Say NO to Automatic Recounts ! Say YES to Mandatory recounts !

[harmonyguy]

I've seen a number of posts calling for Automatic Recounts, and while I'm quite sure I understand what was intended, I'm also quite sure that there's some unintended room left for manouvering.

We need to make sure that no one leaves wiggle-room with the terminology we all use. If one supports a voter verified paper ballot, hand counted with mandatory hand recounts, then one needs to call it that.

An automatic recount could be easily interpreted by some Secretaries of State to be a recount performed using automation, in other words machines and/or computers.

If we let words like paper audit trail, paper reciept, paper backup or automatic recount creep into our vocabulary, we run the risk of diluting the message, and we will end-up with some half-baked, encrypted copy of a cash register reciept to file in your wallet along with your charge card chits, that can only be recounted by each voter returning them to their place of voting.

And while we're at it, most of this has to do with Election Fraud, not Voter Fraud.

There likely are other words with which we need to be careful, but these are a start.
HG

[New Earth]

1. w00t

And while we're at it, most of this has to do with Election Fraud, not Voter Fraud.

good job. :thumbsup:

[harmonyguy]

3. I put that in just for you Faye !

As I was ranting at my keyboard I spotted your earlier post, and figured I'd better include it too. Thanks for the reminder!
HG;-)

[electropop]

25. Good framing! n/t

[New Earth]

2. oh yeah,

and we want PAPER BALLOTS, not 'paper trail'.

[Democrat Dragon]

4. ONE BIG KICK FOR DEMOCRACY!

[LeahD]

5. Do you really think we'll ever get paper ballots

after states have spent a fortune on voting machines?

[Democrat Dragon]

6. that can be arranged

easy, we need money to buy those machienes and examine them for pre-proggramed fraud. The money can be used for whatever the hell the government wants to use it for.

[Bill Bored]

8. Sure, all you need are printers.

And welcome to DU!

I'm not going to get too involved in the debate about whether to go 100% paper or not for the reason you mention. But in states that haven't gone with touchscreens yet, there are paper options such as OpScan and hand counting. If they've bought the DREs, they can attach printers for these ballots.

[In Truth We Trust]

9. Who cares how much we spent. Throw em away. People at one time spent

money to eat off lead plates too. We spent money on asbestos too. We even put lead paint in our childrens bedrooms at one time. By your logic then we should all be eating off lead plates in an asbestos laden house with a toxic paint job.

Paper ballots NOW!
Hand counts NOW

Great original post.

[LeahD]

11. Personally, I'm all for dumping the machines

if retrofitting them with printers like a previous poster suggested isn't feasible. I tossed out the question, only because I foresee serious resistance based on cost issues.

[Amaryllis]

34. Yes! Check this thread. Chuck Herrin insists we need paper ballots and

hand counting. And he has counter arguements for all the arguments about why we shouldn't. He is a e-security expert; hacks for a living and is also a republican.

http://www.democraticunderground.com/discuss//duboard.php?az=show_mesg&forum=203&topic_id=286388&mesg_id=286388

[Bill Bored]

7. Good points

It's "voter-verified paper ballot" (VVPB) and "mandatory random audit", I think. This way you don't make it sound like you're demanding the recount based on some margin in a close election, or other recount law. Audit means an inspection of an accounting procedure and records which is really what you're trying to do.

What do you think?

[Carolab]

12. I don't understand why we would request mandatory audits.

Just count the damn ballots at the precinct.

Let's say there are 1500 voters in a precinct. How long does it take to count 1500 ballots?

At my caucus, we had about 200 people in our precinct. We distributed, voted and counted (twice), signed off in less than an hour. So you should be able to count 1500 ballots in 8 hours or less.

The way it always worked in the past was the pollworkers always hand-counted the ballots. Who cares if we don't have the results for a day or two?

