Counting in Saturday’s Higgins and Bradfield by-elections started with a bang. The first booth to report in Higgins was Toorak West, apparently at the epicentre of the Melbourne’s wealth belt, and it pointed to a stunning anti-Liberal swing of 22%. Such a result would have demolished the existing margin and sent a radically left-wing member to Canberra to represent what had historically been a prestige Liberal Party seat. A rare degree of familiarity with local geography was needed to observe that the booth designated Toorak West was a crucial 1500 metres from that which filled the role at the 2007 election, being situated on South Yarra’s trendy Chapel Street shopping strip rather than in the residential heart of Toorak proper. Only a handful of additional booths were needed to make clear that the evening was not going to be quite as exciting as all that.

Australia’s electoral history records a few occasions when by-elections sent messages that could be heard loud and clear. Bass in 1975 and Canberra in 1995 are the most famous examples, and there could be no misreading the 20% swings Labor suffered at state by-elections in New South Wales last October. However, the story behind by-election results is usually a lot more prosaic, particularly in the increasingly common circumstances where one major party declines to field a candidate. This is certainly true of Saturday’s contests in Higgins and Bradfield, which one of the more neutral commenters on my blog described as “another Rorschach blot” on which partisan commenters were “seeing their own prejudices”.

Certainly the result has provided a psychological boost for Tony Abbott, and it has understandably been reported as such in the media. Before the event, veteran election analyst Malcolm Mackerras had gone so far as to tip a shock Greens victory in Higgins caused by disgust over the Liberals’ sabotage of the emissions trading scheme. Newspoll chief executive Martin O’Shannessy was quoted in The Australian saying he expected the seat would go to preferences, while Liberal insiders were quoted late last week saying they believed Higgins to be “line ball” (I personally said I was “tipping an uncomfortable night” for the Liberals, but expected them to prevail). In the event, the primary vote for Higgins candidate Kelly O’Dwyer was little different from the 54% last recorded by Peter Costello, while Paul Fletcher went equally untroubled in Bradfield on 56% despite suffering a 3% dip on Brendan Nelson’s vote in 2007.

An opposition that loses ground on the primary vote in a field vacated by the governing party would not normally have too much to crow about, but coming in the aftermath of the tumultuous fortnight, the result offers the Liberals limited grounds for comfort. The last time the party went head-to-head with the Greens in a safe-seat by-election was in September last year, in Alexander Downer’s old seat of Mayo. On that occasion the Liberal vote dived nearly 10%, putting the Greens within 3% of a boilover after distribution of preferences. Ten days later, Brendan Nelson was dumped as the party’s leader. By contrast, Tony Abbott has promptly demonstrated that some of the more alarmist interpretations of his electoral prospects have been misplaced.

On one level, seasoned election watchers should have seen this coming. It was often said before the 2004 and 2007 elections that the “doctors’ wives” would give the Liberals rude shocks in their electoral heartland due to the Iraq war and the Howard government’s refusal to sign the Kyoto protocol, but nothing emerged to justify the hype: both elections were decided in the usual fashion in the middle-income mortgage belts. The party again retained the loyalty of most of its wealthy support base on Saturday despite obvious disquiet over some of its policies, providing support for Paul Fletcher’s election night assertion that climate change had been but “one of many important issues”.

Even so, booth results indicate at least some well-heeled voters used the occasion to send the Liberals a message. The wealthiest and safest Liberal areas of Bradfield swung against the Liberals by 6% on the primary vote, compared with break-even results in the more marginal south (Chatswood and Willoughby North) and north-west (Asquith and Hornsby). In Higgins, inner-city Prahran and middle-income Carnegie and Hughesdale swung 2% to the Liberals, while the solidly Liberal remainder went 2% against. It’s tempting to speculate that wealthy voters in Higgins proved more inclined to stick with the Liberals than their Bradfield counterparts due to distaste for Greens candidate Clive Hamilton. Notwithstanding that Hamilton’s 33% was a record for the Greens in a House of Representatives seat, the party’s vote would probably have been even higher if it had opted for a locally oriented candidate with a more middle-of-the-road image.

Even more strikingly, the marginal areas of Higgins swung nearly 7% to the Liberals in two-party Liberal-versus-Greens terms (going off Antony Green’s estimates of the Liberal-versus-Greens results from the 2007 election). It appears a significant number of Labor and pro-Rudd swinging voters felt sufficiently hostile to the Greens to park their votes with other minor parties and send preferences to the Liberals, either out of conscious choice or through observance of how-to-vote cards. These areas gave a particularly strong 6.5% vote to the Democratic Labor Party, whose preference recommendation had the Liberals ahead of the Greens.

As far as it goes, the indications for the Liberals from Saturday’s result are positive. Much of the damage from the switch on climate change looks likely to be concentrated where it can’t do them much harm, and the Abbott ascendancy does not look to have caused them any problems they didn’t suffer already. However, what the by-elections do tell us is much less important than what they don’t, namely the state of the Labor-versus-Liberal contest in marginal seats. Nothing in the results refutes the mass of polling showing Kevin Rudd maintaining a secure hold on the loyalties of urban swinging voters, or suggests the coalition can hope for anything more from the next election than to limit the size of its defeat.