Before it gets too firmly entrenched as conventional wisdom, it’s worth querying the idea that the election campaign, and particularly the anti-union theme, was a miserable failure for the Coalition.

In yesterday’s Age, Katharine Murphy and Michelle Grattan told us that “The ground Mr Howard hoped to make up during the campaign did not eventuate.” This morning Michael Bachelard, while conceding that “it’s likely the negative campaign did scare some back into the Liberal camp”, says “the real victims of the anti-union campaign were the Liberals.”

But in fact Saturday’s result was significantly better for the Coalition than they had any right to expect. Almost all year, the opinion polls have had Labor’s primary vote holding steady at 47% or better, but on election day it delivered less than 44%. The two-party-preferred swing was a little under 6%: large, but far from the massacre that the polls indicated.

The pattern of the Coalition’s recovery is also clear. The outer suburbs swung strongly, but the Liberals’ middle-class heartland stayed put.

Seats like Wentworth, Sturt, Ryan and North Sydney all held off Labor challenges; Kooyong, which was Labor’s best Victorian result in 2004, was its worst this year, with essentially no swing at all.

Prior to the election, Peter Brent theorised that the Coalition would do relatively better in outer suburban seats and continue to lose ground in the heartland. But that didn’t happen; realignment may not have been halted, but it’s certainly stalled. (Unlike most commentators, Brent refreshingly points out his own mistakes, saying on Sunday that his predictions “now look flaky”.)

Almost a century ago, Australians created a two-party system based on class, and it is still with us today. Nothing in the pattern of Saturday’s vote would be particularly surprising to the pre-World War I generation; for all the impact it made on the result, the collapse of socialism might never have happened.

The anti-union campaign may well have turned off some swinging voters, but it wasn’t designed to appeal to them: it was designed to shore up the Coalition’s middle-class base. It seems to have performed that task rather well.

Maybe some in the Coalition thought it could win them the election (misled perhaps by American experience, where voluntary voting means that appealing to the base can be a winning strategy), but my guess is that Howard and his senior advisers knew from the start that the election was lost, and the campaign was all about damage control.

None of this means that the Liberals should hold fast to WorkChoices.

Minimising a loss is not the same as winning; “wars are not won by evacuations”, as Churchill remarked. But we should at least try to set the record straight.