2004 Improvements Saved a Million Votes. Non-Votes Dropped 0.8% !!

[L. Coyote]

Monday, February 14, 2005 · 1:43 p.m. PT

Report: Fixes saved 1M votes in 2004

By ROBERT TANNER, AP

Improvements to voting machines and election administration saved a million votes that otherwise would likely have gone uncounted in the 2004 elections ....

The report released Monday by the Caltech/MIT Voting Technology Project looked at a key measure of election integrity - residual votes, or ballots cast during an election on which voters failed to mark a choice or machines did not record it.......

In 2000, the national residual vote was 1.9 percent of ballots cast for president. The report found a significant improvement this year, with the residual vote falling to 1.1 percent. The analysis examined 37 states and the District of Columbia; figures were unavailable elsewhere."


NOTE: "figures were unavailable elsewhere." What's with this??

http://seattlepi.nwsource.com/national/apelection_story.asp?category=1130&slug=Election%20Changes

[rzemanfl]

1. Rove is tearing out his remaining couple of hairs. Dubyah was

supposed to get ALL of those votes, heads will roll at ES&S and Diebold.

[skids]

2. I'm afraid to even look.

I can easily guess that they just took the SoS on their word and didn't bother to consider phantom voting like what was found in New Mexico.

[L. Coyote]

3. Look at how many states don't know how many non-votes there were!!

This is what I find shocking. I cannot find out how any votes were cast in some states. Apparently, even the SoS does not know in many states.

I have found 5 more than this report used, but 10 of 51 still remain unknown, AND they do not answer e-mail inquiries.

[Wilms]

4. This is awful... n/t

[Awsi Dooger]

5. That's why the newer voting machinery should help us in the long run

And why I've always been in favor of it. The majority of those saved votes were undoubtedly Democratic. Kerry would have lost by a greater margin if using the 2000 standards and technology. That won't be popular here, especially in this forum. Big deal. We blew it on preference, via lousy spring handicapping atop the ticket.

I have not been here in a while, but I see the same blinders-on cynics are desperately reviewing 2004 as if it were supposed to be a replay of 2000, completely ignoring what 9/11 did to party ID and the difference between an incumbent race and an open race.

As long as the tallies from electronic and optical scan machines can be verified, we benefit. I'm sure if you break these numbers down further, the urban precincts will show an even greater improvement than average.

[skids]

6. You sure like to make a lot of assumptions.

First, you make the assumption that the lower number of spoiled votes is because more votes were correctly counted, whereas in some places, they are lower because they were used as a place to put fraudulent votes where they would not be detected. See:

http://www.votersunite.org/info/newmexicophantomvotes.asp

Now it just so happens that, assuming the report cited is correct, in fact the Democratic party did benefit from lower undervote. Marginally. If we don't consider phantoms, that is. If we assume that just because a county has a democratic majority that more democratic undervotes were among the ones saved.

In addition, as it just so turns out, the only major system that generates more undervotes than electronic machines is punch cards, and the improvement was mostly due to the replacement of those. Opscan and paper are the best. Go figure. I guess it is about as safe to assume that DREs benefit us, as it would be to assume that no sane BOE would use DREs without a paper backup. Look where that got us.

Such an "all is right with the world" tact won't get you very far around these parts these days, let me assure you. And we aren't cynics, FYI, we're curious.

[L. Coyote]

7. Interesting observations. Here are a few more.

Given known facts about spoilage rates in Florida, for example, it certainly would have been advantageous to Democrats to replace punch card voting. At the same time, without a paper trail, it remains uncertain if it REALLY was.

That said, the highest known non-vote rate in the nation was in New Mexico on Danather and Sequoia E-voting machines, 2.55 and 2.74 percent respectively (figures including the phantom votes).

Electronic voting and/or tabulation is suscepible to simple manipulation by modifying the code, including instructions to ignore every Xth vote, for example, in select areas.

Is this cynicism? No, it is fact. Cynicism would be, "Why vote, the machine is already fixed." Nonetheless, there is plenty in the way of blinders and cynical posting on this forum, I agree.

If Ohio had had a better SoS and they had updated the punch card counties, the results would certainly be different. Check out this graph of the precincts in Cuyahoga County with over 1 standard deviaion in the already very high non-vote percentage.



This group (with over 3.6% non-votes) represents 11.56 percent of the Cuyahoga precincts and 27 percent of the non-votes (and over 1% of Ohio voters). By comparison, the county average non-vote rate is 1.8 percent, the median is only 1.46. For E-voting statewide, the rate is O.76%.
In Cuyahoga County, people have every right to be cynical, and they certainly do not have equal rights when it comes to having their votes count, especially if they live in Democrat strongholds!! It is very easy to know which machines performed poorly in the last election, and then send them to selected precincts. That's a fact. I'm not offering (non-statistical) proof that that is what happened, so I won't say it did, since that would be cynical!

And check this graph of the precincts at locations with 2 precincts and 2 ballot orders (1/4 of Cuyahoga, 2.5% of Ohio).



It is obvious where the non-voting happened and whose votes counted, whose not.

We certainly have more grounds of agreement than naught. To begin with, counting votes is better than not counting them!! Let's do even better at counting in 2008.

And we are more than curious, we are the fleas on a mean fascist, junkyard dog.

[TruthIsAll]

9. They counted all the hacked optiscans and switched touchscreen votes.

So what the hell is good about it?

No awsie, you haven't been around for a while.
Why don't you go to uscountsvote.org?
Read what the experts have to say.

[berniew1]

8. Touchscreens shown to be unreliable and insecure; didn't count accurately

http://www.votersunite.org messups articles
http://www.flcv.com/ussumall.html
http://www.flcv.com/fraudpat.html

[NationalEnquirer]

10. What a puff story.

Deflects all the criticism of how bad the machines are, sheesh.. MSM, gobble it up and spit it out.. good little doggies..

