Reports of "leaked" internal Liberal polling for next Saturday’s Pittwater by-election have provoked a frisson of excitement among those with an interest in talking up the contest. The poll reportedly has Liberal candidate Paul Nicolaou trailing his most fancied independent rival, local mayor Alex McTaggart, with a 46.5-53.5 split on two-candidate preferred. The results were a gift from "a senior Liberal who did not want to be named" to Lisa Muxworthy of the Manly Times, who smartly cultivated her source by reporting it exactly the way he or she would have wanted – a close race with the outcome to be determined by the undecided, and no luxury of a protest vote for those who normally support the Liberals. Anne Davies of the Sydney Morning Herald spruiked the contest by telling us "the parochial peninsula electorate has shown in the past that it can record large swings when a new candidate is endorsed, particularly if they are not a local". This was presumably a reference to surfer Nat Young’s strong performance against Jim Longley at a by-election way back in 1986, when sewage pollution on local beaches was a national news story. Christian Kerr of Crikey has also invoked the spectre of Nat Young while making plenty of space available for those predicting a Liberal humiliation, citing word-of-mouth evidence to suggest a swing not far shy of 20 per cent.
The Poll Bludger is not persuaded. It is axiomatic that leaked party polls are to be taken with a grain of salt, for reasons which hardly need explaining. Most of those reporting the Liberal figures seem to be conscious of this, with the Sydney Morning Herald providing a nice quote from Alex McTaggart about "a plot to scare the little old ladies into voting Liberal". I have a very particular theory on this occasion, which rests on two pillars – firstly, that the best lies are based on a foundation of truth; secondly, that political parties know this (and much else about the art of deception) better than anyone. On this basis, I suggest that the Liberals have indeed conducted polling that has them on 46.5 per cent, but that this is on the primary rather than the two-party preferred vote. This sounds about right – a sobering 12.9 per cent primary vote slump that would give the Liberals no cause for optimism about the 2007 election, but not enough to threaten their hold on so safe a seat. It would not be too hard to persuade an inquisitive journalist that all minor candidate votes should indeed be added to McTaggart’s two-candidate preferred score, since they will all "give" him their preferences. But in reality, 46.5 per cent would be enough for Nicolaou to win quite easily.
This brings us back to the Nat Young precedent. Leaving aside the fact that some who will vote next Saturday were not even born in 1986, it is worth noting that Young actually fell some way short with 46.9 per cent of the two-candidate vote – an excellent result, but still a clear victory for the Liberals. More importantly, this was in the days when the opportunities of optional preferential voting had yet to catch on. Young gathered 72 per cent of preferences from the 33 per cent who voted for other candidates, with an exhaustion rate of just 4.1 per cent. These days, a third of that vote can be expected to disappear courtesy of those who "just vote one". Even assuming the Liberal poll results are not completely fictitious, it would be quite astonishing if they honestly accounted for this.