1 

2        

3        

4                     U.S. ELECTION ASSISTANCE COMMISSION

5                     PUBLIC MEETING

6 

7 

8               Taken at the Marriott Hotel

9               1535 Broadway, New York, New York

10

11               On the date of Thursday, June 30, 2005

12

13

14

15        COMMISSION MEMBERS:

16                    Gracia Hillman, Chair
                        Paul DeGregorio, Vice Chair
17                    Ray Martinez, Commissioner
                        Tom Wilkey, Executive Director

18

19

20

21        Start time: 11:00 o'clock, a.m.

22        Taken before: ELLEN REACH, a court reporter

1        TESTING LABORATORIES' PERSPECTIVES

2        ON THE PROPOSED VVSG

3        

4                     CHAIR HILLMAN: If we could get

5         assembled, please, so the hearing can begin.

6                     We are going to begin this hearing.

7                     If I could remind you please to turn

8         off your - or at least silence - your cell

9         phones, pagers and any other electronic devices

10        that might make noise and distract from the

11        proceedings.

12                    Thank you very much.

13                    If I could please ask everyone to

14        please stand and join me in the Pledge of

15        Allegiance.

16              (Assembly in Pledge of Allegiance)

17                    CHAIR HILLMAN: If we might have the

18        roll call, please.

19                    MS. THOMPSON: Certainly, Madam

20        Chair.

21                    Commissioners, please respond by

22        saying here or present when I call your name.

1                     Gracia Hillman, Chair?

2                     CHAIR HILLMAN: Here.

3                     MS. THOMPSON: Paul DeGregorio, Vice

4         Chairman?

5                     VICE CHAIR DeGREGORIO: Here.

6                     MS. THOMPSON: Ray Martinez,

7         Commissioner?

8                     COMMISSIONER MARTINEZ: Here.

9                     MS. THOMPSON: Madam Chair, there

10        are three members present.

11                    CHAIR HILLMAN: Thank you.

12                    Commissioners, we have before us the

13        agenda for today's public hearing. If there are

14        no adjustments to the agenda, it would be

15        appropriate to move adoption.

16                    COMMISSIONER MARTINEZ: Move, Madam

17        Chair.

18                    VICE CHAIR DeGREGORIO: Second.

19                    CHAIR HILLMAN: Thank you very much.

20        We have our agenda adopted.

21                    The purpose of today's hearings is

22        for the Election Assistance Commission to receive

1         testimony and comment on the proposed Voluntary

2         Voting System Guidelines.

3                     For those who were with us earlier

4         this morning at our meeting, you did hear several

5         minutes of discussion about the guidelines, our

6         acknowledgement of the work of the Technical

7         Guidelines Development Committee that produced a

8         recommendation and delivered that recommendation

9         to us on May 9th.

10                    Our former Interim Executive

11        Director, Carol Paquette, will make a brief

12        presentation that will put this hearing in the

13        context of where the Election Assistance

14        Commission is today with respect to the proposed

15        guidelines that we have now made public.

16                    Ms. Paquette.

17                    MS. PAQUETTE: Yes, Madam Chair and

18        Commissioners, thank you very much.

19                    Section 202 of HAVA directs the

20        Election Assistance Commission to adopt Voluntary

21        Voting System Guidelines and to provide for the

22        testing, certification, decertification and

1         recertification of voting system hardware and

2         software by accredited test labs, among many

3         other duties that are assigned to the Commission.

4                     To assist the Commission with the

5         work of developing the guidelines, HAVA also

6         provides for the establishment of a body of

7         subject matter experts, called the Technical

8         Guidelines Development Committee, otherwise known

9         as the TGDC.

10                    The composition of this 15 member

11        committee is drawn from a variety of

12        organizations, again, and specified by HAVA.

13        These organizations include the National

14        Association of State Election Directors, the

15        Access Board, the IEEE, the American National

16        Standards Institute, and other representatives as

17        designated in the statute.

18                    The TGDC is chaired by the Director

19        of the National Institute for Standards and

20        Technology, NIST, and NIST is charged with

21        providing technical support to the work of the

22        TGDC.

1                     In this fiscal year, EAC has

2         provided nearly $3 million to fund this work of

3         the TGDC and NIST.

4                     The statute tasks the TGDC with

5         developing recommendations for Voluntary Voting

6         System Guidelines and providing those

7         recommendations for the consideration of the

8         Election Assistance Commission.

9                     The TGDC and NIST have been hard at

10        work at this task for nine months, the time

11        period which is allocated by HAVA, and have

12        delivered a very good product to the EAC for us

13        to begin our review on May the 9th.

14                    The recommendations provided to the

15        Commission by the TGDC essentially augment the

16        2002 Voting Systems Standards, most particularly

17        in the areas of accessibility, usability and

18        security.

19                    This includes new requirements for

20        accessibility, voting system software

21        distribution, system setup validation and the use

22        of wireless communications.

1                     Since some states have required the

2         use of voter verified paper audit trails, the

3         recommended guidelines also include performance

4         specifications for this capability.

5                     We would note for the Federal

6         certification purposes of voting systems, these

7         are optional. They have been included for use by

8         those states that have added this as an

9         additional requirement for their voting systems.

10                    We would also note that paper audit

11        trails is only one of several technical

12        approaches to providing voters an additional

13        means to verify their ballot in addition to the

14        DRE summary screen. And the Commission plans to

15        work with the TGDC and NIST to address the

16        development of similar specifications for audio,

17        video and cryptographic approaches to voter

18        verification.

19                    Having received the recommendations

20        on May 9th, the EAC has been, in the intervening

21        weeks, performing its due diligence requirements

22        as the responsible Federal organization for HAVA,

1         to thoroughly review these recommendations.

2                     In particular, we have performed a

3         legal analysis of the proposed guidelines and

4         recommended guidelines to ensure that they are

5         fully in compliance with the Help America Vote

6         Act.

7                     This has resulted in some changes,

8         particularly in the accessibility provisions.

9         And I would note as an example requirements 2.2.5

10        and 2.2.6, dealing with voter verified paper

11        audit trails, that if the normal procedure is for

12        voters to submit their own ballots, then the

13        accessible voting station needs to provide the

14        same capability for voters who are blind.

15                    That is an example of where the

16        Commission changed a 'should' recommendation to a

17        'shall', to be in conformance with Section 301 of

18        HAVA.

19                    In addition, the Commission has

20        revised selected portions to reflect the new EAC

21        process to carry out the HAVA mandate for the

22        national certification of voting systems. We

1         have also made some updates throughout the

2         document to include the new terminology that has

3         been specified by HAVA.

4                     For example, the testing of voting

5         systems used to be called a 'qualification'

6         process and HAVA now uses the the term

7         'certification' process.

8                     We have made these changes in the

9         document because this is a document that will

10        have a substantial life in being used for the

11        testing and certification of voting systems, both

12        at the Federal level and at the state level.

13        That was the principal rationale for the changing

14        of terminology, to put in an appropriate floor

15        for the terminology that will be used as this

16        document is used as a reference for testing

17        purposes in the coming years.

18                    We certainly expect to have many

19        more changes coming in over the course of the

20        summer and we will certainly be summarizing at

21        the end of that comment period the further

22        changes that have been made to the document and

1         how the various comments have been handled.

2                     The Federal Register Notice on the

3         availability of the guidelines for public comment

4         was published yesterday, June 29th. And that

5         begins the 90 day public comment period.

6                     In the Federal Register Notice and

7         also on our web site we provide fairly extensive

8         information on how parties can obtain a copy of

9         the document and also how they can provide

10        comments.

11                    The document is a very substantial

12        document. It's about 250 pages. And to

13        facilitate ease in getting the document, we have

14        published it on our web page. It is available in

15        both PDF and HTML formats, so that those parties

16        who use assistive devices will be able to access

17        the document.

18                    We have also posted it in a manner

19        so that the entire document can be downloaded or

20        selected sections can be downloaded, again, to

21        make it easier for people to deal with this

22        fairly substantial document.

1                     We will also provide copies in hard

2         copy or on CD-ROM.

3                     If people wish to send requests by

4         e-mail or by mail to the Commission, we have

5         provided a number of mechanisms for people to

6         submit comments.

7                     On our web page we have included an

8         application so that people can log on to the web

9         page and go to the comment screen and provide

10        their comments directly to the web page.

11                    Supporting that comment page is a

12        data base where all the comments will be stored

13        and allow for the comments to be posted and for

14        anyone to also look at comments that have been

15        provided to the document.

16                    We have established a special e-mail

17        address: votingsystemguidelines@EAC.gov, for

18        those individuals who prefer to send us comments

19        by e-mail.

20                    We, of course, will also receive

21        comments by mail. We ask that those be provided,

22        addressed specificaly to Voting System Guidelines

1         Comments, so that they can identified and handled

2         expeditiously.

3                     We are providing, in the data base

4         of comments, the ability for the comments

5         received by e-mail and by mail to also be entered

6         into the data base so that we will have one

7         reference source, where comments from all sources

8         will be collected and posted, both for our own

9         review and analysis, as well as the review by

10        other interested parties.

11                    I would also note that we are having

12        three hearings in addition to the one just

13        convened. We are holding a second hearing on

14        July 28th at the California Institute of

15        Technology, and a third hearing in August in

16        Denver.

17                    And further specifics will be

18        forthcoming as we get that meeting planned.

19                    Are there any further questions?

20                    CHAIR HILLMAN: Before I turn the

21        mike over to my colleagues, can you just state

22        for the record where the California Institute of

1         Technology is located. I know it's in

2         California, but --

3                     MS. PAQUETTE: It's in Pasadena.

4                     CHAIR HILLMAN: Thank you very much.

5                     Commissioners, before I turn the

6         mike over to you, I just want to acknowledge that

7         we have with us today members of the Technical

8         Guidelines Development Committee and from the

9         National Institute of Standards and Technology.

10                    And I would just like to acknowledge

11        from the TGDC Sharon Turner-Buie, who is a member

12        of the EAC Board of Advisors and Director of

13        Elections for Kansas City, Missiouri. And we

14        also have Brit Williams, who is with the Center

15        for Election Systems - I probably messed that up,

16        Brit, but forgive me - at Kennesaw State

17        University, in Georgia. Thank you for being with

18        us.

19                    And from the National Institute of

20        Standards and Technology, which is a part of the

21        U.S. Department of Commerce, we have Craig

22        Burkhardt, who is actually counsel with the

1         Department of Commerce, and has been working with

2         NIST on the development of the guidelines, the

3         proposed guidelines, and Allan Eustis, who is

4         with the National Institute of Standards and

5         Technology.

6                     An important part of the comment

7         period is that the EAC Board of Advisors, which

8         is made up of 37 members, and the EAC Standards

9         Board, which is made up of 110 members, will each

10        go through a process to review the proposed

11        guidelines and offer comments to us.

12                    And we have with us, in addition to

13        Ms. Turner-Buie, we have Wes Kliner from, I

14        believe, Tennessee, but I thought I heard you

15        moved, so I am going to acknowledge Tennessee.

16        Thank you very much. As a member of the Board of

17        Advisors, thank you very much for being with us.

18                    And I think we might have a member

19        of the Standards Board still with us, Mr.

20        Szczesniak - is Ed Szczesniak still here - from

21        the Standards Board. Thank you.

22                    This has been a very wonderful labor

1         of love or labor of something that we have been

2         through to produce these guidelines. So if we

3         get emotional about this, it's because we are

4         very, very attached to both the process and the

5         document.

6                     And, Mr. Vice Chairman, it would be

7         appropriate if you would like to give some

8         comments to or ask some questions of Ms.

9         Paquette. I know your name, Carol.

10                    VICE CHAIR DeGREGORIO: I think

11        labor is an apt description since it was a nine

12        month process to get done. And we know that to

13        have a child it takes nine months. And this was

14        a big child of the EAC to come forth from the

15        Technical Guidelines Development Committee.

16                    And I know I made some comments

17        earlier at our meeting this morning, a few

18        minutes ago.

19                          But I do think that this has been a

20        very important process that the Help America Vote

21        Act spells out to give the voters of this country

22        a greater trust in our voting systems. And it's

1         the first time that the Federal Government has

2         put real money behind it.

3                     Our Chair mentioned that nearly

4         $3 million has been appropriated by the EAC to

5         NIST for this process. And it is the first time

6         we have had some scientists look into this and

7         provide us their guidance.

8                     And I think as we go through this

9         over the next 90 days we will see the fruits of

10        their work come forth as we see the human factors

11        and security and other aspects that have been

12        changed since the 2002 standards and that have

13        been enhanced since the 2002 standards were

14        developed, that have improved this product and

15        made it better for the public out there, election

16        officials.

17                    And I look forward from hearing from

18        the testing laboratories, from the vendors and

19        from the advocates and from the public, and to

20        hear their feedback in this process as we

21        finalize this document and then move forward with

22        this, because we know that it's going to have a

1         major impact out there. And we want it to be a

2         positive impact.

3                     So all the input that we can get to

4         make that happen is important to me as a

5         Commissioner, as has been since we started this

6         process in June 2004.

7                     So thank you, Carol, for providing

8         that history of where we are with this and I look

9         forward to the testimony we are going to hear

10        today.

11                    CHAIR HILLMAN: Thank you.

12                    Commissioner Martinez.

13                    COMMISSIONER MARTINEZ: Thank you,

14        Madam Chair.

15                    I will be very brief. I do have one

16        quick question for Carol. But just as an

17        introductory statement, I guess I said earlier in

18        our public meeting portion of this day that this

19        is, I think, perhaps the most significant thing

20        that this Commission has done thus far in our

21        short 18 months or so of being in existence.

22                    So my thanks to all of our partners

1         in helping to developing this first draft of

2         Voluntary Voting System Guidelines: our partners

3         at NIST and the Department of Commerce and, of

4         course, all of the members who served on a

5         voluntary capacity with the Technical Guidelines

6         Development Committee. We are so grateful and

7         indebted for your service and we thank you for

8         it.

9                     I guess for me, there was a poll

10        that was taken a few weeks after the November

11        2004 election, which was not too long ago, by NBC

12        News and the Wall Street Journal, and it said

13        something like a quarter of all American

14        surveyed, that a quarter of those surveyed in

15        that poll did not believe in the accuracy of the

16        voting systems that they were using to cast a

17        ballot in November of 2004.

18                    And as a Commissioner - really, as

19        an American - we should just be troubled, I

20        think, by such a high level of lack of confidence

21        in the very machinery and the tools that help our

22        democracy to continue to exist.

1                     And that's why, again, I point to

2         this as being among the most significant

3         accomplishments that we have undertaken here at

4         this Commission.

5                     So I hope, Madam Chair and Mr. Vice

6         Chairman, that at the end of this process what I

7         hope and I know will be a very transparent and

8         very open and, yes, at times perhaps even an

9         emotional process for us to go through, because

10        there are so many things at stake here, from

11        accessibility for individuals with disabilities,

12        to folks who believe strongly that there ought to

13        be mechanisms in place to ensure better accuracy

14        of ballots that have been cast --

15                    I think this is one of our

16        panelists, welcome.

17                    So, Madam Chair, I would simply say

18        that I think we have undertaken a significant

19        project and it's one that I know that we will all

20        feel better about at the end of the process.

21                    Carol, if I could, there was a

22        clarification I think that I wanted to make to

1         one of the things that you said.

2                     You were giving an example of our

3         legal analysis and you were talking specifically

4         with the functionality of voting systems where

5         the voter has to cast the ballot aside from, for

6         example, on a DRE you can punch a button that

7         says 'cast ballot'. But there are other voting

8         systems, like optical scan systems, for example,

9         where the functionality of casting a ballot

10        requires a voter to do something other than

11        hitting a button on a touch screen. And you

12        refer to that as VVPAT.

13                    But I think what you meant to say is

14        that we changed this 'should' to a 'shall' when

15        it comes to the functionality of those type of

16        systems and not necessarily when it comes to

17        VVPAT, because VVPAT contemplates that a voter

18        gets to see something behind a screen, but not

19        walk away with a particular ballot and have to go

20        cast it in another location.

21                    So in describing that example, when

22        we made a change from a 'should' to a 'shall',

1         you said that "in the context of VVPAT," but I

2         think what you meant it in was the context of the

3         functionality, for example, an optical scan

4         voting system.

5                     MS. PAQUETTE: You are correct,

6         Commissioner.

7                     COMMISSIONER MARTINEZ: That's it.

8         Thank you, Madam Chair.

9                     CHAIR HILLMAN: Today's hearing is

10        in four sections. We have Panel Number 1,

11        presentations from the testing laboratories. And

12        presenting this morning will be Mr. Joe

13        Hazeltine, Senior Director, Eastern Test

14        Operations with Wyle Laboratories, and Ms.

15        Carolyn Coggins, Director of ITA Operations at

16        SysTest Labs.

17                    Following Panel 1, we will take a

18        lunch break. And when we reconvene at about

19        1:30, the second panel will make presentations

20        and these will be people representing vendors and

21        we will have presentations from people

22        representing ES&S, Vote Here, Sequoia Voting

1         Systems and AccuPoll.

2                     Panel 3 will be presentations about

3         the specific section of the guidelines or

4         sections of the guidelines that address voter

5         verifiable paper audit trails and perspectives

6         will be presented by four individuals.

7                     And then the final section, which I

8         don't know if the mike is - the final section

9         will be public comment period. And we do have

10        people who have registered in advance requesting

11        an opportunity to make a comment.

12                    So, can people hear me? No, I

13        didn't think so. I am not sure. Here we go.

14                    I am going to repeat that anyhow. I

15        was just explaining the make-up of the four

16        panels for today.

17                    So thank you very much to Mr.

18        Hazeltine and Ms. Coggins.

19                    And, Mr. Hazeltine, if we could call

20        on you first.

21                    MR. HAZELTINE: Thank you, Madam

22        Chair and members of the Commission.

1                     I have been asked to speak on the

2         Voluntary Voting System Guidelines, the impact

3         from the --

4                     CHAIR HILLMAN: The mike isn't on

5         here.

6                     MR. HAZELTINE: Thank you, Madam

7         Chair and members of the Election Commission.

8                     I have been asked to speak on the

9         impact of the Voluntary Voting System Guidelines

10        from a testing laboratory standpoint. And you

11        have a copy of my presentation. I am just going

12        to kind of go through that.

13                    I guess, first, in talking about the

14        impact, one is to discuss what's the same in this

15        document as the 2002 Voting System Guidelines.

16                    And when we look at Volume I,

17        Sections 2.1 through 2.2.6 and then 2.2.8 to

18        2.6 - which is System Functional Capabilities -

19        have remained the same.

20                    Section 3 on hardware, Section 4 on

21        software requirements are the same.

22                    Section 5 on telecommunications are

1         the same.

2                     Section 6.1 to 6.6.4.3 on security

3         are the same.

4                     Section 7 on quality assurance and

5         Section 8 on configuration management

6         requirements are the same.

7                     So, the bulk of the document remains

8         the same as before.

9                     In Volume II, we see even more -

10        Sections 2, 3, 4, 5, 6 and 7, which would be

11        Description of the Technical Data Package,

12        Functional Testing, Hardware and Software, System

13        Integration and Examination of Configuration

14        Management Practices - have all remained the

15        same.

16                    Appendix A, the Description of the

17        National Certification Test Plan is the same, and

18        Appendix B, the Description of the National

19        Certification Test Report.

20                    So, again, the bulk of the documents

21        remain close to as it was before.

22                    But there been of some changes and

1         they are significant.

2                     In Volume I, Section 1 has been

3         redone. Primarily it is new text on the program

4         that clarifies the requirements.

5                     In Section 2.2.7 is probably the

6         largest change to the document, where there are

7         many new requirements on Human Factors and

8         compliance with HAVA 301(a)(3) were added. And,

9         certainly, that's an important revision to the

10        document.

11                    Section 6.6.4 to Section 6.8.7.5

12        added many new requirements dealing with

13        security, wireless and the verifiable voting

14        paper audit trail.

15                    Appendix A was added, which is a

16        very good thing, adding a glossary.

17                    Appendix B, adding references,

18        again, a good thing.

19                    Appendix C added best practices for

20        election officials and Appendix D added

21        independent dual verification systems.

22                    We also added Appendix E, which was

1         the NASED Technical Guide, Number 1, dealing with

2         colorblindness and some of those issues.

3                     Volume II had two changes.

4                     Section 1, the revised National

5         Certification Testing Guidelines, shifting over

6         from NASED to the Elections Assistance

7         Commission.

8                     And then Appendix C, where we added

9         the National Certification Test Design criteria,

10        the revision, in the back of that.

11                    So kind of what was changed, in a

12        nutshell.

13                    So if we look at what the impact is

14        to a testing labs, or to us, the Volume I Section

15        1 change, we are --

16                    CHAIR HILLMAN: Technology always

17        challenges us, always.

18                    MR. HAZELTINE: We are back on.

19                    Volume I Section 1, the

20        introduction, there is a new certification that

21        is required for independent testing authorities

22        that would now come through the National

1         Institute of Standards and Technology, through

2         their NVLAP program, so it's a new certification,

3         similar to what we already have. But it will

4         require some additional work.

5                     In Section 2.2.7, Volume 1, Human

6         Factors Assessment, accesibility changes were

7         made. And certainly those are major changes with

8         a major impact to the work that we will have to

9         do.

10                    In Volume 1, 6.6.4 of the software

11        security, that's relatively small changes, which

12        is mostly administrative. Their impact to us is

13        minor.

14                    And 6.7 of Volume I, wireless

15        requirements, would have an impact if that

16        technology were implemented. There are several

17        places where expert staff is needed. There are

18        cryptiographic (sic) issues and other things

19        which would require quite a bit of additional

20        work.

21                    In Section 6.8 of Volume I, the

22        Voter Verifiable Paper Audit Trail, I would

1         consider those to be relatively small. They are

2         administrative type changes. For us the impact

3         is not all that difficult from a testing

4         laboratory.

5                     Certainly, Volume I, Appendix A, B,

6         C, with the glossary, references and best

7         practices, there is no real impact to us. It

8         certainly adds to the clarity of the document and

9         allows you to understand terminology being

10        addressed much better.

11                    Appendix D, where the Independent

12        Verification Systems comes into Volume I, that

13        would be a major impact when implemented because

14        it really requires two systems. So you are

15        basically doing the testing twice, so it would be

16        an impact as far as the work.

17                    Appendix E, the NASED Technical

18        Guide is small, mostly administrative changes for

19        us.

20                    In Volume II, Section 1, the

21        National Certification Program Guidelines, there

22        were some clarification text added. There is no

1         real impact added to us. That is the same

2         practices the 2002 standard required.

3                     Appendix C, the National

4         Certification Test Design Criteria, really no

5         impact to us in as far as conducting a normal

6         test program. If the specimen did fail, there

7         would be some additional testing requirements to

8         work with. But if the system passes, it would

9         pretty much would be a very similar process.

10                    I am talking about possible issues

11        where we get into a little more detail of areas,

12        as we went through it where - I wouldn't say

13        concerned - we were just identifying.

14                    We will need guidance on the

15        summative usability test required for partially

16        blind, blind, persons with limited motor control,

17        persons with limited command of English, and the

18        general population.

19                    I have given some of the references

20        of where that would be in the document. But we

21        will need guidance on how to conduct those tests

22        in the future.

1                     For the accessible voting station,

2         guidance with personal assistive devices is too

3         broad. We would need to know exactly what we are

4         talking about there and what devices could be

5         connected and what our requirements would be to

6         evaluate those.

7                     I noted that the voter control of

8         contrast must be reset after the vote was cast.

9         They do have the ability to reset contrast. But

10        I didn't see anywhere else in the document where

11        it would reset back to some nominal value for the

12        next voter.

13                    Synchronized audio with those

14        scrolling screens, that could be an issue with

15        testing. That would have to be defined exactly

16        what we mean there.

17                    The Auto Tactical Interface

18        requirements will require more testing,

19        certainly, because it's a new feature. Certainly

20        a good feature, but a new one.

21                    The audio qualification to ANSI

22        C63.19 would require more testing. That's where

1         we have to classify the actual microphones and

2         system ratings.

3                     Our recommendation at Wyle would be

4         that we consider EIA/TIA 968, which was the old

5         FCC Part 68 document which regulated telephony

6         (ph) and has much - I don't know - I would say

7         are better rules - that are rules which are well

8         known, have been in place for a number of years,

9         does include hearing aide compatability and does

10        include a process that is somewhat better

11        understood by testing laboratories.

12                    The guidance on testing cochlear

13        mounted hearing devices that we needed in here,

14        that could be an issue because I imagine that

15        there are a number of different types. So that's

16        an area of concern.

17                    The frequency range of 315 Hz to

18        10 kHz, that requires now equipment of the new

19        testing techniques. Once again I would suggest

20        looking back at the ITA/TIA 968, the old FCC Part

21        68 document where the frequency range was not

22        quite so broad and test equipment already exists

1         for it.

2                     Another document would be TIA 470,

3         which would address the acoustic performance as

4         well.

5                     The rate of speech control where you

6         are allowed to, actually adjust or advance how

7         fast or how slowly the audio playback is would

8         require more testing and we would need some

9         guidance on that. It says within reasonable

10        levels. Well, reasonable can be interpreted in

11        different ways, so we would need some help on

12        that.

13                    Confirmation of the blind paper

14        ballot would require some additional testing,

15        additional skill. We need someone who could read

16        those ballots to make sure the system is

17        performing properly.

18                    The actuation force requirement test

19        required in test fixtures so that we can do, I

20        believe that is 22 newtons of force, so we would

21        need some additional devices to run that test.

