How One University Pollster Was Anointed an Election-Night Winner

For most Americans, the runup to the 2016 election was a punishing ultramarathon. But at Quinnipiac and Monmouth Universities, plenty of people were still sprinting to the finish line, powered by jolts of energy from their respected national polling operations.

Quinnipiac’s development office told the Associated Press that it was using updates from its Polling Institute as “earned advertising,” emailing the latest numbers to parents and alums. Running alongside was Monmouth, whose president told The Chronicle’s Steve Kolowich that he had made “a big financial commitment” to Monmouth’s institute, “to play on the national stage.”

Both institutions had reason to feel proud — and a bit competitive. FiveThirtyEight, the data-based news site, assigns grades to virtually every pollster that goes out in the field, and the universities had scored well. Quinnipiac had earned an A-minus; Monmouth was one of just six pollsters — and the only one affiliated with a university — awarded an A-plus. Hopes were high. “Obviously, you look for bragging rights,” the director of Monmouth’s institute admitted.

The rest of that story barely needs to be told. Election night was a “Dewey Defeats Truman” moment for the digital era, a deep shock to a polling-and-quantification industry that had seemed to be ascendant. Neither Monmouth nor Quinnipiac foresaw Donald J. Trump’s victory. No reputable pollster did.

So if election night 2017 was seen as a referendum on many things — whether Democrats are in disarray, whether Trumpism without Trump is politically viable — it was also a moment of reckoning for the polling industry. Had it done anything to rethink an approach that seemed broken? Should it?

Those questions hung in the air at the end of October, when Quinnipiac released a poll of Virginia’s gubernatorial race that gave Ralph Northam, the Democratic candidate, a resounding 17-point lead among likely voters over the Republican Ed Gillespie. Quinnipiac followed up on Monday with its final poll, which cited “yawning racial and gender gaps” in the electorate while pegging Mr. Northam’s cushion at a still-comfortable nine points.

Those were stunning results. The conventional wisdom, and most of the polling, held that Mr. Northam and Mr. Gillespie were running neck and neck, or pretty close to it. Monmouth’s election-eve poll, for example, gave Mr. Northam a two-point edge, well within the margin of error. Outliers are generally treated with skepticism, and Quinnipiac was no exception.

Again, election night happened. And again, experts were caught off guard, this time by a Democratic wave. Mr. Northam rode that wave to a convincing victory. His margin over Mr. Gillespie, as of now: 53.9 percent to 45 percent. Round up a tenth of a percentage point, and you’ve got a nine-point gap — exactly as Quinnipiac predicted. That did not go unnoticed, by the political media …

… or by the polling institute itself.

Quinnipiac also bucked expectations in correctly predicting a rout in New Jersey’s gubernatorial race. So a clear victory for the polling institute, right? In one sense, yes: The university certainly won’t mind the shot of attention it gets.

In a broader sense, it will take time to sort out what last night’s results mean for the polling industry. As a few other polling experts have pointed out, one successful season doesn’t automatically make a pollster the best in the business.

Nor does it mean that polling’s existential crises have been solved.

Today, the Quinnipiac University Poll has reason to celebrate. Tomorrow, it has reason to join every other pollster in the field’s latest election-night post-mortem.

Brock Read is assistant managing editor for daily news at The Chronicle. He directs a team of editors and reporters who cover policy, research, labor, and academic trends, among other things. Follow him on Twitter @bhread, or drop him a line at brock.read@chronicle.com.