Navigation
Topics
Log In
Log Out
:
Special Search
New Today
New This Week
Advanced Search
Tree View
Your Account
Edit Profile
Register
Forgot Password
Tools
Help/Instructions
Policies
CLICK STATE TO SEE:
"WATCH LIST"
Marked with: 
"OPEN & HONEST"
Marked with: 
...
|
| 7-2-07: A preview on the audit debate |
|
| Author |
Message |
    
Bev Harris
Board Administrator
Username: Admin
Post Number: 6397
Registered: 12-2004
Best of Black Box? 
Votes: 3 (A keeper?) | | Posted on Monday, July 2, 2007 - 5:55 pm: | 
|
By Bev Harris
You may find the flaws and throw tomatoes if you like, because I did this quickly. Please weigh in to show me the defects in my reasoning.
So, I've been really stressing out about being late on the empty chair debate on the audits, which was due out on this site two days ago.
I'm late because I want to be sure I'm right, so I've been doing election rigging simulations using tactics that I've seen here and there. And it does turn out I'm right.
The flaw in the Holt Bill (HR 811) so-called audits (I call them spot checks or samplings, not audits) -- anyway, one flaw is the assumption that there won't be ballot substitutions and another flaw is that the same PR spin we've seen for 10 years now will suddenly stop working.
Allow me to explain.
I just ran a simulation on a 2,000 precinct state with a 1 percent spread between candidates, rigging 17,000 votes out of 1,600,000. I first did the rig just the way I'd really do it, which is, I selected a handful of counties where I figured I could have someone on the inside to do a substitution of ballots IF (and only if) it was needed if a precinct that we rigged happened to be chosen for auditing.
I clustered together the precincts rigged, assuming that my handful of operatives would have the most access within a limited geographic range. Not really necessary, but I was using tactics from some of the crookeder states where you can find clusters of election whores for sale.
I rigged a total of 60 precincts, and I rigged them massively and boldly. I assumed each precinct had maybe 800 voters average, but fluctuating amounts, so I manipulated a range of 80 votes to 480 votes, depending on how many total voters showed up in the precinct. I just skewed the heck out of them, knowing that red flags do not trigger audits under HR 811.
Then, because it yielded a 1 percent spread in the race, I chose the five percent audit protocol in Holt, and randomly selected 100 precincts for the audit. I used a randomizer to do that.
57 of my 60 rigged precincts skated through without a hitch. The random selection missed them. Three precincts got selected for manual counts.
Therefore, three guys would be getting a quick call in the middle of the night. One precinct required a substitution of 112 ballots; one needed 190, and one needed 400 replaced.
I knew that if one of my guys couldn't get the ballot switch done, no problem, we'd call in a favor from another operative at the vendor who would explain some technical reason to the media and say it's a glitch.
But it wouldn't affect the election. Even if all three ballot substitutions failed, that's only 702 ballots, easy to spin and everything will be swell.
Woops, there's a reporter, let me just answer his questions:
"There were three minor glitches, out of 2,000 precincts, which affected just a few races on only 702 ballots statewide. That's out of 1,600,000 ballots and frankly, we are seeing improvements across the board. Officials are taking steps to identify and address the problems. This kind of thing happens every election. Fortunately, it didn't affect the results in any race. The audits went smoothly and we're very pleased."
Worst case: News reports will say "97 out of 100 precincts passed the audit with flying colors. There were glitches in three precincts statewide, out of 2,000. One ballot box was misplaced in Pox County, there was a computer malfunction that affected 400 votes in Glitch County, and officials are looking into a small discrepancy affecting 112 votes in Huh County."
[Note that of the 100 precincts selected for audits, only three landed on rigged precincts, giving the audits a 97 percent success rate right out of the gate.]
Better yet, perhaps a couple of the insiders get the job done and swap the ballots. "99 out of 100 precincts passed the audit, but in one precinct, a ballot definition problem was identified which changed 400 votes. But it didn't change the outcome of the election."
Best case, all three of my guys on the inside executed the ballot swap. "100 percent of the audits passed, proving the machines work and fears about fraud were overblown." (Much snickering about the conspiracy theorist activists)
Since the state has "winner take all" electoral college votes, more than enough votes were nailed down to capture the presidential election.
"The audits went smoothly and there were no problems."
Ain't "Voter Confidence" grand?
Bev Harris
Founder - Black Box Voting
* * * * *
Government is the servant of the people, and not the master of them. The people, in delegating authority, do not give their public servants the right to decide what is good for the people to know and what is not good for them to know. We insist on remaining informed so that we may retain control over the instruments of government we have created. |
Bruce Sims
Frequent Voting Rights Forum Participant
Username: Ubetchaiam
Post Number: 987
Registered: 06-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Monday, July 2, 2007 - 6:41 pm: |
|
Bev, do you think you'd have achieved the results you did if:
Because elections are increasingly won by slim margins, we MUST affirm a MINIMUM 99% accuracy standard.
To achieve this we must affirm one essential fact with our anonymous ballots:
>>> That the BALLOTS CAST are the indeed the BALLOTS TABULATED <<<
There is only ONE way to do this with anonymous ballots and that is to sample 5-10% of the ballots (on average 25 to 50 ballots) at the
PRECINCT (before they embark on their anonymous journey) on ELECTION NIGHT while they remain under the citizen-controlled purview promised
in our "government of the people."
All processes post-election are inherently several degrees of separation less secure and based on the rigorous compliance of protocols that may or may not be adhered to and are largely invisible
and uncontrollable by the citizens.
A wonderful by-product of a "BALLOT VERIFICATION" is that when the aggregate of ALL the precinct results are examined it will tell us with 99% statistical accuracy whether an election's official results are accurate or not. If not, then an "AUDIT" post-election is more necessary than ever. |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6398
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Monday, July 2, 2007 - 7:21 pm: |
|
Bruce - You have results for 100 percent (at the precinct) that you are comparing with a sample of 10 percent (at the precinct).
You don't have a sample to match the 10 percent to. You have to match 100% of two identical sets in an audit.
What you're describing is like verifying one check on your bank statement, and then saying "therefore all the rest are correct."
What am I missing? |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3887
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, July 3, 2007 - 1:46 am: |
|
This is great work, Bev. Seeing this example is very helpful. The post-audit "explanations" really ring true. Haven't we heard just these kinds of soothing words after countless elections? And reassuring words are virtually always coupled with obstruction of any closer examination, even where known discrepancies are blatant. There are always "explanations" given for discrepancies (e.g., just call it a "glitch" or a "teething problems" or "inadequate training" or "voter mistakes").
Importantly, official/public concerns are usually limited to those discrepancies that were actually discovered, without any apparent concern about those ballots that were not audited. Anyone who has questions about the rest is treated like a tinfoil hat loony or a sore loser or a troublemaker. "Small" discrepancies virtually never trigger a full hand recount, even where this is supposed to happen (e.g. Ohio 2004).
PR about elections is usually carefully managed because no election official (or vendor or politician) wants to be accused of diminishing public "confidence" in elections. |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1396
Registered: 01-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Tuesday, July 3, 2007 - 5:28 am: |
|
Congratulations, Bev, I think you've proven my point that every voter needs to be capable of verifying the recording of his or her own ballot, as they can't reliably do anything else. And every person should be able to work from a completely published record of the election, to audit the entire election, if he/she chooses to do so.
Examine this carefully, will anything else break us out of this situation, where the evidence is possessed and controlled by people that may be crooked?
And Catherine, it wasn't nearly this tight in Ohio, from polling, it averaged 3% down for Kerry and 3% up for Bush, representing a 6% swing, much larger than this, and it still flew (long enough). |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6399
Registered: 12-2004
Best of Black Box? 