22                    We would need a wheelchair to be

1         provided so that we can evaluate the clearance

2         requirements in Section 2.2.7.4. And there are a

3         number of those. They are good requirements,

4         there are just a number of them.

5                     And for the English illiterate voter

6         machines, would require interpreters be present

7         during some parts of the testing. The required

8         languages should be specified. Right now it's

9         quite broad and that would require some

10        additional work on our part.

11                    Wireless documentation in accordance

12        to 6.7.2.1.3.1 - there are a lot of numbers

13        there - but requires a "subject area recognized

14        expert." So that would need to be identified and

15        provided to us.

16                    If wireless systems currently don't

17        exist, but if they do, that would be be an issue.

18                    The wireless system requires that it

19        be able to operate with or without the wireless

20        capability. So that would require us testing it

21        twice, with and without.

22                    Encryption verification also

1         requires an expertise in the field and we would

2         need some help there.

3                     Again, the Voter Verified Paper

4         Audit Paper Trail testing poses no significant

5         testing issue. It does involve more time, but

6         it's not a technical challenge for us.

7                     Am I going too fast?

8                     SIGNERS: We can't hear you.

9                     CHAIR HILLMAN: If you could move

10        closer to the microphone.

11                    MR. HAZELTINE: When I taught I

12        always had two hours of notes and one hour of

13        speech, so I learned how to speak quickly.

14                    CHAIR HILLMAN: Mr. Hazeltine, we

15        know you live and breathe this stuff, but some of

16        us are running to catch up.

17                    MR. HAZELTINE: I'm sorry.

18                    CHAIR HILLMAN: No problem.

19                    MR. HAZELTINE: Witness IDV Systems,

20        again, will have twice the amount of testing

21        because there are two devices. I had mentioned

22        that earlier.

1                     The bottom line, from our viewpoint,

2         is to be Wyle's viewpoint, we consider the

3         Voluntary Voting System Guidelines to be a

4         significant improvement over the past two

5         documents, the 2002 and 1990 versions,

6         particularly in Human Factors, Accessibility and

7         Security.

8                     It's quite well done. There are

9         some issues, but they are resolvable.

10                    Prior to implementation, further

11        clarification would be required, but I don't

12        believe it's all that difficult to get that data

13        to us.

14                    The current 2005 Voting System

15        Voluntary Guidelines that are out are in need of

16        some corrections and formatting, primarily in

17        correct references and page numbering. Section

18        2.2.7 is probably the one that would need the

19        most work there.

20                    Independent testing laboratories do

21        have some cost requirements through the

22        accreditation through NIST which is not an issue.

1         It's just a thing to be noted.

2                     We will need some guidance for

3         testing laboratories in several areas, which I

4         have identified in the presentation.

5                     There will be some new testing. It

6         will add costs to the certifying of the voting

7         machine.

8                     From Wyle's standpoint we are

9         already training our staff to the new document.

10        I know it's a draft, but it's probably going to

11        be fairly close to the final format. We will be

12        ready to implement by the end of the 90 day

13        comment period. We are quite confident on that.

14                    And we feel that the mandatory

15        implementation date of 24 months after the

16        comment period is both reasonable and do-able.

17                    And I'd like to thank you once again

18        for your time. I have provided my contact

19        information for you on the last sheet and I am

20        certainly willing to answer any and all

21        questions.

22                    CHAIR HILLMAN: Thank you.

1                     Commissioners, if we could just

2         agree that we will let Ms. Coggins make her

3         presentation and then open it up for questions to

4         both. Thank you.

5                     Ms. Coggins.

6                     MS. COGGINS: Thank you. Thank for

7         inviting me here today to provide some comments

8         on the proposed Voluntary Voting System

9         Guidelines.

10                    First let me say that it is my

11        perception that in view of the HAVA January 2006

12        deadline, there is significant concern in the

13        states and the vendors in how the Human Factor

14        requirements stack up against the current testing

15        of accessibility to the 2002 standards.

16                    I think Joe has done a really good

17        job of talking about impact in the long term. I

18        thought I may address my remarks a little to the

19        transition period of providing a little bit of

20        insight as to where the differences are for

21        people are going to have to be making some

22        decisions for January 2006.

1                     First I just want to say this

2         release was made available late last week so

3         these comments are based on my preliminary review

4         of the guidelines and should not be considered

5         complete research.

6                     Also, there has not been really a

7         peer review on the comments, which is our normal

8         process, in view of the time frame. These are my

9         initial interpretations of the guidelines and

10        shouldn't be considered a formal in-depth

11        analysis. We actually are going to make a

12        recommendation that maybe the EAC consider

13        providing this as a service to the states and the

14        vendors in the interim.

15                    Third, in identifying those

16        differences, it does not mean that the systems

17        qualified to the 2002 standards do not contain

18        some of these required features. It means that

19        they weren't required to contain these features

20        when they were tested to the 2002 standards.

21                    A vendor may have incorporated some

22        of these features into their systems. In that

1         case, the ITA would have been tested to them.

2                     The first group of changes is audio.

3         And in the 2002 standards it did not identify

4         specific decibel volume settings for the initial

5         volume or minimum volume control. The VSS did

6         specify maximum, so things were tested to that.

7                     In the 2002 VSS it did not identify

8         specific frequencies over the audible range. So

9         that would be something that would need to be

10        tested.

11                    The 2002 VSS did not stipulate the

12        use of human recorded speech rather than

13        synthesized speech. There are 2002 qualified

14        systems that use synthetic speech.

15                    And there appears to be a

16        contradiction in this requirement to the

17        underlying principle that you don't dictate a

18        design.

19                    And it seems within the discussion

20        on this requirement, it is preferred not to use

21        synthesized speech. But that may be an issue as

22        to systems that are out there.

1                     The 2002 VSS did not stipulate that

2         voters control the rate of speed. And now also

3         the voting system guidelines indicate this as a

4         'should' and not a 'shall'. So I'm not clear

5         whether or not the intention is that that will be

6         a requirement or if that's just --

7                     A 'should' is normally optional when

8         you are looking at a requirement. And there are

9         vendors out there who have addressed this issue.

10                    The next group of changes deals with

11        controls. The 2002 VSS did not stipulate that

12        mechanically operated controls or keys on the

13        voting system shall be tactilely discernible

14        without activating those keys. I'm not sure that

15        that is a really significant issue as far as

16        current systems out there. It doesn't pop to

17        mind that that seems to be a problem.

18        2002 VSS does not stipulate that the

19        status call logging or toggle control keys, such

20        as shift keys, shall be visually discernible and

21        discernible either through touch or sound. That

22        was not stipulated. Nothing comes to mind as

1         being really an issue with that though.

2                     The 2002 VSS did not stipulate that

3         controls shall not require direct bodily contact

4         or for the body to be part of any electrical

5         circuit. Additionally, it did not stipulate that

6         the voting system shall provide a mechanism to

7         enable non-manual input that is functionally

8         equivalent to tactile input. This would be along

9         the lines of CIVNTOP (ph).

10                    And, also, this is stated as a

11        'should' and not a 'shall'. But there are

12        systems out there, 2002 qualified, that did

13        incorporate this type of technology.

14                    And the same with the next

15        requirement, that the VSS stipulates voters who

16        lack fine motor control or use of their hands

17        perform submission of ballot.

18                    Again, that's something that vendors

19        have addressed, some.

20                    Also the guidelines, as Joe had

21        said, are changing the physical reach. They are

22        much are more detailed and the numbers appear to

1         have changed. So that's something that may need

2         to be addressed.

3                     There are systems out there that

4         have really addressed some of these issues

5         through documentation. I have not done a

6         complete analysis of all the changes that went

7         into effect and the measurements. But in some

8         cases this may actually be -- It's an issue of

9         setup and it may actually be something that can

10        be handled from documentation levels.

11                    CHAIR HILLMAN: You will need to

12        speak into the mike directly. You are turning

13        away from it and I think they can't hear you.

14                    MS. COOGINS: I'm sorry. I am

15        usually so loud, they usually tell me to quiet

16        down.

17                    Visual sight impairments is the next

18        group, the final group. The 2002 VSS did not

19        stipulate a system adjusted by the voter or poll

20        worker, including font size, color and contrast

21        shall automatically reset to the default after

22        each voter.

1                     This was actually addressed in a

2         NASED bulletin. So it is my feeling that most

3         systems should have tested that.

4                     The NASED bulletins were part of --

5         The implementation program is this mechanism for

6         having additional bulletins for clarification,

7         were part of the --

8                     CHAIR HILLMAN: Tap it and see what

9         you are getting.

10                    MS. COGGINS: Is that any better?

11                    CHAIR HILLMAN: Yes. Just have it

12        as close to you as you can.

13                    MS. COGGINS: Okay.

14                    CHAIR HILLMAN: Okay. There you go.

15                    Do you want to start over and repeat

16        everything --

17                    No, I'm only kidding.

18                    MS. COGGINS: Do you really want to

19        hear it?

20                    So in terms of the NASED Technical

21        Bulletin Number 1, that was a mechanism that was

22        part of the 2002 implementation program that

1         there would be additional bulletins for updates.

2         So that should have been tested.

3                     The 2002 VSS did not stipulate that

4         all tests intended for the voter during the

5         voting session be presented in sans serif font.

6         It did state that there should be a clear font.

7                     I don't know -- I am under the

8         impression that states do actually stipulate

9         certain fonts on their ballots. So is this

10        perhaps an issue with compliance with state laws

11        or are all states using this font.

12                    If this is a common thing, I'm not

13        aware of that.

14                    The 2002 VSS did not stipulate

15        minimum figure to ground ambient contrast ratios

16        for text to informational graphics. So that

17        would be something that would have to be tested.

18                    And, lastly, the VSS did not

19        stipulate a minimum size for sensitive touch

20        areas. So that would be something that, although

21        I haven't researched the actual sizes, but I

22        don't think that that will wind up being greatly

1         significant to systems that were already

2         qualified.

3                     And, as I said, I understand that

4         the guidelines are just out for public comment.

5         But in view of the phone calls that I have

6         received from states - asking me, oh, my God, I

7         have a system; I want to buy it. Am I going to

8         be in trouble if I buy this system because it's

9         2002 qualified? What does this mean? - I would

10        recommend that in order to keep the states and

11        the local jurisdictions informed and the vendors

12        and public, that the EAC at least initiate

13        publishing a formal gap analysis between the 2002

14        voting system standards and the proposed voting

15        system guidelines.

16                    In doing so, I believe that will

17        help states ask vendors about specific questions

18        about their voting systems. Also, vendors with

19        2002 VSS qualified systems that support things

20        that are proposed in the guidelines could request

21        the lab to provide independent verification of

22        any supported gap-identified functionality and

1         that this could then go as addendums to their

2         reports.

3                     Let's see. There are actually -

4         because of time - there are a number of items and

5         I think Joe went through a number of items.

6                     In an overall comment --

7                     Am I off again?

8                     CHAIR HILLMAN: Yes.

9                     MS. COGGINS: I didn't touch it.

10                    CHAIR HILLMAN: He's working on it.

11                    MS. COGGINS: I think rather than

12        going through each individual issue that we

13        thought, there is kind of an overall comment that

14        I would like to make.

15                    One of the criticisms of the 2002

16        standard was imprecise language. And I will be

17        submitting these comments directly through the

18        comment process.

19                    But there are cases in here where we

20        are seeing imprecise language. In some instances

21        we are seeing - I'll give you an example - the

22        audio system should allow voters to control,

1         within reasonable limits, the rate of speech.

2                     Reasonable limits is not testable.

3                     What's reasonable to you/what's

4         reasonable to me, is a lawsuit.

5                     So that's the one thing that I would

6         say.

7                     We need those kind of things

8         tightened up, providing us with limits.

9                     There is one instance in the Voter

10        Verified Paper Audit Trail, it appears in the

11        guidelines that either the paper or the

12        electronic ballot could be the ballot of record.

13        And in Section 152 of the VSS, which has not

14        change in the guidelines, it actually says that a

15        paper based voting system can't record, store or

16        tabulate selections.

17                    So this seems to be a conflict.

18                    And I may be misinterpreting what is

19        in the new and changed guidelines. But that just

20        seems to be something that maybe should be looked

21        at.

22                    So in terms of just the imprecise

1         language, we would like to see that addressed.

2                     Lastly, there is one item that we,

3         as the labs, in the NASED technical committee

4         meeting, had discussed: issuance of a

5         clarification bulletin that addressed some of the

6         issues encountered in interpretation with the

7         code review standards, and specifically the

8         application of the term in module in various

9         requirements.

10                    But, also, the VSS gave us very

11        broad language on some security requirements and

12        code review. Well, just basically, security

13        requirements. And it allowed the labs on their

14        own to interpret some coding practices as

15        insecure.

16                    The problem is, these practices are

17        not explicity identified and that leads to

18        inconsistent application. A good example of this

19        is the bar coded password.

20                    The labs hope that more explicit

21        requirements for security review would be part of

22        this release. And that's something that I think

1         I would like to see in the security requirements

2         of best practices or here are the things in code

3         review that we want to make sure exist.

4                     Again, as I said, the imprecise

5         language is really one of my concerns, is that

6         all requirements really have testable criteria.

7                     Overall, the standards are good and

8         we look forward to testing to these and we will

9         submit specific comments on these through the

10        comment process.

11                    CHAIR HILLMAN: Okay, thank you.

12                    While the little problems that we

13        are having with the technology here don't disrupt

14        our meeting, you can imagine it wouldn't be funny

15        if we were encountering these little problems

16        with voting machines on Election Day, pointing

17        clearly to the need for good guidelines, testing

18        and certification and re-testing and

19        recertification as necessary.

20                    I thank both of you for your

21        comments this morning. The responsibility for

22        accrediting test labs and the whole certification

1         process is one that the Federal Government,

2         through the Election Assistance Commission, is

3         embracing for the first time. It is a

4         significant and rather awesome responsibility.

5         And it absolutely is important for us to receive

6         your comments against the guidelines as you have

7         noted.

8                     The purpose of this hearing, the

9         purpose of the entire 90 day process, is for us

10        to receive this kind of critical input and for us

11        to have an opportunity to explore some of the

12        comments that you have made and we welcome this

13        opportunity to do that.

14                    Commissioners, I think we have

15        sufficient time. We have about, in total, 30

16        minutes. And so I guess if we divvy it up 10/10

17        and 10, that does include responses back. So we

18        beg your indulgence to confine your responses to

19        questions, if it appears we are bumping up

20        against the time limit.

21                    Thank you.

22                    Commissioner, Mr. Vice Chairman.

1                     VICE CHAIR DeGREGORIO: Thank you,

2         Madam Chair. And thank you both for the

3         excellent presentation and the written comments

4         that you have provided us.

5                     We certainly recognize that this is

6         a first cut for you since these guidelines were

7         just published in the Federal Register and we

8         recognize that over time you will probably add to

9         these comments.

10                    But, initially, they are well done

11        and they certainly raise a lot of important

12        issues for us to look at over the next 90 days,

13        because you are the ones that will have to test

14        the equipment to these guidelines. And you have

15        certainly raised some issues here that are

16        important for us to consider.

17                    One thing that I have heard from

18        election officials around the country about these

19        new guidelines is the time that may be required,

20        the additional time that may be required to test

21        this equipment against these new guidelines for

22        the equipment to be qualified or certified.

1                     And do you see, because you have

2         been in this business for quite a while now, that

3         these guidelines will add additional time to the

4         process to qualify a voting system when it is

5         presented to us and when you have to test it

6         against these guidelines?

7                     Mr. Hazeltine, you can certainly --

8                     MR. HAZELTINE: Mr. Vice Chair, yes,

9         I do. Off the top of my head, I am looking at

10        probably one to two weeks. There are additional

11        activities. They are not tremendous, but there

12        are additional activities.

13                    VICE CHAIR DeGREGORIO: What is the

14        average time? You say one or two weeks

15        additional, so what --

16                    MR. HAZELTINE: Well, the average

17        time for a system, from start to finish, is

18        normally a three to four month time frame. So we

19        are not talking about an order of magnitude

20        change.

21                    VICE CHAIR DeGREGORIO: What about a

22        cost increase to the client? As I understand it,

1         it's going to be the vendors that will pay you

2         all to test their equipment against these

3         guidelines. Is there going to be a significant

4         cost increase as a result of this --

5                     MR. HAZELTINE: I wouldn't say there

6         would be significant. Yes, the activities will

7         go up, so that the time we spend will go up and

8         corresponding new costs.

9                     MS. COGGINS: I would say it will

10        add a small amount of time. It's not really

11        significant. Whether they pass or fail is, a lot

12        of times, whether it takes longer or not.

13                    So if they come and they have to

14        actually retest, then it takes longer and it

15        costs more.

16                    So it's hard to necessarily put a

17        number. You always talk about a number that is

18        successful. But a one week test may require a

19        two month re-engineering effort. So that would

20        be the situation.

21                    And you have a very different, a new

22        vendor has a very different experience from a

1         vendor who is in the market and has deployed

2         voting systems and has been doing this a long

3         time.

4                     So one person it may take just an

5         extra day for a test. It also depends on what

6         the systems themselves contain.

7                     So if we have already tested for

8         some of these things, then there really wouldn't

9         be any additional testing because the VSS

10        requires that we test to their own requirements

11        in certain optional functionality or additional

12        functionality.

13                    There I went again.

14                    CHAIR HILLMAN: It will be fixed

15        during our break so that we don't have to go

16        through this. So our apologies to the two of

17        you.

18                    MS. COGGINS: Not at all.

19                    So that's basically it. There are a

20        lot of variables that are involved in it. It

21        depends on the vendor.

22                    It will add some time; it will add

1         some expense. And I think also in terms of some

2         people, there will be significant times of

3         engineering. That I think would be more of the

4         time that I see, is engineering the products to

5         reach some of these requirements.

6                     But in terms of testing, it probably

7         would be a 5 to 10 percent increase.

8                     VICE CHAIR DeGREGORIO: Previous

9         standards have had an implementation period.

10        When the 1990 standards were developed, there was

11        a several year implementation period. And in

12        2002 there was, I believe, a two year

13        implementation period.

14                    We have proposed in this draft a 24

15        month implementation period, which would take us

16        to the fall of 2007, before these guidelines

17        would take effect and have to be --

18                    Is that a reasonable approach in

19        your view, Mr. Hazeltine?

20                    MR. HAZELTINE: Mr. Vice Chair, as I

21        said earlier, I think that is both reasonable and

22        do-able.

1                     There are some things which I think

2         need to be on the fast track, which would be the

3         Sections 2.2.7, the accessibility requirements.

4                     MS. COGGIN: The market will not

5         wait for them to be put into place. The market

6         will dictate to us that we have to move faster on

7         these.

8                     So while we may have - we will have

9         vendors who will be competing to try and get

10        there before others. So at the lab, we will be

11        trying to get there as quickly as possible.

12                    So two years is, it's a good cutoff.

13        Because as we found in 2002, there were

14        situations where people weren't ready to get

15        there. But I don't know that people will

16        actually --

17                    Somebody will be ready to go three

18        months from now, in my interpretation.

19                    VICE CHAIR DeGREGORIO: If a state,

20        for instance, chooses to adopt a VVPAT portion of

21        this and have its implementation effective for

22        next year, for next year's election, the

1         equipment they are going to use, because their

2         state law or state regulation requires them to

3         use the VVPAT if they use DRE equipment, so you

4         would be prepared then to test this equipment

5         against the VVPAT section of these guidelines,

6         again, if the states chose to adopt that and have

7         a different implementation date?

8                     MS. COGGINS: It depends upon --

9                     In terms of our dealing with the

10        vendor, it's really, we would be working on what

11        their requirements would be.

12                    If they are designing their system

13        to a particular state's requirements, then we

14        would need to design test cases for that.

15                    The good news would be that,

16        ultimately, we probably are going to be able to

17        re-use that with other vendors. But, initially,

18        getting out there --

19                    You actually are to look -- There's

20        very much custom work that's done in every ITA

21        qualification because no two systems are truly

22        the same. So while you can work off a really

1         high level set of requirements, when it comes

2         into an implementation - and, in fact, we have

3         this discussion sitting in a lot of cases where

4         we are sitting in a room and we are testing

5         something on one voting system and we go, oh, my

6         gosh, this is so much like such and such. Did we

7         test them for that? Oh, no, but they don't do

8         this piece, they don't do write-in's the same.

9         Whew, okay, we are covered. Yes, we tested it to

10        the correct standard in that particular case.

11                    So every system actually has a great

12        deal of customization on the test cases.

13                    So I think it's actually our normal

14        process that we really have to be responsive to

15        the system. And while the good news will be that

16        we will be able to repeat processes more

17        consistently across various vendors, at this time

18        we actually have to respond to whatever a vendor

19        design is, because that's what the VSS says that

20        we do.

21                    MR. HAZELTINE: Relative to the

22        Voter Verified Paper Audit Trail, I see no real

1         issues with it.

2                     The major thing that we need - I

3         have already talked to Mr. Wilkey about this - is

4         the compliance matrix. And we have begun

5         discussing having a meeting later this month to

6         do that.

7                     Once that's in place, we are ready

8         to go.

9                     VICE CHAIR DeGREGORIO: Thank you,

10        Madam Chair.

11                    CHAIR HILLMAN: Okay. Thank you.

12                    For your entertainment pleasure, we

13        are raising and lowering the blinds, as you can

14        see. Kind of looks like we are now in a

15        submarine and submerging at any moment now.

16                    Seriously, for the people on this

17        side of the room, even though the sun is

18        filtered, it was getting kind of warm and then

19        little technical gliches with plants and other

20        things being in the way.

21                    So, hopefully, you were entertained

22        for a moment. But we should be squared away now.

1                     Thank you.

2                     Commissioner Martinez.

3                     COMMISSIONER MARTINEZ: Thank you,

4         Madam Chair.

5                     My thanks to both of you as well for

6         making the time and the effort to be here. And

7         as the Vice Chair has noted, we were trying to

8         get this out as soon as we could. And,

9         unfortunately, it took us a little bit longer to

10        do our internal due intelligence. So it was just

11        out last week.

12                    And we appreciate that you have done

13        a rather hasty, but I think a very excellent,

14        analysis for us to begin to think about with

15        regard to --

16                    There goes my mike. Am I back on?

17        Everybody hear me?

18        -- with regard to the Proposed

19        Voluntary Voting System Guidelines.

20                    I want to start with a general

21        question for the both of you.

22                    And that is, one of the things that

1         we have to contemplate right now at the EAC,

2         along with our partners at NIST and of course

3         with the TGDC, is the idea of what happens after

4         this initial, after this first draft is adopted

5         in full or in final form.

6                     And that is to say that there has

7         been, we have deliberated internally and talked

8         amongst ourselves with NIST and the TGDC about

9         whether this is a Version I versus a Version II

10        coming out at some point in the near future.

11                    And I wondered, from your

12        perspective, since you actually have to do the

13        use of the requirements and the actual testing of

14        the voting systems, what life would mean for the

15        testing labs if we were to embrace a chronology

16        that says we go final with VVSG in 90 days with

17        this version and then soon thereafter, let's say

18        six months after that one goes final, we come out

19        with a Version II, another set of, again, revised

20        Voluntary Voting System Guidelines.

21                    Is that problematic from your

22        perspective? What opinion would you like to

1         offer us so that we can be better informed in

2         trying to make a decision about the timing of

3         what comes after this, if anything?

4                     Mr. Hazeltine.

5                     MR. HAZELTINE: Thank you,

6         Commissioner Martinez.

7                     From a testing laboratory

8         standpoint, we can be ready in 90 days as to the

9         standards, and if you decide to revise it 90 days

10        later, we can probably be ready for that one as

11        well. From a reality standpoint, test, the

12        vendors will not be --

13                    CHAIR HILLMAN: Your mike.

14                    MR. HAZELTINE: Once again, we can

15        be ready from a testing laboratory standpoint.

16        But I do think the vendors would have some

17        problems with that. It would be a continuing

18        ratchet of requirements that the systems need to

19        have.

20                    I frankly think that we need to

21        reach some point of stability that people can

22        design to and then if we want to move from

1         forward from that.

2                     COMMISSIONER MARTINEZ: Ms. Coggins.

3                     MS. COGGINS: I would agree also in

4         terms of, in all quality systems, management of

5         change is one of the important factors. So as a

6         test lab, that's one of our mandates, that we

7         need to be able to do that.

8                     I also agree with Joe in that I

9         don't think it's just the vendors. I think it's

10        also the states.

11                    There is paralysis if you have

12        got -- Maybe it's better that it comes in six

13        months or maybe it's better that you defer the

14        first one six months.

15                    But the basic issue is, people don't

16        know what to do. They don't know whether to buy;

17        they don't know what's going to happen; are they

18        going to lose their HAVA funds?

19                    That's information that I am getting

20        from the states where they are calling me and

21        asking me, you know, what does this mean. What

22        does it mean these changes of standards?

1                     COMMISSIONER MARTINEZ: Is there a

2         process right now with the current 2002 VSS and

3         the NASED process where if they want to add

4         something to the existing standards, it is done

5         via an amendment of some kind to the 2002 VSS?

6                     Or is there a process, if they

7         identify something that needs to be addressed - I

8         think you mentioned something about a bulletin

9         that might be issued by NASED or something like

10        that - so there is a process right now.

11                    Can you say something a little bit

12        about that real quick, Mr. Hazeltine.