The fact is, the media have been calling the shots here. It's up to the people to tell the "elected" elections officials what WE want, and the media can wait their damn turn for the results instead of "counting" for us and telling us who won. That's how the fraud has been committed all these years--through the polling companies as an extension of the media.

[LeahD]

13. I agree that we could wait a day or two for election results

It really irritates me that the media is hell-bent on calling elections. It's as if everything has to be wrapped up in a neat little package so everyone can go to bed at 10:00 PM knowing who their President is.

[Carolab]

14. It's a standard they decided for us. I don't remember screaming for this.

Do you? In fact, I think they should STFU until the results are all in and counted. Their "predictions" are a crock anyway and only serve to allow the "adjustments" so exit polls/actual totals can line up. What a con job this all has been, and for so many years. Our voting "rights" were usurped a long time ago.

[claudiajean]

15. Well, if there are 75 contests on a single election day ballot...

...as is the case for many General elections in my home jurisdiction, multiplied by 1500 ballots, assuming 30 minutes of counting per contest for 1500 ballots, then we are talking about an additional 37.5 hours of counting after the polls close.

That's a little over a day and a half of hand counting at the polls, commencing when the polls close.

I can be placed firmly in the camp of endorsing Voter Verified Paper Ballots - my preference is optical scan ballots with automatic recounts (Sorry, HG, I can't endorse mandatory hand recounts. Big waste of tax dollars in many, mnay cases) for close races at a threshhold trigger, random full hand audits of selected races as an aspect of robust auditing, Full audtiting of procedures and balancing of voters to votes, etc.

But hand counted paper ballots for every race in every jurisdiction is just overkill. And grossly unnecessary overkill, which wastes local tax dollars that could otherwise be providing critical local services in danger of being cut.

[Carolab]

16. Optically scanned ballots have a high rate of "error".

In fact they are very susceptible to fraud at the "tabulator" level and it can't be discerned without a full hand recount.

The way our "elections" are going now, recounts are costing tremendous time and money as it is.

Paper ballots, hand-counted, are the way to go, according to every expert on the subject.

[claudiajean]

18. Um, I AM an expert on the subject. Ask Andy Stephenson or Harmony Guy.

I hate having to defend myself like this, but pretty much all of the major players in election reform nationally know who I am in real life, so, while some terrific people for whom I have great respect may disagree with some of my professional opinions and engage me in discussion and debate, I've never before had my opinion dismissed out of hand as unworthy and out-of-alignment with that of "experts".

Indeed, I was the individual who originally uncovered Jeff Dean's criminal background and his activities as related to Diebold elections programming. I was the original whistleblower over four years ago. I did not have the media resources and contacts at the time to publicise the story, so I provided the story to Andy Stephenson, Andy and Bev publicised the information, and the rest is history. Ask Andy. I'm sure he will be happy to promptly confirm.

I would be happy to discuss in-depth the security challenges, and how they can be addressed, of the central tabulator on optically scanned systems. Diebold/Global/GEMS is my particular area of expertise, but I can discuss in-depth the common security issues to any central tabulator.

I have been used as an expert witness and/or technical election administration consultant by attorneys across the nation who have litigated against DRE's in a number of states. I provided expert consultation in cases in Maryland, Colorado, Florida, Washington, California, and most recently, Andy Stephenson had my private line on speed dial for the important work being done by Susan Truitt and Cliff Arnebeck in Ohio.

I have also provided expert assistance to producers for 60 Minutes, MSNBC, as well as over two dozen nationally-recognized print publications.

I went to law school for the sole reason of honing my legal skills to match my election administration skills. One of my first public legal contributions to the debate was the notation that has become Andy's signature line - the legal analysis of what meets the test of being a "ballot", and why only paper meets that definition. I wrote that.

I have a book on election error and fraud in the revision stages that is being considered for summer publication by a major publishing house. Hopefully, it will provide a progressive take on election problems, to counter John Fund's book.

