Page MenuHomePhabricator

Allow blank votes / spoiled ballots
Open, Needs TriagePublicFeature

Description

Occasionally, people have asked to be able to submit their ballots blank, as can be done in real-world elections, generally as a show of protest: https://en.wikipedia.org/wiki/Spoilt_vote#Intentional_spoiling.

For STV elections, SecurePoll currently requires you to rank at least one choice, and I assume this is true for support/oppose votes as well. Is this a design artifact, or an intentional choice? The downside I see is that allowing intentionally blank ballots would also make unintentionally blank ballots possible--we could mitigate this by having the software have it warn you and ask for confirmation if you try to submit a blank ballot, perhaps requiring you to fill a checkbox to make sure.

Event Timeline

Alternatives:

  • "None of the above" option for non-binary polls.
  • Write-in box (to suggest an option, or to write "none of the above" or similar non-vote.

Write-in box (to suggest an option, or to write "none of the above" or similar non-vote.

I think in some elections this might raise the question of what to do if an unofficial candidate wins this way :)

have it warn you and ask for confirmation if you try to submit a blank ballot, perhaps requiring you to fill a checkbox to make sure.

Some people might just click through because of a habit to just click ok on any confirmations or because they don't understand what is written if they use an interface language for which Secure Poll isn't translated yet.

have it warn you and ask for confirmation if you try to submit a blank ballot, perhaps requiring you to fill a checkbox to make sure.

Some people might just click through because of a habit to just click ok on any confirmations or because they don't understand what is written if they use an interface language for which Secure Poll isn't translated yet.

Yes, any time we introduce a new capability, we are also introducing a certain amount of unavoidable risk of user error. For STV elections, this risk is probably dwarfed by the risk that people will rank their candidates backwards (which appears to have occurred in the 2019 ASBS process), but we should certainly design the interface to verify to the greatest extent possible that the voter is doing this consciously.

Further mitigation for this and for voting errors in general could come from having your vote displayed to you and/or emailed to you after you vote--but that's another task (T22023; cf. T20969).

have it warn you and ask for confirmation if you try to submit a blank ballot, perhaps requiring you to fill a checkbox to make sure.

Some people might just click through because of a habit to just click ok on any confirmations or because they don't understand what is written if they use an interface language for which Secure Poll isn't translated yet.

Yes, any time we introduce a new capability, we are also introducing a certain amount of unavoidable risk of user error. For STV elections, this risk is probably dwarfed by the risk that people will rank their candidates backwards (which appears to have occurred in the 2019 ASBS process), but we should certainly design the interface to verify to the greatest extent possible that the voter is doing this consciously.

Further mitigation for this and for voting errors in general could come from having your vote displayed to you and/or emailed to you after you vote--but that's another task (T22023; cf. T20969).

Additionally, you can re-cast Wikimedia votes, so if/when it suddenly closes and the user cries out that they've been robbed of their vote...they can just vote again, striking their "none of the above" vote