Votes: 2 (A keeper?) | | Posted on Tuesday, July 3, 2007 - 5:55 am: |
|
Disagree with your conclusion, Brant.
Examine this carefully and you'll see that hand counts at the polling place in public view using New Hampshire procedures give an election night count with a contemporaneous audit, while preserving privacy, and at about one-fifth the cost of the computerized systems.
Let's not change the way our constitutional republic works solely for the purpose of having technology in elections. Maybe we should look at non-technological options and see how they stack up against technology before saying it can't be done. Perhaps it just can't be done with technology.
Would you make love to your computer? Probably not. There are some places where computers just can't provide the same thing as people. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 64
Registered: 11-2006
Best of Black Box?
Votes: 1 (A keeper?) | | Posted on Tuesday, July 3, 2007 - 1:22 pm: |
|
Bev, the numbers look right to me. In fact one basic sanity check was right on: 1 of 20 precints was sampled and 1 or 20 got caught. Of course statistically none moight have been caught as well.
>A wonderful by-product of a "BALLOT VERIFICATION"
>is that when the aggregate of ALL the precinct
>results are examined it will tell us with 99%
>statistical accuracy whether an election's
>official results are accurate or not. If not,
>then an "AUDIT" post-election is more necessary
>than ever.
*NO*
See:
http://www.bbvforums.org/forums/messages/9707/45345.html?1163928351
It's called a "confidence interval" and has a very real chance of providing absolutely useless results.
Also recall that a basic requirement of this test is that *ALL* ballots have an equal chance of being sampled. Since the sampling takes place based on precincts, or batches, the fundamental basis of this test is *NOT* being adhered to ... In other words the test is *NOT* testing the votes it's testing the precincts although it's presented as testing the votes - It's crutial to understand that point.
(Message edited by ILikeInfo on July 03, 2007) |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1400
Registered: 01-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 7:35 am: |
|
quote:Examine this carefully and you'll see that hand counts at the polling place in public view using New Hampshire procedures give an election night count with a contemporaneous audit, while preserving privacy, and at about one-fifth the cost of the computerized systems.
First, counted by whom? The first neccessity of your argument is that this first count is correct. 'Plain view' has physical problems. If people are chosen significantly in advance, you can't ever be quite sure of that. This requires that the counting be done by unbiased or unswayed or "balanced bias" people.
As for having technology in elections, I don't think that carbon paper qualifies as "technology", do you? And I don't recall mentioning technology here, did I? I won't feel like you're putting words in my mouth if you don't do it. |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6400
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 7:59 am: |
|
quote:First, counted by whom?
Counted by The People.
quote: The first neccessity of your argument is that this first count is correct.
The New Hampshire system is very simple. It has multiple eyes on the ballots. First they are sorted - Bush, Gore, Bush, Bush, Gore, Gore, Gore
Then they are counted. Twice.
Then, in New Hampshire, they have what's called "accessible recounts" -- any candidate can have a recount whenever they want it, of any or all votes, and it's very inexpensive. They do lots and lots of recounts because candidates like this opportunity.
quote: 'Plain view' has physical problems.
The votes are counted at the polling place. I've seen some New Hampshire polling places. They have very localized town meeting halls, places big enough for lots of people to come in. But one good use of technology would be to encourage videotaping, with zoom lens if possible, and to set it up physically so as many spectators as possible can zoom in and capture it all.
Because this happens at the polling place, there are not a huge number of teams in there counting, and the viewing is feasible. You are right that viewing is a problem in the big central count rooms used in other states, and chain of custody is a problem in those situations as well.
quote: This requires that the counting be done by unbiased or unswayed or "balanced bias" people.
Certainly not. The independence of the observers and the physical setup so observers can see the counting is the key. If you can go in and watch the count, with a video camera and zoom in on those ballots, they might be biased but you sure could catch that.
And if you think they did something they shouldn't, in New Hampshire, you can choose any or all locations for a prompt recount at very minimal cost. There are no statistical hurdles there to block your ability to recount the ballots. And unlike a so-called "audit", the recounts, if they produce a change in result, cause a change in who gets installed in office.
Nice system, huh? |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3890
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 9:06 am: |
|
In Ireland, here's a photo of counting observation from our recent national election:
http://www.dieneueepoche.com/pics/2007/05/25/l/2007-05-25-l--20070525092951_IREL AND_ELECTION_LON802.jpg
Our counting system is done in central locations in each county, because of our more democratic PR-STV ranked preference system).
As in NH, any candidate can ask for recounts at any time, as often as they like. There is no charge at all for recounts. |
Sherry Healy
Voting Rights Forum Participant
Username: Sherry_healy
Post Number: 4
Registered: 11-2006
Best of Black Box?
Votes: 1 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 11:27 am: |
|
As to Bev's question about the how the 10% from each precinct would have any meaning, here's my two cents. But first, it's hard to understand this out of context, and so here's a quick synopsis of the two dominant election verification models (one part of the first count and the second a post-election process). Please take a quick peek to see how they each complement each other, as well as see how this would be fairly easy to achieve as a routine "California Two-Step:"
Step 1: Ballot Verification - ELECTION NIGHT
First and foremost a BALLOT VERIFICATION (on average sample 25-50 CA ballots) taken on ELECTION NIGHT at the PRECINCT would affirm that the ballots cast are indeed the ballots tabulated. Secondarily, it would provide at least 99% confidence in the official tally of the election (the validity of the election results) in the statewide and federal races (where it is safe to assume that there would always be a sufficient ballots sampled). The number of ballots required in CA would be 10% (based on the small formula portion of "Titanium Standard" that has graphics which won't port over otherwise I'd just copy and paste, but which is very similar to, and could be replaced with, the Jonathan Simon/Bruce O'Dell sample of ballots method).
Pros: The LESS degrees of separation, i.e. time, location and custody, to affirm at "ground zero" of the election process that the ballots tabulated the better. Because to perform this "BALLOT VERIFICATION" at the PRECINCT on ELECTION NIGHT affords the difficult to subvert decentralized observation for "many eyes" of CITIZEN oversight at every precinct, because it is performed in a territory and time frame CITIZENS can control without interference by the government in power and fulfills the Constitutional promise of a "government of the people." Also, it is notable that this BALLOT VERIFICATION is part and parcel of the "first count," which is essential because we've learned from recent history in U.S. elections that often times that the first count is the ONLY count that really counts. (It's awfully hard to dethrone people once they've been announced victor.) Again, it will affirm 99% statistical accuracy for the statewide and federal races (because there are sufficient ballots sampled and the determination is based on a straightforward math formula). One caveat: This process must be constrained to one evening at the precinct. Because as we learned from the hcpb Philippines the "fix" comes in in the post-election "double checking" of the hand counts.
Cons: It's pretty useless to show problems at the precinct level, and it's pretty useless in small local elections. (Its accuracy is based on a sufficiently large sample of the aggregate of ballots samples from ALL precincts.) The other negative is that it requires an entirely new process, but not difficult process, here, in California, e.g. the hand count of approximately 25-50 ballots on election night. (Of course, less than 50 years ago we used to hand count ALL the ballots, and so I think the precedent has been well-established that hand counts are indeed possible; particularly, a mere 25 to 50 ballots.)
Step 2: Traditional "Spot Check" of Precincts in their Entirety - POST-ELECTION
Pros: This is what we are already accustomed to here in California with our Election Code 15360. It would not be terribly difficult to improve upon it since the bureaucracy for this is already installed. It works magnificently for the small races, because entire precincts are examined. It also works magnificently when it's lucky (based on statistically calculations to improve those odds) to hit upon a precinct where there are shenanigans. It confers a separate methods and oversight, which inherently provides superb redundancy to the first step, the "Ballot Verification," that is a bona fide "checks and balances" in our election system.
