AAP Image/ West Australian/ Nic Ellis
AFTER visiting the governor-general at around midday on Sunday 5 October 2001, John Howard returned to Parliament House to announce that a federal election would be held six weeks later. The signs were looking good for the government: the arrival of the Tampa in August seemed to have added to longer-term fears about “border protection” and the September 11 attacks in the United States had given the broader issue of national security – traditionally strong ground for the Coalition – a dramatic new edge.
But the prime minister’s advisers weren’t leaving anything to chance. Viewers of all the major evening news bulletins saw a visual account of the day’s events in Canberra that had been carefully orchestrated by the prime minister’s principal media adviser, Tony O’Leary. To the usual visual elements – the prime minister driving to see the governor-general at Yarralumla, the prime minister announcing the election, opposition leader Kim Beazley responding – O’Leary added two new moments: the prime minister crossing his private courtyard before leaving for Yarralumla and, later, the prime minister meeting earnestly with his senior advisers. “All networks ran the O’Leary-orchestrated footage, giving Mr Howard the lion’s share of the images on the news,” wrote journalist Dennis Atkins a few days later. “Kim Beazley’s picture was confined to his formal news conference.”
Marginalising the Labor leader on the first day of the campaign was part of a strategy to present John Howard not just as the obvious choice but also as the only choice as prime minister in a time of adversity. Over the next month and a half the Coalition ran a campaign dominated by national security, focusing not so much on the terrorist attacks in America (though they were certainly part of the mix) as on the boats that had been arriving off the Australian coast over the past two years. Immigration Minister Philip Ruddock, the cabinet member most closely identified with asylum seeker policy, received the loudest applause at the Liberal Party campaign launch, and at the same event the prime minister made his infamous declaration, “We will decide who comes to Australia and the conditions under which they come.” The government was returned to office on 10 November and John Howard, who had faced oblivion earlier in the year, began his ascent into the pantheon of electorally dominant Australian political leaders.
It’s this three-month period, from the first sighting of the Tampa to the poll in November, that still has Australian politics in its grip. Heading into another election campaign this year, journalists and commentators are once again using the 2001 election result to argue the case for the electoral potency of “boat people.” As the Sunday Age’s Josh Gordon put it a few days ago, asylum seekers are “a red-hot political issue, particularly for ‘Howard battlers’ in marginal electorates.” But could it be that the Tampa, no matter how controversial, was not a decisive factor in the 2001 election result?
THE GOVERNMENT had started 2001 in a dire state. Polls during January showed that support for the Labor opposition was up to 10 percentage points higher than for the government on the primary vote (not all pollsters were issuing two-party-preferred estimates at this point) and the figures only worsened over the next eight weeks. In early March Newspoll reported Labor “surging to its biggest poll lead over the Coalition in more than six years”; according to ACNielsen, Coalition support had slumped to its lowest level since it began polling in 1972.
A rattled government snapped into action. The controversial new business activity statements were simplified, a plan for tighter regulation of family trusts was dumped, the first home buyers grant was doubled and, as Murray Goot writes in his analysis of the 2001 poll data, the government “embarked on a spending spree that, by the end of the year, would see it listed for the first time by ACNielsen as the biggest advertiser in the country.” Most costly of all, the Coalition cut the fuel excise rate and abolished its indexation.
At first, the polls scarcely moved. It wasn’t until after the May budget – with its generosity to pensioners and self-funded retirees – that support for the government began to pick up. A better-than-feared result for the Liberal candidate in the Aston by-election in mid July signalled some further improvement, but by the time the Tampa hove into view the Coalition was still contemplating the prospect of a significant election loss.
At some point, meanwhile, the government had begun planning for a toughening of policy towards boat arrivals, including a role for the military, and the arrival of the Tampa was a perfect opportunity to test it. As a senior press gallery journalist, Peter Charlton, wrote after the election, “Army sources have told this writer that the [Special Air Service] troopers had been training for such a contingency for some time.” On 29 August, SAS troops boarded and took control of the Tampa, which had gone to the assistance of a boat in trouble and now had over three hundred asylum seekers aboard. The troops’ job was to prevent the captain from disembarking his passengers on Australian soil. Labor had supported the government’s initial border protection measures but baulked at the border protection bill the government produced that day.
