Voter reciepts: A way to stop voter fraud once and for all.

[Hope springs eternal]

You know how when you go to a store you get a reciept for your purchase? Well, I feel it should be a constitutional right to get a reciept of our votes. Yes, I realize that this means people would know how you voted, but chances are, they already know that anyway. That way, you can check for any irregularites and given enough compaints, form class action lawsuits to overturn stolen elections. Do you think it will work?

[EFerrari]

1. HI there, Hope and welcome to DU.

A receipt is a good idea but you can't recount a receipt. That's why we need paper ballots because those can be legally recounted.

[rhiannon55]

2. It works with ATM machines

why not voter machines? To quote the evil Cheney, "Isn't that a no-brainer?"

[bowens43]

4. A printed receipt will do nothing.

The electronic device can easily print a receipt showing the you voted for the Democrat while actually recording the vote for the Republican.

[bowens43]

3. Nope, It won't work.

Just because an electronic voting machine spits out a receipt that indicates that you voted as you intended to vote, doesn't mean that the machine recorded your vote that way.

A receipt would do nothing but give us a false sense of security.

[bryant69]

5. There's no system that can't be hacked

Chicago ran crooked elections just fine in the bad old days.

Bryant
Check it out --> http://politicalcomment.blogspot.com

[bowens43]

6. That's true but nothing is easier then hacking electronic devices.

It's much more difficult to hack paper ballots.

[bryant69]

8. I don't agree.

Each system has weak points and strong points.

Bryant

[karlrschneider]

10. Yeah, you're right. Let's just dump elections and forget the whole thing.

...

[bryant69]

17. I wouldn't really agree with that.

I'm just saying that both paper ballots and electronic voting have strengths and weaknesses, the important thing is to figure out what those are and plan accordingly. Obviously Diebold is not interested in such an examination.

Bryant

[karlrschneider]

29. Well, I think that goes without saying. Not many criminals are interested

in having their activities investigated...
;-)

[bowens43]

13. What are the strong points of electronic voting machines?

[bryant69]

16. Simplicity of use, great ease of uniformity.

Also fewer humans to corrupt, which is both a strong and a weak point.

Bryant

[karlrschneider]

22. I can't agree with that. They aren't 'simple' for millions of people

who aren't techies like many of us are...and there are 50 different 'types' in use. Also, ONE person can affect a entire precinct with half a minute of mischief.

[Atman]

26. You've obviously never seen the workings of an old lever machine

You cannot cheat with an old lever machine. You just can't. It is MECHANICAL. You can't have it flip ONLY one party's votes as you can with an e-machine. There are redundant safe-guards, there are mechanical counters that all must match up -- not at the "end of the day," but every half hour or so. Each lever machine has a "grand total" counter, and two daily counters. Mechanical. Once you throw the lever, the machine is locked. It is not unlocked again until you enter the booth and the poll worker physically unlocks it and readies it for use. The point of this is, while cheating has always existed, it was DIFFICULT and often OBVIOUS under the old system. Even the cheating left a paper trail which could lead to prosecution. To cheat with the old machines you need to do traceable things like have DEAD PEOPLE vote. With e-machines, you need to do nothing but give a huge campaign donor a contract and allow allow him to label his machine's software as "proprietary."

The number of humans needed to corrupt an old lever machine -- and promise to keep their mouths shut -- is infinitely greater with electronic machines.

.

[bowens43]

28. Oh my.

To rig electronic voting nationwide requires only one person per manufacturer.

[SharonAnn]

12. And it would allow people to "buy" and "sell" votes. Here's the way:

It's VVPB with a robust auditing process.

- Voter sees a printout that confirms who/what they're voting for.
- Approves it by "casting" the ballot (hitting a button).
- That ballot automatically feeds into a sealed box (like punch card ballots used to put into).
- Machines and precincts are randomly audited and the printed votes are compared to the tallied votes on the machines.
- Any discrepencies and the entire precint or entire election is manually counted.

- The manual count trumps the machine count.

Without the Voter verifying something that is tangible that can be audited later, and without the audintg of the process, there is room for much chicanery.

I work in computers, financial systems, audit procedures and designing business processes. I have 40 years experience at this, starting with puch cards and checks at banks.

