Overturn this

By Вen Li

(Ed note: This ran as the column "Nugatory" online last week.)


The most extensive democratic exercise at the University of Calgary is now in doubt as the Students’ Union Review Board overturned the wishes of over 6,500 students on the grounds that an undetermined, but not unknowable, number of students’ votes were uncounted.


The ruling upset not only the victories of the winning candidates, but also students’ confidence in the relatively new electronic voting system. Students participating in a democratic process, some for the first time, are unlikely to return to it if they believe their efforts ineffectual.


So it is highly unfortunate that the discontent of so few students could mean rerunning an election in the summer or fall, after many current students have graduated. Doing so deprives more students of their right to vote than could have conceivably been denied in this election, and also denies some candidates the opportunity to once again seek office. The rights of this 20 per cent or so of students would be abrogated in favor of a vocal and unelected few, who do not necessarily represent the interests of all students, to know how closely they lost the election.


Though knowledge is preferable to ignorance, in this case knowing the margin of error comes at the cost of all students becoming ignorant of the composition of the future SU executive and the future of those individuals in it–a difficult bargain.


Despite these negatives, the ruling has utility.


First, it admits that our democratic system, which predates that of many developing nations, is still not perfect and that no election can be perfect. Second, Director Marc Wrubleski of Sorex Software, which created the online voting software, agrees that voters should be prevented from leaving virtual ballots at the voting booth for others to use and has proposed improvements to that effect. Third, the attention generated by the challenge to the results provides an opportunity for the online voting system to be proven and sustained.


However, the panel’s stated reasoning for overturning the election departs entirely from these positives. Three glaring flaws in their ruling illustrate that they know little about the fundamental technology used in electronic voting.


First, their belief that "[a] vote was not registered until the ‘vote accepted’ screen was displayed" is comparable to a young child’s belief that objects cease to exist if they disappear from immediate view. Clearly, the computer server tasked with collecting ballots does not cease to operate when Internet Explorer crashes or ignores the server for a particular voter, much as Hotmail does not stop processing e-mail if a user’s web browser crashes.


Second, their belief that "return[ing] the screen to home or the Students’ Union home page prior to ‘vote accepted’ page being seen, the vote would be destroyed" confuses ballots with actual votes and contradicts repeated statements (solicited by members of the Review Board) from the voting system’s creator about ways in which ballots could be destroyed. Mr. Wrubleski repeatedly stated jumping to any page in the voting process, prior to the ballot being collected or a new ballot being created, would not change the state of that page or the ballot.


Third, in considering the possibility of multiple voting, the board accepts that Mr. Wrubleski "was unable to prove or disprove the actual incidents of any occurrence of this type," admitting any estimate of probable irregularities would be baseless. But the board attempts to use Mr. Wrubleski’s conjecture–that there was a "smaller than one per cent" chance of irregularity–to sustain that "on the balance of probabilities… the result of the election may have been different as a consequence of irregularities." An interesting and needless leap of logic indeed, especially when actual numbers of attempted and successful votes could be determined from logs (which should be) kept by the ballot-keeping computer as an industry-standard practice.


In overturning not only the races determined by a handful of votes but all election and referenda results–including those decided by hundreds or thousands of votes–on misunderstood grounds, the Review Board clearly values knowing the margin of error more than the gross expressed will of the students.