Third Party & Independents Archives

HAVA Series Part Four: Problems with Electronic Voting

Since the 2000 presidential election and the controversy regarding punch-card and lever based machines Digital Recording Electronic devices (DRE’s) have risen in use dramatically.

With the passage of HAVA and $325 million provided to states to replace punch card and lever based machines, many states (44) have further implemented the use of DRE’s. However the increased use of DRE’s due to HAVA funding could easily work against the goals set out in HAVA.

Probably the largest problem with DRE systems rests in the complete lack of knowledge of these systems. This lack of knowledge comes from the fact that the DRE vendors are privately contracted companies. As private entities the vendors of DRE’s do not allow any inspection of the software used, leaving government unable to test the software for potential problems (45). This inability to access the software also means that election officials have little chance to uncover fraud written into the machine, and must rely heavily on the vendors, giving the vendors rather than the government the majority of the power.

Outside of the inability to inspect the software, testing is also kept in secret and the majority of the vendors refuse to talk about any of the testing that is performed on their units (46). The ‘independent’ testing that is done on these systems is often seriously in question as the president of one of the companies contracted to do testing stated, “there’s going to be the risk of a conflict of interest when you are being paid by the vendor that you are qualifying the product for” (47). Vendors can even avoid the certification process that is required by installing uncertified software immediately preceding an election, this occurred when Diebold installed uncertified software in all of Georgia’s 22000 machines right before the 2002 election (48).

Even when attempts have been made to force vendors to provide their software compromise has not been forthcoming. After 4000 votes were lost due to memory problems in 2004 in North Carolina that state passed a law requiring vendors to turn over their software, yet a Dieblod spokesman countered by stating, “the problem is that the law requires us to provide not just our source code but also the code of third party vendors as well as the developers involved. I don’t know who all the developers were for (Microsoft Corp’s Windows) CE. We don’t own it; we license the use of it” (49). More importantly even if the North Carolina law is successful without changes in HAVA vendors will still be able to keep their software private from other states. Thus by supporting the further use of DRE systems HAVA actually decreases security, unless access to the software code is forthcoming from vendors.

Without further guidance by HAVA or the EAC states are essentially free to do away with recounts. As mandated by HAVA DRE’s are not required to have any form of accompanying paper ballot, and with software codes kept secret without such a voter verifiable paper audit trail (VVPAT) recounts, that is recounts outside of the computers system, would be near impossible (50). Because of these concerns as of today twenty five states in all have required VVPAT’s to be attached to DRE’s (51). Despite this occurrence not all states are using VVPAT’s for the same purposes, for instance, Nevada only uses VVPAT’s so voters can be assured the computer correctly reflected their vote and they will not be used in recounts (52). In all only fifteen of the states that require VVPAT’s require their use in recounts (53).

Without VVPAT’s many advocates argue that voters will have no guarantee that the computer correctly reflected their vote. Although HAVA did not require VVPAT’s their was an effort, however to this point unsuccessful, in Congress led by John Ensign (R- Nevada) in the Voting Integrity and Verification Act to implement paper trails which would accompany DRE’s (54). Again because the software is kept secret fraud is possible and without a paper trail for a voter to verify their ballot on a voter cannot be certain their vote was submitted correctly if possible problems in definition in the source code were to exist.

Works Cited:
44. Michael A. Carrier, “Vote Counting, Technology, and Unintended Consequences,” St. John’s Law

Review, 79 (3) (Summer 2005): 646.

45. Ibid, 667.

46. Ibid, 656.

47. Ibid.

48. Ibid.

49. Caron Carlson, “States Scramble to Test Ballot Machines,” eWeek, 22 (48) (Dec. 5, 2005): 20

50. Michael A. Carrier, “Vote Counting, Technology, and Unintended Consequences,” 669

51. Dan Seligson, “Use of Paper Trails Divides Advocates, Election Officials,” Campaigns & Elections, 26

(10) (Jan 2006): 56

52. Ibid.

53. Ibid.

54. Grant Gross, “Bill Seeks to Fix E-Voting Flaws,” Computerworld, 39 (7) (Feb. 14, 2005): 19.

Previous parts in this series can be found at:
Part One: http://www.watchblog.com/thirdparty/archives/004077.html

Part Two: http://www.watchblog.com/thirdparty/archives/004091.html

Part Three: http://www.watchblog.com/thirdparty/archives/004154.html

Posted by Richard Rhodes at September 21, 2006 4:02 AM
Comments
Comment #182996

Richard:

Is it fair to suggest that your solution to the potential problems and/or abuse of electronic voting systems would be to implement a mandatory VVPAT system that would be also used in recounts?

