TESTIMONY OF JOYCE E. THOMANN
On Behalf of members of the
MARYLAND FEDERATION OF REPUBLICAN WOMEN
the
REPUBLICAN WOMEN OF ANNE ARUNDEL COUNTY
And herself as a Maryland Resident and Registered Voter
Re: Senate Bill 392
February 22, 2007
Before the Senate, Education, and Environmental Affairs Committee
Chairman Conway, Members of the Committee: On behalf of the 1,600 members of the Maryland Federation of Republican Women, the Republican Women of Anne Arundel County and myself, I want to thank you for the opportunity to express support for Senate Bill 392.
At its 2006 Fall Convention and 85th Anniversary celebration, the Maryland Federation of Republican Women unanimously adopted a resolution calling for the passage of legislation to require voter verified paper records for use in all Maryland elections.
It is easy to see from the co-sponsors on S.B. 392 that this is a bipartisan issue. We all want to insure that the integrity of each vote, the cornerstone of our Nation, is protected by making certain that all votes are counted as cast by the individual voter.
On January 30, 2007 Florida Governor Charlie Crist announced his intention to recommend that the controversial touch screen machines be scrapped and replaced with optical scanners. U.S. Representative Robert Wexler, Democrat, was quick to praise the Republican Governor for his “bold and comprehensive” plan and went on to say, “We are about to resolve, once and for all, the election integrity problem in Florida, and we are about to realize the dream of creating a paper trail for every voter in the state of Florida.”[1]
A past President of the Association for Computing Machinery (ACM)[2], Barbara Simons, in her testimony before the U.S. House of Representatives[3] Hearing on Electronic Voting Machines, said, “We all want elections that are reliable, secure, accessible, and trusted by the public. Given the known security risks, the possibility that software bugs could generate incorrect election results, or that computerized voting machines may fail during an election, we cannot trust that the results recorded in a paperless voting machine accurately reflect the will of the voters.” Her testimony went on to note, “Most computer professionals oppose paperless voting machines. . . In addition to the technical community, good government organizations have expressed concerns about the security of paperless voting machines. For example, at its 2006 national convention[4] the League of Women Voters passed a resolution on voting machines calling for a voter verified paper ballot or record that would be used for audits and recounts.”
Appearing to echo that comment is the December 2006 report of The National Institute of Standards and Technology[5] which states:
“But many people, especially in the computer engineering and security community, assert that DRE’s are vulnerable to undetectable errors as well as malicious software attacks because there is no audit mechanism other than what the DRE can report on: how many records it has stored, ballots styles, etc. Potentially, a single programmer could ‘rig’ a major election. The computer security community rejects the notion that DRE’s can be made secure, arguing that their design is inadequate to meet the requirements of voting and that they are vulnerable to large-scale errors and election fraud.”
Computer experts at the Center for Information Technology Policy and Department of Computer Science, Princeton University recently published their findings, “Security Analysis of the Diebold AccuVote-TS Voting Machine.”[6] The 4 Main Findings listed in the Executive Summary were summed up as, “many computer scientists doubt that paperless DREs can be made reliable and secure, and they expect that any failures of such systems would likely go undetected.
Dr. David Dill, a Stanford University professor of computer science and a member of the Common Cause panel on election problems noted that votes being switched to another candidate on the electronic ballot, or “vote flipping” had been reported all over the country during the 2006 elections.
New questions have been raised: “Why do Diebold’s Touch-Screen Voting Machines Have Built-In Wireless Infrared Data Transfer Ports?”[7] The lap-top computers which you are using transfer information and data via their Infrared Data Transfer Ports.
I have previously appeared before this Committee strongly supporting a return to the reliable, infinitely cheaper, election judge friendly Optic Scanner which uses a voter verified paper record. This paper record is independent of the machine in its creation and can be used to verify the accuracy of the machine itself as well as to conduct a verifiable recount.
I have been a Chief Election Judge for 26 years and have only missed serving in one election – the 2004 General Election.
As a Chief Election Judge, user of the equipment and a taxpayer, I am an avid supporter of the Optic Scanners for many reasons -- not the least being that ONE Optic Scanner can easily serve the needs of an entire precinct whether that precinct has 400 or 2,000 registered voters because the machine is NOT, what I will refer to as, “voter-linked.[8]” When an Optic Scanner is used, voters are free to mark their ballots in the “privacy booth”, on the floor of the precinct, on the walls, on a table, a chair, or on the backs of each other. They do not need the actual machine in order to mark their ballot. The only wait when the Optic Scanners were used was at the Book Judge check-in table – not at the voting machine itself. The new E-poll books provide for a faster check-in process.
