Last week, a former software engineer named Bo Lipari gave a presentation to a group of Sullivan County citizens arguing against the use of touch-screen electronic voting machines (DREs) and for optical scanners, as detailed in our article. “Election commissioners face crucial choice,” on page 1.
Four years ago, Lipari’s concerns might have been dismissed as extreme. But in the intervening period, evidence has been mounting that electronic voting systems—especially the touch-screen variety—are, at the very least, susceptible to technical glitches and, at worst, highly vulnerable to fraud.
Last Wednesday, Diebold machines used in the election primaries in Maryland suffered such extensive breakdowns (some, but by no means all, due to human error) that judges had to extend voting by an hour in order to accommodate the thousands who had been turned away during the day. And last Monday, a group of independent researchers at Princeton released a study showing that Diebold voting machines can easily be hacked in less than a minute. There is a video of this demonstration at www.youtube.com/watch?v=lwWP-N1HqT0.
What is disturbing about these findings is that many states, including Pennsylvania, either do not have requirements for, or actually forbid the use of, a voter-verifiable paper ballot that can be used to recount votes. Such states have no recourse in case of election debacles, whether accidental or deliberate.
In Pennsylvania’s case, the rationale is that a verifiable paper trail would violate ballot secrecy by revealing how individuals voted. But it is perfectly possible to produce paper ballots that do not disclose the identity of the voter, and it would behoove Pennsylvania voters to urge their representatives to identify such machines, and mandate their use. The optical scanning systems recommended by Lipari would be one way to fill the bill.
Even with an optical scanning system, however, in which votes are manually recorded on paper cards that are then tabulated by machine, those tabulators are subject to breakdown or hacking. The only way to protect against that is to have a robust oversight and review system, in which both software and hardware are rigorously tested for such vulnerabilities. But the major voting machine companies insist that, to preserve competition, they ought to be able to keep their code proprietary, and refuse to submit to full independent reviews from groups such as the Princeton scientists. In effect, they argue that free enterprise is more important than free elections.
The current system does mandate a review of voting machines by one of three federally accredited Independent Testing Authorities (ITAs). Unfortunately, both the ITAs’ independence and their competence are open to question. To begin with, they are paid by the companies whose machines they test. Second, they test machines against a narrow federal standard. If they find a serious flaw in the machine that is not covered by the federal standard, the machine is still passed, and the existence of the flaw is confided only to the manufacturer. This is probably why so many ITA-certified machines have failed both in tests and in actual elections. In California last August, fully 20 percent of machines in a test election broke down despite having received ITA approval. Finally, according to one ITA, Ciber, they do not test for fraud and tamperability, describing these issues as “not applicable.” It is chilling to discover that there is no requirement in the United States of America that electronic voting machines be resistant to tampering.
There is no silver bullet that will solve all these problems, though the choice of an optical scanning system might be a good start. However, to defend against possible hacking or breakdown of the optical scanning tabulators, more will be needed.
First, the hardware and software of all voting machines on the market must be open to public review like that of the Princeton group. Second, the current bogus ITA review system must be replaced with one in which the payment system provides ITAs with an incentive to find flaws, not to cover them up; in which any flaws found—whether included in some pre-set standard or not—are publicly disclosed; and in which vulnerability to fraud is one of the primary issues tested for.
We must rely on our election commissioners to make the best possible choice among the machines eventually certified by the state. But, if we want to make sure our votes count, we also need to push our state and federal legislators to create a strong system for evaluating and enforcing the integrity of our voting system.
We need to make sure that our free democracy is not subverted by our system of free enterprise.
Dr. Punnybone
Trout Limited
Letters to the Editor
[EDITOR'S NOTE: The River Reporter welcomes letters
on all subjects from its readers. They must be signed and include
the correspondent's phone number. The correspondent's name and
town will appear at the bottom of each letter; titles
and affiliations will not, unless the correspondent is writing
on behalf of a group.
Letters are printed at the discretion of the editor.
It is requested they be limited to 300 words; correspondents may
be asked to cut longer letters. Deadline is 1:00 p.m. on Monday.
I’m an unaligned noncombatant in the war between the National Park Service (NPS) and our neighbors with stakes (literal and figurative) along the banks of the Delaware. But I thought a neutral’s view needed airing after all the NPS-bashing over removal of Art Peck’s “lighthouse.” In particular, I have some comments on your editorial in The River Reporter of September 7 and on Cass Collins’s earlier River Muse column, “The Lighthearted Lighthouse.”