And, considering that I am an election administration, American election law, and and election computing expert, the statement "paper ballots, hand counted, are the way to go according to every expert on the subject," is not really accurate.

You can count this election expert firmly in the camp of advocating a voter verified paper ballot, preferably optically scanned as the initial methodology for the count, that can be used for robust auditing and hand recounts in close races. But I don't advocate hand counts for every race. Races that are not close do not require a re-count or re-audit by hand. That's just a waste of time and money that can be used to more carefully audit close races on the same ballot.

Further, the most accurate counts of a close race or contest are hand counts following a machine count that can be cross-audited for accuracy. Not an initial hand count, nor reliance solely upon a machine count, but an appropriate combination of the two.

Elections must balance substantive competing legal, resource, and human demands. Funding resources are not unlimited. Time before certification is short. I always take professional positions that use resources wisely while providing for fair, accurate, transparent elections that ensure a level playing field for all parties and participants.

(Actually, knowing the other national experts I regularly interact with, I think you would find that I am hardly alone in my advocacy position.)

As a final note, don't take my low post count as an indication of anything other than the fact that I rarely comment. But I have been posting for months, and lurked for two years prior to that. I am not a newbie.

[Bill Bored]

26. So what percentage of randomly audited paper ballots

would be enough to detect fraud or error with reasonable statistical certainty? Is 5% enough? There are some bills that say 2% or 3% and some that don't even include random audits?

As I've said before, if you want to count 100% by hand, that's just a 100% random audit, so there's no problem with using the TERM random audit. It's just a matter of degree.

Thanks for your thoughtful comments, esp. the ones about multiple races, etc.

I'm thinking that OpScans can probably be off the shelf programs not even produced by E-Voting companies, can't they? Otherwise, how do they mark all those SATs?

There was also a large study, which I'm unable to cite right now, that showed OpScans worked well. And of course they require paper ballots!

[Carolab]

29. I take it you disagree with Lynn Landes on this, then?

She stood at the first Conyers meeting and made the statement that paper ballots had much better reliability than others. Or did she mean to say optically scanned paper ballots?

Why does VotersUnite propose hand-counted paper ballots?

http://www.votersunite.org/takeaction/handcounted.asp

The "central finding" of a 2001 CalTech/MIT study was that, of all voting systems used in the United States, hand counted paper ballots have the lowest average incidence of spoiled, uncounted, and unmarked ballots.

- Errors in the software, firmware, and election-specific ballot programming of both paperless electronic voting machines (DREs) and optical scan tabulation machines have caused hundreds of election problems in recent years, including high levels of uncounted and unmarked ballots. It is unreasonable to believe that all such errors have been detected.

- Manual recounts of optical-scan ballots have overturned initial, inaccurate machine results in many such cases. It is only reasonable to believe that the outcomes of many other elections (both DRE and optical scan) have been inaccurate, and the inaccuracies were not detected.

- Computer-counting errors have a much greater potential impact than hand-counting errors.

- The electronic voting systems used in the United States, both optical scan and DRE, have severe and unresolved security and accuracy flaws that are not being remedied by election procedures.

- While we advocate the use of computers to assist people in marking their ballots, computers cannot count those ballots reliably.

[claudiajean]

19. Specifically, the error rate is 1/10,000. It's not all that high. n/t

[Stevepol]

20. Audits can do the same thing that a paper election can do, BUT

the audits have to be mandated. There have to be voter-verified paper ballots available for EVERY election and audits MUST BE MANDATED FOR EVERY ELECTION. If a good percentage of the ballots (truly randomly chosen) are counted (maybe 5%), that would be enough to show that the machines are doing a number on the electorate. Then, the law should require that the whole election be hand-recounted in that state. Barring that, I don't see any way to reverse the tsunami of RightWingism that continues to wash over the US, wave after wave.

