Waterloo Internet voting

The City of Waterloo has again rejected the use of Internet voting in its municipal election.

I really liked the framing by Regional Councillor Jane Mitchell: “showing that you are really tech savvy”.  Which is to say, rejecting Internet voting doesn’t show that you are backward-looking, it shows that you actually understand the technology.

Waterloo Region RecordWaterloo rejects online voting, ranked ballot – by Paige Desmond – November 21, 2016

Waterloo ChronicleWaterloo council rejects Internet voting for 2018 – by Samantha Beattie – November 22, 2016

CTV News Kitchener – Waterloo council says no to online voting, ranked ballots – November 22, 2016

The meeting was the November 21, 2016 Council Meeting.  The meeting will be posted to YouTube account citywaterloo.

The relevant report is CORP2016-105 Alternative Voting Methods (Internet Voting) by Olga Smith.  The report can be found on pages pp. 84-93 of the original packet (PDF) for the November 21, 2016 Council Meeting.  (The original packet is also available in the Internet Archive.)  It is an update on previous report CORP2013-053.

The CORP2016-105 report offers an overview of the current unfortunate state of affairs as Internet voting spreads across (small, low-IT-capacity) Ontario municipalities:

Since 2003, there has been an increase in Ontario municipalities introducing alternative voting methods (internet, telephone and mail-in voting). In 2014, 97 out of 414 Ontario municipalities offered paper, vote-by-mail, internet and telephone ballot options, others a combination of paper and internet or, in the case of 61 municipalities, offered all electronic (internet and telephone voting). The municipalities’ population ranges in size from 900 to approximately 300,000.

It then identifies some key concerns

Opponents of internet voting advise of concerns including:

  • security concerns and process vulnerabilities
  • voter authentication
  • loss of transparency with reduced oversight of the voting process by candidates and scrutineers

Next it turns to debunking some Internet voting myths, starting with turnout

The results of an extensive study conducted by Elections British Columbia, and presented to the Legislative Assembly of British Columbia in February 2014, dispel the myth that internet voting increases voter participation in general and participation by young people in particular

The report being referenced is from the British Columbia Independent Panel on Internet Voting, specifically the February 2014 Recommendations Report to the Legislative Assembly of British Columbia (PDF).

For security concerns it paraphrases me (I was not aware of this until now)

Research and information from experts in the computer science field warn of the risks of the use of online voting including:

  • Widespread use of online voting could enable coercion of voters and possibly vote buying.
  • Software and hardware components that would be involved in marking, transmitting, receiving and counting an online ballot represent an unreasonably high risk to the chain-of-custody for the ballot.
  • Canadian government departments have already been successfully cyber attacked.
  • Computer and national security experts warn that online voting is not secure.

The original text from my submission to ERRE reads

Considerations:

  • Widespread use of online voting would enable widespread coercion of voters, including vote buying.
  • The innumerable software and hardware components that would be involved in marking, transmitting, receiving and counting an online ballot represent an unreasonably high risk to the chain-of-custody for the ballot.
  • Canadian government departments have already been successfully cyberattacked by nation-states.
  • Computer security experts warn that online voting is not secure.
  • National security experts warn that online voting is not secure.
  • Social science evidence indicates that online voting won’t increase turnout.

It doesn’t reference my brief (lack of citations is not unusual for city staff reports).

CORP2016-105 also raises examines concerns about Internet voting in the following areas (using the document numbering):

  1. Auditability
  2. Scrutiny
  3. Accessibility (it finds that online voting provides accessibility)
  4. Lack of Federal/Provincial Standards
  5. Additional Costs for Internet Voting

The report next presents data from Guelph and Cambridge, with the following observations:

  • The statistics from municipalities offering a choice of paper ballots or internet voting show that most voters choose to mark a paper ballot.
  • Cambridge did see a minor increase of 1.18% in voter turnout as compared to 2010.
  • Guelph did see an increase of 9% in voter turnout as compared to the 2010 election but the increase could be attributed to the mayoral race and not necessarily that internet voting was offered.

corp2016-105-pp92-93-cambridge-guelph

This is followed by a chart for Guelph’s 2014 election showing a typical voter distribution pattern, i.e. there is no dramatic increase in youth turnout for an election with Internet voting.

city-of-guelph-voter-turnout-by-age-2014
The supplementary information packet (revised packet) for the November 21, 2016 Council Meeting also includes (on page 9) an email from Urs Hengartner, Associate Professor, Cheriton School of Computer Science, University of Waterloo, which I have reproduced below with email address removed:

-----Original Message-----
From: Urs Hengartner
Sent: Sunday, November 20, 2016 1:02 PM
To: Clerk Info
Cc: Jeff Henry
Subject: Comment on item 7.c)
         "Alternative Voting Methods (Internet Voting)"

Hello,

Here is a comment on item 7.c) “Alternative Voting Methods (Internet Voting)” to be discussed during the Waterloo City Council Meeting on Nov 21, 2016.

Given the arguably unexpected outcome of the U.S. presidential election, many people have called for an audit of the election result. As it turns out, about 25% of the entered ballots exist only in electronic form, so auditing them is pointless. Fortunately, the remaining 75% of the ballots (also) exist in paper form and could be audited.

In an Internet voting system, there are no paper ballots that are filled in by a voter him/herself or by a machine in the voter’s presence, so there is no paper record that reliably documents the voter’s intent.
Given that it is each voter’s choice with which voting system to vote, the percentage of votes entered via an Internet voting system could reach a significant proportion of the overall votes. Therefore, auditing an unexpected or narrow election outcome would become impossible. In turn, rumours alleging mistakes in the voting system or attacks on it could not be addressed, leading to voters losing their trust in the integrity of the voting process.

Best,
 Urs

--
Urs Hengartner
Associate Professor
Cheriton School of Computer Science University of Waterloo, Canada

One thought on “Waterloo Internet voting

Comments are closed.