If not, what is your idea on what a solution should be?

Also, if VVPAT is the solution, it seems rather simple—-why is it not being implemented?

Thanks for your thoughts.

Posted by: joebagodonuts at September 21, 2006 7:16 AM
Comment #183015

Richard,

Nice article. I believe the biggest problem with our voting system (no matter what technology we used) is that it is a patchwork of county systems, with different rules and technologies, instead of a single unified voting system, administered by a non-partisan election commission. As long as we have a decentralized network of county Registrars beholden to private partisan voting machine manufacturers, the legitimacy of our vote will always be uncertain.

That being said, I think it is important that voting activist stop pushing Voter Verified Paper Audit Trails (VVPAT) and demand that what ever technology we use produces a Voter Verified Paper Ballot (VVPB). The difference is that in a VVPAT system your vote is stored in a computer database while in a VVPB system the vote is marked directly on a paper ballot (which makes the paper ballot the official primary record of your vote). There are several touch screen devices that will mark a ballot, so ease of use is not an issue.

The reason not to store the primary vote in an electronic database (where it is relatively easy to manipulate numbers) is that recounts are rare and expensive, and a 1% random audit of machine vs. VVPAT’s is only likely to catch irregularities if they are relatively large (while I would have to pull my stat book out, you probably would need to have a discrepancy of 3-5% between VVPAT’s and the machines to have a high probability of detection).

With VVPB’s the votes are counted using optical scanners at the precinct level. Under a single unified voting system, these scanners would be publicly owned machines with code written and owned by the state (code by the way, which is very simple to write).

Oversight by a single authority, in conjunction with state owned machines, would go a long way towards securing our vote.

Posted by: Forrst Hill at September 21, 2006 10:25 AM
Comment #183029

read this

watch the video

I would be very wary of these voting machines. I don’t believe that on a local level, particularly in poorer areas, that there are people knowlegeable enough to research and/or prevent problems.

I believe this jeopardizes our whole voting system. I firmly believe that paper and pen is still the most reliable.

Posted by: womanmarine at September 21, 2006 12:34 PM
Comment #183030

joebagodonuts- This will be addressed in part in future parts. But to answer your question I would personally advocate first that the source code is available for inspection and that officials be involved in the testing of software in some way. Beyond that yes I would advocate VVPATs, but there are far more things that need to be done as well.

Posted by: Richard Rhodes at September 21, 2006 12:35 PM
Comment #183142

Here’s some fresh news from Common Cause:

In Maryland last week, the primary election was beset with problems:

*
Electronic cards to activate voting machines were not delivered to polling places, delaying opening times and causing many voters to miss the chance to vote at all.2
*
When the electronic voting machines could not be used, polling places quickly ran out of paper provisional ballots, and some poll workers resorted to giving voters sample ballots or campaign literature to mark their votes.2
*
Our own IT consultant watched as his wife ‘checked in’ on the new electronic polling book only to have the device crash, be re-booted and then tell her that “She had already voted!”
*
Poll workers were given cell phones to communicate with elections officials about problems - but some of the poll workers did not know how to use the phones.3

—————————

A glimpse of what is coming on Nov. 7.

Posted by: David R. Remer at September 21, 2006 8:39 PM
Comment #183182

There is no way to guarantee our votes will be either counted, or counted correctly, on electronic voting machines. NONE.
And I’m not sure there will ever be a way for us to be certain using them — or having our votes counted on optical scanners, either.
As womanmarine said: “I firmly believe that paper and pen is still the most reliable.”
I agree. And thanks to HAVA whose intention was to reward companies by pushing the nationwide switch to their electronic machines and scanners, the November election, like all our elections since 2000, is very likely to be rigged, hacked or flipped, and so cannot be trusted.
New Rollingstone article by Robert F. Kennedy Jr.:
Will The Next Election Be Hacked?
Fresh disasters at the polls — and new evidence from an industry insider — prove that electronic voting machines can’t be trusted

If you don’t have time to read the whole article, Bradblog has posted some exerpts from it.

Posted by: Adrienne at September 22, 2006 10:53 AM
Comment #183556

Richard,

Come now. What is your problem? Why not give private vendors the power? We live in a corporate oligarchy. They should have the power. If we can’t manipulate the voters into doing the right thing with campaign donations, we need to have a fall back fail safe system.

Posted by: Ray Guest at September 24, 2006 12:11 AM
Post a comment