Cons: This "Spot Check" is currently inadequate as it's presently practiced in California is sloppy, e.g. but our 1% is probably not the statistically optimal (and yet reasonable) number sufficient to the task and it isn't genuinely random and sometimes the "manual" hand count portion is just someone feeding a machine a second time. It plays out in a centralized locations where, despite laws that provide "citizen oversight," the reality is that citizens cannot be certain that the ballots tabulated are the ballot cast. Also, as we've learned from Los Angeles's Conny McCormack, we can be locked out of the process on a whim and left with no legal recourse, despite the laws on the books. It plays out post-election, which as we learned from Ohio's recount, is a process and findings that can be ignored. Also, depending on what improvements are deemed necessary, the "weighting" factors can be disputed.
<b>What to do about discrepancies?</b> When either of these methods determine a discrepancy that would overturn an election, then the "Spot Check" method increases with incrementally larger samples until the problem is determined. Sometimes, this process will lead to a full hand count.
<b>Conclusion:</b> a simple examination of 25-50 ballots on election night, combined with a ramped-up EC 15360 (precinct "spot check") should complement each other to overcome each of their inherent weaknesses. Does anyone have any objections or a better idea? }
<b>P.S. Would hcpb be a better process?</b> If the "New Hampshire method" were used (with at least TWO separate oversight coming to the same conclusion, at the PRECINCT, on ELECTION NIGHT); specifically, if all the ballots could be counted in one night using without changing personnel. Otherwise, I believe this "Two-Step" process has less opportunities for mischief.
(Message edited by Sherry_healy on July 04, 2007) |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 67
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 1:06 pm: |
|
>This requires that the counting be done by >unbiased or unswayed or "balanced bias" people.
I'd also add that so long as the ballots are closely monitored to prevent tampering I see no reason why machines can't be used to count the ballots as many times as you wish *IN ADDITION TO* hand counting.
Both the numbers should match, if they don't there is a problem that needs to be solved.
Again, so long as the physical ballots are monitored.
I don't recall anybody ever saying a machine count is bad, only that it can't be the last word, and that a human count must take place, and "voter fingerprint" must exist on a unique physical ballot.
Cross checks are allways good, it's just a matter of which side of the check is given what level of weight. Clearly the machines are the easiest tamper point (in addition to the movement of the ballots through the system) so humans and 100% recount need to have the final say. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 68
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 1:17 pm: |
|
>First and foremost this would affirm that the
>ballots cast are the ballots tabulated.
>Secondarily, it would provide 99% confidence in
> the official tally of the election in the
>statewide and federal races (where it is safe to
>assume that there would always be a sufficient
>ballots sampled).
I'd like a link please.
These numbers make no sense to the graduate level statistics classes I took in college. Thanks! |
Sherry Healy
Voting Rights Forum Participant
Username: Sherry_healy
Post Number: 5
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 3:39 pm: |
|
About achieving 99% statistical proof of the accuracy of the official tally, again realize it is based on the aggregate of ALL the precinct samples. In other words, it tells you whether the election results may be relied upon or not for the statewide and federal races. And then there's a second phase that will determine the accuracy of local races. And more.
http://www.califelectprotect.net/Titanium.pdf
For the math, see page 9-14 of the above document, or go Google the "Universal Ballot Sample" of Election Defense Alliance by Bruce O'Dell and Jonathan Simon for details. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 70
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Wednesday, July 4, 2007 - 8:46 pm: |
|
Thank you for the link.
No where in that document do I see mention of the *CONFIDENCE INTERVAL* of the proposed measurements. Why is it we yet again see an "authoritative document" that leaves out such a critical value?
From the document:
>The Random ample of allots the “RSOB-Formula”)
>must be ccurate within +/-1%margin of error to
>offer at least a 99%level of confidence.
I'm left no choice but to conclude that elections with voting differences of less than 1% are of no concern since the proposal clearly states it's not intended to be usefull in such cases since it does not have sufficent resolution (ignoring confidence intervals).
Correct me if I'm wrong but the elections with the greatest voting *DIFFERENCES* are least likely to have been manipulated. Throw on top of that the electoral college then we invite manipulation at the very points in the system with the closest races since maniplution can't be detected by the test.
*ANY* batching of any sort, such as that of the electoral college, county by county, state by state, invites manipulation at the close "batched races" since the test, by admition, can't detect the true results or maniplution there of.
So, in turn, I'm left to conclude that the closest of races should actually *REQUIRE* the greatest of scrutiny and therefor a +/- 0% margin or at least the lowest margin that humanity (democracy?) is capable of. Read: Hand count.
>About achieving 99% statistical proof of the
>accuracy of the official tally, again realize
>it is based on the aggregate of ALL the precinct
>samples.
http://www.m-w.com/dictionary/statistics
2 a : a quantity (as the mean of a sample) that is computed from a sample; specifically : ESTIMATE
My head allways hurts when I hear the word "proof" and Statistics used in the same sentance.
Statistics exists for the vary purpose of describing or modeling that which is otherwise impossible, or too expensive, to obtain a closed form expression for. If you have a closed form expression, ie the ballots *EXIST* and are countable, there is no *NEED* for statistics. Statistics are a lazy convenience. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3891
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 12:48 am: |
|
quote:Correct me if I'm wrong but the elections with the greatest voting *DIFFERENCES* are least likely to have been manipulated.
Not necessarily. And particularly not if a known winning margin will trigger audits or recounts. Having a known regulation (if the winner wins by X% or less, then a hand recount/spot audit/audit of x%/all votes is required) means that someone manipulating the system can use that to their advantage by ensuring their win is by a percentage just large enough.
Also, some of the most suspicious wins have been so-called landslides that by any method of scrutiny are simply not credible. (E.g., Chuck Hegel's landslide win, the first time his company's voting machines were used--what a coincidence--in which a popular incumbant was thrown out; or Georgia's Governor race a number of years back in which an extremely popular sitting Governor was supposedly defeated by a landslide, coincidentally also at the same time black box voting was introduced). In these cases, since they were apparrent landslides, no one questioned it. (At that time, we also didn't know about the extent of the vulnerabilities in these technologies, and their ease of being hacked.) These voting technologies--and election officials--typically don't allow real investigation of results, even when the results defy belief.
Once one establishes a set of statistical triggers, some people with the capability of gaming these systems (using electronic voting machines on their own and/or in combination with other election-rigging techniques, such as selective voter disenfranchisement) could fine-tune software triggers so that only the needed amount of manipulation will be carried out, and in whatever combination of precincts is most likely to avoid scrutiny or where, as in Bev's example, local party operatives can be counted on to help out if their area is selected for an audit. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3892
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 12:51 am: |
|
Another example is the Ohio referenda on election reform, where the official results were the reverse of what all polls indicated. Like the examples above in Nebraska and George, we are asked to believe that there was a complete reversal of voter preferences overnight, with no credible reason. |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1401
Registered: 01-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 4:42 am: |
|
Unless you institute a system where "The People" that Bev talks about have custody of the ballots start-to-finish, instead of static elections officials having custody of the ballots, I think you are going to have incidences of "the terrorist threat from the FBI call" or whatever anybody else can dream up to deny "The People" access to the ballots and/or honest counting of the ballots.
Until you remedy this, ab initio, I believe that these problems are going to remain. Don't get me wrong, I hope this works (I know how hard it is going to be to get what Bev is advocating, I shudder to think how bad the chances are of getting something stronger), I just honestly don't think that it will.
And Catherine, having viewed that picture, I think the chances for slight of hand under those circumstances are enormous.