The touch screen machines, or Direct Recording Electronic Devices (DRE’s) ARE “voter linked”. The State Board of Elections (SBE) presently mandates one DRE for every 200 registered voters in the precinct.
In the 2006 election, we had voters standing in lines to use the touch screen machines because they cannot complete a ballot independently of the machine itself. Anne Arundel County’s ballot was long with Federal, state and local offices, 3 Constitutional Amendments; 1 State-wide ballot issue and 3 Amendments to the Anne Arundel County Charter. In the 2006 General Election, one gentleman took over 30 minutes to cast his vote. The minimum time at each machine was 5 minutes per voter. With the Optic Scanners, the time needed to cast a vote ON THE MACHINE was less than 15 seconds. (Obviously a voter could take much longer to mark his/her own paper ballot – with no inconvenience to any other voter.)
In my native state of Colorado, touch screen machines are also used and thus are “voter-linked”. In the 2006 election many voters were not able to cast their votes until 1:00 and 2:00 A.M. These were voters who were already in lines that wrapped around the block before the mandatory 8:00 p.m. poll closing time and thus met the criteria for being able to vote. However, many voters, like my daughter-in-law, were not able to vote because they could not wait for a minimum of two hours in order to get to use a machine. This voter disenfranchisement is the consequence of machines which are “voter-linked”. Today, the City of Denver and surrounding jurisdictions are making plans to do away with the touch screen machines and move entirely to absentee ballots – which are voter verified paper ballots that can only be counted on an Optic Scanner.
Nationwide, the push by voters, by constituents, for voter verified paper records is increasing. Six of the 10 most populous states use or soon will be using voter-verified paper records throughout (CA, NY, IL, OH, MI and NJ). According to the December 1, 2006 report of the National Institute of Standards and Technology, a total of 35 states use voter-verified paper records throughout. Twenty-seven states MANDATE voter-verified paper records statewide; 8 don’t mandate them but use them statewide; 10 states use them on a county-by-county basis and 5 states use only DRE’s statewide (DE, GA, LA, MD and SC).
It is my view that the flood of absentee ballots used in Maryland’s 2006 General Election was a clear statement being made by thousands of voters – we don’t trust these machines you have forced on us.
Maryland’s State Elections Administrator Linda Lamone objects to having printers added to the touch screen machines. I AGREE with Ms. Lamone – but not for the same reasons.
Already the touch screen machines are a nightmare for election judges. We went from having one machine which was easily wheeled into the precinct, set up and monitored throughout the day, to having to deal with 7 to 11 machines. Each of these 55 pound machines must be handled four separate times in each election. Each of these 7 to 11 machines must be separately activated, individually monitored throughout the day and then separately deactivated when the polls close.
Before the touch screen machines were put into use, election judges could arrive at the polling place the day of the election at 6:00 a.m., have fresh coffee made and the polls ready to open at 7:00 a.m. It is now “suggested” that we go to go to the polling place a day ahead of time. Those of us who do this put in at least 3 hours the day before the election getting the machines off the transfer cart, legs pulled out and locked in place and each of the 55-pound machines set up on their legs so voters can use them. (The machines are not opened, powered up or activated until the day of the election. Normally, that activation operation, plus putting up required signage and setting up tables and chairs for election judges takes an hour with 6 people working feverishly to accomplish an on-time opening.)
In order to stem the hemorrhaging of election judges, Maryland’s families (through their taxes) are now paying Diebold employees (hired by the State Board of Elections) to come in to each precinct on election day to help with the set up and trouble shoot the machines throughout the day. I served as a candidate’s Poll Watcher during the January 30, 2007 Annapolis City Special Election. The Election Judges commended the Diebold employee for arriving at the Eastport Volunteer Fire Hall polling place prior to their arrival. I spent 13 hours at that polling place and, except for the two hours when the Diebold employee went to dinner, he remained in the polling place until after 9:00 p.m. Like the election judges he had a long day. Unlike the election judges, Maryland’s families (a/k/a taxpayers) most likely paid him and his counterparts at the Eastport Elementary School, American Legion Cook-Pinkney Post #141 and the Heritage Baptist Church polling places extremely well.
