How to Create a Paper Trail

[G Edward Cook]

As the "Wild Card Candidate from Pennsylvania" I am working on protecting our
vote.
Please check my site GEdwardCook.com

How to Create a Paper Trail
"If we did this in 2004 John Kerry would be President today"
Many of us are trying to get rid of the electronic voting machines and demand a paper trail
with little success. I would like to see a 100% paper ballot; however, this is not going to
happen in 2006. There is a way to create a paper trail without passing any new laws. How?

The Democratic party must set up validation tables at all the polling stations. These validation
tables would be armed with affidavit ballots. An affidavit ballot is an affidavit that says: " I,
(voters name and address), do solemnly swear that I voted on (date of election) as follows:"
We then have a ballot followed by a place for the voter to sign and a place for a witness to
sign. These affidavit ballots would be legal proof that would stand up in court. These ballots
could also be checked with voting records to prove that these voters did vote.

The Democratic party would also have workers at all the polls to pass out and encourage all
voters to protect their vote by filling out an affidavit ballot. I believe we could get nearly
100% of Democratic voters to fill them out.

After the election is over, we would keep these ballots in a locked box until after the official
count so that any person attempting to change the vote count would not know just how many
affidavit ballots there are at any given polling station.

After the official count, there would be a count of our affidavit ballots. For example in Ohio,
if we have a total number of ballots for Paul Hackett that equals over fifty percent of the
votes, we have proof Paul Hackett Won. If we had this kind of proof in 2004, John Kerry
would be President today.

Even if we don't have fifty percent of the votes we could still prove fraud at many polling
stations. It is extremely important to PROVE FRAUD.

Thank you, and please check my site GEdwardCook.com for "How To Exhume an
Election". My e-mail address is [e-mail]

Sincerely,
G Edward Cook

[mazzarro]

1. How feasible is this?

I am in favor of us doing something to counter the rethugs efforts at stealing another election. They sound so cocksure of maintaining their "fraudulent" winning streak that I am afraid our hope, opportunity and efforts will come to naught.
The question we have have to answer is "does the Democratic party have the interest and desire to invest in such a project?"
My view is that the party is too timid and is led by compromised people (lap dogs of corporate influence) that we will not get the support needed to put any safeguards against election fraud in place anytime soon!

[NanceGreggs]

2. Way to go!

I work in the legal profession, so I have a tendency to always think in terms of lawyers, affidavits, etc.

It was my first thought in 2000, after the vote fraud debacle in Florida. If they wouldn't allow a re-vote, why couldn't lawyers and notary publics 'set up shop' and invite voters to swear affidavits as to who they voted for?

I think it's a GREAT idea!

[G Edward Cook]

14. Check Out "How to Exhume an Election

Hi Nance,
Check out "How to Exhume an Election" to be posted on DU soon. It's on my website now. This sounds like what your talking about.
Thanks for your support. I see you live in a Bush Free Zone, I spend a lot of time in Montreal!

[Stevepol]

3. But the voting machine guys would claim the need for a secret ballot.

That's the first thing they say when you mention a VVPB: you're guaranteed a secret ballot. But maybe your ballot could be identified with a number and you could just sign your name to a separate sheet as validation of having voted.

Sounds like a good idea anyway to me.

[G Edward Cook]

4. No secret ballot needed!

You vote then go outside to a private table, not in the control of the voting machine people. Voters could choose to fill out the ballots. No laws would require a secret affidavit ballot.

[Febble]

7. If voters choose whether or not to participate

your sample will statistically worthless.

The only useful outcome (which indeed may be useful) is that you might have a better case for a recount. But it won't tell you the result.

[G Edward Cook]

13. Only Democrats need to vote!

I feel if Democrats take the time to go out and vote they will be willing to protect their votes at any cost. I think we could expect 100% in most precincts in Ohio,
"The Voter Fraud State"

[ISUGRADIA]

28. But Democrats are only about 38% of the electorate

you forget the 1/3 plus that are Independent. Many people who are Democratic voting Independents and Light ID dems are not likely to ID their vote to strangers.