And Catherine, we're not asked to believe it (the huge vote shifts from polling shown to be accurate in parallel races), they don't really give a damn whether we believe it or not, they just want us stuck with the results. As V. Kurt says, the default is always that the vote as it was counted, stands. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3893
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 5:08 am: |
|
Brant, actually the security is extremely tight.
There are pairs of counters, scrutinized closely by multiple persons from various political parties. There's no chance of sleight of hand by the counters. (You'd have to come over and observe in person to understand this. For example, in the sorting of ballots as each ballot is unfolded it is placed in full view so that the observers can see it. It's a fundamentally different mentality--the observers and tallymen are welcomed as a crucial part of the whole process, and it's understood that they need to be able to see what happens. The setup is geared to facilitate that, rather than to obstruct it.)
All counting centers also have lots of police present throughout the count. Security is tight, which is not obvious from the photo. There is no way an observer could reach over and grab ballots or even touch them without being arrested.
Observers and candidates are never allowed to touch any of the ballots. This is even true when questionable ballots are reviewed with candidates or their representative or lawyer (i.e., ballots that are spoiled or that cannot be clearly interpreted for some reason). Only the top election official can touch the ballots as they are shown and discussed. Candidates (or their representative) will have had the opportunity to see every single spoiled ballot. These can also be subsequently re-examined in the event of a recount. |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6401
Registered: 12-2004
Best of Black Box? 
Votes: 3 (A keeper?) | | Posted on Thursday, July 5, 2007 - 6:50 am: |
|
Brant, with polling-place based hand counted paper ballots, "the people" DO have custody from start to finish.
When people frame the hand count at the polling place vs computer counting debate as Luddites vs modern technology (not that you are doing that, but others do) they are really missing the relevant elements:
- Chain of custody with HCPB remains in the hands of The People, as long as the procedure has ballot box in full view all day long, opening the ballot box is done in public, and counting is done in public without the ballots being taken to any other location first
- Freedom of information, right to know is recognized with HCPB, but not with proprietary computerized counts. With open source computerized counts, right to know is subjected to "the literacy test from hell" and chain of custody is not preserved without extraordinary, cumbersome, and somewhat expensive measures.
- With properly administered polling place-based HCPB, a 100 percent audit takes place contemporaneously with the first count. With all computer-based systems, the audit takes place after the fact and with breaks in the chain of custody.
So we have to look at HCPB not just from the framework of technology, but from the framework of chain of custody, authentication/audits, and Freedom of Information. When you do that, it gets more and more expensive and complex to try to bring computerized counting to anything even resembling a par with HCPB. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1109
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 9:40 am: |
|
Hi Bev,
Back from vacation.
I see only a couple of practical problems.
1) In many places, especially northeastern cities, there is an extreme lack of ADA-compliant polling places that have anywhere near sufficient size to host a large number of observers. Some polling places in Philadelphia, PA have literally been abandoned 12-foot wide row homes in terrible neighborhoods. I have never seen these firsthand live, but I have on TV. This is partially a self-inflicted wound because Philadelphia has insisted on having very tiny geographic precincts with few or no public buildings in them many times. Polling sites are MANY times used that BARELY have enough room for a check-in table and a couple of booths to mark ballots.
2) In major cities in the northeast, especially places like Philadelphia, the police are used in parts of the election infrastructure, but are frequently suspect. When you have a place where the police themselves have a strong suspicion of being lawbreaking "thugs", their presence is hardly heartwarming. We have explicit language in the PA Election Code greatly restricting police presence in or around polling places. Philadelphia openly flouts that state law, and no one calls them on it. The Fraternal Order of Police local is among the most politically connected and active unions, and seldom misses an opportunity to formally endorse political candidates.
3) Your system does NOT SEEM to take into account what I call the "Schuylkill County problem". Initial count takes place. All observers agree on the counts. A putative "loser" requests a recount in a court of law. Recount occurs in a courtroom. Putative "loser" now wins, even though the count on Election Night had him losing by several percentage points. In this actually done scenario, the paper ballots had been altered after the fact by the Election Director at the behest of the majority party to change undervotes into valid votes for the favored candidate and valid votes for the opponent had been altered by the same Director into overvotes, and hence cancelled. Remember that the Judge hearing the case was also put into power by the same virtually single-party political machine, and has little incentive to look past the "prima facie" record established in his court? How does the chain of custody STAY protected 24/7 through the entire challenge period? Who acts as "the public" doing the watching during all those several weeks 24 hours a day?
Keep in mind that in many rural places (like where you were in Kentucky, I'm guessing) the police are among the MOST POLITICAL people on the landscape. Their presence frequently diminishes, rather than enhances, public confidence. Frequently not without reason.
(Message edited by Formerelecdir on July 05, 2007)
(Message edited by Formerelecdir on July 05, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1110
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 9:48 am: |
|
One further point:
All recounts are fabulously expensive. The only question is:expensive to whom?
Is it more right to make recounts "free" or cheap to candidates and parties, only to burden the public at large?
As someone once put it on another subject, "If you think health care is expensive now, wait 'til it's free!" |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6402
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 9:56 am: |
|
Kurt,
Very glad to see you back! Thank you for the input on Pennsylvania polling locations.
Wouldn't it make sense to first get data from New Hampshire to see if their recounts are as expensive or less expensive, and then to examine the procedures to learn why?
I will wager their recounts are not expensive to either the taxpayer or the state, because from what I'm seeing, those are some frugal yankees up there. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 71
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 10:12 am: |
|
>Not necessarily. And particularly not if a known
>winning margin will trigger audits or recounts.
>[...]
>(E.g., Chuck Hegel's landslide win, the first time
>his company's voting machines were used--what a
>coincidence
Let me clarify, agreed the electronic systems lack of transparency creates an easy to manipulate system. I was talking to a robust system such as ballots since then the foundation of the overall system is far harder to tamper with. So I should have framed my question in the context of properly administered paper ballots rather than being open ended as I was.
>All recounts are fabulously expensive.
I wonder about cause and effect here. I can allways create more political BS hurdles to increase the costs in an attempt to prevent recounts. Conversely I bet one can reduce / trivialize the cost if one actually had the will to do so - And with such a system in place I'd love to compare the cost to how much has been spent and lost on the purchase of all these electronic systems (ignoring the cost to democracy for the moment). My gut tells me just the interest payments on that equipment could cover the cost of recounts forever ... |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1111
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 10:13 am: |
|
Bev,
Don't get me wrong. Many PA polling sites, especially in upscale suburbs, can be downright palacial! It is the older cities, with aging, more racial minority, and more Democratic constituencies that have the "challenges" in finding suitable polling sites. In addition to all the other problems, building security is an issue to many property owners.
There are advocacy groups espousing the outlawing of using the following as polling sites, fequently the best choices we have in cities: schools (PTA's and others concerned about security), churches (establishment clause advocates), and even lately libraries (interruption of other services).
In my city, if we had had to stop using schools, churches and libraries, and anything else with a religious component to its name, like YMCA's or JCC's, we would have been in great big doo-doo looking for spots to have elections.
Remember, under our code there is only one category of place that can be COMPELLED to host a polling site - public schools. And we have school boards who keep voting to kick the polling place out until their solicitors keep informing them they can't. Yet the PTA's keep protesting.
I have a NH question, then. Do they have "long ballot" comprehensive elections, like we have in PA, or do they have frequent and separated-by-office-level elections, like in Alabama? (The two states I have studied most closely.) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1112
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 10:29 am: |
|
CC,
Let me be clear about MY context in the expense of recounts statement.
In my state the ONLY way any recount is EVER done is via a civil lawsuit. Filing fees and bond postings are prohibitively expensive, all recounts are USUALLY done under strict court supervision, either by the Judge himself or by a "master", and "pro se" litigants go forward at considerable risk to their case, so attroneys get involved. When a litigant brings an attroney, the solicitor to the Election Board frequently must stop all other projects (they frequently are otherwise "lawyering" for the county on other matters).
