Making sense of the polls

Making sense of the polls

Some Queensland Poll Issues

These issues are actually not specific to Queensland, but they apply most to Queensland at this election.

The Sophomore Surge Effect

At the 2010 election, Labor lost seven Queensland seats to the Coalition. Six of these seven seats are currently held by the Coalition on 1-3% margins. However, here the pendulum margin should not be taken literally, owing to the “sophomore surge” effect.

The sophomore surge happens because candidates who have just won an election are usually not well known to the electorate. In the three years that follow, a holder of a marginal seat will make himself or herself better known to the electorate, while the challenger may not be well known. As a result, sitting members will generally perform about 1% better in swing terms than their party as a whole at their first re-election attempt.

After the first re-election, the sitting member’s personal vote is then factored into the margin, and a statewide swing that exceeded that margin would be expected to see the sitting member lose. When a popular sitting member retires, their party will suffer the loss of that member’s personal vote, and so the swing in that seat against the party will probably be higher than the statewide swing.

In 2010, seven of Labor’s Queensland sitting members lost, and now the sitting Coalition members are likely to pick up a sophomore surge. As a result, Labor needs to overcome a double sitting member effect to gain the marginal Queensland seats, and will probably need a 4% statewide swing to make substantial gains in Queensland. Currently, the Poll Bludger’s Bludgertrack is showing zero swing in Queensland, so Labor is well short of where it needs to be. Whether Peter Beattie’s return to politics can boost Labor’s Queensland vote is still to be determined.

Peter Brent has more on the sophomore surge as applied to Queensland.

Will the Pollsters Underestimate Support for Fringe Right-Wing Parties?

At the 1998 Queensland state election, One Nation received 22.7% of the vote; the final pre-election Newspoll had predicted a One Nation vote of 18.5%. Similarly, in 2012 Katter’s Australian Party (KAP) won 11.5%, 2.5% higher than the 9% predicted by the final Newspoll.

At this election, the KAP and Palmer’s United Party (PUP) are two Queensland-based right-wing fringe parties that could have their support underestimated by the pollsters. Respondents are more willing to tell an interviewer that they’re voting for a mainstream party than for a fringe party.

If the pollsters are underestimating votes for fringe right-wing parties, and if that support comes at the expense of the Coalition, then pollsters may be slightly underestimating the swing to Labor in Queensland, though on current Queensland polling this won’t matter much. Votes for the KAP or PUP won’t necessarily come back to the Coalition as preferences.

We produce knowledge-based, ethical journalism. Please donate and help us thrive. Tax deductible.