An Open Letter responding to "The County Dilemma: The Impact of the Help America Vote Act on New York State" [Click here for PDF ~114kb]

I received a copy of your paper on "The County Dilemma" in November, and wish to offer some corrections to your interpretations of HAVA implementation in New York. I do not write to disagree with your overall argument about the dilemma facing the counties in meeting the 2007 deadline. Neither do I presume to criticize your analysis of the districting problems in Albany County. But I am concerned that New York's election commissioners might be confused by some of your statements about some other matters. Following are some of the misstatements and misinterpretations your paper contains:

1. Mistaken interpretations of HAVA.

You say that "HAVA…mandates that every state in the U.S. must implement an electronic style of voting machine system that will allow a voter with any type of disability to vote on the same machine as a voter without a disability" (page 2). This same mistaken claim about HAVA is repeated on page 4.

• HAVA does not require electronic voting machines. It even protects the right to paper ballot systems so long as they provide accessibility to persons with disabilities [See, e.g., Title III.A. Sec. 301 (c) 2].
   
• HAVA does not require the use of the same machine, but only that there is at least one system equipped for individuals with disabilities at each polling place. [Title III. A. Sec.301.a. (3)].

2. Misleading descriptions of direct recording electronic machines (DREs).

You say on page 3 that with a Direct Recording Electronic (DRE) machine, "the voter would make their selection on a touch-screen… and the results are recorded electronically as well as on a paper receipt which is automatically inserted into a ballot receptacle."

• With a DRE, whether touch screen or push button, the vote is recorded directly in electronic circuits. Both the electronic record and the Voter Verifiable Paper Audit Record are produced by the software, rather than directly by the voter. It is possible for code to be written so that these two records are different.

• The VVPAT should not be spoken of as a receipt. This gives the wrong impression to voters, and implies this is a printout which they can take home with them.
   
• The DRE print-out only shows that the print-out has their choices, but it does not show that "their votes were correctly tallied" electronically (p. 3).

3. Undocumented implications of a history of fraud with paper ballots and the lever machines.

You say on page 3 that "The majority of problems encountered [with lever machines] didn't actually include the lever machines, but involved fraud as it relates to paper ballots …"

• Do you mean to imply that there has been election fraud that was caused by paper ballots being associated with lever machines? Or is it just that there are specific problems that need management with any absentee paper ballot system?
   
• Modern camera and heat sensor technologies can protect paper ballots. We should consult bank officers about methods for protecting important legal paper records.
   
• Problems with the lever machines themselves have been reported. See, e.g., William Edelstein, "New York State Law and Lever Voting Machines," www.nyvv.org/doc/NYSLeverMachines.pdf

4. Mistaken interpretations of why the State Board was unable to meet the original deadlines.

You say on page 4 that "In July of 2005 the State Board of Elections was given the implementation task, unfortunately, due to public hearing requirements, the State Board of Elections was unable to meet the original timeline required by the Federal Government."

• New York's governmental structure surely indicated that the State Board of Elections (SBOE) would have the task of implementation of HAVA. The basic HAVA requirements and the kinds of procedures and testing that would be needed were discoverable from the time HAVA was passed in 2002.
   
• The State BOE received the Election Reform and Modernization Act (ERMA) in June 2005, but did not publish draft regulations until December. The first draft was so poor that it would have created more chaos than we now have without the input from experts and citizens during the hearing period.

5. Mistaken claims about Plan B and federal requirements.

You say on pages 4-5 that "In reaching an agreement with the State of New York on the implementation of ‘Plan B' as a way to prove that New York is on the path to becoming HAVA compliant, the Federal government essentially approved the expenditure of its' (sic) own money on voting machine technology that does not even comply with its own laws and which may never be used again."

• The ballot markers approved for Plan B by the SBOE fulfill federal requirements that voters with disabilities be able to cast a private and independent vote. The only federal requirement they do not meet is that they were not made available in each polling place.
   
• The intention of Plan B was to provide pollsite accessibility to those with special needs in 2006.

6. Mistaken claims that DREs and optical scanners are equally vulnerable to deliberate falsification of the count (see your pages 5, 6, 7, 11).

• While research reports from various important governmental and scientific organizations, including the Brennan Center report to which you refer, identify threats to the security of scanners as well as DREs, it is incorrect to imply that they say the DRE and paper ballot scanner systems are equally vulnerable.
   
• The Brennan Center has endorsed the paper ballot scanner system for several reasons, including usability as well as software independence.
   
• Scanners only count votes, while DREs record, cast, store, count, and print records for verification. DRE software is more complex and can provide only a second-hand paper record of the voter's choices. A voter-marked paper ballot is a first-hand record of the voter's choices. Careful procedures allow us to objectively interpret and protect these records.
   
• Your reiterated claim that "both types of voting machines are equally susceptible to attacks from hackers" (p. 7) incorrectly implies that the choice between the two systems cannot lessen the threat. To the contrary, by choosing a system based on voter-marked paper ballots, we can eliminate the computer from the process, and still independently and correctly discern voter intent.
   
• The most serious threat to computer systems is not from outside hackers but from insiders with unobserved access to systems.

7. Mistaken implications that wireless technology is necessary for hackability.

You say on page 6, "Since New York State law does not allow voting machines to be equipped with wireless technology or modems of any type, breaking in and physically attacking the machines is the only way a hacker could be successful in their attempt to alter the results of a machine." However,

• Computers were "hacked" and made to produce false results long before communication technology became common.
   
• Deliberate miscoding of machines requires only a minute of unobserved access by an insider and can be passed to many machines through memory cards.
   
• The law that bans wireless communication does not require inspection or enforcement; nor do any Boards of Elections have the time, skill, or commitment to examine their computerized equipment to determine whether they comply with the law. For this reason we will not know whether or not our computerized voting equipment has wireless communications capability.

8. Mistaken claims that the monetary cost difference between the two systems cannot yet be meaningfully considered.

• Now that vendors' proposals are posted on the NYC website and some costs (e.g., ballot markers) are posted by the Office of Governmental Services, one can meaningfully consider acquisition costs.
   
• In addition, the evidence about the length of time it takes to vote on each machine is now becoming available. Studies such as those by William Edelstein go far toward highlighting how the two systems differ decidedly in their ability to accommodate voters during peak voting times. This, of course, is significant for the number of machines to be purchased and maintained and, therefore, for cost analyses.
   
  See, e.g., William Edelstein, "New Voting Systems for New York—Long Lines and High Cost" www.nyvv.org/doc/votelines.pdf.
   
  Also see Teresa Hommel, "Purchase Cost of New Voting Equipment for New York City," http://www.wheresthepaper.org/ny.html#costReports.
   
• Comparative studies of the operating costs for the two systems are available and have shown that the paper ballot scanner system is considerably less expensive in on-going expenses. (See, e.g., http://www.votersunite.org/info/costcomparison.asp)

I recognize the dilemma New York's counties face, inasmuch as HAVA implementation together with centralization has given them immense responsibility as well numerous unfunded mandates.

I trust you will take a close look at the comments I've made and consider them in your evaluations.

Sincerely,

Wanda Warren Berry, Ph.D.