Our rules of civil procedure state that pro se litigants are to be given wide latitude by trial judges, but they seldom actually get it, so "lawyering up" is a near necessity. In my county, there has only ever been one successful pro se litigant in an election case, and that one was a candidate disqualification case, not a recount, and the litigant was, umm, yeah, me. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 73
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 10:51 am: |
|
Hi Kurt,
That makes complete sense to me. That's also my point as well, barriers are in place to prevent a recount not support it thus driving up the costs.
Putting my "engineering hat" on for a moment, anytime I see something done once I know it's cheaper to do it a second time and even less expensive to do it a third time (and so on). But the current political practice is to craft policies and procedures to give the illusion that costs can't be reduced and even perhaps must increase with each recount.
I guess it's just another reminder that things are broken at an overwhelming number of levels. 
Charles |
Sherry Healy
Voting Rights Forum Participant
Username: Sherry_healy
Post Number: 6
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 10:57 am: |
|
Someone asked if I was giving up on hcpb. Here's my response:
I believe hcpb is preferable if it's practical. It all must be conducted at the PRECINCT on ELECTION NIGHT with two or more groups, with separate oversight, coming to the same answer, e.g. the "New Hampshire" method. It must play out in one night, because as we learned from the hcpb Philippines the "fix" comes in the "double-checking" of the results in the days following the elections. The less degrees of separation from the votes cast, the higher the confidence that the ballots cast are indeed the ballots tabulated.
If your ballots are so long that they cannot be reasonably tallied in one evening with one shift of workers, etc., as specified above, then this "two step" has a higher degree of accuracy.
This particular "CA Two-Step"proposal is for California, and the math has been proven for the first step by Bruce O'Dell, Jonathan Simon, Jerry Adams, and the second step has been prove by Kathy Dopp, Howard Stanislevic, etc.
This is two methods (thus two steps) that must concur within a margin that would not overcome an election lest a full hand count would ensue.
Each of the two models has a "blind spot" but this eliminated by using the two methods AND it also provides superb redundancy which also confers protection.
But if a small state, such as Vermont, can meet the criteria above as far as securing hcpb with something comparable to the "New Hampshire" method, then by all means they should go with hcpb.
And, if you think hcpb is practical in California, that may be so. And if that's the case, this could be the interim step to get us closer to that goal--in the nick of time. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1113
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 11:55 am: |
|
CC,
Here's why the engineering model MAY NOT apply to recounts, depending on how all levels of election admin are staffed.
My points here are based on how business is done in most northeastern / rust belt industial states. All initial vote "counts" whether they are HCPB, or taking results off a machine, or both, as is frequently the case here in PA, have a "marginal cost" (I'm going to put on the hat of MY professional training as an economist) of zero. The same 3-7 polling place officials (usually 5) that have been there all day to sign in voters and keep the records of the polling place do that initial count. They are paid a pitiful sum, especially when you consider that it is AT LEAST a 14 hour day, FAR longer in a Presidential year. Typical pay per worker per day? Less than $100, sometimes FAR less, depending on the county. But the marginal cost is zero, because you already had to pay them all day to sign in voters even if they hadn't counted squat.
Compare that to ANY recount after the fact. There is a County official of some kind involved. A "hired" E.D. in PA, and elected one in FL, a Probate Judge in AL, whatever. He has one sphere of interest. Each candidate has a sphere of interest. Each will likely have attorneys, even if we can do this outside a courtroom, for the sake of argument. The County official will likely have other staff involved, since the REST of the election, the uncontested unrecounted part, still has a legally mandated certification deadline, which does not wait.
Now consider for the moment Florida 2000. Remember the guy with the thick glasses carefully holding a punchcard up to the light? Now think about HCPB. "What is that mark? Is that a vote? What does this mark mean? Is this a ballot 'capable of identification'?"
HCPB has oodles of nuance. It is ripe with opportunities to challenge and dispute. HCPB is many things. But it is one thing above all - an invitation to litigate. And that makes it fabulously expensive, when things get close. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3894
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 12:22 pm: |
|
If the initial count (not a recount!) is done by hand, with good quality observation (e.g. let each candidate's lawyer or other rep observe the counting), including observation of decisions about any unclear ballots, and cross-check of all results (so there are no counting errors), then it becomes less likely that there will be legal challenges to the count, or recounts required. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3895
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 12:25 pm: |
|
OT
Great to have you back, Kurt, I wondered what had happened to you.
There have been quite a number of threads in which many of us were waiting/expecting your input, particularly some threads with new info about the Ohio 2004 elections and anomalous urban vote counts. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1114
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 12:35 pm: |
|
Catherine,
Oh geez Cath.
I don't need to go to Ohio to find anomalous urban vote counts. Every major election in every city in PA in 2006. You can't swing a dead cat (local colloquialism) without hitting a precinct where the total number of voters signed in didn't match the vote totals.
Since DRE's, they almost never match any more. Two main reasons: 1)provisional ballots, which are handled incorrectly by precinct boards the MAJORITY of the time in SOME way or other, and 2) there is about 1% of the population that has no clue how to correctly operate a DRE and are ALSO too proud or stubborn or embarrassed to ask for help.
On edit:
A few months back there was someone on here who stated (I believe vastly incorrectly, but they stated it) that the PA DOS had advised local election boards that if a voter voted a full ballot but had merely failed to press the final button, that the poll workers should cancel that ballot, then the problem would be HUGE. If that last button is not pressed, the public counter will not increment, the vote will not count, but the signature stays in the pollbook. That should make the totals fail to prove out by several dozens per precinct on average, especially when DRE's are first introduced.
If anyone at PA DOS actually instructed localities to act that way, they should be removed, IMO. That is rampant disenfranchisement, and as violative of full suffrage as any poll tax.
(Message edited by Formerelecdir on July 05, 2007) |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3896
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 1:10 pm: |
|
Read the article and you'll get the point. It is significant. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3897
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 1:14 pm: |
|
Kurt, it's here. Didn't mean to hijack this important thread. |
Bob Fleischer
Frequent Voting Rights Forum Participant
Username: Rjf7r
Post Number: 112
Registered: 09-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 1:20 pm: |
|
Re HCPB as "an invitation to litigate" -- an increase in litigation could be a sign that you are doing something right. Litigation could be a sign that there are few alternative ways of snatching victory from the jaws of defeat. It could be a sign that you are not using a system so full of holes and secret spaces that behind-the-scenes cheating is a no brainer, ensuring victory without messy and costly litigation. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1115
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 2:00 pm: |
|
Bob,
With the full and complete appreciation for how much of the world of elections has come around full circle from Florida 2000, it is at least, to me, certain that the American public was repulsed by the spectacle of the recount in Florida. It was the trainwreck that eventually they decided they HAD TO stop watching, and about two years of thought by some really smart people went into avoiding its recurrence.
For all of 2001, and most of 2002, there was a near-unanimous feeling, by everyone "at the table" that the answer to the problem of Florida was to remove the oppotunity to dispute nuance in voting. The "Holy Grail" was to objectify everything and remove the subjectivity. Hence the election community was ripe for the appeal of DRE's. "Either the vote is there, or it is not. No questions." The election administration community nearly had a standing orgasm just thinking about it.
It is STILL what they like best about DRE's. "Oh, you screwed up your ballot? Too bad. Game over for you. You pushed the last button. Gee, wish I could help, but I can't. Better luck next time."
