Hamilton Co. (Cincinnati) finally releases canvass results

[jmknapp]

The precinct-by-precinct results were not released until today.

http://www.hamilton-co.org/boe/inputdata/Electionsresults/Archived/G04OFFCANVASS.xls

Have just started looking at it, but here is an item of interest:


Spoiled ballots (those without a valid vote for president):



This chart excludes one precinct that had a 19% spoilage rate.

[Viva_La_Revolution]

1. Wow. Do you have one for *ush spoilage? nt

[jmknapp]

2. That's for overall spoilage

100% Bush precincts on the left, 100% Kerry precincts on the right

[Viva_La_Revolution]

3. Oh, got it!

Thanks, I've never quite understood what these graphs were showing, or I should say, never quite got the importance of what it shows. Till now, :-)

[Carolab]

4. So the more Kerry the fewer undervotes

That's odd--is this the "default to Bush or to third party/blank" phenomenon? What kind of machines did these people use?

[liberal43110]

18. Wrong

It shows that the more Democratic the precinct, the higher the rate of ballot spoilage.

[Blue22]

5. Whoa! Whose voting machines in Cincinnati?

Well, not surprising from the birthplace of electronic voter
fraud.

See: http://www.votefraud.org/greatest_coverup_of_all.htm



Who's e-vote machines used in this county?

B-)

[roseBudd]

16. Votamatic punch cards we're talking chads

[Hapameli]

6. This is RIGHT ON! I'm doing this for L.A. Co and getting the same

Keep this kicked! We need something constructive to do anyhow.

ETA: Very high numbers of third party votes.

[divineorder]

7. We voted here with punchcards

But punchcards are read electronically, so the question is where the error is.

[newyawker99]

12. Hi Hapameli!!

Welcome to DU!! :toast:

[minvis]

8. Hamilton County Canvass Results

I broke down the undervotes even further. The average undervote percentage for the county was 2.18%. Out of the 1013 precincts, 354 precincts had undervote percentages over that average. 240 or 68% of those 354 precincts Kerry won. I then looked at the precincts which had 80% or more votes for either Kerry or Bush (each of their strongest precincts). I chose 80% as the cutoff because although Kerry had many precincts with 90% or more (48 to be exact), Bush did not have any 90% or over except for one which had 4 votes total (he got all 4 votes, by the way).

Bush had 44 precincts with 80% or more in vote totals. Only 1 out of the 44 had an undervote percentage of over the 2.18% and that one precinct had a percentage of 2.45%, barely over the average. Contrast that to Kerry who had 140 precincts with 80% or more in vote totals. 112 out of those 140 precincts had undervotes over the average. Additionally, 38 of those 112 had undervote averages more than twice the average (4.36%). This does not even include the 19.89% undervote in Cincinnati 5-D (Kerry won that precinct with only 77.84% of the vote, under my arbitrary 80% cutoff).

What does this mean in terms of actual votes lost? Well, I quickly just went through the precincts with the most undervotes, in terms of actual numbers of spoiled ballots. For instance, the precinct with the most actual number of undervotes was 53 in Cincinnati 24-J. Kerry got 83.03% in that precinct. I simply multiplied the 53 undervotes by Kerry's percentage and got 44 votes and then subtracted 44 from 53 and got 9 votes making a difference of +35 votes for Kerry that were lost. I haven't gone through every precinct yet, but just counting the top 30 precincts, there was a net loss of over 550 votes for Kerry.

This result is completely in line with Rove/Blackwell strategy of voter suppression in Ohio in general and in urban areas specifically.

[minvis]

9. Cleves D Turnout

What's up with turnout percentage of 95.7% in Cleves D? The rest of Cleves' precincts are around 67%.

[roseBudd]

24. Undervotes means either the voter did not vote in a race or they were

unable to fully detach the chad. If the chad lays flat, then the card reader interprets a no punch/no vote.

Undervotes in minority precincts are usually blamed on inexperience and lower education levels but undervotes can be caused by chad buildup and chad buildup occurs under certain ballot positions next to braces. Chad buildup also occurs in machines that never have the chad cleared.

Undervotes could also be caused by ballots with an off register die cut.

All punch card ballots in Ohio were supposed to be supplied by Dayton Legal Blank.

[skids]

10. Absentee ratios, FWIW.

Added it to what I had on hand. Man that's a pretty extreme spoilage bias.

