Ontario Premier Doug Ford has seen his popularity dip in the wake of an MPP expense scandal but the total impact appears to be “marginal rather than transformational,” a new survey suggests. Abacus Data surveyed 1,000 Ontario adults on behalf of the Toronto Star from July 23 to July 29. It found that support for the PCs among committed voters has fallen from 42 per cent to 40 per cent when compared to a similar survey that was conducted in June. Support for the Liberals has risen from 30 to 32 per cent while the NDP also saw a two-point jump to 21 per cent. The Greens are at three per cent (down two points) while the remaining three per cent of respondents said that they intend to support another candidate. Doug Ford’s personal popularity also appears to have declined. About 34 per cent of respondents had a positive impression of the premier compared to 45 per cent who had a negative impression, pointing to a - 11 per cent favourability score (down two points from June). NDP Leader Marit Stiles was at 34 per cent positive and 25 per cent negative while interim Liberal Leader John Fraser was at 26 per cent positive and 23 per cent negative. “This is movement in an unfavourable direction for the government, but it is not evidence of a wholesale realignment. The PCs remain clearly in first place, and the opposition vote remains divided between the Liberals and NDP,” an analysis from Abacus Data CEO David Coletto accompanying the latest figures states. Ford came under fire earlier in July after it was revealed that a number of PC MPPs who live in the Greater Toronto Area have submitted thousands of dollars in expense claims for Toronto hotel nights. The scandal prompted former Minister of Tourism, Culture and Gaming Stan Cho to resign from cabinet and agree to pay back approximately $16,000. Ford, however, has continued to face criticism from opposition leaders at Queen’s Park who have suggested that members of his caucus are “living large” at the expense of taxpayers. Abacus Data says that about half of the Ontarians it surveyed indicated they had heard either a lot (21 per cent) or some (31 per cent) about the scandal. However, about one-third of those surveyed (31 per cent) said that they had heard “nothing at all” and 17 per cent offered that they had only heard “a little.” Abacus Data also asked Ontarians whether the scandal has changed their impression of the Ford government. About four out of 10 respondents (41 per cent) said that they had a much more negative or somewhat more negative impression of the Ford government as a result of the scandal. Another 35 per cent of respondents, meanwhile, said that the scandal had no impact and another 17 per cent said that they have a more favourable impression of the Ford government as a result of what they have seen and heard about the story. The numbers were largely split along party lines. Only 28 per cent of those who voted for the Tories in the 2025 Ontario election said they have a more negative impression of the Ford government but that number rose to 49 per cent among Liberal voters and 63 per cent among NDP voters. In his analysis, Coletto said that the numbers show that there has not been a collapse in PC support in the wake of the expense scandal. But he said that “the underlying attitudes point to a larger potential risk” for Ford. “If the controversy continues, if more MPPs or expenses become part of the story, or if the issue expands beyond a small number of individual claims, the political effect could accumulate,” he said. “What is now a marginal decline in support and approval could become more significant, particularly among softer PC voters who are not yet prepared to abandon the party but are increasingly uncomfortable with what they are seeing.” Methodology The survey was conducted with 1,000 eligible voters in Ontario from July 23 to 29, 2026. A random sample of panelists was invited to complete the survey from a set of partner panels based on the PureSpectrum platform. These partners are typically double opt-in survey panels, blended to manage potential skews in the data from a single source. The margin of error for a comparable probability-based random sample of the same size is plus or minus 3.1 percentage points, 19 times out of 20. The data were weighted according to census data to ensure the sample matched Ontario’s population according to the 2021 Census. Totals may not add up to 100 due to rounding. This release also compares the results with a separate Abacus Data survey that was largely completed before the expense controversy broke. Because the results come from separate samples, differences should be interpreted as changes in aggregate public opinion rather than changes among the same individual respondents.