Technology can ease the "Input" process by having flexible ballot formats, multiple languages, and handicapped accessibility. Technology can give a "quick" count of what's in the machine. But there is no way at this time to prove that machine has operated the way it should have, that there are no bugs or that someone hasn't compromised the code. Auditing is REQUIRED!

[karlrschneider]

15. I like your logic but why do we need the machines at all in this case?

A check mark in a box on a paper ballot achieves your sensible goal without any gadgets except a pencil!
;-)

[Atman]

30. Watching the "hanging chad" debacle didn't convince you?

People sitting at desks counting pieces of paper? THAT is what you consider "safe?" YIKES! What is to stop a shill from counting every OTHER piece of paper? I just don't agree that paper and pencil are a good alternative. Seems rather third-world to me, and wide open to abuse. Go back to big, clunky, ugly, PURELY MECHANICAL machines. They keep a tally of vote counts, and you can't flip votes for one without flipping them for all. Like most any other mechanical devices invented at their time, they're rock solid, reliable and secure.

How rock solid, reliable and secure is any Windows-based computer YOU'VE ever used?

:shrug:

.

[karlrschneider]

31. I never came even close to defending a windoze-based machine.

Absent the 'chads', counting a plainly marked paper ballot is completely straightforward and simple, admittedly slow but it beats the shit out of vaporizing electrons.

I'm unfamiliar with the clunky mechanical machines, never saw one. How are their 'tallies' transferred to a master record (and preserved) so it reflects an accurate total?

[Atman]

40. A piano-roll type of paper

You crank the lever to record your vote, it punches it onto a piano-roll, which is later archived. Therefore, there is no way to have only a certain number of votes switch to any one party. You can't rig just a few DEM votes going to the GOP, because EVERY dem vote would have to go to the GOP, and it would be as obvious as getting hit in the nose with a brick.

Each machine has at least three mechanical counters. One is the grand total of all votes ever recorded on the machine. It cannot be reset. The two others are redundant, and must match at all times. If they don't, someone has messed with the machine, and in my precinct, at least, the machine is REQUIRED BY LAW to be locked down and taken out of service...even if only ONE DIGIT off. A poll tender must press a mechanical button which unlocks the mechanism each time a voter enters the booth. If he doesn't, the levers won't function. When the voter pulls the lever to record his vote, it advances all three mechanical counters one digit, and locks out the machine until the tender unlocks it for the next voter.

When each voter enters the precinct, their name and address is checked against a printout of the voter registrations, and if they match, they're allowed to proceed to the voting booth. Every half hour (at least in my precinct) the registrar polls each machine against the log of voters who've checked in at the precinct. All the numbers MUST match up. IOW, if the "checkers" at the entry point record 300 voters, but the counters on the machine record 301 votes, that machine is SHUT DOWN. It is removed from service. There are even two redundant keys on opposite sides of the machine which must be turned, like a nuclear bomb launch you see in the movies. You cannot bring the machine on or off-line without the two keys. Even if you ACCIDENTALLY turn just one of the keys, that machine is considered invalid. These machines are at once incredibly simple, and brilliantly complex. You just can't fuck with them.

ALL cheating must be done at the purely human level, ie, "dead voters" and such. There simply is not way to get the machine to do the cheating for you.

.

[karlrschneider]

43. Thanks for the explanation! I assume 'hanging chads' aren't an

issue with them ;-)

How do the individual totals (for each candidate, issue, etc.) get 'extracted' from the machines?
Is the 'piano roll' put into some other device that does the tally? (I assume there isn't a separate mechanical counter for every possible item on the ballot)

I'm not questioning the machine's integrity, just wondering about some details I never knew anything about.

[Atman]

46. The rolls are read just as a player piano would read notes

Again, since it is all mechanical, all the parameters are set up in advance. IOW, the positions of the counter mechanism is set up to correspond to the position of the candidate's names on the machine. If Candidate X is in Column A, every time the counter reaches a hole in Column A, Candidate X gets a vote. There is no way to make it partial. You can't make the mechanical counter skim off votes, it could only take ALL votes from one column, and that would be so laughably obvious as to be impossible.

.

[karlrschneider]

49. Okay, I'm just trying to learn here. Many player pianos use no electricity,

only mechanically amplified puffs of air...do the voting machines work like that? Are the votes 'counted' by the same machine that 'manufactured' the 'piano roll'? Or is that done with a separate gizmo?