Despite the enormous controversy it generated, and despite the backdrop of rising numbers of boat arrivals, the Tampa gave the government a relatively modest, if useful, lift in support on the primary vote – an average of 3 percentage points across the three main polls, Newspoll, ACNielsen and Morgan. More importantly, though, two out of three of the polls had Labor’s primary vote either steady or down by a statistically insignificant 1 point. Two-party-preferred, ACNielsen recorded a 1 point fall in Labor’s support (from 53 to 52 per cent), while a dramatic reduction in Morgan’s sample size (from nearly 2000 to fewer than 800) might explain the size and fluctuations in the Labor lead (57, 52 and 55 percentage points) it reported over the period of the controversy. Newspoll published no two-party-preferred figures during this period.
These and other polls around this time attempted to gauge which issues were on voters’ minds, and it’s by comparing the two types of question – voting intention and views about “boat people” – that we get a first clue as to why border protection might not be the vote-changer it’s generally regarded to be. A Newspoll conducted two weeks after the Tampa was turned away found that 50 per cent of respondents nominated immigration as a “very important” issue – 20 percentage points higher, Murray Goot points out, than when the same question had been asked in June. Goot goes on: “On the question of whether the government or the opposition was the ‘best party to handle immigration,’ the Coalition opened up a lead of 19 percentage points; in June, the Coalition and Labor had been virtually tied.” Yet overall support for Labor vote, as we’ve just seen, had scarcely changed, with the modest rise in support for the government coming from the minor parties. In other words, the fears about asylum seekers elicited by the direct question about the issue didn’t cross over to voting intention.
What did eventually change the electoral equation were the terrorist attacks a fortnight later, on 11 September. Most polls recorded a decisive shift (Newspoll: a 6 point rise in the Coalition’s primary vote, almost entirely mirrored by a 5 point fall for Labor support; ACNielsen: a 9 point rise in the Coalition’s primary vote and a 7 point fall for Labor) and this new balance of support persisted through to the early weeks of the campaign. It was the enormous shock of the events in New York and Washington rather than the arrival of the Tampa that shifted sentiment in favour of the Coalition. The Coalition strove to link the two issues, but – as we’ll see – even the combination was nowhere near as potent as it’s often been portrayed.
Despite these very recent and dramatic events, something very interesting happened once the election campaign got underway. The government’s border protection bill and the opposition’s wariness had sharpened the previously blurred differences between the two parties to the extent that anyone who wanted to vote for “strong border protection” could choose the real thing, John Howard, or opt for a reluctant and partial border protector, Labor leader Kim Beazley. Yet, as the days went by the trend in support for Labor was upwards. The first Newspoll after the election was called had the party at 43.5 per cent of the two-party-preferred vote and the Coalition on 56.5 per cent. Over the next five polls, Labor’s two-party share worked its way up to 49 per cent and then fell a little, finishing at 47 per cent on the eve of the election. ACNielsen’s findings followed a similar pattern, while Morgan had Labor performing consistently more strongly. On election day Labor’s two-party-preferred vote was 49 per cent.
Beazley’s strategy had been to focus on health and education, and it seems to have worked remarkably successfully in the circumstances. The Coalition government, with all the benefits of incumbency in a time of crisis, managed to attract just 51 per cent of the two-party-preferred vote. In New South Wales – heartland of the Howard battlers, where the Coalition recorded its largest swing, 3.1 per cent – only 51.6 per cent of voters, each of them casting their ballots in the wake of the Tampa and in the shadow of 11 September, preferred the Coalition. In western Sydney, the swing was larger again, but not as large as the swing to Labor at the previous election. Overall, the government increased its majority in the House of Representatives by just two seats, from twelve to fourteen.
WITH changed international circumstances and the Coalition’s punitive policies doing their work, debate about asylum seekers was virtually absent from the following federal election, in 2004. But another three years later there were signs that the Coalition was gearing up for a renewed campaign on the issue. In mid August 2007, with polls showing that the government was facing certain defeat, the immigration minister, Kevin Andrews, announced cuts in the government’s humanitarian quota for African refugees. Conditions in Africa had changed, he said, with the number of conflicts declining and more refugees returning home. (Figures from the UN High Commissioner for Refugees did show a fall in the number of refugees in Africa during 2006, down 6 per cent on the previous year, but Sudan – the main source for Australia’s humanitarian program – remained the third-largest source of refugees in the world.)
Questioned six weeks later about the decision, the minister told listeners in Melbourne that “some groups” – Africans, that is – “don’t seem to be settling and adjusting into the Australian way of life.” Coming just a few days after a fatal attack on a Sudanese man in Noble Park, the comments were acutely insensitive and badly timed; it soon became clear that they were also based entirely on what a member of the minister’s staff called “anecdotal evidence.” As the furore continued, Mr Andrews’s office released a few pages of details later in the week of the interview. One newspaper called this a “dossier,” but it simply listed a series of unsubstantiated allegations about African communities alongside some statistics on the changing composition of Australia’s refugee intake and a brief defence of the humanitarian program.