13                    MR. HAZELTINE: Commissioner

14        Martinez, yes. NASED over the years has issued a

15        number of technical guidances that kind of

16        attaches, kind of like an appendix, if you will,

17        to the standard and provide guidance to us and

18        also to the vendors on what the requirements of

19        the system are.

20                    That is a kind of a stop gap way of

21        adding things in small pieces.

22                    MS. COGGINS: The problem in the

1         process at this point is you have this transition

2         between NASED and the EAC, so we are in

3         paralysis, an example being this bulletin that we

4         talked about about a year ago, can we get this

5         out.

6                     And at this point NASED is

7         anticipating giving this over to the EAC. So,

8         yes, the method is there, but the will is not

9         there, in my belief. I'm not trying to speak for

10        anybody else or anything here.

11                    But there is just, people are in

12        this flux and they are not quite sure. If I put

13        out a bulletin today, what's the impact?

14                    COMMISSIONER MARTINEZ: Right. I

15        guess what I am trying to get at, if I am trying

16        to make a decision about what happens next, is

17        the process working the way it exists now, in

18        that if there is a deficiency that's identified,

19        that you don't have to go through a whole

20        promulgation process again of redoing or revising

21        the entire VSS, but instead you would address

22        those particular deficiencies via a bulletin or

1         via some sort of a more responsive mechanism.

2         And it sounds like that is working right now.

3                     So as we contemplate how to

4         structure ourselves - since we are, by law,

5         supposed to take over this process - that is

6         something that we should be informed about,

7         basically.

8                     MS. COGGINS: The system is there.

9         It's just not currently working -- It's not,

10        nothing is going on, it really is. But, yes,

11        there is a mechanism.

12                    COMMISSIONER MARTINEZ: And one that

13        works, Mr. Hazeltine?

14                    MR. HAZELTINE: Yes, I believe it

15        does.

16                    COMMISSIONER MARTINEZ: In your

17        testimony, Mr. Hazeltine, you have referenced the

18        fact that in some areas - for example, in the

19        BBSG - the section dealing with wireless

20        technology, that you would need to acquire

21        subject matter expertise. Is that available? Is

22        it even out there?

1                     MR. HAZELTINE: Commissioner

2         Martinez, I am not sure and we have not

3         researched that.

4                     I am sure there are experts in blue

5         tooth and 80211 or BT (ph) whatever.

6                     COMMISSIONER MARTINEZ: So there is

7         personnel that you can get.

8                     MR. HAZELTINE: There is. And we

9         have a good bit of expertise in the same thing as

10        well.

11                    Frankly, wireless, I'm not sure if

12        the technology is ready for implementation,

13        primarily from a denial service viewpoint.

14                    COMMISSIONER MARTINEZ: Do you

15        agree? Are there some areas that it's going to

16        be necessary to acquire some additional

17        expertise?

18                    MS. COGGINS: Yes, I think that one

19        of our comments about imprecise language is that,

20        it's saying you are becoming a subject matter

21        expert but it's not identifying the

22        qualifications of the subject matter expert.

1                     That may be something that is going

2         to be addressed in the accreditation criteria.

3         But we would like to see in the voting guide that

4         it actually references a subject matter expert

5         who is accredited by the lab or whatever is the

6         specific thing.

7                     But, yes, that is some of the

8         imprecise language that we would like to see

9         clarified.

10                    COMMISSIONER MARTINEZ: I guess

11        going back to something that my colleague, the

12        Vice Chair, was talking about earlier, one of the

13        things that we had certainly emphasized - and I

14        know that the members of the TGDC were in

15        agreement with this - and that is to try to

16        deliver - with the VVSG - to try to deliver a

17        product that could be responsive to the upcoming

18        Federal election cycle next year, if at all

19        possible.

20                    And, in particular, what we were

21        dealing with was the fact that - I think at last

22        count - something like 16, maybe 17 states, now

1         have madated, via legislative action or

2         administrative rule, the use of VVPAT technology.

3                     And for those states that have moved

4         in that direction - and for perhaps others that

5         will move in that direction between now and the

6         next election cycle - that there would be a means

7         to be able to test --

8                     For example, if I use a DRE in a

9         jurisdiction right now, but the vendor in the

10        area has developed a VVPAT technology that could

11        simply be, essentially, plugged into that

12        existing DRE, that the plug-in, the added device,

13        the VVPAT component, can be tested in an

14        objective repeatable manner, that would happen,

15        obviously, through your test labs.

16                    And what I you hear you all saying

17        is that you can be ready to do that once we go

18        final with the VVPAT section of this VVSG. Is

19        that correct, Mr. Hazeltine?

20                    MR. HAZELTINE: Commissioner

21        Martinez, that is correct. Of the requirements

22        in the document, that was probably some of the

1         most straightforward and easiest to implement.

2                     COMMISSIONER MARTINEZ: Ms. Coggins,

3         do you agree?

4                     MS. COGGINS: Yes. In fact, we are

5         already testing that because vendors have already

6         addressed that.

7                     COMMISSIONER MARTINEZ: So,

8         irrespective of whether a state decides to pull

9         down that section of the VVSG early, a vendor

10        might come to you after we go final with that and

11        say we want to be among the first to be tested

12        against the new VVPAT requirements, you all are

13        going to be ready to do is so is what I am

14        hearing you say.

15                    MR. HAZELTINE: Yes, sir.

16                    MS. COGGINS FOGEL: In fact, I think

17        we have been contacted by a vendor already who is

18        anticipating this and has asked us if they can

19        start testing sometime this month.

20                    COMMISSIONER MARTINEZ: I will take

21        this opportunity, I think we will talk more about

22        this this afternoon, that we go out of our way in

1         this document, in the VVSGV, to add language - I

2         think it's in the appendix right now, in Appendix

3         D - but to say to the general public and to the

4         reader of the VVSG that VVPAT is but one of many

5         ways to achieve what is termed independent

6         verification by NIST and by the TDGC. That other

7         ways to achieve independent verification - aside

8         from using VVPAT technology - is out there.

9                     And at some point perhaps - or

10        perhaps it's usable now - but that there are

11        other means to achieve this.

12                    And at some point down the road -

13        perhaps through an amendment process to the VVSG

14        when they go final, perhaps in the next iteration

15        of the VVSG - there would be other testable

16        requirements for other means of achieving

17        independent verification.

18                    Are you following what I am saying?

19                    MR. HAZELTINE: Yes. If I were to

20        understand what you are saying, I don't think you

21        want to do anything which would stifle

22        technology.

1                     COMMISSIONER MARTINEZ: Right.

2                     MR. HAZELTINE: Let the vendors be

3         creative and come up with other ways of meeting

4         the requirements.

5                     The requirements, I think, are well

6         stated in the document.

7                     COMMISSIONER MARTINEZ: Right. But

8         what you said in your testimony is getting to

9         other means, testing the requirements for other

10        means of independent verification will require

11        additional major effort on your part.

12                    MR. HAZELTINE: As stated in

13        Appendix D, the dual system where you have got

14        the one system and you go back and you repeat the

15        count on the second system, that is two pieces of

16        hardware; that's twice as much testing.

17                    There are other ways to do that.

18                    COMMISSIONER MARTINEZ: Right.

19                    Any comments, Ms. Coggins?

20                    MS. COGGINS: No. In terms of

21        testing, it's going to be nothing really

22        different from what we are currently doing now.

1                     COMMISSIONER MARTINEZ: And I am

2         getting a bad stare from the Chair.

3                     CHAIR HILLMAN: Excuse me, not a bad

4         stare. I am just letting you know we're a little

5         over time.

6                     COMMISSIONER MARTINEZ: I will do

7         one quick, perhaps more of a statement, and you

8         all can do a quick comment, if you want.

9                     The gap analysis that you have

10        suggested for the following reason: the current

11        set of voting system standards, the 2002 VSS, was

12        promulgated by our predecessor agency, the

13        Federal Election Commission. It was promulgated

14        over a two or three year period and it went final

15        sometime - I don't remember the exact month in

16        2002 but perhaps one of you can tell me.

17                    MR. HAZELTINE: Effective January 1,

18        2004. At that point everything had to comply.

19                    COMMISSIONER MARTINEZ: Right. But

20        the FEC Commissioners voted --

21                    MS. COGGINS: It was actually

22        published in April 2002.

1                     COMMISSIONER MARTINEZ: That's the

2         date I was looking for.

3                     MS. COGGINS: It was released in a

4         meeting in May in Colorado.

5                     COMMISSIONER MARTINEZ: The point

6         that I would make, then, to go to your suggestion

7         on the gap analysis, is that, HAVA came along

8         after the 2002 voting system standards went

9         final.

10                    So HAVA, like it or not, HAVA has

11        some voting system standards itself in Section

12        301, which are nothing at all voluntary, in fact,

13        they are mandatory on 1/1/06 - or, actually on

14        1/1/06 - upon all jurisdictions covered by HAVA.

15                    So the gap that you have,

16        essentially, is that you have current voting

17        system standards that were promulgated before

18        HAVA was even close to its final form.

19                    So if you are a vendor, you get

20        tested to the 2002 standards, you still don't

21        have anybody to tell you - other than perhaps

22        your own internal counsel and your own

1         deliberation and your own specifications that you

2         would impose upon yourself as a vendor - whether

3         your system is compliant with what Congress came

4         out with in the Help America Vote Act, in Section

5         301, particulary 301(a)(3), which is the

6         accessiblility language.

7                     Would you all agree with that?

8                     Mr. Hazeltine, go ahead, or whoever.

9                     MR. HAZELTINE: Mr. Martinez, I

10        would agree with that.

11                    I believe the common sense way of

12        looking at this, there are voting systems that

13        are out there now which are complying with the

14        majority of this document other than the HAVA

15        requirements.

16                    HAVA work station, the voting

17        stations, will probably end up being an

18        independent voting station at the polling place.

19        So it is actually almost a separate piece of

20        hardware.

21                    COMMISSIONER MARTINEZ: Ms. Coggins,

22        you were going to say something.

1                     MS. COGGINS: I would say that it's

2         not just that -- You have partial compliance at

3         this point, based upon the vendors. You are

4         correct, yes, you had a gap between the time when

5         the standards were there and when HAVA was

6         imposed.

7                     And, truly, that gap actually is

8         until today, because this is really the first

9         time those standard are there. So it has been, I

10        don't know what, is it three years or something?

11                    COMMISSIONER MARTINEZ: Right.

12                    MS. COGGINS: So that's truly what

13        the gap is.

14                    COMMISSIONER MARTINEZ: Right.

15                    MS. COGGINS: So, yes. And you have

16        had people who have tried to anticipate this and

17        have tried to get there. So in terms of what

18        they are looking - the information - I would

19        absolutely suggest that this information be put

20        out for the benefit of everyone.

21                    COMMISSIONER MARTINEZ: Thank you.

22                    Thank you, Madam Chair.

1                     CHAIR HILLMAN: Okay.

2                     Our Executive Director has a

3         question or two that he wants to ask, so I am

4         going to share my time with him.

5                     But I am going to exercise my

6         prerogative to go first. So if we run out of

7         time, you don't get to ask your question.

8                     I'm only kidding.

9                     What does it mean to you that the

10        language has been changed and we are now talking

11        about guidelines instead of standards? What, if

12        anything, in the field does that mean?

13                    MR. HAZELTINE: Madam Chair, I am

14        aware of a number of documents. Sometimes they

15        are called guidelines; sometimes they are called

16        standards.

17                    They are called -- That's the

18        requirements requirement. So that really the

19        name is no major consequence.

20                    MS. COGGINS: I agree. We are being

21        asked to test requirements and the vendor is

22        actually, when they come to us to initiate a

1         contract, they are specifying test us to this.

2         And so that's what we are testing now, that's our

3         interpretation.

4                     CHAIR HILLMAN: So the quality has

5         no effect on it?

6                     MS. COGGINS: No.

7                     CHAIR HILLMAN: I was momentarily

8         interrupted, I believe, when you were responding

9         to the Vice Chairman's question about the length,

10        the anticipated length of time it would take to

11        test under the guidelines.

12                    Could you just tell me, ballpark,

13        what that was.

14                    MR. HAZELTINE: Madam Chair, my

15        response, right now, our initial review, is maybe

16        one to two weeks. It is not a tremendous --

17                    CHAIR HILLMAN: In total?

18                    MR. HAZELTINE: One to two weeks on

19        a program which probably takes normally about

20        three months.

21                    CHAIR HILLMAN: So we are talking

22        three to four months.

1                     MR. HAZELTINE: It would probably

2         end up adding a half a month to it.

3                     MS. COGGINS: One of the things that

4         is a little difference between the last, that,

5         and this initiative, still kind of up in the air

6         where we are going, SysTestLabs is both a

7         hardware and software test lab.

8                     So our test cycle may be longer

9         because we have a bigger scope.

10                    Currently, Wyle is a hardware ITA

11        and SysTest is a software ITA, the initial

12        software ITA.

13                    So that may be -- There is a

14        different scope of testing that goes on.

15                    For us the test effort is - it can

16        be two months to - we have had people go 18

17        months. So it just depends on the quality of the

18        system that's brought to mark.

19                    THE CHAIR: Did you want to correct

20        something, Mr. Hazeltine?

21                    MR. HAZELTINE: Yes, ma'am.

22                    We would certify for both.

1                     CHAIR HILLMAN: I think I heard you

2         correctly say that there would be an increase in

3         the cost of testing, but it wouldn't be a huge

4         increase.

5                     All things are relative. So, is it

6         5%, 10%?

7                     I know we are ballparking here. But

8         I'm trying to get an appreciation as to whether

9         the cost of testing is going to increase 25%,

10        50%, 10%?

11                    MR. HAZELTINE: Madam Chair, it's

12        not in those ranges.

13                    Just using this straight map, you

14        say it's a three month program - and to your 12

15        weeks, you are adding two.

16                    So 8 to 10%, somewhere in that

17        range, doing it as an engineer.

18                    CHAIR HILLMAN: Even with some of

19        the other things that you said you would have to

20        bring into the testing process because of the new

21        requirements?

22                    MR. HAZELTINE: There are new

1         requirements; they require new tests, primarily

2         the functionality of things.

3                     Once you have gone through it the

4         first time, you have all the fixtures and the

5         knowledge. So it becomes somewhat repetitive.

6                     But the first time through it will

7         always be a bit of a challenge.

8                     The accreditation is an additional

9         cost for us. It's not necessarily something

10        which the vendors would be picking up.

11                    CHAIR HILLMAN: Ms. Coggins.

12                    MS. COGGINS: Would you repeat the

13        question.

14                    CHAIR HILLMAN: It was just about

15        the increase in cost to test against the new

16        guidelines.

17                    MS. COGGINS: I would not see it as

18        being a significant cost, relative to the overall

19        cost of an effort.

20                    It is, again, it's really dependent

21        upon the system that is brought us, independent

22        conditions, as to the length of the time and the

1         cost.

2                     CHAIR HILLMAN: My final question is

3         for you, Ms. Coggins,

4                     You said in the beginning of your

5         presentation, you were talking about the quick

6         time that you used - and we appreciate it - to

7         review the guidelines and that your normal peer

8         review process could not be engaged.

9                     And then you talked about the gap

10        analysis.

11                    You said something about providing

12        this as a service to the states. So my question,

13        if you can remember, is, what is the 'this'?

14                    MS. COGGINS: The gap analysis.

15                    It's basically providing information

16        so that the people understand, here are the

17        likely things you need to be looking for. Ask

18        your vendor, do they have an alternative method,

19        a non-tactile method for entering the voting

20        system.

21                    If they have got that, check off the

22        box; we've got one. And is it in their report.

1         That's basically it.

2                     The other thing, too, is, states may

3         have the opportunity to request additional

4         documentation on specific issues, if they are not

5         seeing these clearly identified in the

6         qualification reports that have been issued.

7                     So that might be something so

8         that --

9                     I know they are all concerned.

10                    CHAIR HILLMAN: Thank you.

11                    Mr. Wilkey.

12                    THE EXECUTIVE DIRECTOR: I will

13        stick to the time limit. I don't want to lose my

14        job before I get my first paycheck.

15                    First, I want to make just a quick

16        comment to both of you - and I wish that the

17        other ITA was here also.

18                    I have had the distinct honor and

19        pleasure of working with you in a prior life over

20        the past number of years and you have always

21        shown a very high level integrity, of

22        professionalism in everything you have done, as

1         well as all of the members of the NASED Board,

2         who have given hundreds of hours of time or

3         talent, without remuneration, without any kind of

4         assistance whatsoever.

5                     I just have one question, and

6         primarily for you, Carolyn.

7                     You made some general comments about

8         it needs to do this or it needs to this

9         throughout your document. Is it your intent,

10        when you go back, to give us some suggested

11        language that you would like to see?

12                    And, Mr. Hazeltine, you could also

13        answer this.

14                    CHAIR HILLMAN: Just a second.

15        Could you please repeat your question, Tom,

16        because the signers are having a hard time

17        hearing you.

18                    THE EXECUTIVE DIRECTOR: I think you

19        made a number of general comments that it needs

20        to do this or it needs to do that, without giving

21        some specific language that you think would make

22        it better.

1                     I ask you this because I don't think

2         there is anyone who has seen this equipment up

3         close and personal as you have.

4                     So I think if you could give us some

5         specific language that you would like to see, it

6         would be very helpful.

7                     Is it your intent to do that?

8                     MS. COGGINS: I think in terms of,

9         in trying to say that something needs to be more

10        detailed, okay, it may be that the language we

11        can provide is, we need a limit. I can't tell

12        you the limit. I can't say - because that's

13        defined in the voting systems. It doesn't allow

14        me to do that.

15                    So we can't actually set the

16        standard, per se, but we can indicate, yes, this

17        language, I need a little more detail on what is

18        the limit here. Or, in some cases, like we have

19        seen, there is issues where it's talking about

20        unacceptable font. It would be helpful if you

21        provided, okay, give us four - I'm sorry, not

22        font, format - give us four acceptable formats

1         and then say you can also do it in publicly

2         acceptable ones. Honing things down a little bit

3         makes it easier not only on us, but makes it

4         easier on the entire community.

5                     Now we have two standard methods of

6         file transfers.

7                     MR. HAZELTINE: I have no problem

8         providing additional guidance. I think we can

9         give you some additional guidance to work with.

10                    CHAIR HILLMAN: Thank you. Thank

11        you all very much.

12                    And we will break now for lunch.

13                    I will ask if there are any

14        announcements before we do this.

15                    MS. THOMPSON: Madam Chair, you just

16        may want to make the announcement again about the

17        restrooms.

18                    CHAIR HILLMAN: Just a reminder that

19        the public restroom facilities are on the 8th

20        floor. They are not on this floor.

21                    And we will begin again at 1:30.

22        Thank you very much.

1                     AFTERNOON SESSION

2        

3                     PRESENTATION FROM VENDORS

4                     REGARDING THE PROPOSED VVSG

5        

6                     CHAIR HILLMAN: The United States

7         Election Assistance Commission's hearing on the

8         proposed Voluntary Voting System Guidelines will

9         resume.

10                    Panel Number 2, presentations from

11        vendors, those companies that manufacture the

12        equipment hardware and software that compose

13        voting systems.

14                    We have with us today Mr. John Groh,

15        President of Election Systems and Software

16        International, also known as ES&S; Mr. Jim Adler,

17        CEO Vote Here; Mr. Alfie Charles, Vice President

18        of business development, Sequoia Voting Systems;

19        and Mr. Dennis Vadura, CEO of AccuPoll.

20                    We will engage with this panel until

21        about 2:45 and then we will take a short break so

22        that we can set up for Panel Number 3.

1                     Gentlemen, thank you for being with

2         us. And without any further ado, I will go in

3         the order that I introduced you. So if we could

4         start with Mr. Groh.

5                     I think what we will do is receive

6         your presentations and then we will ask questions

7         following the fourth presentation. Okay.

8                     MR. GROH: Thank you, Chairman

9         Hillman.

10                    Again, for the record, my name is

11        John Groh. I also serve as Senior Vice President

12        of Election Systems and Software, Inc., which is

13        our American side company.

14                    Today I brought along with me,

15        though, a peer of mine, Ken Carbolito, who is

16        Senior Vice President of product development and

17        software development for our company.

18                    We want to thank you for inviting us

19        to be here today to present ES&S's views

20        regarding the proposed Voluntary Voting System

21        Guidelines, Version I.

22                    I want to acknowledge first the

1         leadership and the hard work that the EAC and

2         your staff has accomplished. It is remarkable to

3         many of us that are in this business that you

4         have been able to make as much progress and so

5         soon.

6                     Now let me get into a little bit of

7         the topic today about Version I. But indulge me

8         a little bit to allow me to share a little bit of

9         history on our company, because I think it sets

10        the stage of why we would be someone that could

11        speak to this.

12                    For over 35 years, ES&S has been at

13        the forefront of driving election systems

14        innovation and working with election officials to

15        enhance the voting experience for all.

16                    We are the largest and most

17        experienced provider of election systems and

18        services, supporting elections across the country

19        and throughout the world.

20                    We are very proud of the fact that

21        twice now we have been the first major election

22        systems vendor to certify our entire product line

1         against the Federal Voluntary Voting System

2         Standards, first in 1990 and again in 2002.

3                     In our opinion, the adherence to

4         standards and the rigor of the certification

5         process is critical to maintaining the integrity

6         of our elections. Therefore, we will tell you,

7         we embrace this process wholeheartedly.

8                     I offer this background because it

9         underscores the fact that ES&S is not new to the

10        standards or the certification process. We

11        understand a dynamic standards process is key to

12        motivating innovation and continued enhancement

13        of the voting technology.

14                    That is why ES&S has been

15        enthusiastically providing input during the TGDC

16        and NIST portion in written form and by attending

17        and participating in all meetings throughout the

18        current guidelines development.

19                    Let me give you some general

20        observations. Through this lens of our

21        experience, we are pleased to offer some initial

22        thoughts regarding Version I. Because the draft

1         guidelines are extensive and were just issued

2         this week, it will take some time to complete a

3         thorough review.

4                     Much like the former presenters

5         today, we too will have a peer review of this and

6         will provide very detailed written commentary on

7         this.

8                     In reviewing the draft guidelines,

9         there is nothing more important than giving the

10        process the time and the deliberation it deserves

11        to get it right. And I want to say this again,

12        because we think this is the most important

13        component of this, giving all of us enough time

14        to get this right and deliberate on it correctly.

15                    One criteria for us has been

16        paramount, as we look at the process we have been

17        engaged in, and that's feasibility, and that's

18        feasibility for many areas.

19                    Technical feasibility because the

20        guidelines must be implemented and have to have

21        any real effect and also in the guidelines, the

22        feasibility of achieving the EAC deadlines of

1         implementation and covering the additional costs

2         associated with substantial upgrades to product.

3                     The EAC and NIST and the TGDC

4         rightly determined that an interim step in the

5         guideline development process may be necessary,

6         with the very specific and limited goal of only

7         addressing significant limitations in the 2002

8         standards.

9                     All along ES&S understood that the

10        objectives of this interim process was a

11        complement to the 2002 standards, not really

12        meant to replace them.

13                    So our first observation about the

14        proposed guideline, particularly given the

15        intention to pursue a second version soon, is a

16        belief that, in sum and content, Version I may

17        have gone beyond the original intent and scope as

18        defined by the EAC.

19                    As proposed, the Version I would

20        impose a substantially new expectation on vendors

21        and election administrators alike. That raises

22        some challenging technical issues I will try and

1         address.

2                     But it also means that the

3         implementation will take significantly more time,

4         add costs and complexity.

5                     On this point we appreciate the

6         Version I effective date provision, clarifying

7         that the guidelines will go into effect 24 months

8         after their final adoption by the EAC.

9                     Certainly, given all that must be

10        done to develop the new product and testing

11        authorities, we believe it will take at least the

12        afforded time to accomplish this task as a

13        supplier or a vendor. In fact, when you consider

14        that it took five full years to develop the

15        implementation of the 1990 standards and then

16        three years to do the same in 2002, this proposed

17        timeline for Version I is aggressive.

18                    However, and again recognizing how

19        firmly we support an evolving standards process,

20        and how committed we are to continued voting

21        system enhancement, know that we will move

22        forward aggressively in that direction.

1                     The effective dates provisions are

2         important because they begin to send messages to

3         states and localities that they can move forward

4         with purchasing new systems to meet the HAVA

5         requirements, especially in Section 301, before

6         these new guidelines are in full force.

7                     In other words, what we are

8         recommending is the need for a strong

9         reaffirmation that voting systems certified to

10        the 2002 standards achieve the objective laid out

11        in HAVA, providing a substantial improvement over

12        older voting technologies.

13                    Every election official who is

14        working to comply with the January 1, 2006

15        deadline must know that they can purchase

16        certified 2002 systems with the confidence that

17        these systems will fulfill the mission and

18        mandate of HAVA.

19                    In addition, they should know that

20        they will be able to continue to use these

21        systems for years to come, that they will not be

22        required to replace HAVA-required technology with

1         new products that meet the proposed Version I

2         until it is necessary.

3                     If, in fact, this could be a

4         requirement in the future, they also should know

5         that funding may be provided to cover the cost of

6         a second upgrade of equipment.

7                     Without sending a strong message to

8         state and local election administrators along

9         these lines, we believe that some may opt for

10        missing the HAVA deadlines to wait for

11        implementation of Version I, and we think that

12        would be a mistake.

13                    Now to the technical issues. The

14        technical context of the Version I, here again

15        with an eye towards feasibility, first, just to

16        reiterate - if the proposed Version I is

17        implemented as it is drafted today, you should

18        know that compliance will or may require

19        comprehensive product development.