[G Edward Cook]

31. Good Point

These independent voters will be encouraged to fill out the ballots as will all voters. Each Independent that fills out a ballot in one less vote we can prove the republicans don't have. I said the Democrats must control the affidavit ballots, not put up a sign saying this is a Democratic thing!

[Nederland]

5. It has to be really close to 100%

In presidential races even if you only get 95% it wouldn't be enough...

[Febble]

6. It's a huge problem with

parallel elections. Unless you get 100% response, you've got a volunteer sample, not a random sample, plus, with the affadavit, you've got a big disincentive to participate, as you are sacrificing the secrecy of your ballot, an important ingredient in fair elections.

My own view is that a well-conducted exit poll would actually be better, with random sampling, anonymity guaranteed to participants, good coverage (at least two interviewers at all times, preferably working in shifts, and well trained) and experienced analysts.

And I say that as someone without much faith in exit polls as a means of auditing elections.

BYU knows how to do it:

http://exitpoll.byu.edu/

And I say that as someone without much faith in exit polls as a means of auditing elections.

[Nederland]

8. Question

What was the non-response (refusal) rate in the NEP 2004 exit poll?

[Febble]

9. Here is a plot:

of refusal rates by Bush's vote share.

In the E-M report cross-tabs for refusal rate are given on 37-39, and are in the 30 percents. One cross-tab has precincts where voters were said to be "co-operative" at 34% and where they were "not co-operative" at 51%. In addition, "miss" rates account for about 11% ("misses" are supposed to be recorded when an interviewer selects a voter but is not able to approach them for some reason (busy with other voters). However, it is possible that misses may also have been voters who managed to avoid being even asked. Between the two, the mean response rate was about 53% (hence the notorious "hypothetical" 56%:50% Kerry:Bush participation rates).

The reason I post the plot, however, is that it shows that the variance was huge. In some precincts the refusal rate is suspiciously low, and one wonders whether interviewers were failing to report refusals and simply selecting a new voter instead (there is anecdotal evidence for this). In others it is very high.

However, despite the potential for non-response bias where refusal rates are high, Nth voter sampling does at least, ensure that voters are selected randomly, and guards against over-representation in the poll by enthusiastic voter groups, which one might, dare I suggest, in a parallel election, tend to be irate Democrats wanting to register their vote somewhere they believe it will be counted. I think the San Diego parallel election bears further study.

[OnTheOtherHand]

10. just to elaborate briefly on the "anecdotal" bit

there is an anecdote, pretty reliably sourced IMHO, of a polling place where the actual participation rate was about 2% of voters. Considering that the interviewers are apparently supposed to interview at least every tenth voter, you can tell that the non-response rate (refusals plus misses) was pretty high.

If I understand the OP, he really is counting on close to 100% participation among Dem voters at least, and I just can't imagine. I don't think any parallel election has come close to that. Of course, if voters become widely convinced that it's essential, then who knows?

[G Edward Cook]

11. That's not the point

We don't need the republican voters, only democrats and as many third party voters as possible. We know the total number of people that voted from the registration that you must sign before you vote. This is public record for each election. We also know who voted.
The total number of votes less the democratic votes and the third party votes that we have affidavit ballots for equals the maximum number of republican votes.
Say there are one million votes and we have 500,000 affidavits for paul and 1,000 ballots for the third party candidate. This means the total republican votes cannot be more than 499,000 that would be:
Paul 500,000
Republican 499,000
others 1,000
Paul Wins and can PROVE IT!
We wouldn't be able to tell in a close election but we could show each precinct and prove some. Many of the stolen elections have been stolen by over 10%.

Check out "How to Exhume an Election" to be posted soon. Also there are voice clips and more on my site GEdwardCook.com

Thanks

[Febble]

15. Good point.

I hadn't heard that methodology before.

Could work. Can't see a flaw offhand. And no statistics required. Glory be!

