Why Are Public Interest Polls Important

By next week, the results of New Jersey's tight gubernatorial race will be in.

Until then, many political analysts and voters alike are reading public opinion polls like tea leaves to try and predict which candidate - Democrat Mikie Sherrill or Republican Jack Ciattarelli - will come away with the win.

But that approach undermines the very purpose of these polls, according to Ashley Koning, director of the Eagleton Center for Public Interest Polling.

"They are never meant to serve as a crystal ball, but rather to help us understand the why and how behind voter behavior," she said.

Here's more of our conversation with Koning about what we are meant to glean from these polls about New Jersey voters.

What is the value of public interest polling? Who do the polls primarily serve: voters or candidates?

Public interest polling serves several different roles in our democracy. During elections, polls provide one of the few systematic and quantitative ways to representatively measure public sentiment in real time - showing how close or not close a race may be at any given moment. They are never meant to predict outcomes or serve as a crystal ball, but rather to help us understand the why and how behind voter behavior.

Outside of election cycles, issue polling allows us to gauge public opinion on topics that may never appear directly on the ballot but remain vital to the well-being and advancement of our democratic system. Taken together, polls give voice to the public - representing citizens as they exist in the population - and connect that collective sentiment to policymakers and candidates alike. As George Gallup, the father of modern scientific polling, once said, polls "make […] a truer democracy."

The gap between candidates in New Jersey's gubernatorial race went from 20 points in early July to 9 points in late August to 5 points in October in three Eagleton Polls. Do you think those changes make people lose confidence in the polls? What do you think people should take away from that shift?

It should actually be the opposite. Races almost always tighten as Election Day approaches - that's the natural progression of any campaign. Undecided voters begin paying closer attention and making up their minds in those final weeks, which we see reflected in the shrinking share of respondents who are unsure. The trajectory of these numbers shows exactly that pattern.

From the start, we knew this would be a competitive race for an open seat between two highly qualified candidates, and that, much like in 2021, New Jersey's increasingly polarized political climate would make for a close contest.

It's also important to interpret and compare polls carefully. Differences in timing, methodology, and sample composition matter - for example, our early-summer poll surveyed self-identified registered voters, our late-summer poll focused on self-identified likely voters, and our fall poll drew from verified registered voters in the official voter database. Those distinctions affect results, but together they tell a consistent story of a race that has steadily narrowed as voters engage.

Last year's polls predicted Kamala Harris would defeat Donald Trump in New Jersey by a wide margin. But on Election Day, Trump lost New Jersey by 6 percentage points - not by the predicted double-digit margin. What factors led to those polling discrepancies? How are pollsters adapting to the political challenges of today?Polling has faced a number of challenges, particularly over the past decade. Response rates have declined sharply and now sit in the low single digits. At the same time, polls themselves have become politicized, and certain segments of the population have grown less willing to participate due to polarization and distrust - something we see especially when President Trump is at the top of the ballot.

Pollsters have been actively studying these issues to ensure that surveys still capture an accurate picture of the public. That includes refining how samples are designed, using multiple modes of recruitment to reach respondents, and applying careful statistical weighting on the back end to align results with the population as a whole. The industry continues to evolve to meet these challenges while maintaining the scientific rigor that makes polling such an essential tool for understanding public opinion.

And ultimately, it all comes down to who actually turns out by Election Day. Voters may tell pollsters they won't vote and then do, they will vote and then don't, or that they are undecided or voting for a different candidate than who they choose at the ballot box. New Jersey saw depressed Democratic turnout in key areas that impacted Harris's overall share of the vote and thus sparked concerns for Democratic turnout in 2025.

Is the number of undecided voters today significant? Who are these voters, and what issues are most important to them?

The number of undecideds has dwindled throughout the course of the campaign, and independents remain pivotal to either candidate's path to victory - particularly for Ciattarelli, given the state's voter registration landscape. Top issues for these groups, unsurprisingly, are taxes and affordability. Independents, in particular, are torn on who will do a better job on a number of these "kitchen table" issues, though they think Sherrill will do better at handling education, transportation, and healthcare and Ciattarelli a better job at crime and safety.

After being seen as a decidedly blue state for decades, are there any demographic shifts that are making us more purple?

New Jersey has always been purple at the state and local levels and once was even a toss-up in presidential elections. We are only classified as a blue state because we have reliably voted for the Democratic candidate for president since Bill Clinton, but our governor's seat often changes parties. If you imagine a map of the state, the corridor that runs through the middle of New Jersey from Philadelphia in the southwest to New York City in the northeast is solidly Democrat, while various counties in the northern and southern parts of the Garden State lean more red. Some say that New Jersey is getting redder. Almost every county grew redder when it came to presidential vote choice in 2024, and Trump flipped Morris, Passaic, Gloucester, Atlantic and Cumberland; Passaic, Atlantic, and Cumberland additionally had lower turnout in 2024 compared to 2020. Much like potential national demographic realignments we have seen in the past year or so, we should be watching to see if Republicans are able to continue to chip away with non-white voters this race - particularly Hispanic males. We should also be watching higher income voters and college-educated voters, given Ciattarelli's success with them in 2021.