20                    For those of us who embrace

21        continually evolving standards and who work hard

22        to meet the challenges of updated guidelines,

1         this is a substantial undertaking. At ES&S we

2         have a passion for this innovation so we don't

3         want to say compliance is impossible. But,

4         without question, it will take time and add

5         significant cost and complexity.

6                     Take, for example, the setup and

7         validation requirements of Section 6.4, calling

8         for hardware and software verification systems

9         provided through third-party vendors which must

10        perform their function without utilizing the

11        voting system software. This will require a

12        whole new approach to voting system hardware.

13                    In addition, to comply, ES&S would

14        have to sacrifice a key security feature of our

15        current voting systems product, the inability to

16        interface with any outside components.

17                    We intentionally developed our

18        products without any ports or other connection

19        points, to eliminate the potential for tampering

20        via connected sources.

21                    Now under Section 6.4, the only way

22        we can comply would be to incorporate the very

1         connection port we have excluded from our design.

2                     Similarly, the voter verifiable

3         paper audit trailer, or VVPAT provision, raises

4         serious concern.

5                     To be clear, ES&S applauds EAC's

6         attempt in the proposed Version I guidelines to

7         provide some guidance on this point.

8                     Our position on VVPAT has always

9         been that while we have absolute confidence in

10        the reliability, accuracy and the security of our

11        voting systems, we know that many election

12        officials and the public, for that matter, want

13        the added benefit that VVPAT brings.

14                    We also understand that this is an

15        important consideration for raising the public's

16        overall confidence in the voting process.

17                    For vendors and election

18        administrators who have already started to

19        address the issues of VVPAT, the proposed Version

20        I is overly prescriptive and could very well

21        erode the progress many of us have already made.

22                    Specifically we are concerned about

1         the requirement defined in Section 2.2 as it

2         relates to 6.8, which relates to maintaining

3         voter anonymity and providing accessibility for

4         the visually impaired.

5                     We support both of those goals.

6         However, to fulfill the specific Version I

7         requirements, including that voter receipts must

8         be shuffled or reorganized in some way, we would

9         have to go back to the drawing board on our VVPAT

10        solution. In fact, we think most, if not all,

11        vendors would be in the same predicament, because

12        many of us - of the available systems today -

13        have used a paper roll.

14                    In addition, there is an issue of

15        general inconsistencies between provisions of the

16        proposed guidelines. In considering the Version

17        I as a whole, we have found several instances

18        that we will comment on where the requirements of

19        one section may conflict with the requirements of

20        another.

21                    Consider, for example, the following

22        two requirements.

1                     In Section 2.2.2.2, it requires that

2         any audio-tactile interface shall allow the voter

3         to have the information provided by the system

4         repeated. But then in Section 5.4.2, it requires

5         that no key or control on a voting station shall

6         have a repeat feature enable. We will need

7         guidance on this.

8                     In our more detailed written

9         responses we will provide some additional

10        examples of this same occurrence. Therefore,

11        before finalizing the Version I, some

12        clarifications and revisions in several instances

13        may be necessary.

14                    Going to recommendations, taking

15        into consideration feasibility, time

16        requirements, and added complexity associated

17        with implementing Version I, we have three

18        recommendations for the EAC.

19                    One, allow election officials to

20        confidently move forward with the current

21        standard; clarify through a safe harbor statement

22        that states and localities that, one, in need to

1         make purchasing decisions today to comply with

2         HAVA, can do so under today's standards and that

3         systems purchased today will not have to be

4         replaced in the future or when the next version

5         comes out.

6                     My second recommendation - I would

7         like to indulge a little bit of the EAC's ability

8         for me to amend my proposal or recommendation in

9         a follow-up further recommendation - but we have

10        stated in two, that because of the time and added

11        complexity and cost of implementing Version I,

12        you may wish to consider to combine Version I

13        with upcoming Version II.

14                    This would eliminate confusion in

15        the marketplace and allow all of us the time we

16        need to effectively develop and implement very

17        clear and understandable standards. Today, at

18        this moment, I know much more from this morning's

19        sessions that would require me to say I want to

20        amend this because I think I have more clarity on

21        what you mean by the 24, and then having Version

22        II follow after that 24 month period and not in

1         series, as opposed to overlapping in great

2         extent.

3                     Our third recommendation is for you

4         to carefully consider the comments from those who

5         must implement and run the elections: the

6         election administrators that are out there and

7         those who design, develop, build - and the ones

8         you have heard from this morning - test, and

9         deliver this very unique specific technology.

10                    So, in conclusion, we thank you

11        again for this opportunity to share our

12        perspective on Version I, its content and

13        challenges around the implementation.

14                    We also hope, as you gather input

15        over the 90 days and consider the feedback, that

16        you will call on all of us to assist in any way.

17                    Certainly there are aspects of the

18        areas that will require modification and others

19        that raise substantial concern. However,

20        overall, this is a step in the right direction.

21        And, again, we applaud you for it.

22                    Again, we will be following up with

1         additional and more specific written feedback

2         almost weekly. In the meantime, please accept

3         these comments as a first contribution to this

4         process.

5                     CHAIR HILLMAN: Thank you.

6                     Mr. Adler, before you begin, I want

7         to make a clarification here for the benefit of

8         the record.

9                     The proposed Voluntary Voting System

10        Guidelines that the Election Assistance

11        Commission has put forth, we do not consider this

12        a Version I. We are not thinking about this in a

13        Version I, Version II.

14                    And I understand there was earlier

15        language that may have led people to believe

16        that.

17                    But our responsibility was to issue

18        guidelines and the Technical Guidelines

19        Development Committee did put forth

20        recommendations that would augment and update the

21        2002 standards, as well as covering some

22        additional components.

1                     We recognize that it was almost

2         impossible, in nine months, for the Technical

3         Guidelines Development Committee and NIST to have

4         done a thorough covering of everything that needs

5         to be addressed. And so we expect and do plan

6         that we will add to the proposed guidelines.

7                     But we do not discuss it -- We

8         aren't, at this point, discussing it as a Version

9         I/Version II. We don't want to confuse anybody

10        that we are going through all this effort on a

11        Version I that might be replaced in very short

12        order by a Version II.

13                    There will be one version of the

14        guidelines and they will be updated and augmented

15        as the Technical Guidelines Development Committee

16        and NIST has time to do work on, but not in any

17        way to have people think that this will be set

18        aside and totally replaced.

19                    MR. GROH: And I thank you for

20        clarifying it. That is one of the major intents

21        of coming to meetings like this, to have this

22        interface.

1                     We get clarification where there may

2         be some, as we have interpreted, a wrong

3         misinterpretation.

4                     CHAIR HILLMAN: No problem. I just

5         wanted to make sure that we were all --

6                     The other point that I meant to

7         address before the panel presentation began is

8         how we determined what the presentations from the

9         vendors would be today.

10                    We were looking for a blend. There

11        are many vendors who make voting systems.

12                    And I am just going to call on our

13        General Cousel, Juliet Thompson, to give a brief

14        explanation as to what we were seeking to

15        accomplish by the diversity and representation on

16        this panel.

17                    MS. THOMPSON: Thank you, Madam

18        Chair.

19                    And, as you aptly pointed out, there

20        are many voting system vendors out there and we

21        expect to hear from all of them during the

22        process of this comment period.

1                     But for the purposes of this panel,

2         we applied several factors, one of which was the

3         types of voting systems that they manufacture:

4         to include optical scan; to include touch screen;

5         and to include full face DRE voting systems, as

6         well as components.

7                     And the second set of factors would

8         be, of course, to be representative of those

9         vendors that are out there in the marketplace,

10        some large vendors, some small vendors.

11                    CHAIR HILLMAN: Thank you.

12                    Mr. Adler.

13                    MR. ADLER: Thank you, Madam Chair,

14        Vice Chair DeGregorio, Commissioner Martinez.

15                    Am I on?

16                    THE CHAIR: Yes,

17                    MR. ADLER: I am now.

18                    THE CHAIR: Just move the mike

19        closer.

20                    MR. ADLER: Mr. Wilkey, Ms.

21        Thompson.

22                    Thank you for having me here today.

1         My name is Jim Adler and I am President of Vote

2         Here, a company I founded in 1998 with a vision

3         to create for more transparent and auditable

4         elections. We provide independent verification

5         technology for both electronic and paper ballot

6         processing, to prove the voting machines, ballot

7         processing and back-end tabulation systems are

8         performing properly.

9                     In the U.S. and abroad, our

10        technology has served more than 12 million

11        voters. I currently coach the IEEE Voter

12        Verifiable Standards Committee. I have

13        testified before the U.S. House Government Reform

14        Subcommittee or TGDC, your TDGC, the National

15        Institutes of Standards and Technology and many,

16        many, many State Legislatures

17                    Since you will be hearing a lot

18        about what is wrong with these standards in the

19        coming months, I thought it might be a good

20        opportunity to tell you what's right about them.

21                    After being involved with the IEEE

22        voting system standards for the better part of

1         two years, I understand what a daunting process

2         this really is. Given the time constraints and

3         contention around some of these topics, I think

4         getting this body of work completed on time is

5         really a huge accomplishment.

6                     It is comprehensive, well-organized

7         and thorough. You, the TGDC, NIST should all be

8         commended on getting to this point. It's a great

9         accomplishment.

10                    To respect the time constraints, I

11        won't make any detailed comments today. But, as

12        many others have said, follow up.

13                    I would like to touch on two broad

14        topics, primarily with regard to the issue of

15        independent verification, one on classifying what

16        you are calling independent dual verification

17        solutions, and the other on component

18        certification. So let me deal with those in

19        turn.

20                    In the April 2005 Voluntary Voting

21        System Guideline submittal to EAC, NIST and TGDC

22        widely settled on defining independent dual

1         verification - and I will use the acronym IDV for

2         that - as a new class of voting system component.

3                     In the original submittal to EAC,

4         TGDC defined independent verification, even set

5         requirements for all IDV systems. And then in

6         subsequent sections they defined requirements for

7         a split process IDV systems, witness IDV systems

8         and cryptographic IDV systems and VVPAT IDV

9         systems.

10                    In the introduction to the VVPAT

11        requirements, the original TGDC standards say,

12        "VVPAT is a form of independent verification

13        system." This was an excellent classification.

14                    It defined a variety of safety

15        equipment, if you will, for voting systems, just

16        like Federal standards define automotive safety

17        equipment - like seat belts, airbag, windshield

18        safety glass, and other safety innovations that

19        improve passenger safety.

20                    Specifically, in the IDV sections,

21        in this instance the TGDC did a great job of even

22        establishing detailed requirements on IDV

1         alternatives.

2                     This classification was consistent

3         with public comments by the EAC Commissioners,

4         previously and today, on the availability of

5         other IDV methods to accomplish the goals of

6         election, confidence and audit.

7                     However, in the proposed VVSG, the

8         classification has been lost. In Section 6 the

9         VVSG now sets VVPAT requirements in a vacuum.

10                    The other IDV alternatives available

11        today are put in an appendix in the back of the

12        bus, and the VVPAT standards are provided outside

13        this appendix with no context.

14                    The VVSG doesn't say what the VVPAT

15        is or what it intends to accomplish. It just

16        says that the requirements are provided for

17        system certification, testing for those states

18        that have decided to include VVPAT.

19                    Understandably, this is a nod to the

20        political activism that has gripped many State

21        Legislatures on this issue, but the standard is

22        needlessly silent on the context.

1                     And there is a larger issue here.

2                     As we talked about today, these

3         standards will establish guidance now and policy

4         for years to come. Several states are now

5         recognizing the need for IDV but are savvy enough

6         to perceive the unintended consequences of VVPAT.

7                     Specifically, they see the erosion

8         of voter privacy through the current reel-to-reel

9         VVPAT designs, the lack of an accessible VVPAT

10        for disabled voters, the dilemma faced when VVPAT

11        ballots are mishandled and don't match the

12        election results, and even question the

13        effectiveness of VVPAT in bolstering voter

14        confidence when in early tests voters are not

15        even checking the VVPAT printout or voters want a

16        receipt they can take home, just like at the ATM.

17                    My point is not that VVPAT is all

18        bad. But the harmful thing about recent activism

19        and legislation is that these issues of efficacy,

20        privacy and accessibility have not been carefully

21        considered, which means that VVPAT may very well

22        undermine the same confidence its supporters have

1         aimed to bolster.

2                     Any election solution, especially an

3         IDV solution, should be required to demonstrate

4         real measurable effectiveness before it is

5         legislated or established as a standard that

6         stands alone without equal.

7                     So I would urge you to revert back

8         to the classification that the technical experts

9         at NIST and the TGDC recommended in their

10        original submission last April. This will allow

11        for best solutions that meet these efficacy,

12        privacy and accessibility requirements needed for

13        all voters, as well as will minimize the costs

14        and administrative burden.

15                    Of course the objective that is

16        often raised is that other methods of IDV are far

17        into the future and have not yet been Federally

18        qualified, certified.

19                    Well, the fact is, IDV approaches

20        have been researched for the last 25 years and

21        have been put into products over the last seven.

22                    The real problem is that IDV

1         solutions can't get independently certified,

2         which brings me to my next comment.

3                     The VVSG establishes IDV standards.

4         But current NASED polcy and VVSG Volume II don't

5         allow for certification of independent

6         components.

7                     This policy allows for a complete

8         voting system to be certified. And I would ask

9         that you reconsider this policy as Federal

10        testing is transferred to the EAC.

11                    The effect of this current

12        system-only certification policy is that

13        components like IDV components can only be

14        certified within an entire voting system, a

15        policy that leads to a classic Catch 22

16        conundrum.

17                    With the current policy, voting

18        system vendors won't integrate and certify best

19        of DRE components unless customers demand them.

20                    But jurisdictional customers are

21        reluctant to demand them or statutorily can't

22        demand them unless they are Federally certified.

1                     The dynamic is especially accute in

2         several states right now that want IDV choices

3         but can't get them because the current process

4         does not allow for component level certification.

5                     I think Carolyn Coggins talked about

6         the paralysis that is out there on this issue.

7         This is one of them.

8                     So I would like to make a modest

9         proposal for component level Federal testing to

10        help rectify the situation and allow market entry

11        for best-of-breed specialties.

12                    Component level Federal testing

13        would be in two phases. The first phase would be

14        component level certification testing, which

15        would require component vendors to deliver a

16        technical data package and a test harness. The

17        test harness would be used to test the component

18        against the component vendor's TDP and the

19        applicable EAC standard. When successful, a

20        component level certification number would be

21        given.

22                    The second phase would test the

1         integrated voting system with the pre-certified

2         component. This would test for a successful

3         integration of the certified component into the

4         voting system.

5                     And when successful, the voting

6         system would be given a certification number that

7         included the certified component.

8                     Assemblance of this process already

9         does exist today when a vendor of a certified

10        voting system submits, say, an audio component

11        for Federal certification. That component is

12        evaluated on its own merits at the component

13        level and then as part of the entire voting

14        system at the system level.

15                    The new policy would remove the

16        current barriers to innovation by allowing

17        Federal certification of components and allow

18        interested states and counties to escape their

19        Catch 22 deadlock. It provides jurisdictions

20        choice and assurance that components meet Federal

21        standards, while getting the bulk of Federal

22        testing done early.

1                     As you know, jurisdictions face

2         looming immovable deadlines. I know, I guess

3         it's a requirement that every panelist must say

4         that.

5                     So anything to expedite testing

6         would help.

7                     So in conclusion, I would like to

8         make a comment about the use of technology in

9         elections. Frankly, there is a distinct

10        anti-technology movement that has gripped

11        election reform. Given the current polarized

12        political climate and general technology fatigue,

13        the backlash is understandable.

14                    However, in my home state of

15        Washington, we all witnessed the recent

16        Governor's race that dragged on for six months.

17                    An election where 90 percent plus of

18        the ballots were cast on paper, it is often said

19        by voting critics that a voting system must prove

20        to the loser that they lost.

21                    Well, the paper trail, as tested in

22        Washington, did not meet that standard.

1                     After watching this, it is clear

2         that election officials simply just don't have

3         the technological tools to deal with the

4         increasingly close races and heated scrutiny and

5         heightened public attention.

6                     It's like we are asking them to

7         accurately weigh a flea on a bathroom scale.

8         It's just not a fair position to put them in.

9                     The current VVSG IDV standards are a

10        step toward providing these tools. But

11        expediency is key.

12                    In many ways Federal agencies are

13        not geared to build airplanes while they fly

14        them, but this is the unenviable position you

15        find yourselves in.

16                    Delays like the 90 day comment

17        period and the 24 month effectivity date are

18        necessary and unavoidable.

19                    The comments I have made today will

20        help immediately by providing guidance to

21        jurisdictions that are currently looking for more

22        effective IDV, and by providing a Federal testing

1         policy that can expediently and responsibly

2         certify them.

3                     As you deliberate on the testimony

4         you hear today, it is important to keep in mind

5         that your policies will have one of two possible

6         outcomes, especially in the area of independent

7         dual verification.

8                     Either IDV competition will be

9         discouraged with jurisdictions being locked into

10        VVPAT, or jurisdications will have choice in

11        their IDV solutions, fostering a climate of

12        innovation.

13                    The VVPAT movement, I am afraid, has

14        largely ignored the competing requirements of

15        privacy, accessibility and voter confidence.

16                    A climate for innovative IDV

17        solutions just won't be able to get away with

18        ignoring these competing requirements because the

19        competitive market pressures just won't allow it.

20                    So, again, I thank you for the

21        opportunity to speak to you today. I certainly

22        will be following up with more detailed comments

1         and, of course, happy to answer any questions.

2                     CHAIR HILLMAN: Thank you.

3                     Mr. Charles.

4                     MR. CHARLES: Good afternoon, Chair

5         and Commissioners.

6                     Sequoia Voting Systems has a 100

7         year history of providing election equipment,

8         supplies and support for state and local

9         officials.

10                    We print ballots, manufacture

11        optical ballot readers and provide two different

12        types of direct recording electronic voting

13        systems.

14                    During the 2004 Presidential

15        election, Sequoia was the largest provider of

16        electronic voting machines in the nation. And we

17        were the first national company to provide a

18        voter verifiable paper record on electronic

19        voting equipment in a major election.

20                    We appreciate the opportunity to

21        participate in this hearing today and commend the

22        EAC, the TGDC and NIST on moving quickly under

1         under tight timetables to assemble the latest

2         draft of the amended Voluntary Voting System

3         Guidelines.

4                     My comments today will focus on the

5         following key areas: first - the content of the

6         guidelines generally; second - the timeline

7         required for compliance with the guidelines and

8         the expectations of the marketplace; the

9         effective date of the regulations; the Federal

10        testing and certification process; the state

11        testing and certification process; and, finally,

12        the need to evaluate and revise the guidelines

13        over time.

14                    While we will reserve detailed

15        comment on the specific components of the

16        guidelines for our written submission, we believe

17        that by and large the draft addresses the types

18        of issues that needed refinement from the 2002

19        standards, particularly in the development of the

20        optional requirements for voter verifiable paper

21        records in the inclusion of human factors and

22        security criteria.

1                     We will, however, provide the

2         Commission with a lengthy set of written comments

3         that address areas that we believe warrant

4         correction, clarification or revisions.

5                     Our primary concern regarding this

6         draft of the guidelines is the inclusion of

7         requirements that are not necessarily testable by

8         the voting system testing labs because they are

9         either ambiguously worded or because they rely on

10        the manner in which the system is implemented, by

11        local administrators, rather than a design of a

12        system itself.

13                    These standards and the systems that

14        result from them are but one piece of the

15        successful conduct of elections. The policies,

16        procedures and people that conduct elections are

17        of equal importance. But to the extent possible,

18        that document should address technology

19        requirements and leave the proper implementation

20        of the system to state and local statutes,

21        procedures and best practices.

22                    Our written remarks will attempt to

1         point out those areas which we believe are better

2         suited for best practices guides and local

3         procedure than inclusion in these technology

4         guidelines.

5                     Through no fault of the Commission,

6         many observers and officials have operated under

7         the false assumption that the adoption of these

8         voluntary guidelines at the national level will

9         somehow be required before they could comply with

10        the statutory mandates of HAVA.

11                    With the 2006 deadline for

12        compliance rapidly approaching, it is important

13        now more than ever for the Commission to help

14        election officials realize that these standards

15        will not be in place and operational prior to the

16        date that equipment purchase decisions will have

17        to be made for compliance with HAVA.

18                    Once the guidelines are finalized

19        and ready to be approved later this year,

20        technology providers will develop and implement

21        any required revisions to hardware and software,

22        the testing authorities will need to be certified

1         to test against the new criteria, and the EAC

2         will need to finalize its process for managing

3         certification requests and interpreting the

4         guidelines as testing gets under way.

5                     Once Federal and state testing is

6         completed and the products are available, local

7         officials will need to acquire the updated

8         systems, companies will need to manufacture and

9         deliver equipment, software upgrades will need to

10        be deployed, poll worker training manuals and

11        courses will need to be modified and voters will

12        need to be educated about the changes.

13                    There simply isn't enough time for

14        that all to happen prior to the 2006 primary

15        elections in many cases.

16                    The last time standards were

17        modified, it was a full three years before

18        equipment tested to the new standards was

19        available in the marketplace. While many of the

20        new features and requirements in this draft are

21        already incorporated into many systems, testing

22        against these standards will not be possible for

1         some time.

2                     Fortunately for election

3         administrators, time required to meet these new

4         standards should not be confused as an impediment

5         to state or county compliance with HAVA mandates

6         for 2006. There are a large number of voting

7         systems available in the market today, both with

8         and without voter verifiable paper records, that

9         have been tested under both the 1990 on 2002

10        standards and will allow election officials to

11        meet the demands of Section 301 of HAVA.

12                    One of the most important decisions

13        facing the Commission will be selecting the

14        appropriate timelines and details associated with

15        the implementing of the new guidelines.

16                    The draft discusses a 24 month

17        period after adoption before the guidelines will

18        be in full effect.

19                    But it is unclear if that will

20        result in the decertification of all non-

21        compliant systems that have been successfully

22        tested to earlier standards, or if the

1         implementation date will simply prevent future

2         certification of non-compliant systems.

3                     The ramifications of the way the

4         effective date is implemented warrant

5         considerable review and discussion with state and

6         local officials.

7                     Continued ongoing changes to

8         standards may be important and useful, but they

9         will also require continued funding to modernize

10        equipment at the local level. It's funding that

11        isn't currently contemplated at the Federal level

12        and I don't think most states and counties have

13        anticipated funding on an ongoing basis.

14                    The EAC should consider the

15        reasonable likelihood of Federal funding before

16        determining the manner in which this and future

17        standards will need to be implemented.

18                    Voting systems are comprised of a

19        series of components that can be tested against

20        existing standards. Components can include the

21        voting machines, the software, ballot activators,

22        et cetera.

1                     When qualified components are

2         combined to create a system, the entire

3         end-to-end system is tested before a Federal

4         qualification number is assigned. When any

5         aspect of a component is revised, that component

6         in the entire system must be tested together.

7                     If components have been tested to

8         different sets of standards, the complete system

9         is only qualified to a particular version of

10        standards when every component of that system has

11        been tested to that same level.

12                    For example, the complete system is

13        only certified to the 2002 standards after all

14        components are qualified to the 2002 standards.

15        If one out of 10 components is qualified to the

16        1990 standards, then that standard still carries

17        the 1990 qualification number.

18                    If in setting the effective date of

19        these guidelines the EAC no longer allows

20        modifications to the older systems in use today

21        without bringing out pieces of hardware and

22        software up to new standards, local officials who

1         have a 1990 or 2002 system could be significantly

2         affected when local/state law changes occur.

3                     If a county needs to modify one

4         component of their system to comply with the new

5         state law, it could be faced with a possible

6         replacement of an entire voting system that would

7         otherwise be perfectly functional and compliant

8         with prior standards, state needs and HAVA

9         mandates.

10                    However, if provisions to individual

11        components can be tested to the latest standards

12        while the complete system retains a certification

13        number reflecting the standards against which it

14        was originally qualified, there shouldn't be any

15        significant fiscal burden imposed on local

16        officials. State law changes will be able to be

17        accommodated more readily.

18                    Some states accept Federal

19        guidelines for certification; other states

20        conduct their own testing process; still others

21        combine Federal testing with state reviews. The

22        multiple layers of often duplicative review are

1         costly, time consuming and delay the latest

2         innovations from getting into the market.

3                     In developing test policies, we

4         encourage the EAC to invite states to provide

5         Federal testing authorities with a checklist of

6         local requirements so that duplicative

7         examinations can be streamlined and state

8         resources can be saved through concurrent state

9         and Federal testing.

10                    Once the guidelines and testing

11        procedures are in place, the EAC will need to

12        evaluate them over time to identify and correct

13        potential weaknesses. We suggest this review

14        should not begin until after the newly qualified

15        systems are deployed and in use for at least two

16        elections.

17                    One of the greatest mistakes this

18        Commission could make would be to revise

19        standards too frequently without providing ample

20        time for a particular version of the standards to

21        be put in place and studied.

22                    Many election officials will tell

1         you that it takes at least two elections to work

2         through all of the logistical and procedural

3         issues that come with the transition to a new

4         voting system.

5                     The quidelines before you today are

6         the result of work that started before even one

7         major election was subjected under the 2002

8         standards.

9                     We recognize that with the passage

10        of HAVA, Congress charged you with the duty to

11        provide for the certification and decertification

12        of voting systems and that state policy decisions

13        created a need for VVPAT guidance.

