Friday, September 16, 2005

One nation, under Diebold (part II)

Well, I received a response from the Secretary of State's office. Of course it doesn't actually address my concerns, but it's a start:

Dear Mr. Smith:

Thank you for your recent email about our voting system. The equipment comprising Georgia's voting system has been used successfully in over 500 elections since its debut in 2002. Three separate independent studies by the University of Georgia have shown that Georgians have extremely high levels of confidence in the accuracy and security of our statewide voting system -- and those confidence levels are much higher than with our previous, antiquated mechanical and paper-based systems. Nevertheless, we take seriously the concerns of those who have reservations about our touch screen voting platform. Achieving an even broader level of public confidence in the elections process is an important goal that we continue to strive for. In addition to implementing the most intensive and robust voting security safeguards found anywhere in the nation, the leadership and staff of the State Elections Division has been aggressively researching and evaluating new voter verification technologies -- including paper receipt systems. Our staff has traveled to Nevada, Maryland and several other states to evaluate these technologies first-hand. Staff from the Kennesaw State University Center for Election Systems has also been involved in this evaluation work as well.

Many of these devices are prototypes, and have not yet been extensively tested in real-world election settings. Some have considerable potential, while others appear to have significant flaws. (It is important to note that the voting system we acquired in 2002 was not an experimental prototype nor would we have ever considered an unproven new technology -- it in fact had been used widely in jurisdictions across the country in numerous elections.) It's also worth noting that there are external audit and verification devices that use audio or video recording approaches instead of paper, and there is considerable debate in academic and industry circles as to which of these technologies would add the most security and value to the electronic voting process.

We believe strongly that federal standards are needed to clearly establish how a paper-based or other voter verification technology should operate and how that record should be stored. We also need to answer the question of what constitutes the official ballot, and how and under what circumstances this external record will be tallied in a recount. The director of Georgia's Elections Division is currently participating on a national panel to help move this process forward so that clear federal standards for new voter verification technologies can be created.

In our view, any new external verification device must not only enhance public confidence -- it must also not jeopardize the huge gains we have already achieved in voter accessibility and accuracy in Georgia. Data from academic researchers at the Cal Tech-MIT Voting Technology Project shows that, with the deployment of our statewide electronic voting platform, Georgia moved from second worst in the nation in voting error rates in 2000 to second best in that measurement in 2004. In fact, Georgia's gain in voting accuracy was the highest in the nation. We take great pride in that improvement and want to insure that it is preserved in future elections. We are also mindful of the need to retain the excellent accessibility features of our current platform that benefits the visually impaired, blind and disabled. Concerns, including some expressed by paper receipt advocates, have also been raised that some receipt systems can compromise the secrecy of the ballot by permitting arriving voters to view the printed choices made by the previous user of the voting terminal.

Before acquiring and deploying a paper receipt or other external verification device, we believe we must have a system that:

  1. Meets new federal standards -- which are currently under development;
  2. Meets all the requirements of federal certification, including extensive testing for functionality and reliability; and passes state certification tests conducted by the KSU Center for Election Systems;
  3. Does not negatively impact ballot accessibility for blind, visually impaired or disabled voters;
  4. Does not compromise the secrecy of the ballot;
  5. Maintains a high level of voting accuracy;
  6. Is user-friendly for voters and is easily deployed, monitored and maintained by county election personnel and poll workers; and
  7. Has been proven highly reliable in real-world elections.

Indeed, all but one of these are the same criteria we applied in our evaluation of candidates for the existing statewide platform when it was acquired in 2002. We believe that, once these reasonable criteria are met, and the state has funds available, an external voter verification system can be deployed as an enhancement to our already highly secure and accurate system.

Adding a paper receipt or other external verification device to our current platform seems, on its face, to be simple and easy. In fact, there are important questions to be answered, standards to be set and real-world experience needed with this emerging technology. There are promising developments occurring in voter verification technology and we are closely monitoring that progress. We are also anxious to see the completion of federal standards, which can set forth clear guideposts for how verification systems should operate effectively in an elections setting.

Finally, it is important to point out that legislation passed by the General Assembly this year makes a very significant change to our absentee voting process in Georgia. Under this provision of HB 244, any voter may now request a mailed paper absentee ballot for any reason. This change effectively creates a paper-based, vote by mail option available to every Georgia voter purely as a matter of choice and preference. For any voter who prefers not to use a touch screen terminal and would prefer, instead, to cast a paper ballot, this option is now available without limitation.

Thank you again for expressing your concerns and for your interest in the security of Georgia's voting system.

State Elections Division

To which I responded:

Dear Mr. Madison,

Thank you for your form letter. While I do believe that maintaining a paper trail is one necessary part of having the verifiable system necessary for maintaining faith in our election system, my question did not refer to that issue at all.

My concerns are these:

  • When did your office become aware of the US-CERT alert issued in August, 2004 concerning the undocumented "backdoor" inserted into the GEMS voting tabulation system by Diebold programmers? (Note: the existence of this "backdoor" has been recently verified by a Diebold employee.)
  • Since becoming aware of the problem, what steps has your office taken to ensure that this problem has been corrected?
  • While this problem is inherent in all of Diebold's GEMS Central Tabulator, releases including, though possibly not limited to, 1.17.7 and 1.18 - whether the votes are loaded on-site or remotely - of particular concern are those electoral districts (such as Fulton and DeKalb counties) where the votes are uploaded to the central tabulator via modem. In these cases, a person with knowledge of the "backdoor" could change the vote counts without having to physically enter the electoral office. What actions has your office taken to warn all electoral divisions within Georgia of the problem and train them on the proper usage of this technology?
  • Since the management of Diebold has been aware of this problem for over a year and has yet to take any independent action, has this company been blacklisted from bidding on future contracts issued by the Secretary of State? And, if not, why not?

Thank you again,
Charles Smith

We'll see how he answers this time.


Post a Comment

<< Home