If you want to know why counties love these things, you don't need to overthink it to the point of altering votes behind the scenes. The appeal of objectivity is enough to pique their interest, and give them a rooting interest in keeping their DRE's. It also explains why they don't want no stinking paper trails. Press the "Polls Closed" button and Ye Olde Fatte Ladie Hath Sung. They love that stuff. |
Bob Fleischer
Frequent Voting Rights Forum Participant
Username: Rjf7r
Post Number: 113
Registered: 09-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 2:00 pm: |
|
re Charles "no reason why machines can't be used to count the ballots as many times as you wish *IN ADDITION TO* hand counting":
In working on our audit bill we were advised that optical scanners have a small but essentially unavoidable rate of undercounting -- missing some marks. We were warned that almost any audit would uncover these discrepancies, and thus the escalation of the audit to full audits would almost always occur. The suggestion was that this would make auditing very unpopular unless a certain error rate could be accepted.
(I don't know if this situation is true, but it is plausible.)
Bob |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1116
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 2:24 pm: |
|
Bob,
Based on my somewhat limited knowledge of optical scanners, based mostly on conversations with those who have used them a great deal in central count operations, they are no more prone to undercounting than overcounting.
They, ironically, are prone to undercount exactly those ballots that have been the most painstakingly thoroughly marked by the voter with a "softish" pencil. You've seen the kind. That dark black mark that has been rubbed in almost glossy black. That gloss itself many times creates more reflectance than a white unmarked oval. Weird, huh?
So, tell me, what do you suppose your REAL SAT score was? Hmmm?
But sometimes the most seemingly benign of stray marks, even creases, get counted as valid marks, and cause overvotes.
Now any vendor worth his salt will claim, "Sure, but it's all in the calibration. You can avoid those problems." Uh huh. When was the last time you saw an election official who could perfectly calibrate hundreds of anythings consecutively with no errors, or better yet, train lots of other somebodies to do it right every time?
Yeah. You, me, a bridge over the East River, a real estate transaction. You gettin me?
(Message edited by Formerelecdir on July 05, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1117
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 2:35 pm: |
|
Catherine,
Quote from you:
"If the initial count (not a recount!) is done by hand, with good quality observation (e.g. let each candidate's lawyer or other rep observe the counting), including observation of decisions about any unclear ballots, and cross-check of all results (so there are no counting errors), then it becomes less likely that there will be legal challenges to the count, or recounts required."
The simple answer is they can't be everywhere at once. In the real world of elections, I have never yet, since I've been politically involved, known a single political candidate who could even round up enough people to have a watcher at each polling place they were running in. So many times, all counting would be done in absolute secrecy in some polling sites, because no one would be there to do the watching. Before I was a county Election Director, I served as a precinct poll worker for about 6 years, including the 1996 Presidential. We never once ever had a Watcher, even in the Presidential year. Had we been so inclined, we could have ghost voted the whole damned precinct. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3898
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 2:57 pm: |
|
That's understandable in the current situation, in which there is no tradition of observing and in many cases outright obstruction to meaningful observation.
Fortunately change is not impossible. It starts with people caring enough to pay attention and get involved. We've already seen this happen when people have access to information. Complacency can be shifted--slowly in some places, faster in others.
(Message edited by Catherine_a on July 05, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1118
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 5:28 pm: |
|
Yes, yes, "tradition" is an especially apt choice of wording. That tradition, in places where the days of economic growth seem past, and reinvention has not yet taken root, like in the rust belt cities, is getting ever harder to regerminate.
Listen to an oldie but a goodie. Billy Joel's "Allentown" is a VERY descriptive song that is still as poignant as ever in MANY places, maybe even in Allentown, even though there are seeds of revival there, if the assimilation of Latinos, or lack thereof, doesn't tear them assunder first.
It's an old Carter term, but "malaise" grips huge portions of the real estate here. I don't know how to overcome that inertia as well as I would like. It will take more combined smarts than I have. Dr. Alan Brau is living near or in another old town, Easton, and he sees something in this region that I frankly don't see. I sincerely hope he's right and I'm wrong, but I doubt it. |
Sherry Healy
Voting Rights Forum Participant
Username: Sherry_healy
Post Number: 7
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 5:48 pm: |
|
Regarding this question:
What is the statistical basis for your saying that a 10% sample of ballots is not sufficient for determining if the outcomes of the federal races were correct? What statistical test are you using as a foundation for your statements?
The following "Details" provide the math for the STEP ONE, the BALLOT VERIFICATION (part and parcel to the first count) to first affirm that the ballots tabulated are indeed the ballots cast.
And as a superb byproduct, it in California it will tell us 99% confidence in the statewide and federal elections whether an election has been stolen. Again it is only STEP ONE of two steps, and so here's the math for it in more detail.
DETAILS:
For races that have less than a 1% difference between candidates, a sample larger than 10% will, in fact, be required in some cases. In addition to that limitation, Congressional District races that have less than 150,000 ballots would also need a larger sample size sample. In California, that has not happened in the past five decades.
Again, we are just talking about California right now. We are just talking about determining if there is a statistically significant difference (as measured by a z-test) between the hand count and the machine count.
Jonathan Simon and Bruce O'Dell conducted an extensive computer simulation to indicate that a 10% sample of ballots would be more than enough for Congressional District races in California:
http://www.electiondefensealliance.org/files/New_UBS_811Update_061707.pdf
For the other federal races, less than a 1% sample will be needed:
http://www.oregonvrc.org/2007/04/oregon_voter_rights_coalitions_proposed_methodo logy_for_election_verification
The above two articles provide the statistical basis for the stance that a 10% sample is sufficient for all but a race of less than 1% difference between candidates (a statistical tie) in California. |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6404
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Thursday, July 5, 2007 - 6:31 pm: |
|
DEBATE WITH A CHAIR - part 3, will audits help or hurt, is up.
It's meaty.
It's here: http://www.bbvforums.org/forums/messages/1954/47822.html |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1119
Registered: 04-2006
Best of Black Box?
Votes: 1 (A keeper?) | | Posted on Friday, July 6, 2007 - 10:04 am: |
|
Bev,
Question, please.
I love the "tone" of this discussion. But as someone who has actually had to "build" elections, from the ground up, account for and create every piece of the infrastructure, or manage its creation, I am confused a little about how a concept like "the people" get intimately involved throughout the whole process.
Granted, some pieces are easier than others. Counting might be the least of the problems. But things like ballot creation, and the very first step, compiling the list of which offices are up for election in a given year, by comparing notes between the level of government that DOES the elections and each municipality, school district, the state, and whomever is electing anybody (this step is BEFORE you ever touch a keyboard to start building or designing ballots).
How do we get "the people" involved from the "get go" and how does that work?
Keep in mind that due to ridiculously late decisions from courts, I have had to build entire ballot design databases personally, ALONE, OVER HOLIDAY WEEKENDS working nearly 18 hours a day in order to get them completed in time to even HAVE the election. I could not compel other staff to help because of their union contract. It is impractical to have public observation of some poor schmuck manually transcribing into a computer at a keyboard 1000 names to build an election database at 3AM on a Sunday morning, but THAT'S THE REAL WORLD OF ELECTION ADMINISTRATION.
How do "the people" keep a solid role in this kind of environment?
And by the way, HCPB would NOT change this need. Even a HCPB election uses software and computers to design the ballots. With 20-some offices on a single ballot with hundreds of combinations of districts, it is impractical any other way.
I guess what I'm driving at is I think elections just MAY BE one of those things that are inherently government's domain, because there is no other practical entity to do it.
(Message edited by Formerelecdir on July 06, 2007) |
Brant Lamb
Frequent Voting Rights Forum Participant
Username: Brantl
Post Number: 1402
Registered: 01-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 10:52 am: |
|
Bev:
quote:Brant, with polling-place based hand counted paper ballots, "the people" DO have custody from start to finish.