[Nederland]

16. You miss the point

In the example you give you are assuming that you can get affidavits from every Democrat and a fair share of third party voters. Given the data Feeble gave above, I'm not sure those assumptions are realistic. Also, real elections contain a much larger number of independents than you are using in your example--and independents may not share your concerns about election fraud. Who knows though--pick a precinct in the next election and see if you can make it work.

Good luck!

[Febble]

17. Febble not

feeble! I wish I wasn't stuck with that stoopid handle. It's from my initials febl.

I think we both missed Cook's point, actually, me being stuck in a statistical groove. I don't think he is trying to get a representive sample, just a minimum number of non-Republican voters, so that if the total non-Rep is less then the minimum non-Rep, it will be clear that non-Rep votes have been either lost (over-votes; undervotes) or switched.

The more complete the participation by non-reps, the smaller the degree of vote-count loss will be detectable. But the nice thing about it is that it isn't inferential.

Subtracting non-Rep votes from total seems a bit tortuous, though. All you want is for as many voters as possible to participate. If the result in a precinct is 500 Rep votes, 450 Dem votes and 50 Green votes, and you have affadavits for 460 Dem votes, 200 Rep votes and 20 Green votes, then you can't say anything about the Rep or Green votes, but you can say that at least 10 Dem votes went uncounted.

Lousy sample, but that doesn't matter. The only problem is, the smaller the Dem vote-loss you want to detect, the larger your Dem participation rate would have to be.

I like it.

[OnTheOtherHand]

22. I think Nederland got this one right

Not that you necessarily have it wrong. But if you reread #16, the point wasn't about statistics at all. It was about whether it is remotely plausible to imagine that one could get, say, 98% of all voters for candidate X to sign affidavits to that effect, so as to limit the possible vote fraud to 2% of the candidate's vote. Or whatever.

And Nederland offers probably the only possible conclusion: "Who knows though--pick a precinct in the next election and see if you can make it work."

[Febble]

23. So it was

my bad.

Yes, the whole thing depends on getting something near 100% participation from voters for the candidates whose votes you want to check.

But I'm more enthusiastic about it as an approach now I know it's not something anyone is going to attempt to do inferential statistics on.

Sorry, Nederland. As you were.

[G Edward Cook]

12. 95% of Democrats would be enough.

We could collect more after the election. We will have the names and addresses of all the voters. It is public record!

[Febble]

18. Yeah, the lovely thing about it is

it isn't statistics.

[G Edward Cook]

19. Correct: it is not a poll

You have that right. It is proof that will hold up in court.

[eridani]

20. Use the search function on this board to find--

info about Judy Alter and parallel elections. Similar to what you are proposing, except she's non-partisan.

[Febble]

21. Judy Alter's San Diego

study was very interesting, but is the reason my first reaction to Cook's reaction was somewhat negative. They did their stats as though it was a poll.

Non-partisan sounds fine, as long as you don't treat it as a poll. But if you aren't treating it as a poll, Cook's partisan approach is probably better - what you want (he wants) is as near to 100% of Democratic voters as he can, and being partisan might help.

If you can get everyone who voted for a particular candidate to sign an affadavit that they did, then if fewer than that many votes are counted, you've got direct evidence of miscount (no inferential statistics required).

But if you want to infer, from a bi-partisan attempt at a sample, that the proportions in your sample reflect the proportions in your count, you are back with the problems of non-participation bias, which are likely to be greater if the anonymity of the voter is has to be sacrificed.

[G Edward Cook]

24. It Must be Partisan!

All are welcome to fill out an affidavit ballot but the democrats must be in charge of this ballot!
The Republicans are stealing the elections so the Democrats must PROVE FRAUD. this MUST be a partisan effort.

Does anyone have another idea that will be set up in time for the 2006 election? We can and have made some headway but the Republicans still control Ohio and will not let laws pass that will hurt their ability to be TRAITORS and steal the 2006 election.

Will we wait until 2007 and talk about how the 2006 election was stolen or will we act NOW?

[Febble]

25. Yes, I get it now.

But you still need VERY high participation from Dem supporters, as Nederland says.

Good luck!

