The graphics compare the current state of the lower houses as of their last full elections (not including more recent byelections). The parliaments are dominated by Liberal and Labor, with a sprinkling of Nationals. Currently all states bar Tasmania use the single-member electorate system to elect their lower houses.
I compared these election results with the actual percentage of votes each party received. Using the number of seats available in that particular house, I calculated what that percentage of votes would correspond to under a proportional representation system. This system disregards geographic location of voters and counts each vote equally into an electorate that includes the entire state (or country) — 10% of the vote will result in 10% of the seats, notwithstanding rounding in small parliaments.
I believe single-member electorates are inadequate for these reasons:
Up to 49% of the electorate prefer the second candidate to be their local MP, and are unlikely to be represented by their MP. This can be much more when considered on first preferences.
MPs value their party over their constituents, meaning that those who did not vote for that party are unrepresented in parliament.
It entrenches power in two large parties, as the race for any single electorate usually comes down to those two.
The system values the geographic concentration of support, as opposed to the amount of support. Parties with relatively high support (~15%) evenly spread nation-wide may not win a seat, yet a party with 0.5% support nation-wide can win a seat if it is concentrated in one electorate.
Marginal seats are given far more attention than safe seats, and parties can take certain groups of voters for granted (for example, the Coalition with rural voters or, arguably, Labor with progressive inner-city voters, although some of them are turning Green).
The proportional representation system is used in the NSW upper house, though only half the MLCs are up for re-election at any one time. It is a system that treats all votes equally, regardless of where they are placed geographically. The system is unlikely to be the best available (New Zealand’s Mixed Member Proportional and Tasmania’s Hare-Clark system are both arguably superior), but it is a result that is easy to model and compare to the current system. It can be argued that parliaments predicted under proportional representation are more representative of the views of the voters than the current system.
The graphic is not so much an argument for the pure proportional representation system, as a means for pointing out the disenfranchisement of the current system. It does not show the ~60% of voters who are represented by an MP they actively voted against. But it shows how the system skews results away from what voters may have wished.
The most disenfranchised party under the current system is the Australian Greens. With broad support across a number of electorates, their vote is not geographically concentrated enough to result in seats anywhere near their level of popular support. Few seats are Green (Melbourne in the federal House of Representatives is one). However, if the Greens’ level of representation matched their popular support, I calculate there would be 36 state mainland Greens MPs and 13 in the federal House of Reps.
The biggest winner out of the current system, conversely, is the National Party. With relatively low national support that is concentrated in a small number of rural electorates, their parliamentary representation outstrips their national support.
The function of parliaments would rapidly change with proportional representation. Instead of being a winner-takes-all scenario, governments would have to negotiate with others. The two large parties may even split.
The most starkly unrepresentative parliament is Queensland’s unicameral Parliament, where the LNP holds a 78-seat majority. I argue the result does not represent the will of the Queensland people, as the proportional representation graph shows. Nor is it, as many southerners like to wrongly assume, an example of gerrymandering. It is simply an example of how unrepresentative the single-member electorate system can be. As academic Dean Jaensch points out, if the single member electorate system produces a fair result, it is the result of luck, not the design of the system.
*Matthew Brown is a biology student and political observer. He is a member of the Greens and has volunteered for the party.