No, some certain "representatives" (more or less representative, depending on method of selection, time of standing in that position, whether there is a partisan nature to their selection, etc) are chosen, certainly not inclusive of all of "the people", representing each and every view. It may be fairly representative, and it may not, but it will certainly only be representative, not complete participation.
And the legal custodial rights are with the elections workers/BOE, not the citizens. IF either of those two parties decide to walk away with the ballots, you tell me what recourse you have, that won't get a cop called down on you. Until you get legalized oversight of the people, from casting to gathering to "safe" storage, where custody is mutual between the elections workers/BOE and "the people" (or their randomly selected representatives), the people are always the red-headed stepchild in this arrangement.
(Message edited by brantl on July 06, 2007) |
Bev Harris
Board Administrator
Username: Admin
Post Number: 6405
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 10:59 am: |
|
Since voters are "The People", all voters in the polling place are, in effect, also privy to visual custody of the ballot box. So are party observers (and many states permit citizen observers, all you have to do is register first so they know who you are).
The first battle is to get public custody, the next would be to improve and expand public participation. As long as it is equally open to everyone, it doesn't need to accomodate 100% of the population inside a single building -- the word "reasonable" is often used in a legal and legislative context.
Let's carve out public access by going after the low hanging fruit first. And Kurt, I think what Stanislevic was referring to was access to the computer files for ballot definition, not sitting there doing it. I'm not sure how meaningful that is if the count is public and sample ballots can be viewed in advance, as is customary, but I suppose in these so-called "audit" concepts it makes sense. And it sure is needed in situations where secret software is used, but of course it's off the table. |
Bruce Sims
Frequent Voting Rights Forum Participant
Username: Ubetchaiam
Post Number: 990
Registered: 06-2005
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 11:23 am: |
|
Kurt, your tale of 'building the election database' may be 'real' for you but out here in CA, that is done by the voting machine vendors; in fact, the 'election officials' can't do it without relying -like a heroin junkie needing a fix- on the vendors.
"I guess what I'm driving at is I think elections just MAY BE one of those things that are inherently government's domain, because there is no other practical entity to do it."
'Government' supposedly represents the 'people'; corporations do NOT. It is the privatization of the election process that puts elections out of the 'government's domain'. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1120
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 12:15 pm: |
|
Bruce,
If I may beat my own chest for a moment, I wouldn't give you a plugged nickel for an Election administrator who needed a vendor to program his database.
If you guys are paying E.D.'s $100K plus, like in San Diego, and these "schmendricks" can't even program an election database without a vendor, then you're not getting your money's worth, by a mile.
Of course, in Pennsylvania, Election Directors are very hard-working "working people", not Press Relations or management "suits" like those I've read about in California.
(Message edited by Formerelecdir on July 06, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1121
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 12:32 pm: |
|
Bev,
re: access to ballot design files (Stanislevic)
There is nothing that can be done with those files without the application program. I've used two separate programs to do EACH election. One for the DRE's (Danaher Guardian EMS) and one for the HCPB absentees (ES&S Unity Suite). Both of thenm produce proprietary format files that are useless without the app.
Remember if you will, at our last attempt, we have NOT been successful in cracking into a Danaher-created JET database, unlike the Diebold ones. Yes, we can get write access to the DB file, but the whole file is encrypted, and you can't see where to edit anything. |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 74
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 12:44 pm: |
|
Kurt,
Marginal costs -> Understood / got it!
How does the "warm body cost" compare to; the cost of the electronic voting machines, thier required setup costs over ballot printing costs, security cost to insure they are not tampered with, as well as the lost interest from tying up all that capital in a closet for most of the year?
(Message edited by ILikeInfo on July 06, 2007) |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3899
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 1:53 pm: |
|
Kurt,
Re: ballot creation process, I imagine citizens/parties/candidates could eagerly proof & offer suggestions on a draft ballot if it were posted to a website, and all candidates, political parties, election/voter groups notified, and a press release sent out. Even with tight deadlines where there might be no time for printing in a local paper--in fact, particularly in those circumstances--people would probably be thrilled to have a look and notify of any errors or problems they see, even if there's only a 24-hour window or less. It would give an election director extra proof-reading eyes for free.
If candidates and parties were notified, I bet they'd check that ballot for errors. And if they didn't, at least people would have been given the opportunity. It's a gesture that could avoid potentially serious problems, avoid costly errors, would cost virtually no time or money to implement, and would empower and involve the public in a meaningful way. Win-win.
All that's needed is an election official to want to have this input. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3900
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 2:08 pm: |
|
Charles,
Here's a link to a cost comparison done by Marian Beddill, complete with a spreadsheet for budget planning:
http://www.bbvforums.org/cgi-bin/forums/show.cgi?tpc=17141&post=17431#POST17431
And Jo Anne Karasek had a post with links to cost comparisons for various solutions suitable for the disabled (see the 3rd and 4th links in the post):
http://www.bbvforums.org/cgi-bin/forums/show.cgi?tpc=73&post=26895#POST26895 |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1122
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 2:31 pm: |
|
CC,
You missed a big one, at least for the old big DRE's like Danaher/Shoup and Sequoia Advantage - the HUGE warehouses that are needed to house them. The smaller ones need less space but need a much more controlled environment than the big honkin' older styles. The big old DRE's have a HUGE environmental range, because they're ALL 1980's firmware, no drives or touchscreens or fragile wussy electronics. Ask Bev. She's now seen a Danaher/Shoup 1242. They're built like Sherman tanks.
But in terms of total cost of doing business, most people I know who've studied it say the old huge DRE's are the cheapest way to run elections, followed by paper based systems, and the MOST expensive are the new generation DRE's because their life cycle is so short. Because of how laws are set up, with multilingialism and everything, plus needing to have 110% print levels even in a 20% turnout election, paper is darned expensive in large jurisdictions.
Paper works great in smaller jurisdictions.
I can't imagine a large jurisdiction (>500,000 voters) in a long ballot state using paper ballots. They abandoned paper decades ago. In some cases SEVEN decades ago, because it was unworkable in the huge jurisdictions.
(Message edited by Formerelecdir on July 06, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1123
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 2:46 pm: |
|
Catherine,
Proofreading for errors is one matter, but as it pertains to paper, there is explicit language as to how paper ballots MUST be laid out. There is no reason to consult about it, because there is only one legal way to lay out a paper ballot in PA. The only decisions are things like sans-serif vs. serifed fonts, whether the ballot contains a line or two to separate offices, etc.
The fields to be included and excluded are explicitly listed in law. Even without the web type listing you suggest, all E.D.'s in PA spend hours listening to people telling them they should have done things differently, when all they have done is followed the law.
The percentage of helpfully offered suggestions that complied with law? About 1%, tops.
Including:
Split the ballot into several sheets - illegal
Don't put that straight party option on there - illegal
List the candidates alphabetically - illegal
Office A should be above Office B - illegal
That language about "electors" is confusing, lose it - illegal
The parties should rotate ballot position - illegal
The name of the party shouldn't be right next to the candidate's name - illegal
The candidate's town of residence shouldn't be on the ballot in the Primary - illegal
You get the picture.
Everybody's an expert except the guy that's paid to be. Too many ideas - only one Election Code.
(Message edited by Formerelecdir on July 06, 2007) |
Charles Christopher
Voting Rights Forum Participant
Username: Ilikeinfo
Post Number: 76
Registered: 11-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 2:59 pm: |
|
Thanks all for the info.
>I can't imagine a large jurisdiction (>500,000
>voters) in a long ballot state using paper
> ballots. They abandoned paper decades ago. In >some cases SEVEN decades ago, because it was
>unworkable in the huge jurisdictions.
I do not understand this. Past on what has been said so far does'nt this come down to how you define the polling locations? The more locations for fewer ballots each has to manage and compile? I understand the costs go up but don't they then approach the cost of rural areas?