Many jurisdictions using the DRE’s have found the “technical support” costs can substantially eclipse the actual cost of the hardware.
I’m sure you agree that a part of your duty is to protect the taxpayers – the Maryland families whose family funds must ultimately pay the bills. I believe you will be very interested in a cost-comparison done between the DRE’s and the Optic Scanners in two Florida counties as well as in four North Carolina counties.
Using the DRE’s, the Sarasota County Supervisor of Elections’ average annual total yearly expenditures for the period 2001-2004, was $12.41 PER REGISTERED VOTER. Using Optical Scanners, Manatee County Florida’s average annual expenditures during the same period, which included paper ballots at 20 cents a piece, was $7.56. In sum, the DRE’s cost 64.1% more per year per registered voter than did the Optic Scanners.[9]
In the North Carolina comparison, “the cost of using touch screen voting or direct recording machines in Guilford and Mecklenburg county was about 30-40% higher than the cost of using optical scan equipment in Wake and Durham county. This means that not only are touch screens more expensive to acquire, they are also more expensive to operate year after year.”[10]
It is only because adding a printer to each of these machines would create an even greater nightmare for the election judges and a significantly larger burden on Maryland’s families (the taxpayers) that I agree with Ms. Lamone’s objection to them. (Perhaps in anticipation of the Maryland Legislature approving printers being added to the DRE’s, the SBE appears to be adjusting the number of required DRE’s down from one DRE for every 200 registered voters to one DRE for every 150 registered voters in the Precinct. In short, Election Judges will have MORE machines to deal with – in addition to the printers!)
The State Elections Administrator usually warns that to change to the Optic Scanners would require a total re-training of Election Judges and reprinting of all the manuals. Prior to Maryland going to the touch screen machines, 19 of Maryland’s 24 jurisdictions used the Optic Scanners. They are infinitely easier to set up and use and I can assure you that the Election Judge’s manuals were much smaller!
In testimony presented to the House Ways and Means Committee regarding House Bill 18 (which is nearly identical to S. B. 392) on February 1, 2007, Michael Sanderson, the Legislative Director of the Maryland Association of Counties expressed concern about the spiraling unfunded mandates which are being passed on the Counties from the State Board of Elections to pay for the new voting equipment AND the additional “technical support.” What had started out to be a 50/50 split between the State and the Counties originally indicated to be approximately $3 Million has now turned into an unfunded mandate which is more than 5 times the original estimate.
I have been present when Legislators have been given incorrect information about the Optic Scanners by the State Elections Administrator, i.e., they are not capable of detecting and therefore preventing either an over- or an under-vote. This is factually inaccurate. The Optic Scanners easily detected either an over- or an under-vote and the voter is given an opportunity to correct their ballot before it was counted. (Around the Nation, some jurisdictions are considering putting “None of the Above” as an option for each candidate/question on the ballot in an effort to eliminate “under voting.” Again, based on my Chief Election Judge experience – particularly in the County Executive race between John Gary and Janet Owens, I can assure you that “None of the above” would have won by a landslide.)
Recently, Maryland temporarily dodged a DRE bullet. On December 5, 2006, the Technical Guidelines Development Committee, by a tie vote, did not pass a proposal to the Federal Election Administration Commission, which would have required all electronic voting systems to be “software independent” and readily audited (which Maryland’s touch screens are NOT) by the 2008 election. The next day, the Committee changed its course and voted UNANIMOUSLY to begin drafting regulations that would require the next generation of voting systems (think 2010) to be software independent.[11]
Already news accounts report Maryland will be facing a $1.4 Billion deficit next year -- which appears not to include the structural pension deficit projected at nearly $2 Billion. It makes little sense to throw another $50+ million after the $100 million already spent on the acquisition of touch screen machines which may soon be banned for use in elections by the Federal government. Furthermore, adding printers to machines, which are already suspect, may only worsen the problem. A News-Record article, “Printers failed on voting machines” noted that “About 9 percent of the printers attached to the county’s voting machines had a jam or other problem.”[12]
Returning to the Optic Scanners would solve many problems at a much lower cost.
The State Elections Administrator has frequently warned that if Maryland were to change its voting equipment, i.e., return to the Optic Scanners, the transition could not be made prior to the 2010 election. The State of North Carolina made the transition in 8 months and conducted an election.[13] Perhaps Maryland’s Election Administrator might consult with North Carolina’s Election Administrator?