[Atman]

50. No electricity involved. Purely mechanical.

The only electricity used is for a light bulb to illuminate the inside of the booth. The tabulator is basically a mechanical counterpart to the voting booth...the machines "write the music" onto the roll, the tabulators play it back.

.

[SharonAnn]

61. Mechanical voting machines can be hacked. They're just gears

and they can be altered. It's tougher, but it's been done before and can be done again.

Yes, people at a desk counting pieces of paper, with other people watching them count the pieces of paper is the best disinfectant there is for vote counting fraud. Remember, sunshine is the best disinfectant.

The only real advantages that electronic machines can have are two:
1. Their flexibility in presenting ballots (short one, long ones, really long ones), presenting ballots in many languages, providing handicapped accessible alternatives for voting (audio, "sip 'n puff")
2. Their ability to do quick counts or tallies of vote counts.

To take advantage of those, you need to provide security at the electronic level but most at the process and procedures level and at the audit level. The audit level is the most critical for finding problems. That's where the "sunshine" comes in. Without that, you've got a corruptible system.

Again, I've got 40 years of working with banks financial systems, suting financial systems, auditing manufacturing systems and processes, computer programming, and I've been the Election Officer for six election cycles now and, with others, guided some changes in our county's election handling procedures.

Please believe what I'm telling you.

[Atman]

67. I suppose you're right. I could machine some new gears...

...with every 57th tooth removed, and hope the missing teeth resulted in votes for my particular candidate. The mechanical skills to work on one of these 1920's-era machines is so readily available :sarcasm:, we actually keep spares around just to cannibalize repair parts (not sarcasm), because no one makes them anymore, and there is only one mechanic in the city with the knowledge to fix them. Seriously. If one breaks during the voting day, we just verify the vote count up until that point, file a report, lock the machine up and roll out a replacement if one is available.

COULD one be altered? Sure. Easily, imperceptibly, and with confidence that the alterations would ultimately help my candidate? No way. For that, you need computer chips.

.

[orleans]

69. thanks for all the great info on these mechanical machines

i never knew anything about them either (except that they looked huge and you pulled a lever)

i have been of the mind that we need to banish the dres

my precinct uses paper and we fill in the circles with a black marker. HOWEVER, when this great paper ballot is filled out the voter slides in through an automatic counter which i just happened to notice in 2004 was manufactured by DIEBOLD

(my heart sank)

my solution is: paper ballots are the best, you fill in the little oval with a black marker or pencil, and every fucking one of them gets hand counted, and counted again several times (just like we used to do in a data research company i worked at--the phone surveys would be taken and then everything got hand counted by three people and the tallies had to match or we had to recount a stack of papers)

but these mechanical machines sound fairly secure as well--

anyway, thanks for the lowdown on them.

[grasswire]

37. the notion of "buying and selling votes" pales...

...in comparison to the damage that electronic voting can do in just moments, with just one machine and one person willing to subvert democracy. Think about the complicated, messy task of trying to steal an election by buying the votes of a couple of thousand people, or even a couple of hundred. Somebody in that group is going to squeal, and somebody's going to prison.

But a crook can crack a voting machine and steal a goddam presidential election in moments.

A possible solution is the introduction of 3-part NCR paper ballot. Everyone knows this stuff. The pink copy goes home with the voter, the white copy goes in the ballot box, and the yellow copy goes into a lockbox held for a number of years in case there's a recount needed -- the paper trail.

NCR 3-part paper is being used for ballots in emerging democracies. Why not here?

[Hope springs eternal]

18. if certain preciencts recorded enough irregulaities

then a class action suit could be had

Think about this:

An average polling place has about maybe 8-10,000 voters using it. Let's say your in a 54-46% democratic district. If your precienct voted red, then you and other democrats get together and tally the votes. If you end up getting serious irregularites, then it's suit time. Every precent has a record of weither it went red or blue (it has to), so if you suspect voter fraud, then you and fellow democrats could take the ovte yourselves.


note: you don't need to know who voted republican to find out weither there was irregualites in the vote, a simple comparison of democratic votes to registered voters should do the trick.

[karlrschneider]

24. Convincing a jury, either in civil or criminal proceedings with that sort

of evidence would be a hill I wouldn't want to climb. The very kind of irregularities you posit were rife in 04, pretty well documented and produced zero satisfaction.

