Now on to to block Diebold DRE

Here is my open records request regarding the Diebold application for certification in Wisconsin:


To: D. Richard Rasmussen
 
Wisconsin State Elections Board
17 West Main Street, Suite 310
PO Box 2973
Madison, Wisconsin  53701-2973
 
Since, Diebold HAS applied for certification in Wisconsin, I would like to make the following open records request.  These may overlap.  If so, one copy of a given document will suffice.
 

1. I would like all records submitted by the vendor, Diebold, pursuant to the application requirements of ElBd7.01
2. I would like all records submitted by the vendor, Diebold, to support the vendor claim the vendor’s voting machinery is certified under the 2002 EAC/NASED voting standards.
3. I would like to know the 2002 certification number issued for the machinery the vendor Diebold has submitted for certification.  For example the certification number for AccuPoll is FEC-N-2-13-22-22-001 (2002). What is Diebold’s certification number?
4. I would like all the reports and documents created by the independent testing authority (ITA) pursuant to the ITA recommendation for the issuance of the 2002 certification number
5. I would like the name and address of the ITA to which the vendor, Diebold, to which submitted its system for testing and certification recommendation.
6. I would like copy of the Physical Configuration Audit submitted by the vendor, Diebold, to the independent testing authority (ITA) which is required by section 8.7.1 of the EAC/NASED standards.
7. I would like to see any test plan employed by the SEB staff to ascertain if the system presented to the Board Staff for the mock election testing under ElBd7.03 is the same system submitted to the ITA for certification.  This is done by comparing the description of each configuration item of the PCA of the ITA certified system to the configuration items of the actual system delivered to the board staff.  This is particularly important.  The vendor, Diebold, HAS lost its certification in California precisely because the system used in 4 elections (2 spring and 2 fall elections of 2004), were NOT the systems certified by the California Secretary of State for use in the state.  If the system delivered to the staff for mock election testing is not the 2002 certified system, there is no point in the staff wasting its time on the mock election testing of Elbd7.03.

 
Most of these documents should already be in the possession of the board staff as the documents are required by ElBd 7.01(1)(a) through ElBd 7.01(1)(g).  If the documents are not in the possession of the staff, the documents are readily available from the vendor or ITA from the EAC/NASED certification process.
 
Please contact me on the number of pages of this request and the cost for creating and mailing the copies requested.
 

Election Law Review Committe in town yesterday

Yesterday, the Special Committee on Election Law Review chaired by Senator Joseph Leibham met in room 502 next to the City of Milwaukee Election Commission.

 

The main proposals desired by the city were:

1)      The ability to print its own ballots.  With the Nader challenge the city did not receive from the county the City order for way too many ballots until just before the statutory 21 day deadline.

2)      The ability to open, examine and tally absentee ballots before election day.  That is bad idea of such proportion as to deserve a separate post.  This was pushed by Senator Spensor Coggs on behalf of the Governor “Diamond Jim”.  More of the wide open (fraud vulnerable) system endorsed by the Governor.

3)      The ability to open, examine and tally absentee ballots on election day, but at a centralized location (the Bradley center for example) and not at the local polling as currently required by statute.  This to is bad idea of such deserving a separate post.  I personal dislike it because it means poll watchers need to examine records in 2 places instead of 1.

4)      The ability

a.       to accepted request for absentee ballots,

b.      distribute absentee ballots to in person requestors, and

c.       Accepted completed absentee ballots,

d.      But not open envelopes or tally votes of absentee ballots

in a satellite location other than Milwaukee City Hall (e.g. the mid west express center).  This is a fabulous idea.  Last November there were 8,000 people who came to room 501 of City hall to do at least one of those 3 things for an absentee ballot.  The Midwest express center or MECCA is designed to accommodate thousands of people at one time.  Senator Spencer Coggs, again on behalf of “Diamond Jim”, insisted 4 or more satellite centers would be better.  His analogy to the DMV satellite locations was a poor example at best.  The City resisted.  The main problem for the City was 8,000 people in City Hall not staffing, not ballots not processing.  The problem was 8,000 people coming to 1 window of a small office.  A convention center floor is much better.

5)      Request statutes be changed so the request for a mailed absentee ballot (mailed to elector) be received no later than 1 week before an election.  In person request for absentee would be still be allowed up until 5:00 pm the day before and election.  This too is fabulous idea.  Mailing an absentee ballot at 5:00 pm Friday, October 29, 2005 for the November 2, 2005 election is pointless.  The ballot is unlikely to even be delivered by snail mail before Wednesday.

 

 

 

 

 

Could not make it to Mad City Today

AccuPoll is in Madison today showing off its wares. Since I have a real job and paying client this afternoon, I had to forego indulging in my hobby of election integrity. Instead I sent the following poor substitute; a sharply worded letter.

Germantown, WI 53022
August 26, 2005

D. Richard Rassmusen
Wisconsin State Elections Board
17 West Main Street, Suite 310
PO Box 2973
Madison, Wisconsin 53701-2973

Dear Richard:

Unfortunately, I will not be able to attend the equipment demonstration today in Madison as I had planned. I have a real, paying client to meet with this afternoon. So, my hobby of election integrity must be subordinated to this.

