Two and a half months before the November 28 Victorian state election, a Resolve poll has the Coalition retaining a clear lead over Labor. Labor is likely to lose the election, but it’s unclear whether the Coalition will be able to govern in majority or need support from One Nation.This week’s federal polls were mixed, with Resolve having the combined One Nation and Coalition primary vote up two points to 51%, while it was down 1.5 point to 46.5% in Morgan. Morgan had the Greens up 1.5 points to a record high 16%, but they were down one point in Resolve to 11%.Labor gained a seat at an NT byelection and there are final results from the Western Australian Secret Harbour byelection.Victorian Resolve pollThe Victorian state election is on November 28. A Resolve poll for The Age, conducted September 6–12 from a sample of 1,000, gave the Coalition 28% of the primary vote (up one since the mid-August Resolve poll), Labor 25% (steady), One Nation 23% (steady), the Greens 12% (steady), independents 8% (up one) and others 5% (down one).No two-party estimate was provided. In August the Coalition led Labor by 53–47, and I am assuming this was still the case given the lack of change on primary votes.This graph will follow Victorian polls in the lead-up to the election. It starts with the polls in late July to early August, when Ben Carroll replaced Jacinta Allan as Labor leader and premier. Labor’s two-party share and the midpoint fieldwork date for the polls are shown.Resolve is the only pollster that has conducted a second poll since July. When pollsters conduct multiple polls, there will be lines connecting polls by the same pollster.In Resolve’s three-way preferred premier question, Liberal Jess Wilson had 28% (up two), Carroll 20% (steady) and One Nation state leader Warren Pickering 8% (down two).By 49–27, respondents did not think the Suburban Rail Loop was worth going into debt for. By 41–35, they thought Labor’s plan to publicly fund dental care if re-elected was not worth going into debt for.Federal Resolve pollA national Resolve poll for Nine newspapers, conducted September 6–12 from a sample of 2,250, gave Labor 28% of the primary vote (steady since the mid-August Resolve poll), One Nation 27% (up one), the Coalition 24% (up one), the Greens 11% (down one), independents 6% (down one) and others 3% (down one).No two-party estimate was given, but applying 2025 election preference flows to the primary votes would give Labor only about a 50.5–49.5 lead over the Coalition.Anthony Albanese’s net approval was down one point to -18 (53% poor, 35% good). Angus Taylor’s was up one point to +9 (39% good, 30% poor). Pauline Hanson’s was up five points to +10 (47% good, 37% poor).In a three-way preferred PM question, Albanese had 29% (down three), Hanson 25% (down one) and Taylor 21% (up two).Asked to choose between cutting spending or raising taxes to address debt and deficits, 48% selected cutting spending, 10% raising taxes and 42% an equal balance. By 59–9, respondents supported bringing house prices down (60–10 in August).On immigration, 40% said Labor’s current net overseas migration target of 225,000 was too high, 32% about right and 11% too low. That target is down from about 300,000 last year. One Nation was thought best to handle immigration by 31%, with 24% for Labor, 19% the Coalition and 5% others.Morgan pollA national Morgan poll, conducted September 7–13 from a sample of 1,583, gave Labor 25% of the primary vote (down two since the early September Morgan poll), One Nation 25% (down 1.5), the Coalition 21% (steady), the Greens 16% (up 1.5) and all Others 12.5% (up two).The 16% for the Greens is a record high for them this term from any pollster, but analyst Kevin Bonham said Morgan polls are better for the Greens than others.By respondent preferences, Labor led One Nation by 54.5–45.5, a one-point gain for Labor. Labor led the Coalition by 55–45, a two-point gain for Labor. By 2025 election preference flows, Labor led the Coalition by 53–47, a 0.5-point gain for Labor.WA Secret Harbour byelection final resultsI previously covered One Nation’s thumping win over Labor at the August 29 Western Australian state Secret Harbour byelection. In final results, One Nation defeated Labor by 57.2–42.8 after preferences, compared with Labor’s 61.5–38.5 win over the Liberals in this seat at the WA 2025 state election.Primary votes were 39.1% One Nation (up 30.7%), 24.1% Labor (down 22.4%), 17.2% Liberals (down 7.8%), 7.8% Greens (down 1.1%) and 3.7% Legalise Cannabis (down 2.1%).Although the Liberals had more than double the Greens vote, the gap between One Nation and Labor closed slightly from 15.0 points on primary votes to 14.4 points after preferences.The distribution of preferences shows that, when Liberal preferences were distributed, they favoured One Nation over Labor by 59.2–40.8. Liberal preferences at this point included 79.4% Liberal primary voters, with the rest from other sources acquired during the distribution of preferences.The Liberal primary vote preference flow to One Nation was probably stronger than when including other sources, but the flow to another right-wing party was weak enough that One Nation could lose seats if the primary votes for One Nation and Labor were close.Labor easily wins Mulka at NT byelectionA byelection in the Northern Territory local seat of Mulka occurred last Saturday after the mid-August resignation of independent member Yingiya Mark Guyula. With only two candidates contesting, Labor defeated an independent by 77.6–22.4. Labor now holds six of the 25 seats in the NT legislature, with the governing Country Liberal Party (CLP) holding 17.At the 2024 NT election, Guyula defeated the CLP by 75.2–24.8 with Labor not contesting. In 2020, he had defeated Labor by 55.1–44.9. Over 60% of Mulka’s population is Indigenous. Turnout is often low in Indigenous-heavy seats, and is only 36% so far at this byelectionThis was NT Labor’s second seat gain at a byelection this term, after they gained Nightcliff from the Greens at a March byelection.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.