[karlrschneider]

7. A receipt by itself isn't useful. What we must have is a paper with

our vote clearly indicated that gets locked into a box IN the polling place and saved for at least
a year. If you really want something to carry home, that's fine but by itself doesn't provide any
fallback for recounts.

[Tess49]

23. Why not print out two receipts? Like they do in restaurants.

One goes in a box, the other goes to the voter.

[karlrschneider]

27. I don't have a problem with that, it's just more paper. But how does

the second one accomplish anything? I guess I don't get that part. If I can deposit a piece of paper that's human-readable in a locked box, showing my self-verified vote, the second one seems to be just more litter to me.

[Tess49]

32. It's for your records -- just like any other receipt.

[karlrschneider]

35. I understand but it isn't the same thing. If somebody says I didn't PAY

for something, I can use the receipt as evidence that I did. How would a 'voting receipt' be of any value to a voter who believes his vote was improperly counted? I mean, what exactly would you actually DO with it?

[Tess49]

68. Check it against the LIST. See post #11 below.

[TahitiNut]

9. No.

A paper record that offers an opportunity for a manual audit or recount is appropriate. A 'receipt' that indicates for whom one voted is nothing but an invitation to "votes for sale" or extortion.

I personally have little problem with optical scanning IF AND ONLY IF there's a manual audit of a STATISTICALLY VALID sample of the optically-scanned paper ballots that affords a 98% level of confidence that there tally is materially accurate. (The sampling method and the term 'materially accurate' are technical - too technical to get into here.) When/if the REQUIRED audit does not provide that assurance, a complete recount should be MANDATED without the need for any candidate to either file a formal challenge or pay for it.

[OnionPatch]

11. I've been thinking of something similar

Each ballot would have a randomly generated unique ID number. Each ballot would have a stub and the main body, both with the number on it. You would tear the stub off and keep it before you drop the ballot into the box. Then, just after the votes are counted, there should be some publicly accessible list on the internet or just a hard copy posted at each precinct with all the numbers and their votes. Each person could verify that their vote was counted as they cast it. No one has to know how anyone else voted.

[karlrschneider]

14. The problem with that is this: Suppose you find out your vote was

'incorrectly' counted...what do you do then? Not much unless your 'stub' also has a record of all your votes, and if it does, we're back to the 'buying votes' problem.

[OnionPatch]

19. Can you explain it in more detail?

I don't quite understand how this would make it easy for voter fraud, or buying votes.

[karlrschneider]

21. Okay I'll try. If your 'stub' only shows you voted,

and not -how-, specifically, you won't have any evidence (other than your undocumented claim) that your vote was wrongly counted. Anybody can claim anything, after the fact. On the other hand, if the
receipts DO show, in detail, how you voted, you have enough direct evidence that you may have voted
the way someone wanted (and paid you) to do. Not you personally, obviously, you see what I mean, nu?
:-)

[Atman]

33. What about the "secret vote" aspect?

If you get a receipt that shows how you voted, how do you stop it from recording WHO voted, and if you have to do that, how do you stop Karl and Diebold from simply programming the machines to give "correct" receipts for "incorrect" votes? Remember, NO ONE IS ALLOWED TO VIEW THE CODE. I can't believe anyone would think computers are safe for ANYTHING like this. A good programmer can write code that makes these machines show anything they want. There is simply NO WAY -- NONE -- to make a computer a reliable voting device UNLESS it is purely, 100% open-source and available for 100% public review. NONE. It cannot be done. Period.

.

[karlrschneider]

36. Are you sure I'm the person you mean to argue with? I have

consistently said the exact thing you have. :eyes:

[Atman]

42. Hey, don't roll your eyes at me!

:hi:

No, I'm not sure you're the person I meant to argue with. Maybe I read your post wrong. Whatever. If we're in agreement, then I guess we agree, right?!

.

[karlrschneider]

45. Haha...I'm not sure, actually. I'm NOT in favor of touch-screen

machines of ANY flavor, as they exist so far. I just replied about the mechanical gizmos, it's down there someplace...
Hey no foul, no harm. :D

[grasswire]

39. which is the greater danger?

The unlikely possibility that a campaign could manage to buy a hundred or a thousand or five thousand votes, considering the mechanics of that and the surety that someone is going to squeal and then someone's going to prison?

