. So if you could give me some idea of how I could get my hands on that I would much appreciate it.

[myschkin]

29. kick

[freedomfries]

30. KICK

[chrisclub]

35. Kerry lost?

http://mysite.verizon.net/myk15/2percent.html

&.view=t&.done=http%3a//photos.groups.yahoo.com/group/geron/lst%3f%26.dir=/%26.src=gr%26.view=t

[Nothing Without Hope]

37. This is important for further exploration. We also need VOTER REGISTRATION

data to factor into the analysis for comparison.

ANYBODY GOT OHIO VOTER REGISTRATION DATA ON A PRECINCT BASIS?

[Time for change]

38. Thank you Hope -- County data would also be quite valuable

Precinct data would be even better, but I could use it only if I had election results by precinct to go along with that.

[Nothing Without Hope]

39. I wonder if Demopedia has these data?

I've just signed up for a password but haven't received it yet.