When the DRE’s were put in place, for the first time in my over two decades of experience as a Chief Judge, Machine Judges were required to put the number of the particular machine used by the voter on each Voter Authority Card (VAC). The State Elections Administrator has consistently stated that an actual paper record of EACH BALLOT CAST can be printed out at the Local Board of Elections. If this is a correct statement, are votes actually secret anymore? I ask this because even prior to the advent of the E-Poll Books it appears it would be possible to reconstruct precisely how EACH VOTER voted by using the sequential number assigned to each voter (given as that voter is logged in to the Poll Books) and tying that sequential number to the number of the machine used by the voter. Thus, the voter with the lowest number (whose name and address is printed on the VAC) would be the first voter to use a particularly numbered machine (and that machine number is identified on his/her VAC), and so on. It would appear that this would be particularly true if the votes are stored sequentially instead of randomly.
The State Elections Administrator’s often notes that there IS an actual paper ballot which can be counted/reviewed because an actual paper ballot can be printed out on the “main-frame” computer back at the Local Board of Elections. Without the actual voter reviewing his/her own printed ballot, it is doubtful that the State Elections Administrator or any member of any Local Election Board could verify that the paper ballot printed out at a different location actually reflected each voter’s intent. The statement that an actual paper ballot can be printed out at another location merely proves the machine can print.
Again, as a Chief Election Judge, I can state without fear of contradiction that each of the DREs can print out the results of the election as many times as I tell it to print the totals. That paper tape only tells anyone that the machine can print – NOT that the totals which it has printed are CORRECT!
The Maryland Federation of Republican Women, the Republican Women of Anne Arundel County – and the League of Women Voters of the United States strongly support the recommendation contained in Senate Bill 392 to have a voter-verified paper record.
I believe that by returning to the precinct-based Optic Scanner and also complying with the HAVA requirements for accessibility for the handicapped, which can be done by the use of an electronic marking device in each precinct, also provided for in S.B. 392, the exodus of election judges will dramatically abate, the costs of conducting Maryland elections passed on to Maryland’s families through taxes will be substantially lower and Marylanders will once again have confidence that their vote IS counted as THEY cast it – not as an unknown programmer and/or computer hacker may have directed.
[1] “Gov. Crist to recommend ditching touch-screen machines, South Florida Sun-Sentinel, January 30, 2007. http://www.sun-sentinel.com/news/local/southflorida/sfl-0130cristvoting,0,6954161.story. (Appendix 1)
[2] ACM is the oldest and largest scientific and educational society of computer professionals with approximately 80,000 members.
[3] Statement of Barbara Simons for the Committee on House Administration Hearing on Electronic Voting Machines, September 28, 2006, (Appendix 2).
[4] June 9-13, 2006, Minneapolis, MN
[5] NIST Draft Report, Posted December 1, 2006, “Requiring Software Independence in VVSG 2007: STS Recommendations for the TGDC.” (14 pages, Appendix 3) http://vote.nist.gov/DraftWhitePaperOnVVPRinVVSG2007-20061120.pdf
[6] “Security Analysis of the Diebold AccuVote-TS Voting Machine,” dated September 13, 2006, 24 pages, plus Executive Summary, 2 pages. (Appendix 4). http://itpolicy.princeton.edu/voting
[7] The Brad Blog report posted February 22, 2006, http://www.bradblog.com?p=2460&print=1 (Appendix 5)
[8] Each voter needs his/her own machine.
[9] March 20, 2005 Comparison of Operating Costs-Addendum, (Appendix 6). http://www.votersunite.org/info/costcomparisonaddendum.asp
[10] “Operating Cost Comparison for Different Types of Voting Systems, 1999-2004, 2005 Study, (Appendix 7). http://www.ncvoter.net/affordable.html
[11] CNET News, December 6, 2006, “Panel changes course, approves e-voting checks by Anne Broache. (Appendix 8) http://new.com.com/Panel+changes+course%2C+approves+e-voting+checks/2100-1028_3-6140956.html
[12] The News-Record, article published December 15, 2006, “Printers failed on voting machines.” (Appendix 9)
http://news-record.com/apps/pbcs.dll/article?AID=/20061215/NEWSREC0101/61215003/-1
[13] “New voting machines see few glitches in N.C, primary,” May 3, 2006 Raleigh News Record article (Appendix 10). http://www.news-record.com/apps/pbcs.dll/article?AID=/20060503/NEWSREC0101/60503019/-1