[claudiajean]

22. We are in agreement...

...I see the key as writing into the election statutes the levels of required random audits for each election (chosen by lot or other means of chance selection), the threshhold that triggers automatic recounts, which should be required anytime a race reaches a certain close level, and requiring full hand audits whenever specific anomalies occur that would indicate a security breach in either the tabulator or with ballot handling and chain of custody.

[Bill Bored]

27. Wait a minute...

The random audits should be on every race, right? You're not saying to rely on MARGINS to trigger audits, are you? Hope not. Margins mean nothing now with hackable machines. The LAWYERS and legislators need to be informed about that. It seems to me they don't have a clue. Look at how Kerry & Co. handled Ohio. They were useless!

[many a good man]

21. Cost is a MAJOR factor; that's why we need federal standards

cj is right about costs. The public sector is so woefully underfunded in this country that even honest public officials are SO restrained by lack of resources that they can't do a good job even if they tried.

The HUGE number of races and questions on the ballots in some places make hand counting everything impossible. We must double spending on elections and/or separate out federal, state, and local elections to make it more feasible. In the current climate of state budgets that's just impossible.

The harsh light of scrutiny on the recent elections in Florida and Ohio merely expose a broken system that's been riddled with problems for generations. The only way to fix the system is to overhaul it. We must re-design the entire system based on state-of-the-art technology, methodology and practices borrowed from the business sector and social sciences. The only entity with the resources and power to do such a thing is the federal government.

The federal government must institute national standards for its elections. It must design the fairest most efficent election system humanly and technically possible and give the states funding to implement it. We shouldn't be afraid of electronic voting or tabulation because it CAN be made transparent and trustworthy.

Electronic systems have numerous advantages over paper-only systems: accessibility, efficiency, and redundancy. In the end it must produce a voter verified paper ballot which is the be-all end-all. Nevertheless, other more efficient means of tabulation should be employed. Multiple redundancy could be designed and used for auditing purposes. Here are some of the ways votes (both paper and electronic) can counted:

1) hand count the paper ballot
2) opti-scan the paper ballot
3) tabulate electronically the memory card from the DRE for each voter
4) tablulate electronically the results written to the DRE's firmware or hard drive

Random cross-checks between any or all of these methods would guarantee reliability. Truly random hand recounts in a set percentage of precincts must employed. Any mismatch would mandate a hand recount of the PAPER ballots- the final arbiter.

The software should be open-source and made as secure as possible. The feds should buy ALL the machines for all the precincts in the country. They should all use identical hardware and software so each one can be re-imaged with the cerified software and audited at all times for tampering. It can be certified and locked down. It can be made more trustworthy than ANY system ever used anywhere in the world.

America is the richest and most powerful nation on the planet. We MUST have the BEST election system of all the democracies in the world. This is crucial to maintain our legitimacy and credibility in the eyes of the world and its own people in the tumultuous century that lies ahead.

[claudiajean]

23. Hear, hear. An excellent summary.

Having Federal funding, and an open-source, publicly owned, scalable system (smallest county in the US has 74 voters, largest has 5 million, so scalability is also an issue) provided to each election jurisdiction by the Federal government would both releive part of the resources burden and make elections less of an unfunded mandate, as well as provide security and consistency, which would be well-appreciated by every honest election official I have ever met.

[In Truth We Trust]

28. I vehemently disagree. You are literally putting a price tag on freedom.

paper and hand counts are the only way to go. It is doen properly and efficiently in most progressive countrie with no problems at all.
Your premise that the time is factored out by the number of contests is somewaht flawed as well. It does not grow the time exponetially as multiple volunteers are used and are responsible for counting certain races which means a simultaneous count. I suggest in all due respect you research further before you disseminate false and misleading information.

Our greatest priority I am sure you would agree is fair and honest elections to which the best possible method must be paper ballots and hand counts period. I hope after you've done some research you would agree that this is the most preventative method of massive wholesale fraWD.

