Open-Audit Voting means a Single Vote Counts

After an incredibly long and busy week of work for my colleagues Olivier Pereira and Olivier de Marneffe, the UCL election, based on Helios, has been verified and tallied. The trustees arrived earlier today and successfully decrypted the result.

Students each got approximately 1/10 of a vote, while Faculty got a full vote. 4000 people voted, which resulted, after weights were applied, to 1111 votes cast. The candidate with the highest number of votes received 555.44 votes, which is barely one vote short of an absolute majority.

If 6 more students had voted, they might have affected the result. If one more faculty had voted, she might have affected the result. Instead, there will be runoff.

But more importantly: because the voting system is precise to a single vote, because every voter can check that their vote was recorded [see a real receipt obtained by a UCL voter], because there was a 24-hour complaint period before the tally during which voters could claim that the system had failed to record their vote (2 people changed their votes successfully during the complaint period), we have confidence in this result.

If this had been a paper-ballot election, and if (only) 0.1% of ballots had gone missing (which is typical with paper ballots, after all, mistakes happen), the entire result of the election would be in doubt and within the margin of error.

So sometimes, elections are decided by one vote. And when they are, it’s really helpful to have individual vote tracing and verification.

UPDATE: and the press reports.

4 thoughts on “Open-Audit Voting means a Single Vote Counts

  1. >Students each got approximately 1/10 of a vote,
    >while Faculty got a full vote. 4000 people voted,
    >which resulted, after weights were applied, to
    >1111 votes cast. The candidate with the highest
    >number of votes received 555.44 votes, which is
    >barely one vote short of an absolute majority.
    Humor me by dwelling on this for a bit. Students had fractional votes? What was the fraction? Why did
    they choose to do this? If they wanted to have vote
    weights why not have them all be integer weights?

  2. >Students each got approximately 1/10 of a vote,
    >while Faculty got a full vote. 4000 people voted,
    >which resulted, after weights were applied, to
    >1111 votes cast. The candidate with the highest
    >number of votes received 555.44 votes, which is
    >barely one vote short of an absolute majority.
    Humor me by dwelling on this for a bit. Students had fractional votes? What was the fraction? Why did
    they choose to do this? If they wanted to have vote
    weights why not have them all be integer weights?

  3. I understand that weighing the votes sounds weird at first, but consider the context. This is the election of the President of the *University*, where faculty stay for a career, and students stay for 4 years. Students outnumber Faculty by a factor of 10 or more. So, to not weigh the votes would mean handing the decision to the Students and the Faculty would have no input on their President.

    Now, whether you’re doing fractions or integers, it doesn’t matter, right? Just multiply everything by 100 if you’d rather have integers :)

  4. I understand that weighing the votes sounds weird at first, but consider the context. This is the election of the President of the *University*, where faculty stay for a career, and students stay for 4 years. Students outnumber Faculty by a factor of 10 or more. So, to not weigh the votes would mean handing the decision to the Students and the Faculty would have no input on their President.

    Now, whether you’re doing fractions or integers, it doesn’t matter, right? Just multiply everything by 100 if you’d rather have integers :)

Comments are closed.