Back with another graph of the day post.
Every election held in Australia post the arrival of Covid-19 has seen a sharp decline in voting on election day and a surge in postal and especially pre-poll voting.
Last October’s Queensland election saw only 27.6% of votes cast as within-district polling day votes, with 43.6% of votes cast as pre-polls and 23.8% as postal vote. (See this post)
March 2021 saw a similar surge in Western Australia with polling day ordinary votes falling to 38.0% compared to 40.2% for pre-poll votes and 14.8% as postal votes. (See this post)
At the November 2018 Victorian election, polling day ordinary votes represented only 48.3% of all votes, the first Australian state election where less than half of votes were cast on the day in district. The rate of Early/Pre-poll voting was 36.8%, then the highest recorded at an Australian election, having quadrupled in 12 years.
Given the trend to voting before polling day has been stronger in Victoria than anywhere else, and given the state’s experience with Covid-19, one can only guess how low the rate of polling day voting will be at the 2022 state election.
The graph below shows the percentage vote by vote type at Victorian elections since 2006.
The 2018 election produced the traditional pattern of two-party preferred vote by vote type. Labor polled 59.3% of the two-party preferred vote on election day, and also 62.6% of the Absent vote. Absent voting is much more common in urban electorates than regional and rural electorates, which is why Absent voting always leans towards Labor given the party's urban base.
Note: Figure in brackets is the vote type as a percentage of the total vote.
Labor polled only 55.7% in Early voting, a feature of elections in the last decade being Labor polling less well with votes cast before election day. And that's especially the case with postal votes, dominated by older and rural voters, where Labor's two-party preferred percentage was only 51.2%.
I've been looking at these numbers in preparation for the Victorian state redistribution, the draft boundaries to be released on 30 June.
At Federal and NSW elections, the Electoral Commissions runs multiple pre-poll voting centres that are also used as count reporting centres. Calculating estimated margins after a redistribution is easier where pre-poll vote totals can be identified by location.
Victoria is one of the states that reports a single figure for pre-poll votes, and also makes no distinction between votes cast within district as opposed to pre-polls cast elsewhere in the state. On election night the Victorian Electoral Commission reports a total for within district pre-polls, but its final statistical returns lump all pre-polls into a single total by district.
Which given less than half of all votes cast in 2018 can be linked to a polling place, is going to leave a larger than normal fudge-factor in estimated post-redistribution margins.
I suspect that the combined pre-poll vote is partly due to Victorian electorates being smaller than Commonwealth electorates and (to a much lesser extent) NSW electorates has resulted in single in-electorate pre-poll centres persisting for longer in Victorian urban electorates and thus the specific need for individual publication is less obvious to the result publication part of the VEC.
COMMENT: Victoria has chosen to make all pre-poll votes ordinary wherever cast. Votes for all electorates go into the same ballot box and have to be sorted before counting. In urban areas many absent votes are also taken as ordinary votes. It massively reduces the number of declaration votes, but has consequences for the count. The VEC opens ballot boxes to start sorting and proposes to start even earlier in 2022.
Depends how easy it is to do a postal. I was going to do one in the March election but couldn’t. Had to do a pre-poll instead.
COMMENT: My advice is always to use pre-poll rather than postal if you can. Pre-polls are counted without further checks but there are numerous reasons why postal votes may never make it into the count. The postal system is slower than it used to be, and all the details on the postal declaration envelope have to be completed correctly. If they aren’t, the vote won’t count.