Victorian election was lukewarm all round

Still, only 9 seats swung to the Liberals, and none by margins that suggested their on-the-ground-campaign was particularly effective, so the ground campaign probably counted for something.

Why our respondents voted them out

I put the data into Leximancer, semantic software that we use to analyse large qualitative samples. Generally when I put voting data into Leximancer it constructs a map with Labor and Liberal voters at opposite ends. The economy and debt dominate the Liberal end and social goods like education and health the other. There is also generally a lot of mentions of the opposing leaders.

This is the map that it constructs for voting intention in the last Victorian election.

Figure 1: Thinking about your first preference vote, in a short sentence tell us why you answered that way?

The parties are arranged in a rectangle with Liberal and Labor sharing one side and the economy is more closely linked to the ALP than it is to the Liberals. Neither of the leaders is mentioned. The Greens are distant from every other political party, and most of the issues of substance such as education, health, public transport and the environment are positioned between them and Labor.

Even though non-Greens minor party voters favour the Liberals on their preferences, they are more closely related to the issues associated with Labor.

In these circumstances the Liberals did surprisingly well in the election campaign as they are not engaged with the deciding issues.

A similar pattern is visible when respondents are asked what the most important election issue was.

Figure 2: In a short sentence please tell us what is the most important issue for your vote this Saturday.

This seems to be symptomatic of a government that failed to connect with electors. They did get marks for good economic management, but voters didn't draw a line between this and delivery of sustainable services.