Qualitative research

Bryan · Monday 19 March 2007 · 6:29 am

Today’s Australian reported on some social research from the Ipsos Mackay Group.

The findings contain a mixed blessing for Labor. While voters are interested in the new Opposition Leader, they still don’t know him. There is also an expectation that Mr Howard will “pull a rabbit out of the hat” at the last minute to secure a fifth term for the Coalition.

The danger for Mr Howard is that his core constituencies of middle-income families and the over-55s are beginning to fret about the future and are considering the switch to Labor.

Offsetting this trend is the rising conservatism of the so-called Generation Y - the younger-than-30 voter bloc that has never known recession.

The report went on to argue that while the opinion polls have Rudd well and truly in front, that support is soft. The report described it is as “registering interest and relief rather than a commitment to vote Labor”. The big downside for Rudd is that the voters still do not feel they know him.

Howard’s weakness is not interest rates as such, but the uncertainty implicit in WorkChoices and his Government’s perceived lack of commitment on the environment.

The report concludes with the observation that “despite the swing to Labor, people expect Howard to win. They are just waiting to see how he is going to do it.”

It is easy — perhaps too easy — for the statistician in me to dismiss qualitative polling out of hand. However, while the qualitative methodology does not allow us to make generalised predictions with a known degree of confidence, they do allow us to better understand why people are responding in the polls the way they do.

In my reading, this report appears quite equivocal. In that, it is consistent with the betting markets, that currently have it as neck and neck between Howard and Rudd. It points to some people considering a change of vote, but not to a wholesale mood for change in the electorate. At the moment, if there is to be a change of government, the report would suggest it will not be with the exceptionally high two-party preferred vote share for Labor that we see in the polls.

Hat tip to rafyMP.