NRC: "A Framework for Understanding Electronic Voting"

From: Ed Kennedy <ekennedyx_at_yahoo_dot_com>
Date: Mon Dec 06 2004 - 20:42:06 CST

Hello All:

I though it might be useful to post Joe's analysis of the NRC report along with the links. I'm doing this with his permission.

Thanks, Ed Kennedy

Post details: NRC: "A Framework for Understanding Electronic Voting"
 Categories: system, elections, certification/testing, accessibility, reform, vendors, standards, news, open source, secrecy, privacy, politics, SIMS, problems, friends
NRC: "A Framework for Understanding Electronic Voting"
The Committee for Electronic Voting - under the Computer Science and Telecommunications Board (CSTB) of the National Academies' National Research Council (NRC) - recently issued a call for papers for input on what questions policy makers should be thinking about given the current state of electronic voting. Here's the loot:

  a.. Preliminary Analysis of E-Voting Problems Highlights Need for Heightened Standards and Testing, Deirdre Mulligan and Joseph Lorenzo Hall, University of California, Berkeley
  b.. Electronic Voting Machines and the Standards-Setting Process, Eddan Katz and Rebecca Bolin, Yale University School of Law
  c.. Illustrative Risks to the Public in the use of Computer Systems and Related Technology, Excerpt: Election Problem Cases as of [November] 25, 2004, Peter G. Neumann, SRI International
  d.. Putting People First: The Importance of User-Centered Design and Universal Usability to Voting Systems, Sharon Laskowski, National Institute of Standards and Technology; and Whitney Quesenbery, Whitney Interactive Design LLC
  e.. Accessibility and Auditability in Electronic Voting, Electronic Frontier Foundation
  f.. Electronic Voting, David Dill and Will Doherty, Verified Voting Foundation
  g.. Electronic Voting Machines in South Carolina, Duncan Buell and Carter Bays, University of South Carolina
  h.. The Need for Usability of Electronic Voting Systems: Questions for Voters and Policy Makers, ACM Special Interest Group on Computer-Human Interaction (SIGCHI), U.S. Public Policy Committee
  i.. Voting, Vote Capture and Vote Counting Symposium: Electronic Voting Best Practices, Jean Camp, Allan Friedman, and Warigia Bowman, Harvard University John F. Kennedy School of Government, June 2004
  j.. Making Each Vote Count: A Research Agenda for Electronic Voting, report of an AAAS workshop on electronic voting, October 2004
  k.. Electronic Voting Systems: The Good, the Bad, and the Stupid, Barbara Simons
There are a lot of good submissions here... I'll be reading them in my copious free time and posting comments here.

Interesting Quotes / My Comments
Here are my random thoughts and things I learned from reading these submissions (The Neumann piece is a tad dense, I'm not sure how to go about evaluating that... I'll use it more as a reference.):

  "For example, the time-stamped logs identifying voters can be cross-referenced against voter-intake rolls in order to verify an accurate count of ballots." -page 9 of SIGCHI submission.

I hope they are referring to keeping track of the total number of votes cast in a polling place compared with the number of voters that have signed registration poll books. Anything more detailed has the potential to compromise ballot privacy and secrecy.

  "Many voters using machines activated with SmartCards assumed that their votes are stored in the card, and were confused by seeing the cards returned to the registration table and recycled: "He told me that there were about 20 cards at the precinct; it was at that point that it really struck me that there was no back up for my vote.'" on page 11 of NIST

This is an interesting point, the machinery of our elections is so complex that voters simply don't understand it; they think that what little technical ability they've had with computer machinery before might translate over to voting technology. Also, we've seen cases where one person will say something emphatic about how a voting machine works in a polling place and everyone will believe them and spread misinformation (most notably when, on Nov. 2, Matt Drudge ran reports from Philly of machines not recording votes... when what was really going on was that a ballot style/precinct number display panel was being mistook as a cast vote incrementor)... would that be groupthink?

  "Further, if we are going to make usability a requirement for any voting system, we must clearly define how we measure usability so that we can evaluate the outcome of the UCD process against those requirements." from page 16 of NIST.

I found this to be a very important point from the earlier NIST human factors in e-voting report... we have to be able to consistently evaluate and compare voting system usability of vastly different technologies. That's not necessarily and easy thing to do...

  "One solution for access to voting by individuals with disabilities that is allowed by HAVA is to have one accessible voting booth per polling place. While we recognize that for budgetary reasons this might be the only feasible solution initially for many jurisdictions, we suggest a critical look at this situation from a long term perspective. This single point solution has a number of problems: [...]" page 17-18 of NIST.

