Today’s Newspoll on the Budget response included the answers to the question:

Overall, would you say the Budget announcements have made you more likely to: vote for Labor/vote for the Coalition/neither?

Though only 12% of voters had rated the Budget as bad for the economy and just 14% of voters thought that they would be worse off as a result (that question is not in The Australian‘s tables, but you can get it from Newspoll’s website), 26% declared themselves more likely to vote Labor as a result, compared to only 19% of respondents who said that they were more likely to vote for the Coalition.

This does seem counter-intuitive and counter-other results. A Galaxy poll reported the previous day had found only 12% less inclined to vote for the Coalition as a result of the Budget and 16% more inclined. And both polls had the two-party preferred bouncing around in the margin of error (Newspoll putting Labor up, Galaxy down) rather than swinging decisively to Labor, as their 7 percentage point Newspoll-claimed premium won from the Budget would seem to indicate.

The trouble with these questions is that some poll respondents answer them strategically, intending to give the party they support anyway a boost. Unsurprisingly, when we look answers to these questions divided by ‘political support’, only 3% of Coalition voters say that the Budget was more likely to make them vote Labor and only 2% of Labor voters say that it was more likely to make them vote for the Coalition.

Newspoll’s ‘more likely to vote Labor/Coalition’ seems to prompt a partisan choice more than Galaxy’s ‘Has the federal Budget made you more or less inclined to vote for the Coalition’, perhaps because it expressly mentions Labor.

But either way I think the question is of little value, especially when it is asked of respondents who have already expressed a clear preference for one or other of the parties.