[davidgmills]

35. I would never eliminate exit polls

They are the only check on the accuracy of the system. I'm in favor of doing what the Germans do.

To me it's like two different instruments to tell time (mechanical
clocks vs. quartz watches).

At the end of the day, you want to know whether they have the same time.

They are two different ways to getting the same final number and they should check out.

You simply can't count 120 million anything without some kind of a statistical back up. We were purportedly off by six million in the last census (according to statistical models) and I believe a census count is much easier to do than a vote count.

I say include "mandatory random exit polling" as well.

[In Truth We Trust]

10. Kickage!!

[claudiajean]

17. The answer is to include defininitions within the legislative intent...

..at the beginning of the statute. that'll clear up any confusion right quick.

(And you know I respect your work, HG, so please accept my comments in the spirit of discussion that I intend them, since you know my background and experience...)

My problem is that "mandatory" sounds to me like it means every race hand counted during the canvass period as an audit process. And that would be a big waste of time and money. Most races are not close enough to require either a hand count, or a machine count. recounting a 90% - 10% race is not a useful exercize.

I also feel that a machine recount of Voter Verified Paper Ballots (as you know, I oppose DRE's, but I'm fine with optical scan ballots as there is a paper ballot for auiting if needed.) for races that are close, but not extremely close, is a fine option. Every recount need not be immediately tipped into hand recount land. The key is audit procedures that will detect individual reader tampering, central tabulator tampering, and will detect compromised ballot handling and security.

[harmonyguy]

24. Woops, I wasn't clear enough myself!

You're right! Mandatory hand recounts addresses the 'essential-ness' and the 'non-automatic-ness' but implies every precinct and every race.

That's what I said but really isn't what I meant.x(
I'll try again.

It should be mandatory that;
In each and every polling location, for each and every ballot counting machine, one or more races be selected at random, and the ballots cast in that race be subjected to a hand-recount, to validate the accuracy of the previous count.

This is not intended to conflict with whatever rules are in place (and probably need changing) regarding close races. It should be viewed as being in addition to.

I think we're saying the same thing, just I didn't explain it very well - I was hung up on Automatic meaning Mandatory.

YOUR expertise is substantially greater than all of the rest of us out here, and while we can come up with all sorts of crazy ideas, you can temper them with the wisdom of reality.

HG ;-)

[Bill Bored]

30. kick n/t

[Carolab]

31. HAND-COUNTED VOTING BEST according to Voting Technology Project

HAND-COUNTED VOTING BEST

Hand-counted paper ballots were found to be the best and most accurate way of voting, according to the Voting Technology Project conducted by political scientists at Caltech and Massachusetts Institute of Technology (MIT).

The Voting Technology Project compared the reliability of voting systems used nationwide from 1988 to 2000 and came to a remarkable conclusion: "The most stunning thing in our work was that hand-counted paper ballots were better than anything else," project director Stephen Ansolabehere said.

This happens to be the exact conclusion reached four years ago by The Spotlight newspaper after its seminal investigation.

The Caltech/ MIT report found that as many as six million ballots were not counted in 2000. Of 800 lever machines tested, 200 had broken meters that stopped counting once they hit 999, but touch-screen machines were even worse.

The evaluation of voting systems found that touch-screen voting systems performed worse than the mechanical lever machines, optically scanned paper ballots and hand-counted paper ballots during the 2000 election. Only punch-card machines performed worse than touch-screen systems, which raises the obvious question: Do we need expensive electronic voting machines at all?


http://americanfreepress.net/08_25_03/Concerns_Over/concerns_over.html

[Democrat Dragon]

33. As for hand counts

Hand counts has easily be rigged too. To avois this I suggest putting high-quality security cameras, the media, and bi-partisan supervisors in the counting rooms.

[LeahD]

32. Thank you, claudiajean, harmonyguy and TorchesAndPitchforks

for your informative posts. Those of us who are new to the electoral reform movement learn so much from those of you with expertise on the subject.