Amen, brother... we raise this in our paper by pointing out that this provision in HAVA will allow the continued sale of very old voting systems qualified against the 1990 standards for years into the future. (Although with Sequoia being sold by De La Rue, who knows... Danaher / Guardian sure as hell doesn't seem to be marketing anything that would meet this requirement of HAVA.)

  "The Electronic Privacy Information Center has submitted a public records request to obtain the full unredacted version [of the MD SAIC report]." from page 3 of Simons.

I didn't know EPC had done this... way to go! I would like to see more.

  "Just as NASA cannot perform an actual test of a spacecraft landing through a Martian atmosphere onto Martian gravity, no election commission can test an election system in advance under completely realistic situations. As such, then, the machines used in elections require engineering standards with at least the rigor we would expect for remote spacecraft." from page 2 of Buell and Bayes.

This is an interesting comparison... I've heard voting software compared to gambling and aviation contexts, but never the pulling off of an election to the pulling off of a space mission. Would a space mission management be a good model for an election? Sounds compelling. The basic rationales in space mission management are careful planning, testing and redundancy. We're missing a lot of that in voting.

  "We believe those who are knowledgeable about software are unimpressed by the certification of reliability by the ITAs. We would be much more impressed if the legions of graduate students around the world had tried and failed to do their worst against the ESS or any other electronic voting machine." page 6 of Buell and Bayes.

;) Hook it up d00d!

  "It may not be possible to obtain absolute assurance that the election process is reliable, but a closed system whose trustworthiness is bought and paid for by private enterprises with a vested interest in the results is neither believable nor acceptable." page 11 of Buell and Bayes.

This is a particularly eloquent way of making a subtle point. The solution is somewhat complex. Should the feds pay for qualification? Should the vendors pay into a pool of money used to qualify voting systems from? It's clear that having the testing authorities being the direct customers of the vendors in the testing process does not result in sufficient testing.

  "3. Paper Backups - In addition to these steps, election jurisdictions should prepare to provide voters with paper ballots upon request. Every state has procedures for absentee and provisional balloting, and those procedures should be extended to voters who choose not to use a DRE that cannot be audited." page 8 of EFF.

This really hasn't been emphasized enough. Counties must have back-up plans in case their voting equipment is completely non-operable. Cities like New Orleans in the recent 2004 election that had no paper back-up plan saw the loss of an untold number of votes when the entire parish's voting machines went down.

(Now I'm just going to quote interesting stuff from the VVF whitepaper)

page 3:

  "New Orleans, Louisiana: 80+ reports with 42 reports of total breakdowns, with long lines and voters turned away from polling places, for example, 'All voters have been unable to vote on touchscreen machines. There are no paper ballots. The precinct official tried to call the Sect'y of State's office for guidance but could not get through. (Not clear if county officials were contacted). Precinct officials don't know what to do.'"

  "Dauphin, Mercer, and Philadelphia Counties, Pennsylvania: Dauphin County had five reports of machine power failure and faulty machine operation. 'There were only 2 machines available for the largest precinct in the area at 101 18th Street, but only one of the machines worked.' Mercer County had 15 reports of catastrophic machine failure. All voter machines were down, and makeshift paper ballots were provided but in some cases were not secured. Philadelphia reports included 28 complaints of misrecording of votes as well as 28 reports of total breakdowns."

page 4:

  "yet another [report] mentioned a machine that would only take votes in Spanish."

  "'When a handicapped voter can't get inside the polls, the machine is brought out to the voter. But because all the machines are connected, everything inside stops until the handicapped voter is done voting.'"

(Note that usually two poll workers have to accompany a terminal outside of the polling place for this type of "curbside" voting. If this requires the number of election staff in the polling place to drop below a number mandated by procedure, they have to temporarily close the polling place to allow the disabled voter to vote. I've spoken to Doug Jones, the expert on the machine this county was using (the ES&S iVotronic) and removing one terminal from the daisy chain does not impair the other voting terminals connected to it via serial cable. This must then be a personnel problem.)

page 6:

  "Some regulatory agencies require incident reports and routinely investigate potential safety problems. Should there be a mandatory or voluntary incident reporting system for election problems? If so, what sort of problems should be reported, at what level of detail? How should reports be investigated? What agency or agencies should have the responsibility of collecting the reports and investigating them?"

page 10

  "But there may be opportunities for innovative use of technology as well. For example, San Mateo County Registrar of Voters Warren Slocum used a 'web-cam' to broadcast the testing of the machines in his warehouse over the internet."

10777 Bendigo Cove
San Diego, CA 92126-2510
"Let us all tend to our gardens."  Candide - Voltaire

OVC discuss mailing lists
Send requests to subscribe or unsubscribe to
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain

icon_minipost.gif us.gif
Received on Fri Dec 31 23:17:06 2004

This archive was generated by hypermail 2.1.8 : Fri Dec 31 2004 - 23:17:22 CST