Flawed poll gives Lamont oversized lead
You might have missed a poll from Sacred Heart University that showed Gov. Ned Lamont beating GOP challenger Bob Stefanowski by a hefty 18 points, 48 percent to 30 percent, “if the election for governor were held today.”
The online poll, taken between March 24 and April 11 and released last Friday, was not widely reported despite high interest in the election. That’s because the fraternity and sorority of political chatterers and power brokers in Connecticut — on both sides and in the middle — believe the result is, um, not even close to believable.
Most accounts have Lamont slightly ahead of Stefanowski, nothing remotely comfortable despite the advantages of incumbency and a 37 percent-to-20 percent Democratic Party registration advantage. Unaffiliateds account for 42 percent of the electorate.
So how can a poll by SHU, a generally credible public policy operation, with an established partner, Glastonbury-based GreatBlue Research, turn out a result well into double digits for Lamont?
The answers tell us a lot about polling in the fractured, all-digital, postpandemic world of 2022. It’s extraordinarily hard to reach a representative sample of the population at a slice in time and accurately project their opinions of the actual world.
Possible problems
First, maybe SHU/ GreatBlue is right. It’s possible we politicos, a smallish bunch currently enmeshed in a legislative session, are all wrong and Lamont’s message about a Connecticut on the rise is reaching voters in a way we’re missing.
We’ve all messed up before, notably with the Trump-Clinton election of 2016. Although I called that one correctly, my record is spotty like everyone else’s.
Inconsistencies point to other problems in the SHU poll of 1,000 Connecticut residents. It’s an all-digital poll, meaning respondents were asked to mark up 39 questions without a prompt from a real person, which can cause problems. And it was an existing panel of respondents, basically self-selecting, not a fresh, new, random sample — the gold standard.
The poll includes 19.7 percent who are not likely voters, by their own account, and yet everyone was invited to opine on who they might vote for in the race for governor. Who cares what nonvoters think in an election poll? Not me.
The 1,000 respondents were heavily weighted white — 80 percent as compared with 69 percent of the state’s population. It was underweighted for Black people by a couple of points and woefully underweighted for the Hispanic and Latino population — 6.3 percent compared with a reality of 17 percent.
When the poll asked who respondents were voting for in the governor race, it did not state the names of Lamont and Stefanowski but “they were explicitly listed as the answer options/choices,” an executive at GreatBlue told me in an email.
That might not be enough of a prompt to separate support from name recognition. In other words, there may be people who will vote for the Republican nominee but who didn’t easily identify Stefanowski as that person.
The poll shows that 74 percent rated their “overall quality of life in Connecticut” as good or excellent, up from 69 percent
in October, when a similar poll asked the same question. That’s a big number and it raises questions about how an incumbent governor in a strong majority party would only poll 48 percent.
It also appears to skew toward New Haven County and underweight unaffiliated voters, which matters, since that’s the group both parties are fighting to win.
Would those factors tend to make a poll swing incorrectly to Lamont? Maybe, maybe not, but those discrepancies raise questions. Something has to be amiss somewhere and neither SHU nor GreatBlue explained it when I asked.
The candidates react
Lamont, wisely, told reporters he places no stock in the poll whatsoever. That of course is in keeping with Politics 101 for incumbents seeking reelection, thou shalt disavow all polls publicly.
“None. None. Seven months is a long way away,” Lamont said when asked about the poll. “I’ve got a lot of work I gotta do between now and then. I did like the fact that there’s an uptick in people who have a renewed appreciation for the quality of life in Connecticut. I think that’s really important .... You see that reflected in some people moving to the state.”
And he was off and running, talking about the new, improved Connecticut on his watch — an allowable reference to polls in the handbook.
The Stefanowski camp sees things differently. This from Liz Kurantowicz, Stefanowski’s strategy consultant:
“Even in a significantly flawed survey, the governor still doesn’t get 50 percent of the vote and that’s because while he’s bending over backwards to make people believe his version of the truth, they know Ned Lamont has made Connecticut less affordable and less safe while his administration is mired in scandals.”
That’s the great thing about polls. Right or wrong, they give us a lot to talk about. Lamont beat Stefanowski by 3 percentage points in 2018 in a high-turnout election.
Ben Proto, the Republican state chairman, noted that President Joe Biden fetched a 39 percent approval rating in the poll, which raises questions about Lamont’s supposed margin over Stefanowski.
To repeat: Maybe Sacred Heart University and GreatBlue are right, that Lamont is on his way to far outperforming the national Democratic Party. The state’s political observers on both sides are all locked in for a tight race and Lamont is on the air with TV ads seven months out, meaning his camp sees a dogfight, not a walk in the park.
Same ending as my earlier column this week about Lamont’s and Stefanowski’s wildly different views of the Connecticut economy: This is going to be fun.