How often do Tasmanian MPs lose their seats?

One of the almost unique features of Tasmania’s Hare-Clark electoral system is the potential for voters to choose individual candidates over others in the same party, and on occasion achieve the outcome of an individual MP losing their seat to a fellow member of the same party. Even when that doesn’t happen, it’s common that multiple members of the same party are competing for the same seat.

While there are quite a few proportional electoral systems in Australia, most of them feature “above the line” voting, which effectively imposes a party list element. If a party wins three seats, the first three candidates on their list win a seat. While below the line voting theoretically allows for candidates to be elected outside of their party’s order, in practice this doesn’t happen.

But in Tasmania (and the ACT), the absence of a party list or above the line voting, and the enforcement of the party list by rotating candidate names within each party’s group on different ballot papers (known as Robson rotation) means that candidates compete with each other. No seat is safe on its own. Even if a party is guaranteed of winning at least one seat, the individual identity of that MP could change (at least in the major parties).

So in this post I thought I would consider how often this happens. How often do MPs lose their seats, and how often does it happen to a member of their same party or another party? How often do they retire, and how often do they retire early? And how have these rates changed over time, in particular with the contraction of the Assembly in 1998?

To keep it simple, this first chart shows the proportion of MPs who were re-elected at each election since 1986. It features two metrics: one just covers the proportion of MPs elected at the previous election who were re-elected, and the other includes those who had entered parliament through a casual vacancy and were then re-elected.

[EDIT: I had originally started these charts in 1989, but I shifted it back to 1986 after advice that the 1986 election had seen a remarkably high rate of MPs losing their seat, mostly to members of the same party with little change in the partisan balance.]

It’s a regular dilemma I’ve found in posting this data simply – do you analyse the data for the people at the start of a term, or the end of the term?

1986 stands out as having the lowest rate of re-elected MPs throughout this period, but this wasn’t because a lot of seats changed hands between parties.

There was a drop in re-election rates in 1998, which can be largely explained by almost 30% of seats in the Assembly disappearing.

There was an odd spike in 2006 when almost every sitting MP was re-elected. You’ll notice 2006 stands out on every chart in this post. And then in 2018 and 2021 we saw the rate of re-election climbing, but only if you include casual vacancies.

This next chart shows that, in the case of MPs who didn’t continue into the next parliament, how they left.

Bear in mind that from 2002 onwards there was a significantly smaller pool of MPs, and it does seem to have reduced the number of MPs retiring at the election or losing their seat, but the number of casual vacancies has been steady at four per term through the last three parliamentary terms, which is in line with the 1989-1996 average in a larger Assembly.

The number of defeated MPs in 1986 stands out clearly, with almost twice as many defeated MPs as in 1998.

This next chart shows the result at the election for each MP. In this case defeated or retiring part-term MPs are treated the same as full-term MPs.

There was a spike in retirements and defeats in 1998, when the contraction in the Assembly size made it harder to win re-election. A bare majority of those elected at the 2006 election were re-elected in 2010, and 2002 also saw a lot of turnover. Since 2010 there has been declining rates of turnover.

But again 1986 stands out – a bare majority of 18 MPs were re-elected, with 15 MPs being defeated and two retiring.

Finally this chart shows who did the defeating in the case of MPs who lost their seat at an election.

In 1998 there were six MPs who lost their seats because their seat didn’t exist anymore. That is obviously a once-off, and they can’t be counted in either category since nobody replaced them. Another four MPs retired by choice and their seats were left vacant, adding up to the ten abolished seats.

In 1986, amazingly there was a relatively low rate of MPs losing their seat to other parties. Just two members lost their seat to another party, but 13 members were defeated by fellow party members.

As mentioned before, there were no defeated MPs in 2006. Over this entire period, there was the same number of MPs defeated by a member of their own party and defeated by someone else.

It’s been slightly more common for MPs to be defeated by another party since 2010, but there have been plenty of exceptions. In 2010, four MPs lost seats to fellow party members, and in 2021 there were three. Indeed there were no intra-party defeats in 2021. There was only one change in the party seat totals in 2021, with Labor losing a seat in Clark to an independent. But since Madeleine Ogilvie had filled a casual vacancy previously held by Labor as an independent and went on to be re-elected as a Liberal, this didn’t result in any MPs losing their seat.

The 2024 election should be another unusual one. All 25 incumbent MPs are running for re-election, although three of them were elected as party members in 2021 to later leave their parties and run this time as independents. I suspect we will see relatively low rates of MPs losing their seats, either to their own party or to others, with the extra vacancies creating more space for talented new entrants without having to push out others.

Liked it? Take a second to support the Tally Room on Patreon!