Election polling numbers that don’t reflect election results have contributed to a mistrust in both public institutions and the polling industry at large, said Patrick Murray, director of the Monmouth University Poll.
“Those times when we don’t get it right, which are honest mistakes, are no longer perceived as honest mistakes and contribute to a real deepening of this public district in institutions,” Murray said on Hill.TV’s “Rising.”
Differences between horse race polling predictions and election results have also undermined trust in polling more generally, Murray added.
“Nowadays we’re in a situation where those misses are used as evidence that there’s something nefarious going on and that undermines everything else that we do in polling that has nothing to do with election polling, but has to do with measuring the fundamental state of our republic and our polity,” he said.
In a piece published on NJ.com in November, Murray wrote he “blew it” after the Monmouth University Poll predicted New Jersey Democratic Governor Phill Murphy held a double digit lead over Jack Ciattarelli, a Republican challenger, days before the state’s election. Governor Murphy instead won reelection as governor by three points.
Murray said a shift away from horse race polling in elections can help to legitimize polling again.
“What value do we bring to the table when we’re saying one candidate is ahead of another candidate by 6 or 7 points, does it really contribute to the dialogue?” he said, “There are other ways that we can do election polling that don’t necessarily need to use the horse race, such as candidate favorability ratings, issue ratings those kind of things.”