< Back to table of 'First Report - December 2004 Contents'
Part 5 – Accuracy and Secrecy of the System
5.3 Secrecy
The major issue in relation to secrecy that emerged in the public submissions received and in the Commission’s own testing relates to a matter deemed outside the Commission’s terms of reference, namely, the secrecy accorded to a voter who wishes to “spoil” a vote, or to register a completely blank ballot. Setting this aside for now, a number of other issues arose in relation to secrecy.
There is some limited infringement of ballot secrecy arising from the audible “beeps” made by the machine during the act of voting, both as preferences are being selected and to signal voter errors. This among other things allows limited inferences to be drawn by those outside the polling booth about the number of preferences cast. In particular, voters voting for a single candidate would be easy to identify by those in the vicinity of the machine and in this sense their ballots would not be fully secret.
There is also reduced voting secrecy for people with certain disabilities, as well as for people who are unfamiliar with or unsettled by new technology. Such people may need third-party assistance in using the voting machine, thereby infringing their ballot secrecy. While the Commission acknowledges that voters with a disability have the right to secrecy, it has been established by the courts that this right is not an absolute one8.
Publication of the ballot results in full but in random order, as happened after the 2002 pilots, is a very valuable aid in ensuring the accuracy of the results, since anyone is free to recount these for themselves. Nonetheless it has been submitted that this can in theory reveal deliberate and distinctive voter “signatures” of low-preference votes (highly improbable rankings of the candidates ranked 11 to 15 in a 15-candidate contest, for example), which could allow voters to identify themselves in a context of corruption or intimidation. This may be deemed improbable, but it remains a possibility.
In this context it is worth noting that, in general, the issue of the publication of the full set of votes cast highlights a conflict between the objectives of secrecy and accuracy. Publication, as noted, greatly enhances accuracy by allowing anyone to check the count process. At the same time it diminishes secrecy to the extent that a particular individual’s vote might in some way be distinctive. If votes are to be published for small local areas – to replace the information hitherto available on the basis of tallies – then potential infringements of ballot secrecy become more likely. In general, however, the gains in both real and publicly perceived accuracy arising from publishing the full set of votes and making these available for recount seem likely to outweigh any potential resulting loss of secrecy.
Because the storage of votes on the ballot module is pseudo random, not truly random, it may be possible for an expert insider to overcome the randomness of the storage of votes on the ballot module. Combined with knowledge of the order in which people voted using the voting machine in question and the knowledge of precisely how the first person to vote on the machine had cast their vote, it might be possible to deduce how everyone using the machine had voted. This also may be deemed highly improbable, but remains a theoretical possibility.
8McMahon v Attorney General [1972] IR 69, (1972) 106 ILTR 89
< Back to table of 'First Report - December 2004 Contents' |
