New polling reveals hypothetical match-up between two potential CT gubernatorial candidates

WTNH – Friday, August 22, 2025
By Mike Cerulli

HARTFORD, Conn. (WTNH) — Neither New Britain Mayor Erin Stewart nor Governor Ned Lamont are officially candidates in the 2026 gubernatorial election, but new polling shows that Stewart would trail by eight percentage points in a hypothetical match-up between the two.

A survey conducted by Stewart’s exploratory campaign for the state’s highest office of 600 likely general election voters yielded 50% support for Lamont, 42% support for Stewart and 8% undecided. Survey recipients were presented with the following five options:

Vote Stewart (39%)
Lean Stewart (3%)
Vote Lamont (47%)
Lean Lamont (3%)
Undecided (8%)

“The poll shows that Mayor Stewart, even in an exploratory phase, starts where our last Republican candidates finished,” Morgan Wilson, a senior advisor to Stewart’s exploratory campaign and a vice president at the firm that conducted the survey, said.

In the 2022 gubernatorial election, Republican candidate Bob Stefanowski captured 43% of the vote compared to Lamont’s 56%.

The Stewart campaign’s survey also gauged Lamont’s favorability and found that he holds a net positive rating of seven points. Among independents, the survey found his favorability ratings effectively tied, with 46% reporting an unfavorable view and 47% holding a favorable one.

Wilson characterized these numbers as a sign of weakness for the two-term incumbent Democrat, who has said he’s grown “more inclined” to seek a third term next year.

“If I were his advisor, I’d be worried,” Wilson said.

One of Lamont’s closest political advisors signaled that the governor is ready for a fight.

“Governor Lamont is proud to put his record on affordability and opportunity up against anyone,” Rob Blanchard, the governor’s chief spokesman, said.

The results of the poll were compiled in a campaign memo which was viewed by News 8. Additional information, including the method of gathering responses and “crosstabs,” which display detailed information on poll responses, were not included in the memo.

OnMessage, the company that conducted the poll, is a GOP campaign firm that has been operating for two decades in races across the country. They’ve polled for Republican candidates including former Governor Rick Scott of Florida, Senator Josh Hawley of Missouri and Governor Henry McMaster of South Carolina.

The firm’s memo also included survey results testing support in a hypothetical Republican primary between Stewart and several other GOP politicians, including State Sen. Ryan Fazio and First Selectwoman Jen Tooker of Westport. Fazio and Tooker are already declared candidates in the race.

Fazio, a Greenwich native who has built a public profile as a crusader against high energy prices, launched his campaign earlier this month. Stewart formed her exploratory committee in January and has been kicking the tires on a run for governor for years.

The survey found Stewart with a sizable lead over Fazio. She garnered 42% (38% “Vote Stewart” and 4% “Lean Stewart”) to Fazio’s (10% “Vote Fazio” and 3% “Lean Fazio”). The results of the primary survey touched off a back-and-forth between the Stewart and Fazio camps, with Fazio’s team characterizing the results as “disingenuous.”

“Clearly Ryan’s competition is sensing his momentum, as just one week into his candidacy they are shopping disingenuous push polling,” Jim Conroy, a campaign strategist working for Fazio, said.

The term “push poll” generally refers to polls with questions written with the intention of swaying a respondent to a certain viewpoint.

Conroy and the Fazio campaign are not alone in questioning whether the poll was designed with an intended result in mind. Bill Evans, a Republican campaign operative in Connecticut, took to social media to lambast the Stewart poll.

Evans cited several screenshots of poll questions that were circulated among political insiders and widely assumed to be from the Stewart-commissioned survey.

One question asked voters if they would prefer “a candidate that is a mayor in Connecticut who believes in fiscal conservatism and common sense” or “a candidate that had a career in private equity driving business growth and served in the Connecticut State Senate for four years representing lower Fairfield County.”

The latter description fits Fazio, the former fits Stewart.

Another question contrasted “a candidate that embraces Trump and his energetic style as a fighter” with “a candidate that is constantly focused on being bipartisan, he doesn’t endorse the Trump MAGA agenda and avoids saying Trump’s name publicly.”

The wording and gendered language of those questions led some, including Evans, to believe that they were written with the goal of nudging respondents toward Stewart and away from Fazio. Evans questioned whether the comparisons were presented prior to the question asking respondents about their candidate preference.

“It sure looks like today’s poll for governor was clearly a push poll,” Evans wrote on Facebook on Thursday, when the Stewart campaign announced the results of their primary survey.

Wilson, the Stewart advisor, said the candidate preference questions were asked up front. The campaign’s memo did not list all the questions asked in the survey or the order in which they were presented.

Roy Occhiogrosso, a Democratic campaign operative who strategized former Governor Dannel Malloy’s three gubernatorial runs, said the questions that appeared to contrast Stewart and Fazio had the hallmarks of a “benchmark poll” rather than a push poll.

In a benchmark poll, respondents are typically asked about candidate preference before being given a list of questions intended to test the effectiveness of various political messages — including lines of attacks on potential opponents.

“In a true survey, you ask a head-to-head at the outset,” Occhiogrosso said, “you then ask a battery of questions introducing positive and negative information about both candidates, so that you’re filling the respondent’s head with knowledge and then you retest the head-to-head.”

Wilson agreed with the characterization of the survey as a benchmark poll.

“These are credible polls, credible pollsters,” Liz Kurantowicz, a Republican strategist who has run multiple campaigns in Connecticut, said.

She is not working for either Fazio or Stewart.

Like Occhiogrosso, Kurantowicz has overseen dozens of polls conducted on behalf of candidates. Those polls frequently find their way into the hands of reporters and into campaign fundraising pitches.

“Some of it is to try to create a narrative in the press,” Occiogrosso said. “Other times, people will put out information like this because they think it helps rally their supporters.”

Ultimately, Occhiogrosso questioned the veracity of a head-to-head poll this far out from the election.

“I look at those numbers at some level as being meaningless,” he said. “If at this point, only eight percent are undecided, I just find that hard to believe. Nobody’s focused on this right now.”

Kurantowicz noted that the results of the Lamont-Stewart head-to-head survey roughly maps onto recent historical averages for Republicans in gubernatorial elections.

“There’s a lot of time left, though,” Kurantowicz added. “This is a baseline assessment of Erin’s political future.”

Click here to read the entire article.