The controversy continued over the next few weeks, with several senior police officers disagreeing publicly with the minister’s comments. But the most interesting aspect of this controversy, coming just a few months ahead of the 2007 federal election, was the fact that it suddenly disappeared from sight. In mid October Kevin Andrews fell silent, and the Coalition largely steered clear of asylum seekers and refugees during the election campaign.
After Labor won the election on 24 November I asked a Liberal MP why the Coalition hadn’t run harder on an issue that was still generally seen as having clinched the 2001 election. He described a combination of factors, including disaffection with Kevin Andrews over his mishandling of the Haneef case and concern among those Coalition MPs who were already known as dissenters on asylum-related questions. But, he said, the vital factor seemed to have been the fact that polling had revealed that the issue had “lost its purchase” in the electorate. The party might also have been anxious not to revive memories of its role in the “children overboard” affair.
IN 2001 the Howard government thought it won because of the Tampa, but the evidence shows otherwise. In 2007 party polling revealed that attacks on African refugees weren’t gaining any traction in the electorate, which must have raised questions about the longer-term potency of asylum seekers as an election issue. Surely the parties realise now that the anxiety about “boat people” in 2010 won’t necessarily translate into shifts in electoral support?
As speculation about the date of this year’s election intensified in June, George Megalogenis and Rosanne Barrett set out to discover what happened to the million or so voters who once supported Pauline Hanson. “Neither side will say this out loud,” they wrote in the Inquirer section of the Australian on 19 June this year, “but the Hansonites were part of the reason Kevin Rudd won so handsomely in 2007. They joined the rest of the nation in rejecting Work Choices, and in embracing the idea of Rudd as a younger version of Howard.”
You might expect worries about boat arrivals to galvanise that million-strong group of potentially footloose voters and give Tony Abbott an opportunity to win back support from Labor. But as Megalogenis and Barrett point out, the opposition leader recorded “his biggest drop in personal rating so far” in a Newspoll conducted over the two days after he released the Coalition’s new Pacific Solution policy on asylum seekers. The rating only returned to its previous level after Mr Abbott kept a low profile for most of June.
After interviewing voters in the electorate of Forde, part of Pauline Hanson’s south-east Queensland heartland, the two journalists reported that healthcare, the cost of living and taxation are the key issues, “not boat people or climate change.” They go on: “The Liberals can get [Forde] back with a swing of 3.4 per cent on the new boundaries. But they won’t do it on the issue of border protection, which was barely mentioned when Inquirer tested public attitudes this week.”
This might fall under the heading of “anecdotal” evidence, but it points to one reason why there’s such a gap between what we think we know about attitudes towards asylum seekers and refugees and what we think that means for elections. When people are asked by polling companies what they think about “boat people” or “refugees,” or even about immigration in general, they express strong feelings. But if they’re asked which issues are in their minds when they’re making a decision about how to vote, these issues retreat into the background. They will certainly influence some individuals to vote either way, but overall they don’t register in relative levels of party support.
Even in 2001, when the borders were at the centre of the election campaign, this pattern was clear. The Australian Election Study, a detailed survey of just over 2000 voters, posed both sorts of question. In response to the question, “When you were deciding about how to vote, how important were refugees/asylum seekers to you personally?” 48.6 per cent said “extremely important” and a further 32 per cent said “quite important.” (Of course, saying the issue is important doesn’t mean you’re hostile to asylum seekers or refugees.) When respondents were given a list of twelve issues, from education to worker entitlements, and asked to nominate which were either the most or the second-most important factor in their voting decision, much lower numbers put “refugees and asylum seekers” first (11.9 per cent) or second (10.8 per cent). Health, education and tax all scored more highly, in that order. And if respondents had been asked to assemble their own list of important issues (as the voters of Forde were, in effect, by Megalogenis and Barrett), refugees and asylum seekers might barely have figured at all.
But despite all this evidence, the issue of asylum seekers and refugees has hung onto its air of electoral potency. The reason is probably quite simple. The conventional narrative of the 2001 election resonates for many supporters of refugees just as much as it does for people who want to close the borders by whatever means are necessary. For those who are sympathetic to these new arrivals, the assumed electoral benefit of closing the borders fits into a longer-term narrative about Australian racism as a political force; for those who are hostile, it lends support to policies like the Pacific Solution. On the available evidence, both views are wrong, which means that this time around the government should have resisted the pressure to announce hasty solutions to an electoral problem that doesn’t exist. •
Peter Browne is editor of Inside Story and author of The Longest Journey: Resettling Refugees from Africa (UNSW Press, 2006).
