Peter Martin argues that the Liberal Party is the natural national party and the Labor the natural state party. He writes that the Liberal Party risks making itself the permanent party of opposition if it competes with Labor for services based government. I strongly disagree with that analysis. As little as seven years ago there were four Liberal state governments.

I have created a graph to display it visually. It takes a middling year of when a party was in power at the national level and then counts how many Labor and anti-Labor (Liberal, Country, Progressive Nationals etc) there were at the state level the same year in June.

If there is a modern reading of that data it is that the Liberals have an advantage in state contests too. Though the first part of the latter half of the 20thC saw a split in Labor which could account for the poor showing in 1960 but even so the Liberal Party has done well.

Despite the wall to wall Labor parties currently at the state level it is only a new phenomenon that has occurred in the last five years. In 2000 there were four Liberal state and territory governments. This is pretty much slap bang in the middle of Howard's tenure.

Those graphs suggest the state elections are competitive despite Australia having a history of slow oscillation between parties holding power. This may have more to do with the advantage incumbency brings than anything else. If Rudd wins the election later this year then Labor has every expectation that they will be in power at the federal level for approximately nine years and three elections.

Australian voters are not very retributive. Then again, most of the state Labor governments are getting as old and stale as the Liberal government at the national level is. If there is a Labor government at the federal level I would not be surprised to see many states change government too.

The Trends in Australian Political Opinion is a must read for graph junkies. Tonnes and tonnes of interesting data that is cleanly laid out. One such fascinating graph is voting considerations.

Voters are policy driven and a larger percent is more interested in the party than the party leader or the local representative when considering how they will vote. This paints a far more sophisticated picture of the Australian voter than the celebrity horse race the media prefers.

Another interesting graph from the Trends in Australian Political Opinion on the interest of voters in politics.

If you sum the voters that are some[what] interested and a good deal interested in 2004 then it becomes 79% of respondents. This is probably why there is a vibrant Australian political blogopshere and also why the op-ed columnists in the daily rags have a ready audience for their meat throwing trolls.

The republic issue from the Trends in Australian Political Opinion.

Which is consistent with what we know already. Australians are predominantly a republican people (and democratic people from the other graphs). There is enough flag disillusionment that a pluralist flag response would work as a policy; and finally the Queen has lost her position of having civic meaning or purpose in Australia. The graph is not really showing anything new.