This fax is to document my concerns with the AccuPoll system. First and foremost, I do not see how AccuPoll can meet the requirements of 5.91. Second, there is no documentation of the certification to FEC 2002 standards. Third, the statement regarding open-source is likely to be false.

AccuPoll as described on the website and in literature cannot possibly comply with 5.91(1) and 5.91(18) simultaneously. As long as the elector can handle the ballot receipt printed by the machine, there is no guarantee all the receipts will be available for a manual recounting of the election. AccuPoll is quite adamant that the scrap of paper printed by the machine is NOT the official ballot. This is deliberate and understandable on the part of AccuPoll. If the scrap of paper were the ballot then all the restrictions of chapter 5.64 and other ballot provisions would apply to the printed scrap of paper. (By the way, this may be a problem for Populex as well, since the paper printed by Populex is the official ballot). But, since this is not the ballot the elector is free to take the ballot receipt home. This is similar to the elector taking home a sample ballot home. The ballot receipt, the sample ballots and the “I Voted” stickers have the same status under the law; election aids and souvenirs. If even 1 elector can or does leave the polling location with a receipt or fails to deposit the receipt in a central collection box, the requirement of 5.91(18) is not met. The election cannot be counted or recounted manually.

The logical solution to get out of this box is to require the elector to hand the ballot receipt to an election official and the official then deposits the receipt into the central collection box. The first problem with this approach is there no way to legally compel the elector to give his election souvenir to anyone, let alone an election official. Putting this concern aside for a moment, leads to the deeper statutory problem of this approach. Paragraph 5.91(1) requires the privacy of elector voting be maintained. Since the ballot receipt contains voting information for a particular elector, this document (even though not a ballot) cannot be given to an election official. The votes of an identifiable elector would not be private if the scrap of paper called a ballot receipt is given over to an election official. Printing all of the digital images stored in the AccuPoll machine fails to meet 5.91(18) because the digital was never verified by the elector. The scrap of paper called a ballot receipt was presented for possible verification by the voter, not the digital image.

The printer and hardware of the AccuPoll machine could be altered so the elector can view and verify his ballot receipt without handling it. This is the receipt-under-glass approach. The problem here is that if the AccuPoll Machine is altered to provide a receipt-under-glass, then the certification number FEC-N-2-13-22-22-001 (2002) does not apply to the newly configured, receipt-under-glass system. Chapter 8.7.1 (Physical Configuration Audit) of the 2002 NASED/EAC/FEC standards is quite explicit that the certification and the number assigned to that certification is for the single and particular configuration of hardware and software described by the Physical Configuration Audit (PCA). The PCA is submitted by the vendor to the independent testing authority. Any change to the physical configuration (new requirement for existing hardware, new version of some software component, the addition of software or hardware or the deletion of hardware or software) requires a new PCA. The system with the new PCA must be resubmitted for certification and the issuance of a new certification number.

So on the 5.91 issues here are the salient questions for AccuPoll:

1. What is the official ballot? The screen or the paper? If the screen how does this system simultaneously comply with 5.91(1) and 5.91(18)? If the paper, then does the paper conform to the ballot regulations of chapter 5, e.g. 5.64?
2. Is there a receipt-under-glass version? If so, what is the certification number for that system? If not, how long will it take to get the receipt-under-glass system certified and a new certification number issued for the receipt-under-glass system?
3. Where is the physical configuration audit (PCA) associated with certification number FEC-N-2-13-22-22-001 (2002)?
1. When will the staff of the board get the PCA and other certification documentation created and supporting certification number FEC-N-2-13-22-22-001 (2002)?
2. Is the system here today in Madison the same system described in the PCA?
3. If not, when will a copy of the system actually described by certification number FEC-N-2-13-22-22-001 (2002) be provided to the staff of the election board?
4. How does the staff of the election board compare the AccuPoll system before them to the PCA of certification number FEC-N-2-13-22-22-001 (2002) to confirm or deny the system before them is certified? The staff needs to do this before it wastes its time entering test ballots as part of the staff testing via mock election.

Related to the above questions is my second concern. I can find no copies of the underlying documentation to certification number FEC-N-2-13-22-22-001 (2002). There is no information or links on the AccuPoll, Ciber, inc. or SysTest web-sites.

5. Which ITA was the ITA for FEC-N-2-13-22-22-001 (2002)?
6. Where are the ITA and AccuPoll documents submitted and created during the certification related to the issuance of certification number FEC-N-2-13-22-22-001 (2002)?

A third, relatively unrelated issue with AccuPoll is the lie that the AccuPoll software is Open-Source.

7. Where is the source code repository for the open-sourced code?
8. Since the minimum, legal definition of open-sourced requires delivery of source code upon sale, how much disk space does the delivered source code occupy? What is the deliver media: ZIP file, TAR ball, CD-ROM, access to an FTP site or other? Since Erath County, Texas purchased the AccuPoll Voting System on June 21, 2005, AccuPoll has provided source code at least once. The answers to these questions are if nothing else, what did AccuPoll do while delivering source code to Erath County, Texas?

