Howard the campaigner: legend or reality?

There is a narrative in Australian political commentary that the polls must narrow before the election day. This song has been sung by journalists and opinion writers ever since Kevin Rudd took over the Labor leadership nearly a year ago: “When would the honeymoon end?” mused The Australian’s Paul Kelly to Barrie Cassidy on the ABC’s Insiders program back in March. Writers like Glenn Milne and Andrew Bolt have been wondering ever since. Wind the clock on six months to late September and the Insiders crew is still at it: Malcolm Farr, Annabel Crabb and Gerard Henderson all agreeing with Cassidy that Rudd “could not defy gravity for ever”.

But he seems to be. Is it a certainty that he comes back to the field?

The received wisdom relies on a “ground hog” theory of voter behaviour where each election is viewed through the prisms of the past. One of the enduring myths of John Howard’s premiership is that each election win is a carbon-copy of the one preceding it - the wily Howard wins over the pragmatic and risk-averse voter with a hip-pocket campaign focused on economic record and national security.

But any objective assessment of Howard’s election campaigns shows that each were shaped by their own set of unique circumstances. Back in February this year I warned in this journal that the “polls must narrow” narrative could not be taken as a given. I offered three reasons then, and I believe they have held true throughout 2007:

• Beazley had Labor in front on the polls for most of 2006, which was a stark contrast to other election lead-in years. This cost Howard a good deal of political capital; a significant number of voters had already shifted to Labor prior to Rudd’s appointment;
• Rudd is not Mark Latham;
• the “cost of government” makes winning each successive election more difficult.

Now that an election is almost upon us, it is appropriate to look a little closer at the question of what happens to voter opinion during modern Australian election campaigns. The good news for John Howard is that an analysis of the eight federal elections held over the past 20 years shows that, on average, the gap between the parties does narrow during the campaign - the bad news is that it is nowhere near enough.

A comparison between the Newspoll taken one month out from each election with the actual election results since 1987 (acknowledgement to Peter Brent for some of the data), shows an average contraction of 2.7 percentage points in the two-party preferred vote during the campaigns. Currently there is at least ten points (probably more) separating the parties so, if this election runs to average, a 7.3 per cent margin is well and truly a comfortable Labor win.

To make the news worse for the government, in the one example from the sample of a big campaign movement, in 2001 when the gap narrowed by eleven points, it was the opposition that was behind and made up ground to make the elections close. There is no example in recent history where an incumbent has had to narrow a double-digit gap to win.

Another important element of the received wisdom is that John Howard is a good campaigner and is sure to improve the Coalition’s position by dint of his superior campaign skills. After all, commentators like Paul Daley argue, Howard is “good on the stump” and sure to win a campaign. The problem is that much of this argument relies on some sort of group faith and not much on analysis.

Again, looking at Newspoll one month away from the five elections Howard has contested as leader, and comparing each against the respective election two party preferred, gives an average variation of … zero. (1987, +2; 1996, 0; 1998, +.5; 2001, -5.5; 2004, +3.)

On average, Howard as leader neither loses or gains during his campaigns. Notwithstanding, trying to find something positive, of his five elections he only went backwards in one, but it was a big 5.5 drop in 2001. Being charitable and leaving that one out, it might be concluded that Howard, on his record, is a modest campaigner with the capacity to make up a couple of percentage points or so. It is hard to say anything more than that.

That brings us to the scariest question (from the government’s viewpoint) of all: what happens if Rudd improves from here?

If it is not a given that Howard will win the campaign, surely it is equally conceivable that he could lose it? The September monthly aggregate of the four main opinion polls (Newspoll, Morgan, ACNielsen and Galaxy) has the Coalition on 42.5 per cent of the two-party preferred vote. If you plug that result into Bryan Palmer’s election calculator the Coalition win 44 of the 150 House of Representative seats. And that’s just if Howard holds his current ground. Ouch.

The inference from the data is that John Howard’s record is against him winning this election. That is not to say he can’t, just that to do it he will have to live up to his legend rather than his reality.