According to the NSW Electoral Commission iVote is technology assisted voting and has been used in a New South Wales general election and five by-elections since 2011 for certain classes of electors - those with impaired vision or a physical disability requiring assistance, the profoundly illiterate, persons living more than 20 kms from the nearest polling station and those out of the state on polling day.
Voting is done on the Internet using a standard web browser or by call centre operators taking phone votes. Originally the second voting method was by phone using a standard handset and DTMF tones, but this was changed after the 2011 general election.
The largest group of iVote users were electors voting outside NSW on election day (over 43,000).
Of those electors who registered to iVote, 4,239 or 8.30% did not eventually use this system to cast their vote and 1,438 or 2.90% did not vote at all.
Over one thousand electors (1,335) registered to iVote during the 19 November 2011 by-election for the state seat of Clarence on the NSW North Coast and, most of these were first time uses of this voting system.
It has been proposed that the iVote system be used again for the March 2015 general election and during the September 2016 local government elections.
However, there are some issues with the iVote system that are not generally advertised by the NSW Government.
Below are excerpts from the Federal Government Joint Standing Committee on Electoral Matter’s Second interim report on the inquiry into the conduct of the 2013 federal election: An assessment of electronic voting options, with my red bolding:
For example, the lost vote rate in the 2013 West Australian Senate race (1370 out of 1,348,797, slightly over 0.1%) was about the same as the demonstrated vote misrecording rate in Australia’s largest Internet voting trial, the NSW iVote project (43 misrecorded electronic votes out of 46,864, slightly under 0.1%) (PWC, 2011). The WA Senate incident received much more attention because it impacted an election outcome, not because the system was inherently much less reliable. Even more importantly, the paper-based Senate process retained paper evidence of the 99.9% of votes that weren’t lost; the iVote system produced no meaningful evidence of the correctness of any of the votes.
the ‘weak point’ in a paper-based voting system, resulting in a lost box of ballot papers, may lead to an unverifiable close result (such as in WA): but one ‘weak point’ in a wide-ranging electronic voting system has the potential to expose an entire election’s vote data to manipulation, corruption or attack, undermining the parliamentary system supported by the electoral process.
The NSW iVote system (outlined in Chapter 3) used in the 2011 state election had an average cost per vote cast of $74 compared to an average cost of all votes cast of $8.
While the iVote system is relatively secure, due to the fact that it utilises telephone systems for blind or low vision voting transactions and encrypted internet data architecture, the vote data on the voter’s computer or in the NSWEC’s servers is still open to potential manipulation.
In response to criticisms of the system’s security, the NSWEC has commissioned a third-party provider to strengthen the security of the system software prior to the 2015 state election, along with other hardware and data transmission improvements.
Vision Australia made a submission to the Joint Committee concerning telephone assisted voting during the 2013 federal election which included these observations:
It was anonymous, but not truly secret. People felt uncomfortable about verbalising their voting intentions to another person, and expressed the view that no-one else in the community would regard it as acceptable to be required to do this. Some clients in residential facilities and other places with limited privacy also expressed concern that their conversation with the call centre staff would be overheard and their voting intentions revealed…..
Clients who had voted using the iVote system in the NSW 2011 election were especially aware of the lack of independence involved in using the call centre option.
Some clients noted that they had no way of verifying that their voting intentions had been notated accurately and lodged correctly. While they did not necessarily mistrust the call centre operator, they were nevertheless aware that any human-mediated process introduces the possibility of errors, and such errors are more likely to occur when the process becomes complex, such as when a voter is voting “below the line”.