In Liberty,

John Washburn

AccuPoll to be in Madison Friday

A DRE touch screen by AccuPoll is to be demonstrated in Madison this Friday. I would encourage any election integrity advocates/activists to attend. The election integrity crowd is severely under-represented at these demonstrations.

The claim by AccuPoll to be open-sourced is unsupported by any documentation. There is no SourceForge, CSV, RCS or other source-code repository to be found. Given this, I am skeptical of any other claim made by the company. If a company will lie about it Unique Selling Proposition is is capable of lying about anything.

As has been made clear by the SEB staff AccuPoll has NOT applied for state certification. Thus, none of the statutes of chapter 5 nor any of the regulations of ElBd7 need to be complied with at this time. This open meeting is just a meet and greet / dog and pony show by AccuPoll of their wares to prospective interested parties.

Milwaukee City Election Commission is back

Sue Edman, Executive Director of the City of Milwaukee Election Commission, contacted me Monday regarding several loose ends left by the interim director, Sharon Robinson. There are 4 loose ends as relates to me. I have 2 pending open records requests for documents from the Election Commission Task Force Report. A response for each is being drafted. One of the open items is the illegal poll inspector, Carol Tellef, for ward 13 during the February 15, 2005 spring primary. The fourth open item is the illegal (closed to the public) canvass of ward 188 during the April 5, 2005 elections.

These 4 open items are in addition to my complaint before the Wisconsin State Elections Board regarding the failure to conduct 7.51-compliant local board canvasses during the September 2004, November 2004, or February 15, 2004 elections. The SEB complaint is still winding its way through the SEB labyrinth with claims, responses, rebuttals to the responses, and rebuttals to the rebuttals currently being exchanged.

Help America Vote Act FAQ's

Myth Busters by VotersUnite! is one of the more succinct analysis of the requirements of Help America Vote Act (HAVA) and the actual (as opposed to claimed) impositions by HAVA on the several states.

For Wisconsin the most germain impositions of HAVA are the creation of a central list of citizens and improved access for those with vision of mobility impairments.

Accenture is working on the Statewide Voter Registration System. It is now 6 weeks late from the delivery date of July 1, 2005. Accenture still wants a copy of the SVRS data. Why would Accenture wants a list of 3 million names, addresses, driver's licenses and social security numbers (at least last 4 digits)? Wisconsin is rightly fighting this. Hopefully this data ownership dispute is not impeding progress on completing the roll out of the SVRS.

For paper ballots (either hand counted or optically scanned) the Rhode Island tactile ballot and/or braille ballots meet the vision impairment requirement. I have seen the RI tactile ballot template, it is an absolutely elegant solution to the problem. The ballot marked with the aid of the tactile template is indistinguishable from a ballot marked by any other person. Click the link to order a sample Rhode Island tactile ballot.

Hand Count Paper Ballots

Checkout How to Count Hand Ballots for how Canada counts by hand. I especially like the perforated ballot and stub system. If used in Milwaukee, Scott Walker and the City Election Commission could very precisely know and document how many ballots were printed, delivered, used, not used and returned during an election at any given polling place.

My only innovation here is to use a high-end counting scale in order to count the stacks of sorted ballots. The worst case in Wisconsin is the Village of Brown Deer. Here is what it would take to count 7080 ballots of 10 partisan races by hand in one night.

The current trend is there may be no certifiable equipment available on the market by midnight of January 1, 2006. If so, hand counting paper ballots may be the only HAVA-compliant answer available to the State of Wisconsin.

The more I know about voting machinery, the more I am drawn to the conclusion the most trustworthy, most secure, and least expensive system is hand counting paper ballots during a local canvass performed on election night at the polling location by adversarial, partisan poll inspectors where the local canvass is open to the public. Wisconsin statute requires all of these elements except the hand counting.

Now, getting the Cities of Milwaukee, Madison, Racine, and Kenosha to actually obey Wisconsin Chapter 7.51 on local board canvassing is another matter altogether. The Wisconsin system for local board canvassing would work if the Wisconsin-style canvasses were actually performed in every municipality in the state.

Diebold Direct Recorded Election Equiptment is out

The DRE election machines by Diebold, ES&S, and Sequoia are not certifiable in Wisconsin. This is excellent news for Wisconsing election integrity.

Wisconsin chapter 5.91(18) requires a ballot record which allows for manual recounting. The board legal council, George Dunst, has issued a legal opinion which asserts it is the position of the Wisconsin State Election Board that the requirements of 5.91(18) can only be met by printed ballot and not by a receipt or a receipt under glass. It is the opinion that a ballot receipt fails both the audit requirement and the recount requirement of 5.19(18).

This is excellent news for Wisconsin in that our state will not need to deal with the new kinds of frauds presented and executed on DRE touch screens by Diebold, Sequoia and ES&S.