Voice support up in Essential poll, but it is still behind

The referendum on the Indigenous Voice to parliament will be held on October 14. A national Essential poll, conducted September 27 to October 1 from a sample of 1,125, gave “no” to the Voice a 49–43 lead, a narrowing from a 51–41 “no” lead a fortnight ago.

This is the first time since June that “yes” has gained ground between two separate polls by the same pollster. On voter strength, 42% were hard “no” (steady), 30% hard “yes” (up two), 13% soft “yes” (up one) and 7% soft “no” (down one).

The graph below has been updated with additional results from Freshwater and Morgan (see below) as well as Essential. The Essential and Morgan polls are the best pollsters for “yes”, but it is still behind with these polls. “Yes” is much further behind with other polls, including a 20-point deficit in last week’s Newspoll.

There was a large difference between Resolve and Essential’s Voice polls in June, when Essential gave “yes” a 60–40 lead but Resolve had “no” ahead by 51–49. I thought that Resolve was more likely to be right, and this opinion hasn’t changed.

Read more: Resolve first national poll to have 'no' ahead in Voice referendum, but Essential has 'yes' far ahead

In other Essential questions on the Voice, 42% of those who were “no” or undecided said their main reason for voting “no” was that the Voice would divide Australia in the constitution on the basis of race, 26% said there is not enough detail, 18% said it won’t make a difference to the lives of ordinary Indigenous Australians and 14% said it will give Indigenous Australians rights that other Australians don’t have.

By 49–26, respondents expected the referendum to be defeated.

Labor recovers in Essential voting intentions

In Essential’s two party estimate that includes undecided, Labor led by 50–45, after reaching a low for this term of 49–45 last fortnight. Primary votes were 33% Labor (up two), 32% Coalition (steady), 14% Greens (up one), 6% One Nation (down two), 2% UAP (steady), 7% for all Others (down one) and 5% undecided (down one).

The gains for Labor and the Greens on primary votes suggest that respondent preferences were better for the Coalition and cost Labor a larger lead.

Respondents were asked to look back at the response to COVID. By 42–30, they gave the federal government a good rating. Western Australia was the best state government with a 59% good rating, while South Australia and New South Wales were tied at 50% good. Queensland was 42% good, with Victoria trailing on 39% good.

By 62–11, voters agreed that the recently announced COVID inquiry should examine actions by all levels of government during the pandemic. On the death toll of around 23,800, 46% thought we had done well to keep it that low, 38% thought it was too high and 16% thought COVID was overstated and didn’t believe the numbers.

By 40–37, voters in this national poll thought Daniel Andrews had made a poor contribution to Victoria rather than a good one.

Freshwater poll: Labor only ahead by 51–49

A Freshwater poll for The Financial Review, conducted September 22–24 from a sample of 1,003, gave Labor just a 51–49 lead, a one-point gain for the Coalition since May. Primary votes were 37% Coalition (steady), 33% Labor (down one), 13% Greens (up one) and 17% for all Others (steady).

The two most recent Freshwater polls have favoured the Coalition relative to other recent polls, with last week’s Newspoll giving Labor a 54–46 lead.

Albanese’s ratings were 41% unfavourable and 38% favourable for a net approval of -3, down eight points since May. Dutton’s net approval improved two points to -10. Albanese led Dutton by 46–37 as preferred PM, down from 51–33 in May.

The Liberals led Labor by 38–29 on economic management and by 32–30 on cost of living.

On the Voice referendum, “no” led by 50–33 (50–35 in a Freshwater poll in early September). With undecided excluded, “no” led by 60–40.

Morgan poll has best result for ‘yes’ since August

A national Morgan poll, conducted September 18–24 from a sample of 1,511, gave “no” just a 44–39 lead. While this is a reversal of the 46–36 “yes” lead in the previous Morgan Voice poll in May, it’s the best result for “yes” from any pollster since an early August Essential poll gave “no” a four-point lead.

This poll was conducted using online methods, whereas previous Morgan Voice polls used SMS. The pollster expects a bigger win for “no” than the 53–47 “no” lead after excluding undecided voters as undecided are expected to break for “no”.

Morgan’s weekly federal poll last week gave Labor a 54–46 lead, unchanged on the previous week. This poll was conducted September 18–24 from a sample of 1,393. Primary votes were 35% Coalition, 32.5% Labor, 14% Greens and 18.5% for all Others.

Jacinta Allan replaces Daniel Andrews as Victorian Premier

Daniel Andrews resigned as Victorian Labor Premier and Member for Mulgrave on September 27. Former deputy Premier Jacinta Allan was elected Labor leader and premier unopposed at a September 27 Labor caucus, and former Public Transport Minister Ben Carroll deputy premier.

Had the leadership been contested, a postal ballot of the Labor membership would have been required, and this was expected to take at least a month. Allan was in Andrews’ left faction, while Carroll is on the right.

A byelection will be required in Mulgrave, which Andrews won by a 60.2–39.8 margin against the Liberals in 2022. An independent, Ian Cook, beat the Liberals but lost to Andrews by 60.8–39.2.

Andrews became Victorian premier after winning the November 2014 state election. In his nearly nine years in power, he did nothing to reform the Victorian upper house’s electoral system. The Victorian upper house is the only parliamentary chamber in Australia that still uses the group voting ticket system for its elections.

Read more: How Victorian Labor's failure on upper house electoral reform undermines democracy

Adrian Beaumont does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.