Or the ease of one person alone with one voting machine for just a moment -- long enough to throw millions of votes with no chance of detection?

[karlrschneider]

41. I assume the question was rhetorical...and I don't think buying votes

is actually a big problem but I brought it up as response to some posters who expressed concerns about the possibility. But then, one machine -> millions of votes might be a tad hyperbolic, eh? J/K, obviously I consider it extremely serious.
KS

[grasswire]

51. might be hyperbolic

But the information that one strategic entry into the system at one terminal can spread the "glitch" to the whole system via the network means that arguably a whole state's system could be infected.

[Tess49]

25. I was thinking this very same thing. Couldn't figure out how to

explain it as well as you did without making it seem too cumbersome.

[Atman]

20. I don't understand why anyone thinks a receipt will do anything

Often I can't even get a printout on my Mac to match what is on the screen.

RECEIPTS and PRINTOUTS are not the problem. The machines themselves are the problem. Period. The entire receipt issue is a waste of time. Imagine if there was a voting debacle...as if that's going to be tough! What will they do, ask everyone to bring in their receipt? Are you SERIOUS?

Just eliminate the e-machines and you eliminate the problem, and we go back to good old fashioned "dead people" voting and all the other nonsensical bullshit. The old lever machines are iron clad.

.

[napi21]

34. I've only heard of one system that sounded like it would work.

I believe it was Sequoia, but not sure.

The voting machine is a touch screen computer, but linked to it was aprinter that was sealed under glass. The voter selected his choices, checked the summary screen and hit cast me vote. THEN the results of that vote was printed ON PAPER for the voter to see to make sure his vote was correctly recorded. The printed "receipt" (if that's what you want to call it) then drops down into a sealed collection bin for use ONLY if there needs to be a recount OR to perform an audit on the machine.

The problem I've heard about actually giving the voter a paper receipt to walk away with is that some unscrupulous politicians (Imagine that!) might pay voters for proof that they voted a certain way when they bring their receipt to them as the proof.

Isee no reason the paper printout that is sealed but visible to the voter wouldn't put your mind ad ease about the votes being accurate, and at the very least, could be audited!!!!

[karlrschneider]

38. In Oklahoma we have a system that I'm fairly comfortable with:

It's the "Op-Tech" optical scanner. We mark a paper ballot (they're printed on some decent weight cardstock) with a marker like a Sharpie, drawing a short line between 2 boxes. It is eminently simple and straightforward. Then we stick the ballot into the machine which sucks it in, tallies the votes then the paper ballot is fed into a locked box. I need to find out how long they keep them, though.
I have watched the machines which display total ballots counted and they appear to be accurate on
that particular count (they don't show how many votes the various candidates have received, obviously, on the front panel)

[Atman]

44. If it computerized, it is hackable. Period.

Even the system you describe relies upon computer chips to tally the votes. Computer chips in any way shape or form are simply NOT reliable. ANYTHING can be hard-coded onto the chips, and as long as the government allows the manufacturers to claim their chips and code are "proprietary," in contradiction with all US voting laws, then there is no way to examine what code has been placed on those chips.

.

[karlrschneider]

47. I'm not sure these machines claim proprietary software but I'm gonna

find out this week. I guess I have a little more faith in electronics than you apparently do (NOT ignoring how they might be configured by fallible or crooked people) since I fly an airplane that's
largely operated by 'chips' and I don't much think about potential failures. Of course there's a lot of redundancy built into our avionics and flight controls. In any case, I absolutely think the voting machine code must be open sourced.

[Atman]

52. Your plane's chips are different in one way...

Unless someone stands to gain a brazillion dollars in insurance money by rigging a chip which will bring your plane down, I don't see why anyone would bother. Not to mention getting the access to the plane's on-board computers. OTOH, since e-machines are portable, and have been proven to have been taken home by poll workers, and since one person in a remote location can screw with the chip on a voting e-machine in order to help gain control of the government of the United States, the threat is much more real.

.

[karlrschneider]

57. I get what you're saying but I think you might be confusing pure hardware

with firmware and/or software. Obviously any system that isn't 100% hardwired can be compromised
by the alteration of memory (whether machine instruction or data). I'm just saying I trust the 'chips' to do what they are built or told to.

[napi21]

54. Yes of course. Any MACHINE is hackable if you want to look at it

that way. The onlyaccurate system would be total paper with PEople counting them, and even THENI question the accuracy of ability of a payoff.