14                    But we strongly suggest that once

15        these standards are adopted, they are given time

16        to be tested in the field before a new batch of

17        requirements are developed.

18                    The rapidly changing standards

19        process has required companies in our industry to

20        spend an inordinate amount of time revising

21        technology to meet changing guidelines, when that

22        time could have well been spent adding the

1         features and benefits, like VVPAT, that have been

2         demanded by the market in many places.

3                     Sequoia Voting Systems believes that

4         when completed, the voluntary systems guidelines

5         will continue to ensure that voting technology is

6         reliable, secure and accurate.

7                     To assist the EAC to make the most

8         of this project, we respectfully offer the

9         following recommendations:

10                    First, as you are doing today and as

11        you have done throughout the policy decisions

12        this Board has made, solicit comprehensive verbal

13        and written comments from all stakeholders;

14                    Second, provide a detailed public

15        timeline to help local officials understand how

16        we get from the draft regulation phase of this

17        process to the evenutal use of VVSG compliant

18        equipment at the polls;

19                    Third, rather than waiting 24 months

20        to implement the proposal, allow the regulations

21        to take effect immediately upon adoption or as

22        soon as testing authorities are able to review

1         the systems, provided, however, that individual

2         components in the systems may still be qualified

3         against prior versions of standards, as long as

4         they are appropriately noted as such on testing

5         reports and on published lists of certified

6         equipment;

7                     Fourth, facilitate concurrent

8         Federal and state testing through the development

9         of a checklist of state-specific criteria which

10        can be tested by the federally approved

11        laboratories, as necessary, to help streamline

12        the state certification process;

13                    Fifth, allow equipment tested under

14        these regulations to be deployed and monitored

15        for at least two elections before initiating a

16        new set of Federal criteria.

17                    We appreciate the invitation to

18        provide our thoughts to this hearing and welcome

19        the opportunity to continue to work with the

20        Commission on this project and we plan to submit

21        our detailed written comments on a line-by-line

22        basis within the next couple of weeks.

1                     Thank you.

2                     CHAIR HILLMAN: Thank you very much.

3                     Mr. Vadura, we are at somewhat of a

4         little handicap here because we don't have copies

5         of your --

6                     MR. VADURA: I understand.

7                     CHAIR HILLMAN: So if you would just

8         be so kind in making your presentation to bear

9         that in mind.

10                    I mean, like he speaks fast but I

11        can read fast, so I was able to keep up with him.

12        So just bear with us as we take notes from your

13        presentation.

14                    MR. VADURA: That's fine. Can

15        everybody hear me?

16                    Flight schedules last night

17        prevented me from printing them, so I will

18        provide copies to the staff.

19                    Good afternoon, Madam Chair and

20        Commissioners. My name is Dennis Vadura. I am

21        co-founder and chairman of AccuPoll and AccuPoll

22        Holdings. We are the only public company in the

1         space that is dedicated solely to voting.

2                     And I would like to thank the EAC

3         for inviting AccuPoll and myself to participate

4         in this panel.

5                     AccuPoll has been promoting a voter

6         solution that includes a voter verified paper

7         audit trail, or what's now called VVPAT, since

8         AccuPoll's inception.

9                     Our view is that VVPAT is to

10        electronic voting what stairs are to tall

11        buildings. It is the essential safety net that

12        allows for emergency action in the event of a

13        disaster.

14                    The key to a legitimate democratic

15        government is a trusted election process.

16        America works hard to promote democratic ideals

17        all over the world and with HAVA we are making

18        sure that we provide trusted and private

19        elections for all of our citizens.

20                    AccuPoll is all about promoting

21        trust, accessibility and practicality in the

22        voting process.

1                     In this regard we welcome the

2         updated standards for voting systems and would

3         like to offer the following comments with respect

4         to these new proposed standards.

5                     We believe that it is possible to

6         create a VVPAT that is both accessible and

7         private. The disability community has fought

8         hard for HAVA and have gained the right under

9         HAVA to vote unassisted and in private. We see

10        no reason that a voter verified paper trail needs

11        to be engineered such that it removes any of

12        these rights. We welcome the inclusion of these

13        HAVA requirements in the revised standards.

14                    AccuPoll has recently made changes

15        to its voting system such that we can demonstrate

16        today a practical VVPAT solution that is both

17        accessible, private and, based on our current

18        understanding of the standards, essentially

19        compliant with the proposed standards.

20                    We will be demonstrating our

21        solution at the upcoming NASS conference and

22        encourage all that are interested to come and

1         view our approach.

2                     The key features of our approach is

3         audio feedback of the VVPAT for every voter,

4         whether they are sighted or not, provisions for

5         privacy of the voting process and of the VVPAT

6         for the voters that require additional assistance

7         as they complete the voting process, and still

8         maintain their privacy.

9                     AccuPoll essentially agrees with the

10        disability community that optical scan ballot

11        marking systems are not accessible voting systems

12        and are, therefore, not complaint with HAVA

13        accessibility requirements.

14                    Optical scan systems on their own do

15        nothing to help a disabled voter vote in private

16        and provide no easy and accessible confirmation

17        that a voter's intent was correctly read or

18        recorded.

19                    Ballot marking systems that

20        complete an optical scan ballot for a disabled

21        voter do nothing to assist a voter that is not

22        able to handle a paper ballot on their own and,

1         therefore, may create a privacy issue for a

2         disabled voter.

3                     AccuPoll does, however, have issues

4         with certain aspects of the proposed voting

5         system standards. Given the recent update, like

6         our colleagues here, we will be providing a

7         detailed set of comments regarding specific

8         issues that we see with the current draft of the

9         standards.

10                    Most notably, we see some possible

11        ambiguities and a need for additional clarity

12        with respect to certain certification issues and

13        limits as were discussed by Carolyn earlier.

14                    In keeping with the practice of the

15        Federal Voting System Standards that set

16        requirements and not implementation choices, the

17        current standards for voting systems must set

18        requirements rather than mandate engineering

19        solutions. In this way the standards would allow

20        the marketplace to develop the most cost

21        effective and practical solution while ensuring

22        compliance with the standards and with HAVA.

1                     We are also concerned with too many

2         changes or evolution cycles and amendments to the

3         standards.

4                     While it may take the ITA's a short

5         period of time to ramp up and test against new

6         standards, it takes at least six months for

7         vendors to update and certify their systems to be

8         compliant, after which there are renewed state

9         certification requirements before the revised

10        changes can be made available to customers.

11                    The state certification process may

12        take in excess of an additional six months.

13                    We, therefore, suggest that the

14        revised standards be finalized, issued and not

15        altered for at least two years. This excludes

16        interpretation bulletins that would clarify an

17        interpretation of the requirement, rather than

18        modify or add new requirements.

19                    We owe it to our customers and to

20        the ITA's to maintain stability in the

21        requirements so that products can be engineered

22        and marketed on reasonable timelines and costs in

1         what really is anything but a traditional

2         marketplace.

3                     In our view, it is better to wait

4         six months and issue a document that is well

5         thought out, rather than issue early and amend

6         later. Thank you.

7                     CHAIR HILLMAN: Thank you very much.

8                     Commissioners, we are running a

9         little bit late on this panel. So what I propose

10        we do is to make up our time into our break

11        period.

12                    We will still plan to have the third

13        panel start its presentation at 3 o'clock and we

14        will need a few minutes to just set the table up

15        for the third panel and that would give us the

16        ten minutes per Commissioner questioning, that

17        will allow you time to engage with the panelists.

18                    So, if that works, Commissioner

19        Martinez.

20                    COMMISSIONER MARTINEZ: Thank you,

21        Madam Chair.

22                    My thanks to all of you for making

1         the time and the effort to be here and to testify

2         and certainly I look forward to receiving more

3         comprehensive, as I know you all will provide,

4         written testimony during the 90 day comment

5         period on the proposed VVSG.

6                     John, let me start, if I could, from

7         your perspective. I think one of the things that

8         you have identified is the difficulty in the

9         language in the current draft of the VVSG, which

10        would require randomization of the VVPAT paper,

11        as opposed to, I guess, the current technology of

12        reel-to-reel.

13                    Just, if you can talk a little bit

14        on that from your perspective, why that is

15        such a significant change from the vendors'

16        perspective.

17                    MR. GROH: Well, again, I will speak

18        for to us and not try and speak for others. But,

19        you know, all of us see each other's technology.

20                    And as we approach this and try to

21        target getting a solution out there as we

22        anticipated this coming, we all looked at what

1         would be a best practice from the standpoint of

2         making it easy for the poll worker to use, very

3         easy for the voter to interface with and look at

4         and understand that there were limitations with

5         how to interface with our technology.

6                     One of the major concerns we had is

7         how do you handle this at a precinct when you run

8         out of paper or have a paper jam. So I think all

9         of us chose to use a reel type of voting system

10        where paper is going to roll up in a rolling

11        process.

12                    If we are required to shuffle it, we

13        will have to basically scrap that and start with

14        something else that will be much different than

15        what we initially perceived and that will take --

16        that takes some time and a lot of expense to it.

17                    COMMISSIONER MARTINEZ: Is the

18        difficulty that you perceive one of technological

19        feasibility or is it more that you designed it in

20        a way that you thought would be more poll worker

21        friendly, if you will?

22                    MR. GROH: As Mr. Adler has talked

1         about, there are other methodologies of doing

2         this. So this is one. And understanding we do

3         not adhere to this as the only prescriptive way

4         to do it, but if you are going to offer it as one

5         of the solutions, you also must recognize that

6         poll workers and election administrators need to

7         manage this and handle it.

8                     And, as we know, loose pieces of

9         paper have a tendency to get lost.

10                    Paper rolls then that are going to

11        have to be cut and also be kept for cancellation

12        on ballots if somebody wants to cancel that

13        ballot or that record on there, when you start

14        having them now cut and commingle, we just saw

15        the inevitable problems with those.

16                    COMMISSIONER MARTINEZ: I appreciate

17        that, John.

18                    Anybody else on the panel would like

19        to comment?

20                    Mr. Adler.

21                    MR. ADLER: I think that every

22        election system has to meet simultaneous

1         requirements of privacy and accessibility and

2         effectiveness and election administration.

3                     And if you relax any of those

4         requirements, you run into problems.

5                     I mean, one of the objectives we had

6         was to simultaneously solve all of these issues.

7         And if you start to relax them, you start to

8         either infringe on people's civil rights or their

9         expectation of privacy or you start to create a

10        system that can't be administered.

11                    And, hence the plea for a climate of

12        innovation.

13                    COMMISSIONER MARTINEZ: Mr. Charles.

14                    MR. CHARLES: We deployed a system

15        in Nevada that had reel-to-reel design. And what

16        we found was that it was much easier for poll

17        workers to administer that approach and more

18        prone to successful use throughout the day.

19                    I think that the language in the

20        existing draft is permissive. It may permit you

21        to or -- It doesn't state that you have to cut

22        those ballots and shuffle them.

1                     But I think if you added language to

2         it that permitted administrative procedures to

3         ensure the protection of the anonymity of those

4         ballots, that would go a long way to resolving

5         that concern.

6                     Because what we see is a

7         self-randomization of those records. When the

8         voter comes into the polls, the voter can use any

9         machine in the polling place.

10                    When those records come back, they

11        are sealed and stored and not accessible to the

12        people who may have seen how someone voted.

13                    So if you can incorporate

14        administrative procedures into that or at least

15        allow the use of administrative procedures to

16        help assure that randomonization and privacy, I

17        think you accomplish two goals.

18                    One, you allow the most funcational

19        and reliable technology to be used. But you also

20        solve the concerns about voter privacy.

21                    COMMISSIONER MARTINEZ: Mr. Vadura,

22        any comment?

1                     MR. VADURA: Our current solution

2         doesn't use a reel-to-reel, but still maintains

3         the voter privacy issues.

4                     And I think the particular

5         requirement requiring the shuffling of the pieces

6         of paper is both probably unnecessary and

7         overburdensome and also probably too costly.

8         There are cheaper solutions to get that done.

9                     So, administratively, in the polling

10        place, I believe in another section of the

11        standards it says that you can't, polling

12        officials can't have access to that paper trail

13        during the election day.

14                    But if you have a paper jam, you

15        have to have a practical way to clear that paper

16        jam in that kind of scenario. So I think there

17        are some issues with the standards with respect

18        to that particular requirement.

19                    MR. GROH: Commissioner Martinez,

20        this morning -- I want to answer a question that

21        you asked this morning and make sure all of us

22        would respond to it.

1                     You had asked in the responses, will

2         you make recommendations. From our perspective,

3         as a developer, we will make recommendations in

4         ours for language. So we will try and give you

5         some words or language around that that will help

6         you get your arms around it.

7                     COMMISSIONER MARTINEZ: Thank you.

8                     The next question, I want to go back

9         to something that I was talking about this

10        morning with the testing labs. And I don't know

11        if everybody, if all four of you were in the

12        audience. I think Jim and John, I think you were

13        both here, so you all heard the exchange.

14                    The topic that I brought up was the

15        issue of a Version II, which our Chair talked

16        about a little earlier.

17                    And I just want to get - because

18        actually the response from the system, from the

19        testing labs was - it really doesn't impact us

20        greatly. But they thought certainly the vendor

21        perspective and perhaps even the state and local

22        election administrators, et cetera, that it could

1         have a greater impact.

2                     Just a quick comment or two about

3         the issue of -- Because I know I am having to

4         contemplate what happens next.

5                     Already we are thinking about, all

6         right, so if we go through the next 90 days and

7         we end up with the final product that we adopt as

8         the next iteration or the latest voting system

9         guidelines, then what happens?

10                    Obviously, the Help America Vote Act

11        creates this umbrella organization - the

12        Technical Guidelines Development Committee -

13        which I think has brought tremendous expertise to

14        the table.

15                    The question becomes how do we

16        utilize that, as well as our partnership with

17        NIST? And what happens next? Do we begin

18        working right away on a full blown version or

19        full blown revision, if you will, of what we have

20        just adopted in final form? Or do we --

21                    That's why I was trying to get

22        myself educated to the current process, and that

1         is, that when we find deficiencies in the current

2         product, that we address them to an amendment

3         process, for example. And is that perhaps easier

4         to deal with from every perspective - vendor,

5         election administrator, et cetera?

6                     Any comments?

7                     Mr. Charles, you have heard the

8         exchange. Any comments about that whole topic?

9                     MR. CHARLES: I think it's important

10        to distinguish between the ease of the testing

11        laboratories in testing the equipment and the

12        difficulty it takes for us to make the

13        modifications and submit them to the test.

14                    So there is a much lengthier time

15        for to us to do the work we need to do before we

16        get it to them.

17                    I think the first step ought to be,

18        develop a way to judge any gaps in these

19        standards by observing the elections after these

20        standards are adopted in the field.

21                    I think develop a criteria for

22        reviewing what happens in elections, determine if

1         there are pieces of these regulations that are

2         overly burdensome or areas that are missing and

3         develop that framework so you can study.

4                     And then once you have studied them

5         in practice in an election, then you can apply

6         that to the next round.

7                     But I think revising them before we

8         see them in the field is working more on theory

9         than on practical application. And that may not

10        be the best use of the time.

11                    COMMISSIONER MARTINEZ: Mr. Groh,

12        any comment on that?

13                    MR. GROH: Well, the competitive

14        nature of this market means that many of us are

15        already working on these.

16                    But, again - and somebody used the

17        analogy of, we are trying to build a plane while

18        it's running down a runway.

19                    And for us, as a vendor community,

20        there is one component I want to remind

21        everybody. We are going to run, from our

22        company's perspective, about 5000 elections

1         between now and next November. And that also

2         keeps us very busy.

3                     So running these parallel paths, I

4         don't have the luxury of stopping, working on the

5         elections, fixing small things that are changes

6         in election rules.

7                     So we are aggressively pursuing

8         them, but it will take us time.

9                     I agree with Ms. Coggins's comment

10        this morning, that the vendors that have gone

11        through this numerous times are at a great

12        advantage because we understand it.

13                    But I can tell you the difficulty

14        that we have of submitting something that we feel

15        fits and meets what the requirements are and then

16        when it is tested there are things that are found

17        that we hadn't thought of. And that's part of

18        the process that you go through to flush those

19        out.

20                    But that's what takes the 18 months,

21        the 16 months to do, it's the back and forth.

22        And we can't just stop and say let's hold

1         elections for four years, get all this done, and

2         get a final product out there.

3                     We are at the same time trying to

4         supply the election sites.

5                     COMMISSIONER MARTINEZ: Mr. Adler,

6         Mr. Vadura, any comments?.

7                     MR. ADLER: I think there is two

8         buckets here. There is quite a bit of election

9         technology that is well understood and is going

10        to change in a much less frequent rate than say

11        the side of business that, where you are focused,

12        which is on voter verified paper audit trails and

13        independent verification systems, where things

14        are theoretical right now, quite frankly.

15                    And these devices are just starting

16        to move into the market and having a mechanism to

17        be able to adjust, where in many respects the

18        states are now laboratories in this field, in

19        this area. And being able to have guidance and

20        being able to draw on NIST experts and your own

21        experience to help states navigate that and have

22        a real mechanism to help the certification

1         process would be very valuable.

2                     COMMISSIONER MARTINEZ: Mr. Vadura,

3         any comment?

4                     MR. VADURA: Yes. I will speak for

5         us. We just completed our 2002 certification on

6         our complete system with, including a VVPAT. And

7         we got through the Federal certification process,

8         that took some amount of time.

9                     We then have to now go through state

10        certifications in various other states, which is

11        fine; that's the way the business works.

12                    But the issue is if you make too

13        many changes too frequently, we are in a constant

14        state of certification somewhere at that point.

15        And I don't think that benefits anybody.

16                    It doesn't benefit the customers; it

17        doesn't benefit the vendors.

18                    COMMISSIONER MARTINEZ: My final

19        question, then, again going back to something I

20        brought up this morning, and that is - maybe it

21        will be more in the form of a comment.

22                    If somebody wants to jump in, you

1         can do so very quickly.

2                     But the whole notion, there is

3         clearly a gap here. I just want to make sure

4         that you all don't disagree with the exchange

5         that we had this morning that I had with the test

6         labs, and that is, the gap is that the current

7         voting system standards that we operate under -

8         the 2002 VSS - were developed before HAVA was

9         even close to being in its final form. So we all

10        agree with that.

11                    To get a system through the testing

12        process and to say we are certified to the 2002

13        VSS does not necessarily mean that you are HAVA

14        compliant, because the 2002 VSS does not test to

15        the language of Section 301 in HAVA.

16                    Now I understand that many vendors

17        will take a look at, obviously, what is in

18        Section 301 and build a system to be, from their

19        perspective, compliant with HAVA. But, in fact,

20        there are no voting system standards currently

21        that test to the language of Section 301,

22        specifically to the language of Section 301(a)(3)

1         dealing with accessibility.

2                     Any comment about that? And I think

3         my time is out, so any quick comment about that.

4                     MR. CHARLES: I think, very quickly,

5         there are systems that meet the plain language of

6         HAVA without being tested as a standard. You can

7         look at those and determine that they are

8         accessible, they do those things, that you can

9         check and change your ballot, all those pieces,

10        under the 1990 standards as well as the 2002.

11                    But, you are right, there is not a

12        HAVA certification of a system.

13                    MR. GROH: I would concur with that.

14                    COMMISSIONER MARTINEZ: Thank you

15        all very, very much.

16                    Thank you, Madam Chair.

17                    CHAIR HILLMAN: Thank you.

18                    Mr. Vice Chairman.

19                    VICE CHAIR DeGREGORIO: Thank you,

20        Madam Chair.

21                    Actually, to follow up on what

22        Commissioner Martinez just brought up, and Mr.

1         Groh, you specifically, in one of your

2         recommendations, your first one here, you talked

3         about the EAC should clarify through a safe

4         harbor statement that states and localities that

5         want and need to make purchase decisions today to

6         comply with HAVA can do so under today's

7         standards and that systems purchased today will

8         not have to be replaced in the future.

9                     Are you suggesting, through some

10        kind of statement that we would issue, some kind

11        of safe harbor statement, that these systems

12        would not have to be compliant with 301(a)?

13                    MR. GROH: No, not at all.

14                    But, again, we are in that gray area

15        which Commissioner Martinez just pointed out: we

16        are not testing to that.

17                    But I think a common sense

18        approach - and there has been a DOJ opinion that

19        goes back to, I think, 2003, on DRE with the

20        VVPAT, of providing and meeting the level of

21        compliance that 301 requires. And I think that

22        that is do-able, but it hasn't been tested to it.

1                     VICE CHAIR DeGREGORIO: And I

2         recognize there is no guidelines for this because

3         it is new to HAVA.

4                     But I don't see 301(a)(3) as a gray

5         area. It is in the statute and states have to

6         comply with this.

7                     MR. GROH: And the gray area to me

8         is it hasn't been tested under the 2002. But it

9         doesn't mean that some of the systems don't have

10        the elements in it that if there were a test, it

11        would pass it.

12                    VICE CHAIR DeGREGORIO: Mr. Adler,

13        could you just tell us, your IDV system, your

14        independent dual verification system, how does

15        that work?

16                    MR. ADLER: It's a unit that is

17        connected to an electronic voting system and it

18        gives, when the voter votes, the voter goes

19        through their confirmation process and then once

20        they confirm their vote, that vote goes into the

21        unit, and as the VVSG talks to, maintains a

22        separate copy of the vote.

1                     The voter then has an opportunity to

2         get a receipt, to probe that ballot to make sure

3         that it did, in fact, reflect their intent. And

4         then they could actually, if the jurisdiction so

5         desires, give them an opportunity to use that

6         receipt to make sure the vote is actually in the

7         count.

8                     So the model is very similar to an

9         ATM transaction. You get a receipt. You take

10        out your $20; you make sure the receipt matches

11        the $20. And then you take that receipt and

12        clear it against your statement at the end of the

13        month.

14                    The statement at the end of the

15        month in the election context is the transcript

16        that the results are certified against.

17                    So it actually provides the ability

18        for the voter to do that.

19                    There is also the ability for the

20        election officials to gather these receipts and

21        verify that, in fact, the confidence in the

22        election results are there.

1                     And this really goes back to what we

2         discussed at the NIST symposium back in December

3         2003, I believe, where, talking about trust and

4         confidence in election systems, where you can

5         actually prove that the voting system - either on

6         the back end - tabulation data base - or the

7         voting machine itself is not making mistakes.

8                     You can prove it. You don't need to

9         assert it; you can actually prove it.

10                    VICE CHAIR DeGREGORIO: Is your

11        system a standalone system or meant to be used

12        with another system?

13                    MR. ADLER: It's meant to be used

14        with an electronic voting system, a minimal level

15        of interaction.

16                    VICE CHAIR DeGREGORIO: Just wanted

17        to make sure I understood how it worked. Because

18        you make great reference to it in your comments

19        here regarding the guidelines that the TGDC came

20        up with.

21                    And I want to make sure I understand

22        exactly how you envision this working.

1                     MR. ADLER: The appendix, talking

2         about end-to-end cryptographic verification,

3         actually has a comprehensive set of detailed

4         requirements that goes through how the system

5         like this one works.

6                     VICE CHAIR DeGREGORIO: Mr. Charles,

7         you, in your comments, talk about recommending to

8         us that instead of waiting 24 months to implement

9         this, that we should do it immediately upon

10        adoption or as soon as testing authorities are

11        able to review the system. But you still want

12        individual components to be qualified under the

13        2002 standards.

14                    Are you suggesting that instead of

15        testing a whole system, that a system, part of

16        the system was qualified under 2002 and tested by

17        and ITA to be qualified, if something was added

18        to it, just that component should be tested

19        against these guidelines and that's something

20        that could be done right away instead of waiting

21        the 24 months?

22                    MR. CHARLES: It would be faster and

1         less expensive, especially if it relates to older

2         systems where a county or state may have a 1990

3         model system, a 1990 standard system, and they

4         want to add a printer to it.

5                     Some of that hardware may not be

6         economically upgradable to the 2005 or 2002

7         standards. But you could add a printer to it and

8         it's functioning for some time and be able to

9         test.

10                    Right now you can't test a component

11        and add it to a 1990 system to help a state meet

12        state law without putting the entire system up to

13        2002.

14                    That places a pretty significant

15        burden on states or counties that may have a

16        statewide 1990 system and cannot afford to

17        implement the hardware at the necessary level.

18                    It's really not just for this

19        version but, in going forward, I don't think that

20        there is a reason, that the different versions of

21        standards cannot co-exist, as long as when you

22        are testing things, you are trying to test them

1         to bring them up and you qualify - and the

2         certification numbers or components of systems -

3         what level that system has passed so that

4         purchasers of equipment will know what level.

5                     But that way, if possible, it would

6         be nice to help states avoid replacing entire

7         systems of hardware because the standards may

8         change in the future.

9                     VICE CHAIR DeGREGORIO: But when you

10        add new components like that, even if it's just a

11        printer, sometimes doesn't that affect the whole

12        system or could affect the whole system?

13                    Sometimes you add a new printer and

14        it's the wrong driver because the driver is ten

15        years old with the old system. And you are

16        getting in there and you are changing the driver

17        on the software on a system that has been

18        certified.

19                    MR. CHARLES: But you still test

20        that system end-to-end. So you would confirm

21        that that 2002 standard printer matches up with

22        the 1990 equipment and you could do an end-to-end

1         test from software to election setup to election

2         operations to printer, you can confirm that that

3         works properly within that system.

