Losing The Republic Like It's 1999: The Polling

The current Voice referendum with its rapidly declining support has brought back memories of referendums past.  Surprisingly, given that it's our most recent referendum, there has been little information available about the polling trajectory of the 1999 Republic defeat.  The received wisdom is that support for the Republic started off very high and ended up low, as happened with several other referendums (especially the 1988 quadruple failure) and as is so far happening with the Voice.

It turns out to be not so simple.  The story of the 1999 Republic referendum polling is one where only the last few months of data are all that usable, and in public polls at least, Yes was never all that much in front.

A summary of Republic polling is available in one paper available online but I thought, surely there was more?  I could not find any usable online archive of Republic referendum polling but what I did find, to my surprise, was some polling commentary in an article I published in Togatus April 1999.  1999 model me - not then a polling analyst so please cut some slack for misusing "push-poll" for a seriously bad polling question - wrote this:

"Poll results are confusing. In the last week of January the Age/AC Nielsen showed a national 41% yes vote, but Newspoll showed 58%.[*] The latter was a virtual push-poll because it included the statement 'this will most likely mean that the head of state will not be a politician' and ARM won't get even that simple message through to a thick republican public that mistakes an extra election with people-power."

[* 59% actually. This Newspoll wording was said to have been sourced from the ARM, and was roundly and rightly condemned by opponents as out of step with other polling at the time.]

In fact, question wording was a major issue in 1999 polling, and a case where public polling itself influenced the ballot paper design for the better.

The prelude and the question question

The 1998 Constitutional Convention endorsed a model in which the Queen would be replaced as Head of State, and the Governor-General as her representative, by a President appointed by a two-thirds majority of a joint sitting of Parliament.  The President could not be a current MP nor a member of a political party when appointed (defined how? verified how?  Were these people dialling it in?)   

Polling through the early part of 1999 continued to show clear majority support for a Republic, as had been the case since about 1994.   But the few polls that referred to the proposed Republic structure found very poor results, because most Australians continued to believe that if there was to be a Republic at all, the President should be directly elected.  Questions referring to a Republic with a President appointed by a two-thirds majority of Parliament had very different results to the generic polling: Nielsen late January 1999 41% Yes 45% No, Newspoll early March 33-55, Newspoll late March 39-51, Nielsen late July 40-50.  

In early August the question of how the Republic referendum ballot paper was to be worded came to a head.  Nielsen released a poll with a 54-32 lead for the generic Republic.  But when respondents (I assume in an A/B sample test) were asked "This year you will be asked to vote in a referendum for a republic in which the president would be chosen by a two-thirds majority of the federal parliament. Will you vote yes or no in the referendum?" the result crashed to 31-55.  When they were asked "* This year you will be asked to vote in a referendum for a republic in which the Queen and the Governor-General would be replaced by an Australian president. Will you vote yes or no in the referendum?" the result was similar to the generic question, 57-33.  

This was at a relatively early stage of engagement with the referendum and I suspect that closer to the date numbers would have been much closer.  Nonetheless this polling fueled the belief that a referendum question that only mentioned how the new President would be chosen, and didn't mention that they would be a replacement for the Queen and Governor-General, would be unfair.  Shortly after, Prime Minister John Howard agreed to a compromise in which the ballot paper question would mention both aspects.  The ballot paper question was:

"A Proposed Law: To alter the Constitution to establish the Commonwealth of Australia as a republic with the Queen and Governor-General being replaced by a President appointed by a two-thirds majority of the members of the Commonwealth Parliament.

Do you approve this proposed alteration?"

This was a very fair referendum question.  

The polls thereafter

By searching NewsBank articles using the term "republic poll" I have managed to find media references to fifteen national polls that were in field after the actual question wording was set, and that appear to have asked about voting intention at the referendum as opposed to generic support.  These were five Newspolls, four Nielsens, three Morgans, two Taverners and one Quadrant.  One of the Taverners is reported in the Newsbank text capture as having a sample size of 89 (yes, eighty-nine); I have given this the benefit of the doubt and assumed that is an error.  I have also given all the polls the benefit of the doubt in that any of them might have had design issues that would cause me to exclude them if I knew, but I can only rely on media reports of them here.  

The following is a graph of the polls on a two-answer basis.  Until late October polls generally had Yes slightly ahead to even on the national vote, but in the final two weeks all the polls by the most regular players had Yes losing and the only polls with Yes level or ahead on the national vote were polls by the minor players Taverner and Quadrant, which may well have had house effects.

Even in the late Quadrant poll Yes while ahead nationwide was still losing because of a 3-3 state voting breakdown, and difficulties with state breakdowns had been flagged repeatedly in commentary, with quite a few state-specific polls available to that effect.  I've considered the state polling outside the scope of this article, but I think it's fair to say that the 1999 Republic proposal was never really winning - at least on the public polling.  And there were some very gloomy internal polls in mid-October (not included in my graph), with one by Rod Cameron getting a 39-56 result and an internal Labor poll on 38-50.  

Here's a graph with a simple trend line that will look very familiar to those who have followed my Voice work! 

    

Key to colours: Green - Newspoll, Dark blue - Nielsen, Black - Morgan, Burnt red - Taverner, Yellow - Quadrant, red star = final result

While I have again used a humble order-2 polynomial to fit these 15 data points, I think the picture is more complex, especially taking the internal polling (published post-hoc including some in Malcolm Turnbull's Republic diaries) into account.  I think a better interpretation is that polled Yes support may have crashed as low as about 43-44 with a week and a half to go, then risen to about 47 (which turned out to be an overestimate, with all of the final polls having Yes a bit too high).  It might have been something more like this:

 

Key to colours: Green - Newspoll, Dark blue - Nielsen, Black - Morgan, Burnt red - Taverner, Yellow - Quadrant,  red star - final result

A factor here that appears to have moved the major final polls and that I think can be treated as a discontinuity is that on 1 November 1999 Opposition Leader Kim Beazley said that if Yes won the referendum and Labor won the 2001 election, then Labor would hold a referendum on the direct-election model in its first term in office.  Beazley did not commit to Labor voting Yes in such a referendum, leading, as Michelle Grattan observed, to the potential for two terms in a row to have Prime Ministers holding referendums and voting No in them.  Anyway, whether it moved the polls or not, it was too little too late.

Worst Poll Ever!

This article would not be complete without mentioning the "Deliberative Poll".  All manner of media hype - around one hundred articles - accompanied the bizarre spectacle of 347 already Yes-leaning respondents (53-40) being transported to Canberra to hear two days of arguments for and against the Republic and then state their intentions.  After this 73% of them now intended to vote Yes, and their former enthusiasm for the direct election model was sharply reduced.  Initial media enthusiasm for the concept (an invention of some American professor, because of course) faded when the result was obviously silly, with claims that the Deliberative Poll was "brainwashing" and a "push-poll" getting quite a friendly run in the thing's wake.   We will not see its like again, at least I hope not.