Now that the dust has (mostly) settled, it’s time for some analysis. Just before the election I published a post discouraging people from voting informally. Now that the results are pretty much finalised, I’d like to demonstrate what could have happened if the informal lower house votes from this election had instead been cast formally.

I’m using data gathered from the AEC website on the 13th of July at 11AM GMT+8. At this point in time, the ABC has called 76 seats for the Liberal National Party (LNP) (with another predicted, bringing the total to 77), 68 seats for the Australian Labour Party (ALP), one seat for the Greens (GRN), one for the Nick Xeonophon Team (NXT), one for Katter’s Australian Party (KAP), and two for independent candidates.

Combining the two party preferred and informal counts available from the AEC website for each electorate, I was able to calculate the proportion of informal votes required to overcome the margin in each seat, i.e. if that percentage of informal votes were instead cast for the second place candidate, they would have won. For many seats, this percentage was larger than 100. That is, even if every single informal vote had instead gone to the second place candidate, the outcome would not be different. However, 24 seats could have been upset if enough informal votes in that electorate instead went to the candidate in second place. These seats are listed below:

As the table demonstrates, all of the following scenarios could have eventuated if enough informal voters had chosen to instead cast a formal vote for the second place candidate:

  • LNP majority government of up to 87 seats (currently the LNP is predicted to form majority with 77 seats).
  • ALP majority government of up to 80 seats (currently the ALP is predicted to be in opposition with 68 seats).
  • The Greens win their second seat, taking it from the ALP (currently they have only retained their one seat from last parliament).
  • Nick Xenophon Team win a second seat, taking it from the LNP (currently they are set to take only one).
  • And of course, various combinations of hung parliaments.

One of the most interesting seats here is Herbert, having the closest margin of all the seats in this election. The table shows that if just a little over one in every hundred informal votes in Herbert had instead been cast for the LNP, the tally would show the LNP in the lead instead of the ALP. Another thing to note is that this analysis only takes into account the first and second candidates in each electorate. Some seats, such as Melbourne Ports saw a very small margin between the second and third candidates. It’s possible that redistribution of informal votes in such electorates could see the current third place candidate move into first place.

A slightly different scenario emerges if we consider the hypothetical case of all informal votes becoming formal, cast either for the first or second place candidate. This is more realistic, as we would not expect every single voter who decides against voting informally on their way to the polling booth to vote for the same candidate. With competition from some of the previously informal votes now going to the first place candidate, second place candidates require a higher percentage of the informal votes to pull ahead. However, all 24 of the previously mentioned seats are still theoretically able to be upset under this model:

The next time you are contemplating voting informally, remember that 16% of the seats in this election could have been held by a different candidate if enough informal votes were instead cast formally. Please, don’t waste your vote.