4                     And that's how we moved from the

5         1990 to the 2000 standards, each piece or each

6         component would get upgraded, but they would be

7         tested as a complete system. So as long as you

8         do that complete system test, you would be able

9         to ensure the reliability and durability of that

10        component.

11                    VICE CHAIR DeGREGORIO: I know that

12        the three of you provided us with written

13        testimony - and Mr. Vadura you are going to

14        provide us with yours - and I appreciate your

15        testimony.

16                    I have been in this business a long

17        time and know many of you for many years because

18        I was a Director of Elections in St. Louis County

19        and have been coming to IAKVIAC (ph) meetings

20        since 1986 and going to the vendors' show and

21        seeing all the new and innovative equipment and

22        meeting new people in the last few years through

1         my work that I did with IFIS (ph) internationally

2         and then in this job.

3                     And I have to say, I am amazed every

4         year - and especially in the last two years - to

5         see the new products on display and how you are

6         trying to respond to the marketplace.

7                     And, you know, the fact of the

8         matter is that from 2000 to 2004, 25% of the

9         country saw new equipment.

10                    From 2004 to 2006 we have this

11        challenge that 30% of the country, though, is

12        using lever machines and punch cards,

13        particularly, are going to be changing over into

14        new systems, so it brings new challenge. These

15        guidelines bring a challenge to you.

16                    But the Federal Government, for the

17        first time, is providingly $3 billion to get part

18        of this, part of the equipment part of it done.

19                    Are you all up to the challenge and

20        your colleagues - because you work with your

21        colleagues I know. You are friendly - you are

22        competitors but you are friendly competitors, I

1         always felt so.

2                     Are you up to this challenge, do you

3         think? Is there enough vendors in the

4         marketplace to meet this demand and to serve not

5         only the needs of election officials - but, let's

6         face it - we are serving the needs of the public

7         out there, and a skeptical public - there is a

8         chunk of them.

9                     And the survey data, Commissioner

10        Martinez talked about it this morning, of about

11        one-quarter of the voters that don't quite trust

12        the systems. And there is that skepticism out

13        there.

14                    Are you all meeting that challenge

15        and is there anything that in these guidelines

16        that we can add, in addition to what you said

17        here, that might be helpful to you in helping you

18        to meet the challenge?

19                    We will start here with Dennis and

20        go this way (indicating).

21                    MR. VADURA: I think we are all

22        working really hard to meet those challenges.

1                     Having a staple set of standards

2         helps to meet those challenges because you can

3         predict what needs to be done to get it certified

4         in time.

5                     There is an effective barrier

6         to entry for new vendors coming into this

7         marketplace, as of today, which is certification.

8                     I don't believe anybody can come

9         into this marketplace starting today and get up

10        to speed to deliver anything by 2006. So I think

11        everybody else here would agree with that.

12                    So, is there enough vendors? We

13        will find out. I hope so.

14                    MR. CHARLES: I think there is. I

15        think that we are not going to sleep much in the

16        next 24 months. I think we are all going to -

17        every county election official, every state

18        official and every company - will be pretty

19        heavily taxed on this. But we have been

20        preparing for it for some time, not just on the

21        manufacturing side, but making sure that there

22        are enough people trained to support the local

1         elections.

2                     My biggest concern is for small

3         rural jurisdictions that really haven't invested

4         a lot of time and may not command the market size

5         to get the level of support that they may need.

6         I hope that every election official out there

7         realizes that time is of the essence. The sooner

8         they can make decisions and start, the better

9         they will be and the better that we as an

10        industry will be able to support them and adjust

11        resources accordingly so that they get the

12        support they need.

13                    MR. ADLER: I think there is a level

14        of passion for elections on no matter what side

15        of this table you are on.

16                    And that translates over to a level

17        of engagement with the public and putting in the

18        sleepless nights, not because you have to, it's

19        in your blood. You just want to and are driven

20        to deliver and make sure every vote counts.

21                    And engage the skeptics, quite

22        frankly. I know that we have been very open to

1         doing that, during this more difficult period.

2                     So I think the process is

3         invigorating. I think it stresses all of us out,

4         we all know that. But it's also, we are doing

5         good work here.

6                     And this is not a -- This is a

7         contentious process. I think that's not all a

8         bad thing. But we just have to be honest,

9         respectful and engaged.

10                    The work load over the next 24

11        months is going to be horrendous - I will tell

12        you that - no matter where you are.

13                    But I think that's not surprising.

14        I am certain it's not a surprise to you, Paul, or

15        anybody else, having done elections for so many

16        years.

17                    Having 30% of the country change

18        election equipment is going to be tough. And I

19        agree with Alfie here, for small jurisdictions

20        it's going to be especially difficult.

21                    MR. GROH: We look at this from many

22        perspectives. And, again, this is not just a

1         single thread where if we deliver the hardware,

2         that solves the situation. From my company's

3         standpoint, we know we can deliver the technology

4         and get it there.

5                     But what we are concerned about is

6         can the election administrators and the poll

7         workers and the voters feasibly take this

8         dramatic change that's going to happen very

9         quickly? And it is not equal across the United

10        States.

11                    We have election laws that are

12        different in different places that put an added

13        burden.

14                    We heard New York City, one of the

15        election administrators this morning, talk about

16        that, of trying to phase this in.

17                    So our concern as a company is we

18        can deliver the best technology. But if we

19        haven't done the next step, which is really

20        shared by us, by the election administrator, it

21        can look like a failure, because it doesn't work

22        on Election Day and it's an administrative or a

1         poll worker or a voter issue.

2                     But the first finger that will be

3         pointed will go back to us as the election

4         supplier. So we work very hard on the

5         implementation.

6                     So we can do it. We will put in a

7         tremendous number of hours into accomplishing

8         this. I do not see us having a constraint in

9         that we couldn't deliver all that we could get

10        orders for.

11                    But it's the project management

12        piece of it.

13                    So the other component that you are

14        addressing in HAVA is that voter education, a new

15        way of looking at poll workers that need to be

16        brought on through your college program and that

17        you have that.

18                    We applaud all of those, because

19        that's what is going to be required.

20                    And then I think that there is a

21        common sense or a reasonable man approach to ask

22        the question, if we deliver all this at one big

1         time and cause a major change-over, can we

2         swallow that.

3                     And I guess the analogy I would give

4         is, we all know we can start on an exercise

5         program, but that first day I don't do 50 sit-ups

6         because I won't do a sit-up for about another

7         week.

8                     We can't go into this on day one

9         thinking, I am going to enter the exercise

10        program full bore. I have to work my way up into

11        it.

12                    VICE CHAIR DeGREGORIO: Thank you

13        all very much.

14                    CHAIR HILLMAN: Thank you.

15                    In the remaining few minutes, let me

16        just ask a question and then I will give a nod to

17        my colleague, Mr. Wilkey, if he has a question.

18                    I have a specific question for you,

19        Mr. Groh. No, maybe it wasn't your question. I

20        think it was for Sequoia.

21                    Somebody indicated, made a general

22        reference to - it was Sequoia - ambiguously

1         worded and without any specific reference.

2                     And could I just ask, not for you to

3         comment here, but in your follow-up submission,

4         if you would be as specific as possible, because

5         I don't know what to respond to with that.

6                     You were all in business and had

7         systems in the market in 2000; is that correct?

8                     Did you have something in the

9         market?

10                    MR. ADLER: No.

11                    CHAIR HILLMAN: You started in '98,

12        but you certainly were in the field in 2000.

13                    MR. ADLER: That's right.

14                    CHAIR HILLMAN: And, therefore, were

15        following the year's, following the 2000

16        Presidential election, I'm sure, with great

17        interest with respect to the discussions about

18        various voting systems.

19                    So I would ask you, it's sort of a

20        two part question. But it's the second part of

21        my question that's probably where you should

22        emphasize.

1                     One would be what your companies

2         have done to address some of the deficiencies

3         that are now, that we are trying to address in

4         the guidelines.

5                     But the other part of that question

6         is, is there anything in here that's a big

7         surprise to you, given what you have heard and

8         what you have been engaged in over the past four,

9         five years? Is there anything in here that

10        surprises you that you don't think you are ready

11        for or that your company wasn't expecting or

12        preparing for?

13                    MR. GROH: I will go ahead and jump

14        into this.

15                    No, it hasn't been. And part of why

16        it's not a major surprise to us is because we

17        have been involved throughout the process.

18                    Are there little unique things in

19        there? Yes, they are. But they don't rise to

20        the level that I am surprised or that it knocks

21        me out of my chair.

22                    All of us that I think have been in

1         this have been trying to work on it before we

2         actually could see the target. We knew there was

3         going to be a target around the corner. I may be

4         off a few degrees and that's going to cause me to

5         have to recheck it.

6                     I think our concern, from Election

7         Systems & Software, is that we not be so

8         prescriptive or come up with a specific

9         methodology to solve something that could have

10        other means of solving. Allow us, as the

11        developers of techonology, to provide those.

12                    That's been one of the challenges,

13        that ES&S has done as an innovator, we have had

14        our customer base say I want a paper verifiable

15        audit trail.

16                    We have attempted to do that in a

17        format that we think is usable and we want to

18        introduce that and have that be tested.

19                    We have had our customer base say we

20        want to stay with paper base. Can you provide us

21        with a solution that will allow us to meet 301

22        and allow accessibility but still maintain and

1         utilize a paper base system. We have made that

2         effort to do that.

3                     And we are learning and are trying

4         to make adjustments to that. But we are driven,

5         all of us are driven, by the competitive nature

6         of what our customers will pay for, buy and want.

7                     CHAIR HILLMAN: Any comments, Mr.

8         Adler?

9                     MR. ADLER: I would agree with Mr.

10        Groh, that the process has been collaborative and

11        open and has built upon the work done previously

12        in a very professional way. So there were no

13        surprises really, other than those few degrees

14        that we are addressing through this comment

15        period.

16                    With respect to what we have been

17        doing since 2000, I think we are in a paradigm

18        shift in elections. I think that's what

19        generated HAVA, in many respects, and we are

20        grappling with that tectonic shift.

21                    And you are leading this industry

22        through that tectonic shift. And those shifts

1         are not often pretty.

2                     But there has been a tremendous

3         amount of innovation. There will continue to be.

4         And what I would hope and what I see is that

5         there is a real penchant for encouraging

6         innovation as we move forward.

7                     We know this is not the last set of

8         election systems jurisdictions procured and there

9         will be no changes going forward. We know there

10        will be. And this is a roadmap to take us there.

11                    MR. CHARLES: Just briefly, I think

12        there isn't anything in this document that is a

13        tremendous surprise and I think that the credit

14        for that goes to the TGDC and NIST in having

15        experts on there that really understand the

16        elections process and the work that they put in.

17                    I know that the hours they put in

18        were incredible under a tight timeline.

19                    If there was one recommendation that

20        I would have for the next version of this, it

21        would be to invite representatives of the vendor

22        community to participate on the TGDC so that

1         there will be somebody in addition to the testing

2         authorities and the election officials, the

3         designers of the equipment, so that we can have a

4         more direct involvement in the discussions that

5         take place.

6                     But even without that, I think they

7         did an excellent job in keeping this to something

8         that was relatively expected by the industry.

9                     MR. VADURA: I would argue with

10        that. There is no real surprises in this other

11        than, I think on the next round, I agree with Mr.

12        Charles, let us contribute. Because I think some

13        practicality can be brought to the process from a

14        manufacturing and design standpoint.

15                    It will make a better document and

16        something that we can all live with and, frankly,

17        jurisdictions can procure at a cheaper price.

18                    MR. GROH: I would like to make one

19        additional point so we can get this on the

20        record.

21                    All of us need to remember and

22        recognize that when HAVA was being worked out on

1         the legislative format and it was being costed

2         out or priced out, it was based on the technology

3         that existed then.

4                     This technology we have now has made

5         a leap forward, which is positive.

6                     But at the time that they scaled

7         this and scoped it, it was not in today's terms.

8         It was in 2002 technology terms.

9                     CHAIR HILLMAN: Thank you.

10                    EXECUTIVE DIRECTOR WILKEY: I think

11        this is working.

12                    Just shifting gears on a subject for

13        one minute.

14                    Anyone who has heard me speak over

15        the last twenty years since we started the

16        development of the 1990 standard, I hate to admit

17        I have been around even longer than that, has

18        heard me say on more than one occasion that it's

19        great that we have these standards for hardware

20        and software testing.

21                    But what we need to have is the

22        second half of that. That is 50% of the battle.

1                     The other 50% is the management

2         guidelines. And I have talked about that until I

3         am blue in the face, as blue as that cap is on

4         that bottle of water.

5                     Hopefully during this coming year we

6         will begin a venture with the National

7         Association of the State Election Directors to

8         finally get that moving. And we have set aside

9         some money out of our budget to begin that

10        process,

11                    I would like your reaction of how

12        that fits into your plans and how you can help us

13        make that document the best that it can be.

14                    MR. VADURA: I think we all have to

15        have operational guidelines as part of our

16        certification documents. And some of what's in

17        there is good stuff and can contribute to this

18        process.

19                    And I think any set of formal

20        management guidelines can only help us because it

21        will help to mold further the shape of the

22        system, how it works from an operational

1         standpoint, not just a technology standpoint.

2                     MR. CHARLES: I think it's an

3         excellent idea. It's, as you know, a very

4         difficult task to integrate requirements that

5         address multiple types of systems, multiple

6         vendors of those types of systems, as well as the

7         different state laws and requirements.

8                     So it's not an easy task, but an

9         important one. And I think we can assist in that

10        process by offering our expertise as project

11        management and people with experience across

12        state lines. So we can help refine those and

13        develop them as they go forward.

14                    But it would be very helpful - as

15        with the hardware and software, it would be

16        helpful to have state specific requirements that

17        could be tested so there is a one-stop-shop for

18        approval of voting technology.

19                    MR. ADLER: I think that, too, this

20        issue of managing for change, being able to

21        implement new systems and dealing with the real

22        management challenge of conducting elections with

1         temporary workers and under tough scrutiny, that

2         takes real leadership at multiple levels.

3                     So anything that could enhance the

4         leadership and management tools that our election

5         directors/election officials have, is a huge,

6         huge benefit to the public.

7                     MR. GROH: Election systems and

8         software, if you approach this like you did with

9         TGDC and NIST, we would be more than happy to

10        participate. Because, again, this is that

11        missing link that I explained earlier.

12                    I can deliver excellent hardware.

13        Now I am beholden upon the best management

14        practices of the election administrator. And

15        when that fails, I have then failed by virtue of

16        the public exposure to that. And I don't want

17        that.

18                    So we are working hard. But as you

19        have eloquently explained, Tom, when this has not

20        be broadly embraced and there is not something

21        that is there that uniformly lays this out so

22        best practices methodology procedures can be held

1         to by County Commissioners holding their election

2         administrators to this, you don't have a way to

3         measure it.

4                     The Election Center I know would be

5         a wonderful partner in this, along with the

6         vendor community.

7                     We know what it takes to run our

8         technology and what it takes to run and manage it

9         well. So let us at least advise in that.

10                    We will come behind it and do the

11        training in that. But it's difficult to do

12        training for people who don't want to be trained.

13                    I know so many people that want to

14        go out and pick up a golf club and start hitting

15        a golf ball before they have had any lessons and

16        it just doesn't work.

17                    EXECUTIVE DIRECTOR WILKEY: Thank

18        you very much.

19                    CHAIR HILLMAN: Thank you very much

20        to the panelists.

21                    We are not going to take a break. I

22        am just going to ask the EAC staff to move

1         quickly to reset the table and escort our third

2         set of panelists up to the table.

3                     Thank you very much.

4        

5        

6         * * * * *

7        

8        

9                     PERSPECTIVES ON PROPOSED

10                    VVPAT TESTING GUIDELINES

11       

12                    CHAIR HILLMAN: If we could get

13        reassembled, please. I just need to get

14        everybody.

15                    I will remind everybody to make sure

16        that their cell phones, pages and other

17        electronic devices are silenced. We would

18        appreciate that.

19                    Our fourth panelist is here and he

20        will be joining us, running a few minutes late.

21        And I don't want to cut into the end of the day

22        by short cutting anybody else's time for

1         presentation.

2                     This is our third panel. It is

3         Perspectives on Proposed Voter Verifiable Audit

4         Trail Testing Guidelines.

5                     Presenting this afternoon will be

6         Peter Kosinski, Chief State Election Official

7         from the State of New York - welcome; Harvard

8         Lomax, Clark County Registrar of Voters from

9         Nevada - welcome, thank you for traveling; Jim

10        Dickson, who will be joining us in just a few

11        minutes; and Avi Rubin, who is a Professor at

12        Johns Hopkins University in Baltimore.

13                    Thank you all very much.

14                    It will be, I think, okay. I know

15        that Jim Dickson will be interested in what the

16        other panelists have to say. But Mr. Kosinski,

17        if you would be so kind as to get started with

18        us.

19                    And I don't believe we have any

20        written remarks.

21                    MR. KOSINSKI: No.

22                    CHAIR HILLMAN: I know we invited

1         you very late so this is not --

2                     MR. KOSINSKI: Quite all right.

3                     CHAIR HILLMAN: I just want to make

4         sure I'm not missing anything.

5                     MR. KOSINSKI: You are not missing a

6         thing.

7                     CHAIR HILLMAN: But ask if you would

8         also indulge us to speak slowly so we can take

9         notes.

10                    MR. KOSINSKI: I will do my best.

11        If I am speaking too quickly, let me know.

12                    CHAIR HILLMAN: And I would ask that

13        people please summarize, particularly if you have

14        submitted written statements, so that we might

15        have ample time for questions.

16                    Thank you.

17                    MR. KOSINSKI: Thank you,

18        Commissioner. I will try to summarize, even

19        without my written statement, to move the program

20        along.

21                    First of all, I'd like to thank you

22        for inviting me today and I'd like to welcome

1         everyone to New York.

2                     I know my friend John Ravitz was

3         here this morning to welcome everyone. But I

4         would also like to welcome you on behalf of New

5         York and hope that you can take advantage of some

6         of what we have to offer in the City.

7                     I'd also like to congratulate the

8         Commission on their wise choice of their

9         Executive Director.

10                    And I certainly would like to note

11        that Tom Wilkey used to be the Director of our

12        State Board of Elections and we feel you made an

13        excellent choice in picking Tom as the Executive

14        Director.

15                    And, of course, we worked closely

16        with Tom. I personally worked closely with Tom

17        for many, many years and he is excellent in his

18        job and we have a great deal of faith in his

19        ability to carry forward on this.

20                    I'll just spend a couple of minutes

21        speaking a little bit from the administrative

22        standpoint, I think, on how the voter verified

1         receipt affects us, I believe, and a couple of

2         comments regarding this.

3                     The State of New York I think

4         decided early on in this process that they would

5         enact and require, as part of any voting system

6         in New York that would be part of the HAVA

7         project, would have a voter verified receipt

8         attached to it.

9                     Our legislation, as I am sure you

10        are aware, has just recently passed our State

11        Legislature. But in the very first drafts of the

12        legislation that was proposed in New York, way

13        back two years ago, the voter verified receipt

14        was part of those pieces of legislation in both

15        House of our State Legislature.

16                    And I don't think there was a great

17        deal of discussion or certainly dissension

18        regarding that as being a component of any new

19        voting system in New York. And, in fact, the

20        final version of the legislation that was adopted

21        by our State Legislature - and while it hasn't

22        been signed by our Governor, we are hopeful that

1         it will be in the very near future - that that is

2         a part of our legislation.

3                     Now, what it does not do, as you

4         might imagine, is give a great deal of detail as

5         to exactly how that is to work. And that is why

6         we are paying such close attention to these

7         guidelines as you are issuing them.

8                     We think this is a very, very

9         important component, of course, of any voting

10        system. It does create some issues, though, as

11        we go through the voting process.

12                    New York is a state, as I am sure

13        you know, that has had lever machines for its

14        entire voting history. And because of that,

15        paper ballots have been a very small part of

16        voting in New York State.

17                    Up until the changes that we are

18        anticipating, the only paper ballots that were

19        really available in New York were absentee

20        voting, the provisional ballot voting, military

21        voting. That kind of voting was all done on

22        paper. But other than that, it was all done on

1         lever machines.

2                     And at the end of an election cycle,

3         there would be very, very little paper; maybe 8%

4         of our total votes would be cast on paper. The

5         remaining 92% were cast on the lever machines.

6         And there was no paper that was at issue

7         regarding that particular election, at least as

8         far as the votes cast at the precincts.

9                     With the new system, needless to

10        say, this will change. And having the voter

11        verified receipt as part of that system, it will

12        now greatly increase the amount of paper that the

13        state will be dealing with.

14                    We think it's important that any

15        system that incorporates this kind of a program

16        have as a component to it that we ensure that the

17        voter actually does verify the votes that are

18        being cast.

19                    And I think that, as I read through

20        the standards that are being put out, that it is

21        important to try to build in to any system that

22        would be engaged by the voter, that not only we

1         create the voter verified system, but that we

2         also try as best we can to ensure that that voter

3         actually verifies that ballot.

4                     Because we realize that at the end

5         of the process, the official ballot - if there

6         becomes a dispute in an election - will be the

7         paper ballot. And so it needs to have some

8         component that, at the end of the vote, when the

9         voter - prior to leaving that voting booth - is

10        asked to verify that particular ballot, that in

11        fact that voter does that.

12                    So that we have some comfort level

13        that, in fact, that voter has looked at that

14        piece of paper, has verified that piece of paper

15        as their actual votes. So if those votes are

16        contested, that we are counting the actual votes

17        of that particular voter.

18                    The other issue that, of course,

19        comes up, which I think was touched on - and I

20        think Tom, in fact, touched on it as well - is

21        the management of the system. And the management

22        of the system will largely come about not only

1         during the election itself, but post election.

2                     So once you have created this piece

3         of paper in New York State - we can have upwards

4         of seven million to eight million votes in a

5         particular election - when you are creating that

6         kind of paper, that there be a system in place

7         that will allow for the protection of that paper,

8         that will allow for the careful counting of that

9         paper, if that becomes necessary, and that much

10        of New York City's law, in fact, is devoted to

11        that.

12                    For example, the amount of audit

13        that must be done to verify that the votes cast

14        on the electronic machine are matching up with

15        the votes to the verified pieces of paper. And

16        that is a very important component of the system,

17        that not only are you creating this piece of

18        paper, not only are you verifying this piece of

19        paper, but that you also have a system in place

20        for ensuring that these pieces of paper are

21        protected and they are also being utilized in a

22        way to make the public aware that the system is

1         being monitored and that the integrity of the

2         system is not being compromised.

3                     We would certainly hope, and would

4         anticipate, that any system that would be created

5         would be accessible to the disabled. We know

6         that that's a very, very important concept. That

7         really was a lot of what drove HAVA. It was, I

8         think, a critical element in the development of

9         HAVA itself.

10                    And we think it's very, very

11        important that as we create these systems, that

12        we keep that in mind.

13                    And while the voter verified

14        receipt, I believe, was a component of this, that

15        really emerged following the enactment of HAVA,

16        it was not necessarily anticipated during the

17        discussions in Washington of the HAVA legislation

18        itself that since this has emerged as such an

19        important part of this particular process, that

20        we make sure that the disabled community is

21        protected; that we do not in any way compromise

22        these systems, so that they are fully accessible

1         to the disabled.

2                     Other than that, I think that the

3         creation of this system is something that as we

4         work through it - it will take some time - I know

5         that that's something that has been spoken of

6         earlier here today - that as we work with these

7         systems and see how they are being utilized by

8         the people who are actually in charge of

9         utilizing them - our election inspectors, as well

10        as our voters - we can maybe fine tune these, to

11        make sure that they are accomplishing the goals

12        that we have set out for these particular

13        systems.

14                    And we know that that's important,

15        that we get some opportunity to use the system

16        before we make necessarily final adjustments to

17        that system to make sure that it is being

18        adequately monitored and adequately utilized.

19                    Other than that, I will just accept

20        questions.

21                    CHAIR HILLMAN: Thank you very much.

22                    We will hear next from Mr. Lomax.

1                     And then following Mr. Lomax we will

2         hear from Jim Dickson, who is Vice President for

3         Governmental Affairs with the American

4         Association for People with Disabilities, and

5         then from Professor Rubin.

6                     Mr. Lomax.

7                     MR. LOMAX: I go by Larry, so I am

8         Larry Lomax from Clark County, Nevada.

9                     And I was asked here because we are

10        one of the few counties that has actually used

11        the VVPAT. We used it for the 2004 primary and

12        general and then for our 2005 municipal

13        elections, primary and general elections.

14                    So we do have experience. And I'd

15        like to talk about some of that experience

16        first-hand because, again, that's why you

17        actually asked me, and compare it to the

18        standards as I read them.

19                    And I will point out, as I

20        understand, what's important is that the vendors

21        and the technical people understand these

22        standards, not so much that I do. But I see some

1         potential problems, anyway, that I think need to

2         be addressed.

3                     First, as a user, I think it's very

4         important that we don't ever forget that the

5         standards have to be practical. And I think

6         there is some ideals that we would all like to

7         attain, but I'm not sure in some cases that the

8         ideals are realistic, or at least they may be

9         overly burdensome upon those of us that have to

10        use the voting machines and implement these

11        things.

12                    Second of all, and I just point this

13        out, that Volume I, the very beginning of it

14        says, its purpose is to describe the requirements

15        for electronic components of voting systems.

16                    And there's a lot of standards in

17        here that really go way beyond that into the

18        administrative procedures, especially, I saw in

19        the area of security. What they are really

20        requiring is us to perform certain administrative

21        functions that have little to do with electronic

22        components.