The Idea is that in the optical scan and the machine I described above, there IS a documented paper trail. Similar to what you get from an ATM or the cash register at your local store. The problems would have to be severe enough to cause a question, but at least you have some hard copy to go back to!

[Atman]

56. No, ANY machine is NOT hackable.

Please read my posts about mechanical machines.

I'm a poll worker; I started as a machine tender, which required (by law) a training session in the operation of the machines. I was an Assistant Registrar of Voters in 2004. I don't proclaim to be an expert, but I know these machines pretty well. I've examined them literally inside and out.

.

[karlrschneider]

59. Do you think there might be one of those for sale up there in CT

somewhere? I'd buy one just to tinker with it...and I'm willing to give odds I could 'hack' it.
:D
(I'm serious)

[Atman]

62. I'm serious...you couldn't

Unless you simply wanted to switch all dem votes to GOP, or vice versa. I'm not sure you're getting the point about how mechanical these machines are. What would you hack? If you mess with the counters, you'd have to make the Master Counter, which has recorded every lever pull since the machine was manufactured, and is recorded in the registrar's log, match the Daily Counter, which has a redundant mechanical counter and they all must jibe at all times. The machine can't be opened to mess with in the first place without TWO keys on opposite sides of the machine being turned. I'm glad you're confident of your abilities, and you can "tinker" all you want, but there is simply no way to make such an old mechanical machine do anything out of the ordinary that would not be readily, instantly apparent to anyone with any familiarity of their inner workings.

But I'd love to see you buy one off of E-bay and give it a try. You could probably make a lot of money selling your hacking attempt, a la Geraldo and Al Capone's vault. The results would be about as spectacular.

.

[karlrschneider]

63. I'll check on Ebay. I've done some very interesting things with

teletype machines (both model 26s and 28s)...they do have one tiny electrical component (the selector magnet) but are otherwise 99.9% mechanical, VERY complex. :D

[Atman]

66. You do realize these machines are the size of Volkswagons, right?

This ain't no teletype machine! It's made of thick steel plate, and the shipping alone would probably cost you a fortune.

But I'd love to see you try!



We actually still use these exact same machines my Connecticut precinct.

http://americanhistory.si.edu/vote/votingmachine.html

.

[meldroc]

48. Not just No, but HELL NO!!! Welcome to vote buying and extortion!

We have to preserve the secret ballot, or we will see extortion and vote buying.

Ron Rivest, cryptographer and security guru, and the R in RSA, came up with a clever way to try to deal with this problem: the ThreeBallot Voting System.

http://theory.lcs.mit.edu/~rivest/Rivest-TheThreeBallotVotingSystem.pdf

I wouldn't mind seeing a system like this in use, though I would want the known weaknesses addressed first. ThreeBallot is damned clever, but there are ways to cheat.

[RC]

53. The best way to vote is to apply the KISS principle.

Keep It Simple Stupid.

A mark put on a piece of paper by the voter's hand.
Count the marks in public to find out how many people voted for a candidate or for/against whatever else in on the ballot.

If it takes a week or more for the results, so be it. This having to know who won 10 minutes after the polls close is just another invitation for election fraud.

Electronic voting has no place in a fair and open election.

[napi21]

55. Now you sound like our old Chairman when the VP's were trying to talk him into installing our first

computer. He believed in paper and pencil too.

[karlrschneider]

60. Computers are great for doing things people cannot.

Many people can still count integers.

[napi21]

64. Having been a Director of accounting for a number of years,

there were many times I question your statement! You can't believe how many PEoPLE can't seem to do that right!

[Mayberry Machiavelli]

58. The key to any system is a 3d party audit, accepted by all. This would

require an acknowledgement that there are problems in our electoral system, or potential problems, large enough to warrant such oversight.

I don't see such acknowledgement coming from Republicans, corporate media, or the average complacent American.

[napi21]

65. You're absolutely right! That's why I favor the Sequioia machine

with the sealed printer attached. The voter doesn't get a piece of paper to take home, but that paper remains with thevoting machine which gives you the ability to perform an audit!

Right now, there is very little to audit. All I can think of is do the results from each precinct total the results reported from the final collection point? And do the number of signed in voters equal the number of votes at the precinct level. I'm not even sure if anyone is checking THAT!