>Ask Bev. She's now seen a Danaher/Shoup 1242.
>They're built like Sherman tanks.
I'd also bet tampering is far easier to detect. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3901
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 3:33 pm: |
|
Kurt,
It would be worth doing just to get the proof-reading. If there's any other area where it would be legal to make changes, then specify what that would be.
And if the existing regulations are lousy, and there has already been input, then why not consider legislating changes to enable changes? Or let people know what actions need to be taken to get these changes put in place? E.g., is it state legislators they need to influence? State AG? Who makes the rules in question, and what are the procedures for changing them? Inform the people, and then the ball is in their court.
Why limit our future to "the way things are" today? |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1124
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Friday, July 6, 2007 - 7:23 pm: |
|
Catherine,
Remember, you asked. 
In most of the states that have had few core recent changes in their codes, except to implement HAVA, (examples: PA, NJ, CT, NY, probably OH, KY, AL, TN, VA, and a few others) the answer is legislatures, almost exclusively.
There is one end run around that, but as Lehto has seen, it can be treacherous, and that involves litigating from a "rights" perspective. Where legislatures are intractable, that may be the last option left. Of course, federal legislation can trump states, but too much federal role in elections could pose Constitutional difficulties if not done delicately. HAVA was a bold step toward a federal role that MAY ALREADY be unconstitutional if tried in our current courts. HR811 almost certainly is unconstitutional, in my view. It essentially ends state sovereignty in an area where the federal government has no legitimate role under the Constitution, unless a squirelly Commerce Clause argument can be made. This is EXACTLY why the latest HR811 limits EVERYTHING to federal offices; it's the only way it could ever CONCEIVABLY pass a Constitutional challenge.
In a few states, the chief state election official has wide policy and discretionary authority akin to lawmaking. Mine is NOT such a state. Neither is NJ or NY or AL. Based on what Blackwell pulled, OH may be. CA certainly seems to be from this distance, so in states where an SoS has wide discretionary authority, one's lobbying can be "concentrated".
If anyone is at all unsure, call your Secretary of State's office. Ask for a "policy director" or similar official. Ask them this question: "Your boss is a member of NASS, right? Tell me, compared to other states, how much authority in setting election policy is REALLY with his office, as opposed to closely held by the legislative branch?" Any policy guy getting a call like this HAS TO BE so impressed with your grasp of the issue, he'll likely be VERY candid with you. (BTW, our guy, Cortes, is being groomed for VERY high office in NASS, and not without cause. He is SUMMMOOOOOTH! Best, most savvy Latino political figure I've EVER known.)
I expect to find that as you move west, you'll TEND to find more administrative authority. In the "old country" (East Coast) you'll tend to find legislatures are where it's at. Exceptions will abound, though.
If it's an old line industrial state, legislatures tend to keep power close to themselves. Florida is a weird and wonderfully fluid state, having seen more changes more often in elections than anywhere in the last 7 years. It's almost a "what are we doing THIS year?" state. Very unique. Almost a class of their own. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3903
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Saturday, July 7, 2007 - 4:05 am: |
|
Kurt,
The point was not to tell me anything on this website. The point was that if election directors would tell their voters what tactics to take--invite suggestions, and give them information about how to change regulations they disagree with (even if the election director has a different point of view)--then voters would have the tools to create change, to engage with the system, to make it their own.
Teach people how to fish, in other words. Instead of just telling them "this is the way it is, so get used to it".
Teach them how the system can be changed, if they want, once you've explained the current rules. Both parts are important: understanding the current state of affairs, and learning about some of the pathways to create change.
Inspire change, inspire involvement. Don't dampen it by implying everything is etched in stone for all eternity. Otherwise you become part of the problem instead of part of the solution. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1125
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Saturday, July 7, 2007 - 9:05 am: |
|
Catherine,
It's all out there for anyone to use. There's nothing new or unique about lobbying for change in this issue as contrasted with any other.
The problem is one of what I call "The Expectations of the Deeply Committed". (May have to copyright that. ) In any issue, especially this one, there is a deeply committed core of activists who have become utterly convinced that they have "The Answer".
If you take out of this movement those for whom it really is all about left-leaning "progressive" liberal politics, and "election reform" is merely a means to that end (a sizable chunk), you have left a core of individuals who have seriously digested this down, and dissected it to the point that they have come upon "The Answer" for them. I submit to you that that "Answer" for this core is "Hand Counted Paper Ballots". And for damned good reason, too. It is "The Answer" to the key elements of the definition of the problem, as they see it. It is an entirely appropriate solution to what they see as the most important problems.
Here's the rub.
There are OTHER problems, that this core sees as LESS important or peripheral, that other key constituencies or "stakeholders", as the popular term has arisen, feel are not only "important", but perhaps "The Most Important" or even "Crucial".
And some of these other constituencies have some "problems" from the perspective of the HCPB core of the election integrity movement - they are better organized, better politically connected, less "balkanized" in their thinking, and have been at the "legislative table" far far FAR longer than election integrity activists.
I speak of the accessibility lobby, mostly in the form of "Independent Living Councils" and "Disability Law Projects", the National Association of Counties, NASS, NASED, the Election Center, and about a half dozen other fully organized and Washington K Street office located lobbying organizations that will deliver the "ANYTHING BUT HCPB" message every hour if need be to kill the idea.
We simply do not have the level of political clout that combination of forces bring to the table.
Unfortunate, but true.
On edit:
Here's a sad irony for you. The people for whom it's all about partisan politics, and not serious election reform, have a HUGE spot at the current table, while the true reformers do not. Lesson? Washington today is ALL politics, and sound policy can go take a hike. It's so bad right now that if you do have an idea for legislation that is politically neutral but a sound public policy, you have to find a way to sell it as a partisan advantage, or you pretty much can't even get heard.
(Message edited by Formerelecdir on July 07, 2007) |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1126
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Saturday, July 7, 2007 - 9:32 am: |
|
Catherine,
Now as far as local or County Election Directors in any way "helping" people lobby for change - that will get them fired.
It simply is NOT their role. With a few exceptions around the country, local election directors have ZERO policy making or lobbying authority. They are strictly ministerial in their function. Even the legislatures almost NEVER want to hear from Election Directors about election law policy. They want to hear from those Election Directors' elected political bosses exclusively. No one wants to hear from the people who actually run the business of elections. Their bosses don't. The legislature doesn't. The Congress doesn't. Even election integrity activists don't.
So if you're not getting much from E.D.'s, that's why. They can't help you, and you can't help them. All talking to you will get them is unemployed. |
Catherine Ansbro
Frequent Voting Rights Forum Participant
Username: Catherine_a
Post Number: 3907
Registered: 12-2004
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Saturday, July 7, 2007 - 12:59 pm: |
|
Kurt, I wasn't suggestion EDs should "help" anyone make any change. Rather, I suggested that if a voter complains, that an ED explain the process by which changes are made. |
V. Kurt Bellman
Frequent Voting Rights Forum Participant
Username: Formerelecdir
Post Number: 1132
Registered: 04-2006
Best of Black Box? N/A
Votes: 0 (A keeper?) | | Posted on Saturday, July 7, 2007 - 1:39 pm: |
|
Catherine,
THAT MUCH any E.D. owes ANY voter. It's simply common courtesy. But please understand if SOME E.D.'s have to do that quietly and out of the public view as much as possible.
It is fraught with political peril to say anything that their overtly political bosses may disapprove of for any reason. E.D.'s (the appointed ones, not the elected ones so much) are among the most "on the perpetual hot seat" people in all of government service.
I'm sure most people here who have dealt with one have almost smelled the fear. |
|
|