[Land Shark]

11. Report cites 0.5% is problem "residual" voting, yet fails to talk

about New Mexico with undervoting rates of several percent that go much higher in high-minority districts. (see database on New Mexico at www.votersunite.org )

Although it is possible, it is unlikely that Caltech and MIT did not get a copy of the Snohomish report which specifically talks about NM undervoting rates and raises the issue clearly that undervoting does not appear to be a sign of health, whether low or high. If this idea is not discussed in the Caltech.MIT paper, the paper is incomplete. If it was known and omitted, that raises even more questions about Caltech.

Snohomish Report downloadable at:
www.votersunite.org/info/SnohomishElectionFraudInvestigation.pdf

Chief author Charles Stewart's email address: [e-mail]

Stewart's temp election home page with state breakdowns and other links is found at: http://mit.edu/cstewart/www/

(edited to add links)

[Land Shark]

12. HAVA; 1 million votes admitted lost in past is a revelation

to most people, but the investigation of 2004 is just begun.

What the reader will take away from this article, and therefore what it "says", is that the 2004 election was "better than the past". Stewart,the author, says something similar to that in the quote attributed to him.

A similar "communication" style is used when talking about HAVA and DREs. They like to say "HAVA requires DREs" or "HAVE requires touch screens" both of which are untrue. HAVA requires accessibility for the disabled, but these definitely need not be DREs or touch screens.

But perhaps the most important part of the above deceptive statement is what's behind it: The implication is that DREs are a good advanced technology that a liberal reform bill would want the disabled to have.

Our response should be that the disabled should not be MISTREATED by being forced to vote on inferior, malfunctioning technology. It's a nightmare that one would tell a disabled person they are "helping" them vote and them give them a DRE. what is this? Blind voters should vote blind with untouchable vapor ballots?

[mgr]

13. Whats with California?

California was to have required a paper trail, but was generally successful in implementing HAVA, yet has an increase in residuals?

Mike

[Land Shark]

14. This "report" sets up a logical circle:

1. DREs *set out* from the beginning to reduce undervotes, as an advertised product feature.
2. Seemingly reduced undervotes are then the only evidence cited in order to claim DRE success.

Gee, isn't anything else relevant to the success of a voting technology? Counting? Not breaking down? Not creating bottlenecks? Not costing a lot of money? Not having phantom votes? not counting votes secretly where citizens can't observe? Not creating problems of public confidence? etc.

Again, the underlying but unmistakeable message is that DREs are successful and good. That needs to be challenged, as well as the evidence being actually that undervotes were increased in many places on DREs, and apparently assigned to the wrong candidate (as prevoting was witnessed by some voters) in others.

Because this underlying message is surmised by many readers, they don't question it: it's "their" thinking, not "outside" info to be evaluated.

[minorjive]

15. Spoilage rates really only matter at the county and precinct levels

State and national averages do not show anything about the disparitites that exist from precinct to precinct. We all know that spoilage rates tend to be highest in minority precincts. You can have states, like Virginia, that as a whole fell below the national rate in 2000, but had counties with rates above 10%. See the the Harvard Civil Rights Project, Democracy Spoiled for a whole study on this matter. Their basic conclusion is that pretty much as a rule, the more African Americans in a county, the more ballot spoilage there is. Hmmm, wonder why that is. What follows is an adapted excerpt from a longish thing I wrote about this over the summer. You can read my whole essay here:
http://minorjive.typepad.com/hungryblues/2004/08/we_who_believe_.html

Georgia, Illinois and Texas all had more spoiled ballots than Florida. Georgia, Illinois, Indiana, New Jersey, North Carolina, South Carolina and Wyoming all had higher rates of ballot spoilage than Florida. And there were significant rates of ballot spoilage in many other states.

Charts from Democracy Spoiled:
AL-MT

NC-WY


But consider this warning from the authors of the study:

This “where the missing ballots are” approach is useful in considering where improvements in election system performance will contribute most to assuring the greatest possible number of citizens that their votes will be counted. On the other hand, this is a very different agenda from ensuring that each state is as successful as they should be at treating all voters equally, or that voters in different states are treated equally. (2)

The more blacks, the more spoiled ballots.





But also notice that as the percent of blacks increases, the variance among counties also increases. When you get to counties where the spoilage rate is upwards of 14%, you also have counties in the same state with almost no votes lost. States like Florida that have among the worst rates of spoilage also have counties with with excellent, near zero spoilage rates. To be clear, the Harvard study finds that "as the white voting age population in a county increases, the spoiled ballot rate correspondingly decreases."

Examining the 100 counties with the worst (highest) spoilage rates nationwide, our analysis also found that 67 of these have black populations above 12%. Of the top 100 counties with the best performance (lowest spoilage), the reverse is true – only 10 had sizeable black populations, while the population of 70 of the counties was over 75% white. (8)

Thus you can have states like Virginia. Virginia's state-wide spoilage rate is 1.91%, just slightly below the national average of 1.94%. That's a lot better than Florida, whose state-wide average is 2.95%. But these state averages are virtually meaningless. In Florida there are predominantly white counties with rates as low as .27% and predominantly black counties with rates as high as 12%. Though Florida is above average and Virginia is "below average," the disparities in Virginia are similar to those in Florida.

<>

In Virginia, there are some counties with rates as low as .37% and others with rates as high as 10%. (5-6)

And the problem does not abate merely by taking income, education and other factors into consideration. In Florida, racial disparities in ballot spoilage across counties persisted even when comparing counties with identical income, education, and other factors. Taken together, these statistics reveal the persisting vote dilution of blacks on a county, state, and national level. (8)

[bleever]

16. Kick.