The polling yields multiple dividends for the schools involved. It boosts their name recognition, through intense media coverage of the ups and downs of the presidential horse race, and it helps them stay engaged with public policy issues.
Paul Trible, president of Christopher Newport, a public university of 5,000 students in Newport News, Va., said students learn a lot about the country and its politics when they make calls for the university’s polls. “It is a powerful educational opportunity,” he said. In addition he said, polling serves the university’s public mission.
“Obviously, when people see the name Christopher Newport University, it expands the reach and reputation of our young and vibrant university,” Trible said. “That’s a plus for sure, but that’s not the primary purpose.”
Christopher Newport has published two Virginia polls during this presidential cycle, the latest in April, according to the website Real Clear Politics. Roanoke, a private college in Salem with 2,000 students, has published four. Its latest, in May, showed Democratic presidential candidate Hillary Clinton tied with Republican Donald Trump. Hampton, a private university of 4,300 students, published one in July that also showed a tie.
But Marist College in New York published a Virginia poll in July with NBC News and the Wall Street Journal that showed Clinton with a solid lead.
Polling carries some risks for these nonpartisan institutions of higher learning. The biggest: What if they get it wrong?
Pollsters enjoy diving into the minutiae of survey technique. They relish surprising findings. But they really love it when their predictions come true on Election Day.
For Doug Schwartz, Quinnipiac University’s pollster since 1994, two solid calls stand out from heated elections in his first several years.
On Nov. 2, 1998, the university released a poll showing then-Rep. Charles Schumer (D-N.Y.) with a lead of 8 percentage points over Sen. Alfonse D’Amato (R-N.Y.) in his quest to unseat the incumbent. Schumer won by a little more than 10.
On Nov. 6, 2000, the university released a poll showing Clinton ahead of Rep. Rick Lazio (R-N.Y.) by 12 points in a closely watched campaign for a U.S. Senate seat from New York. She won by 12.
Such results helped build Quinnipiac’s reputation as the university in Hamden, Conn., was expanding its political polling in the Northeast region and nationally. “I’m a perfectionist,” Schwartz said. “I want to get it all right.”
He can’t, of course. Polls have margins of error. In pure tossup races they will sometimes miss the mark. But one national analyst, at the website FiveThirtyEight, calculates that Quinnipiac calls races correctly 87 percent of the time.
The success rates were 88 percent for Marist, 82 percent for Suffolk University in Massachusetts, 75 percent for Roanoke — and 100 percent for Christopher Newport and Hampton. The latter figure appears to be an outlier because the number of polls from those two schools that the website analyzed was very small. (FiveThirtyEight calculates that polls from The Washington Post/ABC News called races correctly 78 percent of the time.)
Marist’s breakout polling moment came during a Democratic gubernatorial primary in 1982, a closely watched contest between Mario Cuomo, who was then the lieutenant governor, and New York City Mayor Ed Koch.
Many polls and pundits predicted Koch would win. “Koch was this overwhelmingly popular downstate mayor,” said Marist pollster Lee Miringoff. “The line then, and throughout much of the campaign, was a David-and-Goliath kind of thing.” But Miringoff released a poll well before the primary that showed Koch had significant weaknesses. The pollster doesn’t recall the numbers from the survey, only that he felt out on a limb at the time. “Every now and then you go against the conventional wisdom,” he said. “Nowadays we would call it an outlier.”
As it happened, David beat Goliath. Cuomo won the primary and went on to win in the general election. And Marist got a boost.
“Ever since then, the poll has kept growing,” said former Marist president Dennis Murray, who recently retired after leading the college for 37 years. Marist started presidential polling in 1983 in the New Hampshire primary, and it expanded in 2008, polling in eight battleground states. Since February 2015, it has done 36 state polls with NBC and The Wall Street Journal.
Marist’s fortunes were intertwined with all of those phone calls to voters. “The poll grew with the college,” Murray said, “and the poll helped the college grow.” Marist had about 1,200 students in 1979, Murray said. Now it has 6,400.
At Suffolk in Boston, pollster David Paleologos credits one of his students for a critical development as Massachusetts approached a gubernatorial election. In early 2002, Republican Jane Swift was the acting governor and considering a run for the office later that year. But Mitt Romney had just completed a successful tenure as leader of the Salt Lake City Winter Olympics. Paleologos said a student suggested that Suffolk poll voters to gauge support for Romney against Swift.
The poll found Romney would win handily, Paleologos said. Not long afterward, Swift dropped out of the race and Romney jumped in. “That was the inflection point that put us on the map,” he said. In 2008, Suffolk started polling in battleground presidential states. (FiveThirtyEight estimates Suffolk calls 82 percent of races correctly.)
Polling has obvious marketing benefits for colleges with ambitions to become better known. But it also holds educational potential, pollsters and college officials say. Paleologos, pointing to the Swift-Romney poll, said that it is “an amazing educational experience” for students to be able to shape questions posed to voters. “Where else can a government student have that kind of impact?” he said.