It was 2002, and Lynnfield resident David Paleologos began taking Massachusetts’ political pulse.
Paleologos had been teaching at Suffolk University by that time for seven years. But his goal of establishing a political research center at the university had not come to fruition.
He was teaching a class when a student made an intriguing suggestion: Why not conduct a poll to gauge how a forthright businessman named Mitt Romney might square off against Massachusetts gubernatorial candidate Jane Swift?
It was a weird idea, but Paleologos went with it. The poll results were echoed in the Boston Herald, and showed Romney beating Swift by a large margin.
The rest is history: Swift dropped out of the race, Romney entered — and won — and the Suffolk University Political Research Center, now one of the preeminent academic polling institutions in the U.S., was born.
“Romney’s political career was born,” Paleologos said. “And at that moment, I realized the value of a student, someone who thinks differently than I and thinks outside the box.”
The Suffolk University Political Research Center initially partnered with Channel 7 news in Boston, then with other prominent media partners, such as the Boston Globe. Paleologos employed this mantra: To get an accurate poll, the demographics of those polled must be proportional to the area the pollsters are trying to predict.
For example, when trying to figure out who will win a gubernatorial race in Massachusetts, where 14.3 percent of the population is 25 to 34 years old, 14.3 percent of those polled must be 25 to 34 years old. It goes the same for race, gender, and geography, Paleologos said, and “demographic proportionality” is the hallmark of a good poll.
That rule of proportionality today guides all good polling institutions, Paleologos said, including Suffolk University Political Research Center, which famed statistician Nate Silver has said is correct 80 percent of the time.
“If I do a poll that’s mostly young people, of course the Democrats are always going to be ahead. If I do a poll with people who are mostly rural and white, Republicans are always going to be ahead,” he said.
Despite its hitherto acclaim, Suffolk University Political Research Center is still evolving. For example, Paleologos said, just recently Suffolk polls have been polling according to nonbinary and transgender demographics, not just basing its “demographic proportionality” of male or female. He said it’s a very small percentage of people who don’t identify as male or female — less than 1 percent — but, to get an accurate poll, counting them in a poll’s demographics is still important.
“It’s important to set the bar,” Paleologos said.
In 2002, the majority of respondents did so using a landline telephone, until cell phones “became a sliver, then a slice, then a majority, now almost 90 percent,” of respondents, Paleologos said.
There isn’t concrete data to answer why, but poll respondents also have “a much shorter attention span” these days, Paleologos said. Two decades ago, a “long” poll would have taken about 30 minutes on the phone with a respondent — and that wasn’t uncommon. Today, “long” means about 12 minutes.
And yes, it’s true America is much more polarized politically.
“People are far, far more polarized than they were in 2002. In 2002, you had a small group of people who were far-left and a small group of people who were far-right, and then a whole bunch of people in the middle who could swing either way,” Paleologos said. “Now, you have mountains of people on the left and mountains of people on the right.”
“In the upcoming election, campaigns are going to have to really dig deep for that very small group in the middle that can swing either way,” he said.
But there have been polling failures in the past. In 2016, most polls predicted Hillary Clinton would be the next president of the U.S., not Donald Trump.
Paleologos, who is ethnically Greek, always does his last interview of the year with a Greek newspaper, and he remembered telling one in 2016, “There are a high number of undecided voters and a high number of third-party voters, and those are both red flags.”
Contrary to the public perception, that national polls got the 2016 presidential race wrong, the national pollsters got it right, Paleologos said. National polls “poll the popular vote, and she won the popular vote.”