4 Comments
Boat arrivals feed into other concerns – eg, law and order, security, terrorism, control -and in 2001 we had a government encouraging this blurring of boundaries. A Newspoll conducted in September 2001 (just after Tampa) found that 72 per cent of people rated ‘immigration and boat people’ as most likely to influence their vote, 13 per cent said they had changed their vote on the basis of ‘the refugee boat people coming to Australia’, 66 per cent now saw defence and terrorism as influencing their vote. There can be no doubt that boat arrivals are influential but there is always more to the story.
Arthur Sinodinos has said that talk of the children overboard affair again in the 2004 election campaign only helped the Howard Government. Every time the subject was talked about in those years it favoured the Coalition (and still does). It gives them a platform to sound tough and they can tap into a powerful urban mythology. They are not in Government at the moment so perhaps it has less potency. The obsession with which people want to focus on this subject cannot be underestimated.
In 2007 Andrews was talking about African refugees who had come under the humanitarian program, not boat arrivals, and it was a time when we were having few boats (hence his choice to turn to African refugees). It is only when the boats are coming that the hysteria rises and anger on a whole range of subjects becomes magnified around the boats.
I think it is clear that the subject does have the power to attract and disgust voters either way, but by how much is debateable. For a busy person it is hard to distinguish between the parties in complex policy areas but on boat arrivals everyone thinks they are the expert and perhaps they feel they have a more simple choice.
Thanks, Susan. I think what you’re saying in your first para reinforces my point. The key word here is ‘influential’ and the only real test is how people actually vote. Some people certainly told pollsters in September 2001 that they were influenced by ‘border protection’, but when they were faced with a ballot paper in November 2001 they didn’t show much evidence that it was on their minds. If the issue did influence votes then it must have influenced a more or less equal number of votes either way. The Coalition managed a swing of around 2 per cent, but that was from a very low base (due to the GST in 1998) and in a climate heavily influenced by the September 11 attacks. People like Peter Reith tried to link the attacks to boat people but if you look at the pattern of support for the government in the polls it’s Sept 11 that made the difference.
Talk of children overboard might indeed influence the tough-guy image of the Coalition, as Arthur S. says, but I can’t see any evidence that it influenced the 2004 or 2007 election. In 2007, I’d argue, the African refugees were being sized up as proxies for asylum seekers – the niceties of onshore versus offshore applicants didn’t really figure.
Cheers,
Peter Browne
I do agree with what you are saying Peter on a number of levels but I think it is impossible to actually separate voting intentions when those factors all blend together. I think it does take more than one issue to create a perception of toughness and I guess after many years of discussion on this subject I am not convinced of the effects on voting. Goot and Sowerbutts make a point in relation to the 2001 election in their paper ‘Dog Whistles and Death Penalities’
quote: ‘Law and order, like border protection, leverages a concern with security that is both widespread and wide-ranging. This helps explain why, during the election, the Coalition campaigned hard in many electorates against drugs, and why Labor, some months after the election, created a shadow portfolio that covered not only border security, but ‘gang violence’, ‘organised crime’, and ‘drug abuse’’
Just before the Tampa incident Philip Ruddock was making connections between ‘illegal migration, heroin dependency, some instances of prostitution and associated criminal activity’.
I agree that the African refugees were being used as proxies by Andrews, but it didn’t work for a number of reasons, including that they were not boat arrivals and we were not experiencing arrivals at the time. The debate seems to have less bite when we talk about people already in residence, although there is still some resonance as expressed by the tabloid commentators.
Thanks for writing this. We have enough mythology operating in the debate and it is refreshing to read some deeper analysis that doesn’t take anything for granted.
Murray Goot has kindly sent me “Explaining Howard’s Success: Social Structure, Issue Agendas and Party Support, 1993–2004,” an article he co-wrote with Ian Watson, which includes an analysis of the 2001 election using Australian Election Study data. The full article, in the June 2007 edition of the Australian Journal of Political Science, is here (PDF), and the relevant section is quoted below. Although it tackles the issue in a different way from my article, above, it throws light on why the 2001 election result was closer than I suspect it would have been if the Tampa had been more important and health, education and unemployment – and, interestingly, the GST – had been less important. – Peter Browne
2 Trackbacks
[...] ties with the Liberals was a delusion. The 2001 election was when this myth really took off, but any objective reading of the polls of that time would have exposed the reality that it was 9/11, not Tampa that changed the [...]
[...] working class voters and, especially, in western Sydney. But as this blog has argued before, and Peter Browne and Murray Groot have set out in more detail, this supposed ability of Howard to tap into the darker [...]