1                     Third - and I think this has already

2         come up today in two preceding panels - I don't

3         think the standards are sufficiently clear in

4         several cases. And I will identify a few here.

5                     And a good example that has already

6         come up, the gentleman from ES&S apparently

7         interprets the standard to mean that the paper

8         trail printer cannot be a spool-to-spool or reel

9         to reel printer, or it must somehow chop the

10        ballots at the end of that.

11                    I talked to Alfie from Sequoia. He

12        believes just the opposite, that that is allowed

13        under these standards.

14                    So these things have to be clear

15        enough so we all understand exactly what they

16        mean.

17                    The standards is guidance. We don't

18        want to have to create guidance to clarify the

19        guidance because that's kind of where we are

20        after HAVA. HAVA was in there; now we are trying

21        to figure out exactly what that meant.

22                    Another example I can give you is

1         there is a standard that says the paper record

2         shall be sturdy, clean and of sufficient

3         durability to be used for clarification and other

4         things.

5                     Well, I had my staff read these

6         standards and they immediately came to me on that

7         issue because we use thermal paper, which is

8         sufficient for what we need this paper for. But

9         then is it really very durable? But, again, what

10        does a standard such as this mean?

11                    So I don't want to find us in a

12        position later on of not being sure exactly what

13        is required and what isn't.

14                    A few examples - again, looking at

15        it from the practical level, somebody who has to

16        administer these things - font size, one of the

17        first things mentioned in there. It's not

18        realistic to ask for large font size, I don't

19        believe, anyway. The bigger the font size, the

20        more paper each ballot uses up, the larger the

21        printer is going to have to be, the more

22        difficult the logistical problems are for those

1         of us that have to deal with it. And also the

2         greater the expense.

3                     I think the goal would be, at least

4         from my perspective, is a printer where the paper

5         did not have to be replaced on the date of the

6         elections, that has sufficient capacity to deal

7         with all the voters who are going to use that

8         machine. And that means relatively small print.

9                     We used a printer that printed in 10

10        point font; that's small. However, the secret to

11        this problem is magnification. We also had a

12        magnifier at every machine. So that somebody who

13        had difficulty reading 10 point font could hold a

14        magnifier up and it would bring it up to 20 point

15        font, which is pretty large.

16                    I could honestly say we have had now

17        approximately a half a million voters vote in

18        Clark County using the paper trail printers and

19        this has not been an issue. People comment that

20        it's small, the print is small, but that's it. I

21        haven't had a single person complain or make an

22        issue of the fact that 10 point font is what we

1         are using.

2                     Now, these standards require 3

3         millimeters I think, just to make it tough on me.

4         We measured that and that apparently is 12 point

5         font, which is a little bigger and that's fine.

6                     But the standards also talk about,

7         you should have two choices of font there for the

8         voter. And I'm not sure in standards what

9         'should' means. I'm not sure what the vendors

10        then have to do.

11                    If it says 'should', maybe that's an

12        atta boy. But it seems to me you ought to just

13        set what the minimum requirements are.

14                    It looked to me that 6.3 millimeter

15        font in height is 28 point font, and that would

16        require paper this size (indicating), which is

17        out of the question.

18                    There is a section in the standards

19        that talk about approve or spoil the paper record

20        and then it talks to having the ability to match

21        the spoiled paper records with the spoiled

22        electronic ballot. There are no spoiled

1         electronic ballots.

2                     One of the beauties of electronic

3         voting is, the voter goes through the machine.

4         At any time he or she wants, they can change;

5         correct any mistakes they have made.

6                     After they have made all their

7         selections, they get to a review screen, which

8         shows them the selections they have made and

9         shows them all the contests in which they made no

10        selection. They can immediately go back and

11        correct those.

12                    There is even a warning notice that

13        comes up at the end that says, you didn't make a

14        choice at every place you could have. And they

15        have to go through all of that before they get to

16        the point where they print their paper record.

17                    They print their paper record and

18        and then they review that. They can either

19        accept it or reject it.

20                    There is three reasons they might

21        reject it.

22                    First, we made a mistake programming

1         the machine. And there we got a disaster and

2         that would be brought to our attention very early

3         election morning and we would have big trouble.

4         But that would be a programming error on our

5         part. And if we have done our job, that's not

6         going to happen.

7                     The second reason is, despite all

8         those advantages or opportunities to correct a

9         choice, they didn't catch it until they saw the

10        paper record, in which case, they can reject the

11        paper record. And remember, the electronic

12        record hasn't been recorded get.

13                    They can go back and correct the

14        electronic record again, print the paper record

15        again and say okay, it's okay now, and then print

16        and finalize and cast their ballot.

17                    The third reason, and this was the

18        most common in my opinion in Clark County, they

19        just want to see what happens if they hit the

20        reject button and see what the printer will do.

21                    And in our case it would print VOID

22        and then they could go back and print it again.

1                     As I've said, we have had about half

2         a million people vote and very, very, very few

3         people ever void a ballot.

4                     Because the thing you have to

5         understand is the printed record will always

6         match what they picked on the voting machine, if

7         we have done our job right. It always matches.

8                     I have a lot of people call me up

9         and ask what was the match, how many mistakes

10        were there? There are no mistakes; it always

11        matches. It always matches.

12                    So most of the time when they are

13        doing this, they are just playing around.

14                    Preserve voter privacy and

15        anonymity, another issue here. Reel to reel

16        printers, from a practical standpoint, in my

17        opinion, are the only reasonable way to go here.

18                    In a county my size which is about,

19        we have about 750,000 active registered voters -

20        and we have a two week early voting period - an

21        election will generate in the vicinity of 8000

22        tapes. Now, these aren't the ballots chopped up;

1         these are tapes.

2                     Each tape has to be maintained and

3         catalogued in a manner so that we could find any

4         tape for any given machine on any given day if we

5         needed to audit it against electronic results,

6         because in Nevada we are using an auditing

7         process for the tapes, a randomly selected

8         auditing process.

9                     If you chop those tapes into little

10        five or six inch strips of ballots that will then

11        curl up because they came off a reel, and I have

12        got a half a million of those I have to track and

13        account for and catalogue in some manner so that

14        I can get them, either to audit, or if a state

15        was using them as ballots, they would have to

16        maintain those, you are really creating some

17        administrative challenges, which are going to

18        lead to problems if you chop those things up.

19                    It would also, if you chop them up -

20        as far as I can imagine, anyway - it would

21        require you to manually do any recounts of those

22        things. If you are just dealing with these

1         little curlied up pieces of paper, someone is

2         going to have to hold each one up to do a manual

3         recount.

4                     At least with a reel-to-reel, there

5         is a possibility, using bar codes below the

6         ballots, of an entirely independent system going

7         through and reading, doing your recount on a

8         reel-to-reel printer.

9                     So I could go into this more, but I

10        really recommend that you make your printers

11        large enough to hold everybody that is going to

12        vote on one day. And that's the way we do it

13        now.

14                    They can be sealed up so that

15        workers have no access to them. They attach them

16        to the machines at the beginning of the day; they

17        just take them apart, take them off at the end of

18        the day and they bring them in. But they don't

19        have to mess with that paper; you don't have to

20        worry about storing votes at the voting place.

21                    And just, administratively, I think

22        that's really key in my opinion.

1                     There is a couple other areas in the

2         area of anonymity which I just think reflect a

3         misunderstanding, again.

4                     There's a discussion about

5         protecting the privacy of people who use

6         alternative languages. In the discussion it

7         suggests that we have at least five people who

8         vote in another language on a machine.

9                     We have no idea what language they

10        are going to pick when they go into a machine and

11        we certainly are not going to ask them what

12        language they plan on voting.

13                    Their identity is protected

14        administratively in the same manner we protect

15        everyone else's.

16                    It always amazes me, I hear a lot of

17        concern, for instance, on a reel-to-reel printer,

18        protecting the voter's privacy. And yet no one

19        ever brings up a concern that on an election in

20        Clark County we will get about 60,000 absentee

21        ballots, and those ballots will be recived in an

22        envelope with the voter's name right on it.

1                     In fact, the voter has to sign and

2         we make sure that that was the right voter.

3                     And, yet, no one is worried about

4         the voter's privacy then. And they don't need to

5         be. We have administrative procedures to make

6         sure those ballots are separated and nobody knows

7         who they are.

8                     But we maintain administrative

9         procedures in the county to be sure that we

10        protect the electronic voter's privacy also, by

11        the way in which we do not record the order in

12        which they sign in in a roster book.

13                    We have multiple machines in our

14        polling places and the voter picks where they are

15        going go and nobody knows where they are going to

16        go or tracks where they go. So there is no way

17        to go back and reconstruct who used what machine

18        or what order they voted on.

19                    I mean, it's very easy to do and

20        everybody's privacy is protected.

21                    I will just point out a few

22        additional things because I don't want to take up

1         too much more time.

2                     Just an administrative thing. In

3         Section 2.2.7, it does say that if a state is

4         planning to use the paper record as the official

5         ballot, then they have to be sure that a visually

6         impaired person or any other disabled person

7         would be able to review the paper record as the

8         same manner that a somewhat sighted person would

9         be able to do.

10                    And on that, that same language

11        should be brought back to the VVPAT Section,

12        because right now it isn't in here. It's only up

13        in front.

14                    Another example of something - I

15        guess this is almost laughable in a sense. It

16        says the voting station should be physically

17        secure from intentional damage. Good luck.

18                    I don't know what that means or how

19        you would ever enforce that. But if someone is

20        going to come in with a hammer and wants to smash

21        our machine, they are going to be able to do it.

22        So I'm not sure what that standard means.

1                     There is another requirement in

2         there - and these are just examples of things I

3         just don't quite understand or if they have been

4         thought through - that says we are supposed to

5         put a seal to seal the connection between the

6         printer and the voting machine.

7                     All right, if I have 5000 voting

8         machines that means I got to have - remember,

9         where a printer fits in, in most cases it's just

10        a plug with the two little screws that are

11        screwed in there - you know, the little twisty

12        things like on your own computer at home -

13                    Why am I sealing it?

14                    First I have got to buy all those

15        seals and then when they install it they have got

16        to seal it and they have got to track it and

17        somehow log it.

18                    If the printer becomes disconnected,

19        the voting machine will lock up. And that's one

20        of the standards it requires. It says the voting

21        machine has to alert you if the printer is

22        disconnected.

1                     I don't see any - there is no logic,

2         that I can think of, as to why you are wasting

3         your time sealing it. It's an administrative

4         requirement that sounds good, but I'm not sure

5         there is a reason behind it that I could figure

6         out.

7                     I'll wrap it up.

8                     But, in general, I do think that

9         there is more clarity required in these so that

10        we really understand why these procedures have

11        been developed.

12                    And I do think that much of the

13        security, as Mr. Wilkey pointed out, much of the

14        security in an election, in the privacy and

15        anonymity, is a result of administrative

16        procedures and oversight at the polling place.

17                    And we have accepted that for years

18        with paper ballots and we seem to have forgotten

19        that.

20                    And with the electronic, it is also

21        true. And I think that's an important thing to

22        keep in mind. Thanks.

1                     CHAIR HILLMAN: Thank you very much.

2                     Mr. Dickson.

3                     MR. DICKSON: How much time do I

4         have?

5                     CHAIR HILLMAN: Ten minutes.

6                     MR. DICKSON: Thank you, Madam

7         Chair, Mr. Commissioners for inviting me to

8         come.

9                     You don't have my paper testimony

10        because my paper testimony is in my suitcase

11        which is somewhere lost in Delta Airlines.

12                    We will be submitting this short

13        version, plus additional comments.

14                    I first want to talk about the

15        context in which Congress passed and the

16        President signed the law that says people with

17        disabilities have the right to vote.

18                    CHAIR HILLMAN: Get closer to the

19        mike.

20                    MR. DICKSON: I want to put into the

21        context, I want to discuss the context out of

22        which Congress passed a law and the President

1         signed, that says voters with disabilities have

2         the right to vote privately and independently.

3                     Twenty-one years ago Congress passed

4         a law that said polling places should be

5         accessible to people in wheelchairs, 21 years

6         ago.

7                     As best we can tell, somewhere

8         around 65% of the nation's polling places are not

9         accessible to people in wheelchairs - 21 years,

10        more than half.

11                    Missouri just finished - and good

12        for Missouri - a survey to see how many of their

13        polling places were not accessible. The survey

14        reports it's 71%. There is no plan in Missouri

15        for how they are going to fix them, when they are

16        going to fix them by. What good is the report?

17                    And this phenomena in Missouri

18        repeats the phenomena of 18 years ago when

19        reports were sent in to the Federal Election

20        Commission that said these are the places that

21        aren't accessible.

22                    Twelve years ago Congress recognized

1         that people with disabilities are registered to

2         vote at a rate considerably below the rest of the

3         public. Congress also recognized when it passed

4         motor voter, of the National Voter Registration

5         Act, that many people with disabilities have --

6         I'm not going to go to the Department of Motor

7         Vehicles. I have no reason.

8                     Many of us can't even get there.

9         Many of us don't have the money to pay for the

10        license if we could get there.

11                    We are sitting here in New York and

12        the NVRA said twelve years ago that community

13        agencies that offer services to people with

14        disabilities need to do the same thing that motor

15        vehicles do. Twelve years ago.

16                    One state - thank God for Kentucky -

17        is obeying the law. One state, it's twelve

18        years.

19                    We are sitting here in New York City

20        in New York State, and I think these numbers are

21        right. In New York City, about 56% - I'll get

22        the precise numbers if my memory is off - of

1         people do not have driver's licenses.

2                     In upstate New York, well over 90%

3         have driver's licenses.

4                     The agencies that conduct services

5         to disabilities in the city, and upstate -

6         particularly state Medicaid - are not obeying the

7         law.

8                     So we come to a law now that says we

9         have the right to vote privately and

10        independently. Now, that ought to be pretty

11        here, I least I thought it was clear when

12        Congress voted on it.

13                    Yesterday I was at a County Council

14        meeting in Valusha County, Florida. The

15        president of Verified Voting for Valusha, County

16        said on the record, people with disabilities have

17        75% privacy and 75% secrecy. What more do we

18        want?

19                    I was not able to ask her what is

20        75% privacy. I wasn't able to ask her what is

21        75% percent of secrecy at this.

22                    But that statement reflects an

1         attitude that three Federal laws have tried to

2         remedy, that we are treated like 75% of a

3         citizen. And this is going to stop.

4                     And HAVA gives us the tools to stop

5         it.

6                     I want to commend the Commission

7         because this version of the standards is

8         considerably improved. But there are still some

9         things that need to be done.

10                    In the standard there is I think a

11        structural problem with how you have handled

12        disability that in 28 years of disability policy

13        I have never seen anything structured like this.

14                    Usually the disability sections are

15        in one place. Makes it clear; it makes it fair;

16        it's easy for industry.

17                    In this version of the standards, we

18        are scattered all over creation and gone.

19                    And one result of that is that the

20        standard says that a voter who is blind shall

21        have a right to vote privately and independently.

22        In another section it says that a voter who is

1         paralyzed doesn't have that same right.

2                     Now, I hope that that was just a

3         piece of confusion.

4                     The ADA makes it crystal clear that

5         when we are talking about disabilities, we are

6         talking about all disabilities. HAVA makes it

7         crystal clear that it's the ADA that defines what

8         disability is.

9                     And we cannot have standards for

10        voting systems that allow one group of disabled

11        citizens to vote privately and independently and

12        another group not to do that.

13                    It's interesting to me that we just

14        had two examples of this voter verified paper

15        trail Neanderthal nonsense, and it is nonsense.

16                    I understand that state legislatures

17        have passed laws. We are comfortable with what

18        Nevada has done, because in Nevada the piece of

19        paper is not the ballot.

20                    We got real problems with New York

21        because the piece of paper is the ballot and I

22        know of no way that you are going to make that

1         piece of paper accessible with current

2         technology.

3                     I'm going to say something that has

4         been on my mind for a long time that I have not

5         said before.

6                     We have got this paper trail thing

7         because there is this small segment of the

8         computer world that's attitude is, you can't

9         trust computers to do anything. I think of them

10        as the cunieform school of computer science.

11                    They cooked up an idea of a voter

12        verified paper trail without testing it, without

13        measuring it, without seeing if it will work.

14        It's supposed to give the voter verification that

15        the vote was casted right. That's great. That's

16        a good idea.

17                    But let's put something in the

18        polling place that actually will do that.

19                    Professor Selker (ph) at MIT just

20        did a study, a scientific test, the first one,

21        that took computers and did this fearmongering

22        nonsense of the ballot is going to change from

1         the machine to the paper. So he had MIT graduate

2         students and ungraduate students vote.

3                     And the computer did eat their vote.

4         This isn't real. And the paper printed an error.

5         Okay. So the fearmongers, this is the theory

6         that the fearmongers have said is going to

7         happen.

8         7% of the voters in the test found

9         the error on the paper - 7%. When he was there.

10        And these are MIT graduate and undergraduate

11        students.

12                    Now, how effective a form of

13        verification is that going to be for the public

14        at large?

15                    Professor Selker also used audio

16        verification for the same voters. Same thing

17        happened, computer ate the vote, the paper

18        printed something different. 85% found the error

19        when it was audio.

20                    Now I am tickled with that. I am

21        blind. I live by audio.

22                    I am not going out saying we need to

1         pass a law that says let's put audio into the

2         polling place.

3                     I think the Commission could do the

4         nation a service, a great service, by making

5         clear that we need standards that are based on

6         science and data, not on theory that generates

7         contracts.

8                     HAVA is very clear - private and

9         independent by 2006. And the reason for 2006

10        it's 21 years and we do not have wheelchair

11        accessibility, it's 12 years and we are not being

12        registered to vote under law.

13                    There are some -- If a place wants

14        to have the verified piece of paper on the side

15        of the machine, that's fine, as long as it's not

16        the ballot.

17                    When it becomes the ballot, it is

18        not accessible. It does not meet HAVA's

19        criteria. It does not meet the ADA. You can't

20        spend Federal money on it.

21                    And I hope the Commission will make

22        that clear.

1                     Thank you very much.

2                     CHAIR HILLMAN: Thank you, Mr.

3         Dickson.

4                     And we will now have Dr. Rubin.

5                     DR. RUBIN: Thank you, Madam Chair

6         and Commissioners. Good afternoon.

7                     My name is Avi Rubin and I am a

8         computer science professor at Johns Hopkins

9         University. My area of expertise is computer

10        security. And I have been doing research in

11        industry and in academia on computer security for

12        14 years now.

13                    Since 1997 I have been applying

14        computer security towards studying voting systems

15        in elections and in 2004 I became a poll worker

16        working as an election judge in Baltimore County

17        in the primary and the general election.

18                    Election security is part of

19        national security. We look around the world and

20        we see just how important and sensitive the

21        ability to vote with integrity is.

22                    And I believe that many states have

1         come a long way since the Presidential election

2         in improving of the security of the voting

3         process. But I also think a lot of work remains

4         to be done.

5                     As somebody who specializes in

6         security, I view the world through a special lens

7         of being interested in security. And I imagine

8         the reason I'm here today to talk to you is to

9         talk about security issues, that is what I do.

10                    Reading through the proposed

11        guidelines, I have several comments about them

12        and most of them are very detailed and somewhat

13        technical and I'll be putting them in writing and

14        giving them to you.

15                    I only had a chance in the last few

16        days to read them very thoroughly and I have

17        about four or five pages of notes to give you on

18        those, which I want to clean up.

19                    But I think several important things

20        should come out now. And one of them is a

21        distinction that I would like to make, which is,

22        to distinguish between a voter verified paper

1         audit trail, VVPAT, which is discussed in the

2         guidelines, and a voter verified paper ballot.

3                     These things, when we distinguish

4         them, both of them can be viewed as a voter

5         verified paper record, which is something that

6         can be used to audit an election.

7                     But paper ballots are very, very

8         different from paper audit trails. And I think

9         if that distinction were made explicit rather

10        than implicit - clearly everyone understands that

11        there is a distinction - but if it were made

12        explicit, it might make the standards read a

13        little more cleanly.

14                    And, in particular, I am concerned

15        about the possibility that a voter verified paper

16        ballot, which I view as a more secure system,

17        might not meet the standard because of the

18        wording specifying a voter verified paper audit

19        trail.

20                    And so I think looking at the

21        standards again with an eye towards separating

22        out those two terms would be a useful thing to

1         do.

2                     I am in the class of computer

3         scientists who believe that a fully electronic

4         system that's made up of software can be made to

5         do whatever the programmer does.

6                     I have conducted many experiments

7         with my graduate students; I have written a lot

8         of software myself. And I know that when you

9         program a machine, that machine will do whatever

10        you program it to, with a caveat - a big caveat -

11        that there will always be a lot of bugs that you

12        didn't intend, no matter how good a programmer

13        you are.

14                    Microsoft is a very well resourced

15        company, puts a lot of effort into making

16        software and they do a great job. But they

17        release security updates once a month, if not

18        more often, because it's impossible to write

19        software without having bugs in it; that's just a

20        fact of life.

21                    So, as a result, many - especially

22        computer scientists - but many other individuals

1         as well, have thought about what can we do about

2         the fact that one of the tools that we have at

3         our disposal, which is electronics and computers,

4         is not entirely reliable from a security point of

5         view.

6                     We can still utilize it. Computers

7         are great. I use them every chance I get for

8         everything that I can think of. But I know their

9         limitations.

10                    And one of their limitations is that

11        it's very, very difficult for an outside auditor

12        or inspector to look at a large software package

13        and know everything about what it does. And it's

14        very, very easy for someone to program some other

15        functionality.

16                    I have papers and experiments that I

17        have done to demonstrate that.

18                    So what do we do? We say let's come

19        up with this independent dual verification.

20                    When I read this proposed guidelines

21        I loved that section. I think that's the right

22        idea, let's come up with totally independent dual

1         verification.

2                     My feeling is that the best one of

3         those options is paper, but not necessarily the

4         reel-to-reel variety.

5                     Let me outline some of what I view

6         as the disadvantages of the reel-to-reel VVPAT

7         over the verified paper ballots, which are

8         another form of paper.

9                     I think that we have to always keep

10        in mind what the purpose of the paper was. The

11        purpose of the paper is that we don't necessarily

12        trust these machines, so we want to do manual

13        audits.

14                    It's important that we not only do

15        manual audits, but random audits.

16                    If we do that and they match, we can

17        have confidence that wholesale fraud is not

18        likely to have taken place. With that in mind, I

19        think that the reel-to-reel systems place a

20        tremendous burden on election officials and

21        voters who have to perform those manual counts.

22                    That's because I also don't believe

1         that there should be anything that's not humanly

2         readable on the ballot. I don't like bar codes

3         and that's a controversial statement. Even

4         within the verified voting community, there is

5         not uniform agreement.

6                     The simple reason that I feel that

7         way is that as a voter I am a plain and simple

8         person and I see a mark that I can't read on my

9         ballot, if I'm suspicious, how do I know that's

10        not my name? It's probably not, right. But we

11        should have a voting system that is completely

12        transparent to the average citizen.

13                    If you think about an oppressive

14        government that is out to get you - which is kind

15        of the point of a democracy is to avoid that

16        being possible - then we don't want to have a

17        marking on the ballot that citizens don't know

18        what it is.

19                    And if we don't put bar codes and

20        other markings on the ballots in a reel-to-reel

21        system, they are very, very difficult to count

22        manually, as was stated earlier.

1                     That's why I think that what we

2         really want to have are voter verified paper

3         ballot systems where the ballot itself is

4         something like what we might find in an absentee

5         ballot, something that can be marked with a

6         machine with accessibility features, can be

7         verified with machines with accessibility

8         features, or visually or through audio.

9                     But what I worry about with the

10        reel-to-reel systems is that the path of least

11        resistance for an election worker in a busy

12        hectic election is not going to be to do the

13        manual recount; they'll skimp on it. And I think

14        we should design voting systems where the path of

15        least resistance is the best possible path.

16                    Not saying that poll workers won't

17        do it. I'm saying we should make it as easy as

18        possible.

19                    I disagree with my fellow panelists

20        about the importance of the privacy in the

21        reel-to-reel. I don't believe in reel-to-reel

22        systems that are chopped up because of the

1         reasons that were mentioned. But I also don't

2         like the idea of the reel keeping the order,

3         preserving the order.

4                     If we are suspicious of the

5         government - in the poll site where I worked, we

6         assigned people to the stations - so I think

7         voters would accept a situation, even if it

8         wasn't the right procedure, where poll workers

9         assigned them to the machines. And at that point

10        we have got a list of how those people voted in

11        that paper trail.

12                    I believe that reel-to-reel VVPAT on

13        DRE's are the unfortunate product of a

14        misunderstanding of the arguments against

15        paperless voting. But I do believe with a lot of

16        effort and vigilance they can produce a more

17        secure election than without a paper trail at

18        all.

19                    I would like to give some guidelines

20        for maximizing the security and auditability of

21        elections regardless of what kind of paper trials

22        that are used.

1                     Random audits have to take place

2         where the tallies are compared to electronic

3         totals. These have to be like truly random. I

4         am thinking like ping pong balls coming out of a

5         machine like you have in the lottery that nobody

6         could have known in advance which sites would be

7         picked.

8                     If somebody could know in advance

9         which sites will be picked for the audit, that

10        manual recount is losing most of its value.

11                    I think that the manual audits

12        should be done by hand, where possible.

13                    And the most important point that I

14        can make about this is that we have to have pre-

15        established, in advance, before the election,

16        procedures for what to do in a discrepancy.