[skids]

11. The undervote is not new -- a look at past years.

The bias in the undervote may be a bit worse this year, but is nothing
especially new. It's been going on at least since 1996.



However, before anyone says "well that's just how the poor/minority population vote" let's take a look at Leon County, Florida (29% black, versus hamilton's I think 23-27% IIRC.)



So either there's undervote stealing in FL, or a systematic, long-term, unaddressed bias in the voting system in OH. Given what I've seen of press quotes/stories surrounding the Leon county BOE, my money is that Hamilton is the one with the problem.

The sad thing is, it isn't an abberation.

[roseBudd]

26. The undervote in minority communities is explained away...

by saying that education and experience are the reason for the high spoilage rate. What if it is just more "dirty tricks"

Punch card machines also have a guide layer that is produced for every election. The sylus can only punch holes in areas wher there is a hole in the guide layer.

http://www.cs.uiowa.edu/~jones/cards/chad.html

[mordarlar]

13. talk to me like i am 5. Wouldn't Kerry's be more? I am confused

If there was fraud to be seen against Kerry i would think he would show a higher spoilage rate.

[skids]

14. Depends on the method.

A lot of potential fraud methods don't involve spoilage at all. Some (like tuning opscans to favor one party) could actually reduce spoilage.

What we are seeing here though IMO is more the fact that, by and large, the poor neighborhoods get the crappy machines and the least trained poll workers. Do keep in mind that in a close election, every little bit counts, and a few percentage points of spoilage is a lot of headroom to work with.

[mordarlar]

15. Thank you for replying. I understand op scan but...

If Bush has a higher rate of spoilage and we try to determine whether some of Kerry's are valid wouldn't we possibly have a large rate of finding them for Bush too? As his numbers are larger and the ratio is greater? Another thing, who compiled these stats. they know what we are looking for. They have kept the info from us. Could they have altered? Could we prove it?

[skids]

23. WRT looking at spoiled ballots

Generally speaking, if the spoilage rate were even (the line was flat) each candidate should pick up additional votes in proportion to the
number of votes they had in the county.

But in Hamilton the line is not flat, and the race was not a blowout in either direction, so in this case Kerry could pick up votes by a full hand recount.

As far as the trustworthiness of the data, usually they try to obscure data, not fabricate it. For example a lot of the counties gave us numbers that did not list the spoilage, and those that did did not usually separate bad votes out from blank votes. And then there is the fact that all but one county (Fairfield) that I am aware of rolled the Nader/Disqualified numbers into the spoilage figure.

[mordarlar]

17. they all look like higher spoilage the more % Kerry except...

the bar graphs those are throwing me.

[skids]

22. The bar graphs aren't undervotes/spoilage.

They are absentee versus booth voting. The higher the bar, the larger
a percent of the votes which a candidate received came from absentee ballots.

Generally speaking, each party has a pattern of ballot "rolloff" where they undervote more on lower races on the ticket. Depending on the county, the overall absentee/booth ratio may vary but usually the candidates perform about the same in this metric relative to other candidates from the same party.

That falls apart in places where there is a lot of cross-over voting, because the patterns of the cross-over voters become signifigant and interfere with the other two patterns.

[mordarlar]

25. thank you so much.

[liberal43110]

19. Ballot spoilage does NOT equal an undervote

Ballot spoilage refers to anything that prevents the ballot from being counted, including the paper getting folded an unable to be read in a scanner.

An undervote is a ballot that does not contain any vote for a certain office, in this case, president.

Ballot spoilage could include undervotes, overvotes, handwritten notes in the margin (in Cleveland, some voters punched the ballot for Kerry and then wrote "Kerry" in the margin, which caused their ballots to be rejected), folding, and all other things that prevent the ballot from being counted.

[mordarlar]

20. Lack of HAVA stickers offered at poll or the rule change about birth date

issued AFTER the election for example? :(

[berniew1]

21. In 2000 Gore had thousands of legal vote "spoiled ballots" not counted by

machines and not allowed to be looked at and counted as per Florida law. Gore had thousands of legal "overvotes" in Palm Beach, Duval, Hillsboro, and Gadsden counties where butterfly ballot type design allowed votes the opportunity to vote for Gore twice, clearly a legal vote under Florida law, but rejected by machine readers.