|
Electronic Voting Machines
|
VERIFIED VOTING NEW MEXICO |
No. While fraud is a worry, promoting public confidence is the primary issue, graceful recovery from the bugs in the system is a close second, and avoiding costly errors a distant third. The basic rule of thumb is that one wants to avoid any situation where one is unnecessarily placing trust in a person, business, or machine. Even the maker of Australia's touch screen voting system has said as much: "Why should you have to trust me?" While there are in fact plenty of reasons to question the craftsmanship, security competence, and altruism of voting machines sellers this is largely irrelevant: a properly designed voting process can remove the need for blind faith.
Currently available touch screen voting systems all have two fatal flaws: 1) they are designed in such a way that the software, hardware and operator training must be flawless 2) the process unnecessarily makes us trust that the manufacturer is a saint. All engineers know it is simply not possible to write flawless complex software or, at least, to ever know that you did. Instead mission critical systems are designed to be fault tolerant: they fail safely and have a backup.
In voting this means voter-verifiable, recountable, hard copies of the ballots. We can eliminate the need for blind faith in the manufacture's good intentions and/or competence through the use of open source software, in much the same way we require open-meetings laws.
Logic and accuracy tests can't find subtle flaws and neither will code reviews. Open source has the advantage of not only more eyes on the code, but because there are no non-disclosure agreements, the most skillful eyes can look at it.
No, organizations more qualified and less vested strongly believe there are serious outstanding issues. Up-to-date analysis by the Library of Congress (CRS), General Accounting Office, The Ohio Attorney General, Maryland's SAIC commission, and qualified technologists have found the Federal Election Commission and NASED certification and testing process egregiously outdated, and have suggested that touch screens bring unprecedented and not yet addressed vulnerabilities into the election process.
Confirmed reports, most recently in Virginia, Florida, and North Carolina, of serious touch screen voting system malfunctions prove that certification and testing is not sufficient to eliminate design errors in complex systems. NIST has not yet revamped voting standards, and current machines have not even qualified for lesser standards used in health-care or Military hardware.
About a third of United States citizens live in states that have or are considering requiring Voter-verified paper ballot systems. Other states have voted to delay implementation.
Problems and concerns motivated Secretaries of state in California, Ohio, Nevada, and Louisiana, as well as the NY state Assembly, to request a costly replacement/retrofit of their current Touch Screens to permit voter-verifiable balloting or other improved standards. Indeed California halted the use of existing systems after a vendor, Diebold, was caught making unauthorized changes to certified software.
Diebold Systems get the most abuse only because, due to a computer security lapse, they accidentally made public confidential problems with their systems. There is no reason to believe other companies are any better; they just have not exposed themselves to the same degree. Indeed the Sequoia data collection software uses the notoriously insecure MS Windows.
In fact, any company that won;t make their source code and hardware open should be assumed to have security problems. It would be very hard to cite a counter example.