17                    If it's the case that we get a

18        different tally on the paper than we have in the

19        electronic machine, what to do.

20                    And you have to keep in mind that it

21        may be the case that the paper ballots show

22        something that's statistically ridiculous

1         compared to the expectation and the electronic

2         tally matches the expectation.

3                     If that were to happen, there has to

4         be a procedure.

5                     Also it may be the case that the

6         paper matches the expectation and the election is

7         all for one candidate, which is impossible.

8                     Whatever it is that we do, it has to

9         be established in advance what the procedure is

10        for handling it.

11                    I think it's important to have

12        comprehensive security reviews of the system in

13        addition to guidance or standard compliance

14        testing. If the ITA's are simply looking at the

15        standards or guidelines document that you

16        produced and matching it up with that system, say

17        yes it meets those, that's not enough. I think

18        you want to have a red team or a security review

19        that's required to go in and do a risk assessment

20        of the system in the face of various different

21        kinds of attacks. And I think that that should

22        be mandated.

1                     I know that I am using up a lot of

2         time and you already have my written comments.

3                     So I think that I will just wrap up

4         with a couple of things.

5                     Let me say that the properties that

6         I would like to see a voting system that utilizes

7         paper have are that it should maximize the

8         probability that the voters will actually verify

9         their vote. And one way to do that is to have

10        them use a marking system that produces a piece

11        of paper that they then have to scan.

12                    It's important that the order of the

13        votes in the paper trail be randomized. And the

14        procedures in place for what to do in the case of

15        a discrepancy, as I just said.

16                    It's important that every step of

17        the way be accessible to disabled voters, such as

18        blind voters and deaf voters.

19                    And, in my opinion, the paper record

20        should be the authoritative vote.

21                    Now, let me just talk about two or

22        three things that I found in the standards

1         themselves.

2                     One of them that troubled me the

3         most was the requirement that there be real time

4         clocks in the machines.

5                     Maybe there is a benefit to audit,

6         but I think that has to be weighed against the

7         security risks of - if I was an attacker and I

8         wanted to write malicious code for a voting

9         machine, the clock would be a great help to me.

10        It would let me know if I was in an election or

11        if I was in a test.

12                    The other thing is that I read in

13        there that there should be a time stamp on every

14        record in the machine. Well, that is now an

15        electronic version of the reel-to-reel that

16        preserves the order in which people voted. And I

17        think that needs to be weighed against the

18        possible benefit of the audit.

19                    There were other technical things

20        that I found that I didn't think were actually

21        enforceable, like the requirement to know what

22        the values and the registers and the various

1         variables are at setup time. I'm not sure

2         technically how that would be accomplished.

3                     And I agree with the statement that

4         five people in the language minimum doesn't do

5         much if you only have three people in a precinct

6         that vote in a particular language. What do you

7         do?

8                     So, I will write up all of my

9         specific comments for you that I had mostly with

10        respect to security. And I'll wrap it up with

11        that.

12                    CHAIR HILLMAN: Thank you so much.

13                    Okay. In keeping with our schedule

14        for today, we will begin the public comment

15        period at 4:30. However, once again, we are not

16        likely to take a break between this panel and the

17        public comment period.

18                    So, Commissioners, in keeping with

19        that, we will have to stick to our ten minute

20        allocation of time so that we can do justice to

21        both this panel and those individuals who are

22        signed up for the public comment period.

1                     And I am going to give myself a

2         treat and start with the line of questioning so I

3         can get all of mine in this time.

4                     Professor Rubin, when I first became

5         aware of the Election Assistance Commission

6         having been organized under the Help America Vote

7         Act, I did not hear much, a little bit of

8         conversation about paper trails. And even then,

9         I'm not so sure the term V V P A T had been

10        codified.

11                    I think people were sort of

12        struggling with what to call it. Different

13        things mean different things.

14                    And I really appreciate in your

15        testimony the way you have sort of identified

16        three different terms for verified paper.

17                    What has evolved in the past two

18        years around this discussion, this discussion

19        being verified paper, and what of that has

20        surprised you, the change in the conversation

21        from where we were at this time 2003 to where we

22        are today?

1                     DR. RUBIN: You know, it's true.

2         When I started working in elections, I was

3         working with the country of Costa Rica. They

4         asked us to help them design an electronic voting

5         system.

6                     And I came to the problem initially

7         from the point of view, well, that's a great

8         application, we can use our computer security

9         knowledge to do that.

10                    And every time when you do security

11        work, you always do it in the face of a

12        particular adversary. So let's say the average

13        adversary does this or does this or does that.

14                    And every time we tried to design a

15        system that would be fully electronic and secure,

16        we hit a wall. Because we just couldn't figure

17        out what to do if the people who programmed the

18        computers themselves were the adversaries.

19                    The idea of a paper trial came along

20        because it's something that's not part of the

21        electronic system.

22                    So in order to cheat, it would

1         require much, much, much more effort and what I

2         think is a retail level of fraud, than it would

3         in an electronic system.

4                     What I have seen involved is I have

5         seen conferences and workshops - many of which I

6         have attended in the computer science community -

7         where people have tried to come up with solutions

8         that wouldn't require paper. And some of them

9         are very novel - the cryptographic schemes. And

10        I'm a big fan of those schemes.

11                    However, along with all the other

12        requirements of accessibility, security,

13        auditability, is one of transparency. And I

14        think that's where some of these schemes fall a

15        little short.

16                    What I think has evolved is the

17        understanding that we need to move with the

18        speed - not move too quickly before we understand

19        how to assess the risks that we have.

20                    The reason I really enjoyed the dual

21        verification section here is that I think it's

22        exactly the right idea. That if we had --

1                     The more independent ways we have of

2         recording and storing the votes and truly

3         independent - meaning you don't take it into the

4         computer and then store it in audio and then

5         store it in hard drive somewhere - you actually

6         have to capture them separately, like the camera

7         idea does.

8                     I think the more we open our minds

9         and look to ways of doing this, the more hope

10        there is to have good auditable elections.

11                    I think that paper is the one that

12        we understand the best and the one that average

13        citizens understand the best.

14                    If you have got a camera taking a

15        picture of the screen, of your confirmation

16        screen, one, was that camera controlled by the

17        software that's the software that you are worried

18        about or is it control by something else?

19                    How do you store these; how do you

20        do the manual recounts?

21                    These are all things that need to be

22        explored, not to mention the possible reflection

1         of the voter off of the screen and people's

2         paranoia about having a camera in the voting

3         booth.

4                     But when I read that section, I

5         think, now people are thinking about this the

6         right way: how do you verify an audit

7         independently from the electronic tally.

8                     CHAIR HILLMAN: Thank you. And I

9         have to say that you must be a pure scientist at

10        heart, because nobody else said that they have

11        enjoyed reading any section of this document.

12                    So for you to say you enjoyed the

13        section, makes me feel quite giddy. Thank you

14        very much.

15                    Mr. Lomax, how many times has Nevada

16        used its new system with the verified paper?

17                    MR. LOMAX: We have used it for four

18        elections, it would be the primary and the

19        general election in 2004 and then we used it for

20        the primary and general municipal elections in

21        2005.

22                    CHAIR HILLMAN: And you were the

1         administrator over all four elections, were you?

2                     MR. LOMAX: Yes, I was.

3                     CHAIR HILLMAN: Of the errors that

4         occurred, what percentage of those would be

5         assigned to human error?

6                     MR. LOMAX: I'm not sure what you

7         mean by errors that occurred.

8                     CHAIR HILLMAN: Whether, you know,

9         we are talking about jammed paper, something not

10        quite --

11                    MR. LOMAX: Errors related to the

12        printer, is that what we are talking about?

13                    CHAIR HILLMAN: Yes, yes.

14                    MR. LOMAXL There were no errors in

15        the sense that the paper record did not reflect

16        what the electronic record did, either as

17        presented to us by some voter complaining or when

18        we did our random audits. It always matches.

19                    Out of the half a million

20        essentially records that were or ballots were

21        cast and then records that were created, there

22        were less than 25 instances where there was a

1         paper jam in which some portion of that paper

2         record was unreadable because of the way it got

3         caught up. We didn't lose the whole ballot but

4         we lost some portion of it.

5                     So that comes out to one in every

6         20,000.

7                     And that also was almost always as a

8         result of human error in the sense of the way the

9         people that threaded the paper in the printer.

10        The printer is unbelievably reliable, what we are

11        using. I mean, it works. But you do have to

12        thread the paper in there correctly.

13                    So what we found when we had paper

14        problems - and they were very rare to begin

15        with - but they were generally training problems

16        because the paper hadn't been put in there

17        correctly. So it is all human error is what is

18        comes down to.

19                    CHAIR HILLMAN: Okay, thank you.

20                    Mr. Kosinski, I believe you said

21        that the State of New York has always used a

22        lever machine pretty much. There may be some

1         areas using a paper ballot, but punch cards or

2         optical scans have you used?

3                     MR. KOSINSKI: We have used punch

4         cards and optical scans for absentee voting,

5         central count voting only.

6                     CHAIR HILLMAN: So for the voters of

7         New York, an experience with either an optical

8         scan or a touch screen with verified paper would

9         be a first time experience that they would be

10        voting on something at a polling place in a

11        voting booth that would produce a piece of paper,

12        whether marking a ballot or an optical scan or

13        verification on an electronic machine; is that

14        correct?

15                    MR. KOSINSKI: Yes, yes, that's

16        correct.

17                    CHAIR HILLMAN: Have the draft

18        guidelines that you have seen and today's

19        discussion helped you in your thinking about what

20        New York is going to be confronting with the

21        transition?

22                    And I'm not going to talk about the

1         delays or 2006 or '07, or just whatever length of

2         time it takes for New York to get to the point

3         where it replaces the lever machine.

4                     MR. KOSINSKI: I think it helps. I

5         think, however, a lot of what we are faced with

6         in New York is what you have probably identified,

7         which is this is something new to our voters; it

8         is also something new to our election workers,

9         you know.

10                    So that without having any

11        experience in this area, it's very hard to know

12        really how these guidelines will help us until we

13        actually start to use it and see how people react

14        to it.

15                    The guidelines may assist us in

16        developing a system that will work, technically

17        work. But I think then you are faced with a

18        system that has to work with the people that you

19        are going to use it on.

20                    And whether that is inspectors that

21        have to use it or it's voters that have to use it

22        for the very first time, that's obviously an

1         unknown factor.

2                     And I think that the primary thing

3         that we are looking at in that area is the

4         training and the education of the voters prior to

5         the systems being introduced into our state.

6                     CHAIR HILLMAN: Thank you.

7                     Mr. Dickson, you will indulge me. I

8         really don't have any questions for you.

9                     I had the opportunity to hear your

10        presentations and, as usual, you are clear and

11        right to the point. And there is not much

12        confusion in my mind over what you had to say.

13        Thank you.

14                    Mr. Vice Chairman.

15                    VICE CHAIR DeGREGORIO: Thank you,

16        Madam Chair. I think we have a very diverse

17        group of folks in front of us. We have Mr.

18        Kosinski, who is going to be new to VVPAT and his

19        state hasn't used it yet; Larry Lomax has had

20        half a million voters use it; we have Mr. Rubin

21        who advocates for it; and Mr. Dickson who is

22        opposed to it.

1                     So we have had a very diverse group

2         of opinions.

3                     Let me ask Mr. Lomax, though, some

4         technical questions about this and about your

5         system.

6                     First let's talk about the font

7         size, because you suggest that a 10 point font

8         size - using Times New Roman as the font, I

9         assume - your statement here is in 12, I'm pretty

10        sure it's 12. If I have to look at a 10 I have

11        to make sure that I've got these on and the

12        bottom part of my glasses on. It still seems

13        pretty small to me.

14                    Is there any evidence, when you have

15        such a small font size, that people will just

16        ignore the VVPAT and not even look at it because

17        it's too small, whereas if the font size was

18        larger, if it was 15 or 20, then they might even

19        look at this and pay attention to it and compare,

20        whereas if it's a small font size, they might --

21                    Is there any evidence or anything

22        that you have looked at --

1                     MR. LOMAX: I have no data that

2         would reflect that.

3                     I do know that in Clark County it

4         had been advertised, so to speak, that the voters

5         did not look at the paper trail.

6                     There was a CD made - after we

7         caught somebody filming up there for a couple of

8         hours and they got thrown out - they did their

9         own analysis and that was their determination.

10                    But I would point out that in Clark

11        County we have been doing DRE voting since 1996.

12        So adding the paper trail to the touch screen

13        machine was not nearly as revolutionary as it's

14        going to be, for instance, in New York, when they

15        go from a paper system to this whole setup, the

16        electronic voting and the printer.

17                    So our voters were very comfortable

18        with the electronic machines to begin with. This

19        paper trail issue was not a big deal in Clark

20        County.

21                    So I think a lot of the voters, in

22        fact, in a lot of polling places we had a lot of

1         our older machines without the printer and our

2         new ones with the printer, the majority of voters

3         chose to use the old machine, just because they

4         were comfortable with that and they didn't have

5         the printer.

6                     They had their choice which machine

7         they wanted to use.

8                     So I think we had a comfort level

9         that makes us, perhaps, a bad example to relate

10        to how many people look at the printer.

11                    I don't doubt, though, that the

12        smaller font might discourage some people from

13        looking.

14                    VICE CHAIR DeGREGORIO: Because I

15        was in your warehouse in February '04 and saw

16        your two different systems. And what you just

17        said here confirmed what I thought, you have two

18        systems there and voters were given a choice, if

19        they wanted to use the one with the verification

20        of the paper and ones that did not have that.

21                    What percent used the verification?

22                    MR. LOMAX: As I said before, we do

1         early voting for two weeks. During the early

2         voting period we only use the machines that have

3         the paper trail printer.

4                     So during that time period 271,000

5         people voted using the printer.

6                     On Election Day we had at least two

7         of the paper trail machines in every polling

8         places. And in some polling places we had two of

9         those and 20 of the others.

10                    But on Election Day, about 70,000

11        people chose to use the paper trail machine and

12        that makes about 150,000 that chose to use the

13        other one.

14                    VICE CHAIR DeGREGORIO: You allow

15        observers in your polling stations, party

16        observers or somebody who can sit there all day?

17                    MR. LOMAX: In the Presidential

18        election, we had up to 14 poll watchers from the

19        various parties.

20                    VICE CHAIR DeGREGORIO: So someone

21        could come in and make a record of every single

22        person who voted and the time they voted and

1         the --

2                     MR. LOMAX: You can say that's true,

3         but that's absolutely not true.

4                     First of all, if you come into one

5         of our polling places, you don't know who the

6         people are as they go up.

7                     We vote multiple precincts in a

8         polling place. Just because you are standing in

9         there doesn't mean you have the foggiest idea who

10        all these people are going off to vote. They are

11        going off to all the different voting machines on

12        their own.

13                    VICE CHAIR DeGREGORIO: Well, some

14        states have the requirement that the political

15        party observers or the candidate observers can

16        sit behind the poll workers and actually have to

17        be told the name of the person that is coming to

18        vote.

19                    It may not be true in your state,

20        but some states have that requirement.

21                    I just wanted to see if Nevada had

22        that requirement.

1                     Let make get to a couple of things.

2                     You make a very good suggestion here

3         on this reel-to-reel issue where you have two in

4         one polling place and people go use either one

5         and there is no way to track the sequence of

6         voting and who voted.

7                     And that is, you know, a reasonable

8         solution to allow a reel-to-reel that prevents

9         anyone from - and allows, I think, the privacy of

10        the voter - that you can't determine someone's

11        vote by looking at these pieces of paper

12        afterwards because you have two and you are using

13        both of them.

14                    It seems like a very reasonable

15        accommodation.

16                    But if you did have one, just one at

17        a polling station, and it was reel-to-reel and it

18        was a state that required or allowed some

19        observer or someone to sit there and take a name

20        of the person and then in some kind of a

21        discovery in a recount or election contest, we

22        can actually go in then and determine who voted

1         for whom.

2                     MR. LOMAX: I would never argue

3         that.

4                     VICE CHAIR DeGREGORIO: That could

5         be problematic.

6                     MR. LOMAX: I mean, you have to have

7         appropriate administrative procedures.

8                     As I said before with the absentee

9         ballots, the ballot comes with the voter's name

10        on it.

11                    You have to have procedures that you

12        break it apart so you can't track it.

13                    You can overcome any of these

14        systems, depending on what you want to do.

15                    VICE CHAIR DeGREGORIO: Thank you.

16                    Let me ask the Professor here, who

17        does not believe in a reel-to-reel system - has

18        made that clear - but believes in paper.

19                    And I appreciate your discussion

20        here about trying to make a distinction between

21        the VVPAT and what is a ballot and what is a

22        non-ballot, and that is helpful.

1                     Are you suggesting here, Dr. Rubin,

2         though, an optical scan type of system would fit

3         the definition of the VVPB, the voter verified

4         paper ballot, marked by the voter, so that

5         definition would - an optical scan system would

6         qualify to that?

7                     DR. RUBIN: That's right. However,

8         it doesn't mean that the optical scan ballot is

9         necessarily filled out by hand.

10                    VICE CHAIR DeGREGORIO: Right.

11                    DR. RUBIN: But the nice thing is

12        that whatever machine is used isn't part of the

13        trusted computing base of the system, so it

14        doesn't have to be trusted.

15                    VICE CHAIR DeGREGORIO: Do you have

16        any comments on font sizes that's in these

17        guidelines when it comes to the paper trail?

18        Have you looked at this issue? Have your

19        students done any studies or anything like that?

20                    DR. RUBIN: No, we have not looked

21        at that issue and the human factors.

22                    VICE CHAIR DeGREGORIO: Jim, I want

1         to second our Chair's comments. You are a very

2         articulate advocate, spokesperson.

3                     MR. DICKSON: I want to, if I

4         could --

5                     CHAIR HILLMAN: He wants us to ask

6         him a question, so, Jim, please.

7                     VICE CHAIR DeGREGORIO: I am going

8         to allow you the opportunity to comment on that.

9                     MR. DICKSON: I really appreciate

10        the effort that Larry and his colleague took with

11        the magnifying glass and I'm sure that was

12        helpful for some.

13                    But, again, we have got to really

14        look at fact and reality. The reality is that

15        sight loss and those that use/need magnification,

16        there isn't a single magnifying glass that's

17        going to work for everybody. There are too many

18        variables about the light, where the damage in

19        the eye is.

20                    So it is great to have it there, but

21        we need to really understand that it will fit a

22        few, but it's not going to fit most.

1                     And in terms of the size of the

2         population that we are talking about, for every

3         one person who you can tell is blind because we

4         use dogs or canes or are essentially totally

5         blind, there are nine whose vision is such that

6         they can't read print. So it's a lot of people.

7                     CHAIR HILLMAN: Thank you.

8                     Commissioner Martinez.

9                     COMMISSIONER MARTINEZ: Thank you,

10        Madam Chair.

11                    I will be brief in my questions.

12                    I guess I want to start by

13        reiterating, at least for the purpose of

14        discussion, just what this panel is not.

15                    This panel is not a discussion about

16        the security or lack thereof of DRE systems. I

17        think that discussion has to continue. I have

18        certainly had discussions with Professor Avi

19        Rubin about continuing to engage - who is

20        actually, you have been, I think, very accessible

21        to the EAC in talking about that particular

22        issue.

1                     But, again, we are not here to

2         discuss the security or lack thereof of

3         electronic voting systems.

4                     We are not here, either, to discuss

5         the pro's and con's as to whether VVPAT is an

6         appropriate method to increase voter confidence

7         or to increase security.

8                     The position that the EAC has taken

9         is that that is a decision left solely to the

10        discretion of state and local governments.

11                    What we felt was appropriate was,

12        given the fact that at least 16 jurisdictions -

13        and I guess I am learning, Mr. Kosinski, that in

14        this legislation that is pending here in New

15        York - that perhaps 17 states now, will be

16        requiring - either through legislative action or

17        administrative rule - that those states use VVPAT

18        by a date certain, most of which are 1/1/06.

19                    Though some states, like Colorado,

20        which recently passed a VVPAT requirement, says

21        they don't have to implement it until 1/1/08.

22                    And so, from my perspective, that is

1         a decision appropriately left to state and local

2         jurisdictions. And they, in turn, will have to

3         be held accountable to the citizens that they

4         serve.

5                     Having said that, having stated from

6         my perspective what this panel is not, I want to

7         start, if I could, Mr. Lomax, with just a general

8         perspective.

9                     You are, among other things that you

10        do, you are obviously the administrator there in,

11        is it Clark County?

12                    MR. LOMAX: Yes.

13                    COMMISSIONER MARTINEZ: -- in

14        Nevada, of the elections.

15                    And I wonder if you can give us the

16        perspective of before you had VVPAT and where you

17        stand now, in terms of just a personal

18        perspective about what purpose it serves for you

19        as a local election administrator.

20                    MR. LOMAX: Prior to having the

21        VVPAT, every election there was a small - as it

22        is nationally - small but very vocal minority

1         that complained to us, sent me e-mails, telephone

2         calls about the lack or of a way to verify

3         electronic voting.

4                     None of this occurred prior to

5         Florida, but a lot of it occurred after Florida.

6                     Since we have the VVPAT, all of that

7         has gone away. I didn't have any complaints in

8         the last election.

9                     So I would still suggest that the

10        great majority of people in Clark County really

11        didn't care one way or the other.

12                    But for that small minority of

13        people who passionately did care, it's made a big

14        difference. They have been very comfortable and

15        we really had no issues related to voting

16        machines in the last election.

17                    COMMISSIONER MARTINEZ: Is there a

18        state law in Nevada, a requirement that you audit

19        a certain number of the paper records to the

20        electronic ballot, because that's what it is, I

21        assume?

22                    MR. LOMAX: The Secretary of State

1         has put out guidelines in what we call the

2         Administrative Code. And what he requires, that

3         1% of the machines after the election are

4         randomly selected. And what we do then is we had

5         teams of people manually recount the paper

6         records and we match those results against

7         electronically recorded results from those

8         selected machines.

9                     COMMISSIONER MARTINEZ: Does the

10        paper that is produced with respect to the VVPAT

11        have any legal standing?

12                    I know it's not the official ballot,

13        but is it the official record for recount

14        purposes in Nevada?

15                    MR. LOMAX: No, it is not. Right

16        now our recount procedures are all based upon the

17        electronic results.

18                    COMMISSIONER MARTINEZ: Professor

19        Rubin, the Vice Chair alluded to this a few

20        minutes ago, but Mr. Lomax has talked about some

21        of the administrative protocols that they

22        implement to ensure that the reel-to-reel aspect

1         of this VVPAT technology that they use in Clark

2         County accounts for trying to maintain the

3         privacy of the voter with the use of at least two

4         machines. And I guess they don't have a sign-in

5         requirement; they don't keep a record of exactly

6         where each voter signed in, et cetera.

7                     Does that help to alleviate, from

8         your perspective, some of your concerns about

9         reel-to-reel technology?

10                    DR. RUBIN: To some degree. But I'm

11        still concerned about the possibility, not of

12        being able to match every voter with their vote,

13        but of being able to target a particular person.

14                    If you are an observer and you want

15        to know how did that person vote and you get to

16        participate in a recount and you watch when they

17        came in and how many people went before them, you

18        will be able to tell how they voted.

19                    I think the privacy you want is

20        having a sequential record of the order of which

21        people voted is a bad idea, because I think we

22        can have voting systems that are just as good in

1         every possible way that don't do that.

2                     COMMISSIONER MARTINEZ: Thank you.

3                     Mr. Dickson, I do have a question

4         for you, so you can have another shot at giving

5         us what I think is a very eloquent perspective.

6                     I guess, from my - and, Jim, you and

7         I have had many discussions over the course of my

8         short time on this Commission, and certainly over

9         the past few months - and I think that the

10        dilemma for me, obviously, is that I certainly do

11        recognize the language that has been written very

12        intentionally by Congress.

13                    And we have to keep in mind that

14        HAVA was passed overwhelmingly in both Houses of

15        Congress and signed quickly by President Bush

16        back on October 29th of 2002.

17                    There is language, clearly, that

18        represents significant civil rights strides with

19        persons with disabilities when it comes to being

20        able to vote.

21                    So I certainly recognize and give

22        due credit to the eloquence that you bring to the

1         table at all times in reminding us of that

2         particular language.

3                     I guess my question is also, it

4         plays along those lines. And that is, as

5         somebody who has to help determine an

6         interpretation of an important Federal statute,

7         like the Help America Vote Act, is it

8         appropriate, from your perspective, for us to

9         temper our decision or be influenced by the

10        technology, the capabilities that currently

11        exist, from a technology perspective?

12                    MR. DICKSON: I think you have to.

13        We have a deadline that is the law and it was put

14        there consciously. We can't--

15                    There is a notion abroad in the land

16        that, well, we can just wait until something

17        better comes.

18                    And I think it would be very helpful

19        if the Commission would say, very clearly, you

20        know, you got to buy it by the first of the year,

21        it has to be accessible. You can't wait for

22        something that's better.

1                     I also think that there is an

2         underlying assumption by those who say we got

3         something better coming down the road, and that

4         is that there is going to be money to buy that

5         something better.

6                     Well, I have been working for 25

7         years to try to get government to put more money

8         into elections. We have got one pot of money. I

9         hope there will be additional money coming. I

10        will certainly do - and the disability community

11        will do - everything we can to get the additional

12        money to election officials.

13                    But you got to buy what's available

14        now is absolutely clear. And we can't be waiting

15        for something in the future.