[AmBlue]

34. With MoveOn's Leave No Voter Behind project...

...in 2004, we asked our voters before election day to come see our poll worker outside the poll and let us know they voted so we could check them off our lists. No one had a problem with that. If Dem pre-election efforts included a request to follow up by checking in with the Dem rep outside, I think it would be doable.

One question: If we are only tracking Dems, how is that going to help when they jack up the Rep vote numbers to counter this effort? Are we also going to be balancing # of voter sigs with machine register totals? That is the only way I see that this could work.

[G Edward Cook]

35. Yes, you have that right!

We use the total with voter reg Sig's. We would like to get third party candidates to fill them out to. We could get more affidavit ballots signed after the election. We will know the names of all the voters that did not sign an affidavit. We need only to prove the count wrong in parts of the state and we prove the whole election is wrong! Polls are not proof. See How to Exhume an Election on my site GEdwardCook.com
Thanks, G Edward Cook

[OnTheOtherHand]

26. I have to wonder whether the best use of 10,000++ Ohio activists

on Election Day is to tie them down collecting affidavit ballots. (I'm not sure how many polling places there are -- probably fewer than 10K because many polling places have multiple precincts -- but then, if you station just one person per polling place, you are gonna have huge miss rates.)

Folks, this is an example of why I think 2004 is actually worth being analytical about. If you're convinced that the Dems won Ohio in 2004 but votes were actually miscounted or tampered with, then this may be an excellent use of people power. If you're not convinced, then in an off-year election, actually getting people to vote would probably take precedence. If you think the election was stolen through registration purges, then people power should be focused on getting people registered and making sure the registrations get processed. And so on. Not to force a false choice, but I haven't yet seen the movement that had an excess of volunteer resources to devote to all possible purposes.

(EDIT: Of course, if we're talking Ohio, there wasn't a single competitive House race in 2004. I don't see any Dem candidate who got between 42% and 60% of the vote. We'll see what shakes out in 2006.)

[G Edward Cook]

30. Too Late

If we wait for 2006 it will be too late. We can now EXHUME parts of the 2005 election that were way off. See How to Exhume an Election on my site GEdwardCook.com
Thanks

[EFerrari]

32. Then, the pool has to grow.

In Ohio, do you want to bet we could get help from the NAACP, for example?

We know that every tactic was used in OH and that Blackwell invented some new ones.

We have to mind registrations AND protect Black and student voters AND mind the polls. Because all three are at risk. I suppose, we could say, protect minority voters, period -- there appear to have been problems in Indian Country and in Hispanic precincts out west.

What use is it to get people to the polls if they can't cast their ballot or if that ballot isn't counted? The Thuggery is very greedy. Once they have stolen your vote, they will not want to return it.

I think this is a great idea.

[G Edward Cook]

33. We need only to prove fraud in a few precincts!

Thanks sfexpat2000,

We have not been able to PROVE FRAUD anywhere in Ohio yet we know there is massive fraud. The affidavit ballot will prove fraud even if it is only in some precincts. If we prove fraud in one precinct we prove fraud in the whole election. If we cannot do the affiliate in the whole election we must set it up in some precincts.

Please check out How to Exhume an Election! If you can't find it here it is on my website GEdwardCook.com

Thanks, G Edward Cook

[onthebench]

27. Back in 2002 the mafia in Italy tried this

They tried to buy thousands of cell phones with picture capability. The system would work like this.

1. Mafia guy stands outside polling area.
2. Mafia guy offers money to voter to vote for his candidate.
3. Voter is given cell phone with camera.
4. Voter takes picture of ballot just before voting.
5. Voter hands camera to Mafia guy in exchange for money.

They got caught when the distributor of the cell phones noticed the unusual order. No with many that have small cameras in phones or even one as small as a Chapstick (I have one this small that I got at a trade show for free), it could be easy to do this. The voter takes a picture of the ballot or voting machine screen. They send the pic to a central website that can then do a parallel tally spot check.

[G Edward Cook]

29. We are protecting the votes

This cannot be done inside the voting booth because in there the vote is secret. After you leave the voting machine and go outside your free to fill out an affidavit ballot.