Rutgers polling expert explains what voters should know about polls ahead of 2026 midterm elections

NEW BRUNSWICK, N.J. (Middlesex County) — With the 2026 midterm elections approaching Tuesday, Nov. 3, voters are seeing a steady stream of polls measuring which party could control Congress and how candidates are performing in competitive races across the country. However, a Rutgers University polling expert cautions that election surveys should not be viewed as predictions of who will win, particularly in closely contested congressional districts where voter turnout can significantly influence the outcome.

Ashley Koning, an assistant research professor and director of the Eagleton Center for Public Interest Polling at Rutgers University, said voters should focus on polling trends, methodology and the people surveyed rather than relying on individual results.

“Polls aren’t meant to predict election outcomes. A poll is a snapshot in time: It tells us how people feel when they’re interviewed, and it can only reflect what the people it reaches tell us. During an election, who votes and how they vote can keep shifting right up through Election Day. Likely voters are an unknown population until the votes are counted, so every pre-election poll is making its best guess about who will turn out.”

Koning said election polling has performed better in recent midterm cycles than many people realize, although some previous elections produced significant polling errors. She noted that polling organizations have continued refining their methods, including using voter registration records and adjusting survey results to better reflect voters’ education levels and past voting behavior.

“The bigger misses came when Donald Trump was on the ballot and Democratic support was overestimated. Pollsters have also kept refining their methods, from voter file samples to weighting by education and past vote.”

Koning said credible polling organizations that follow established research practices and publicly disclose their methods deserve considerable trust, even though no survey can perfectly predict who will vote.

National polls do not determine congressional outcomes

National polls frequently ask voters which political party they would prefer to control Congress, a measure commonly known as the generic congressional ballot. According to Koning, those surveys can provide useful information about the country’s political mood, particularly when considered alongside presidential approval ratings and historical midterm voting patterns.

However, national polling results do not directly translate into the number of seats either party will win in the House or Senate. Individual congressional races are influenced by factors including incumbency, candidate quality, local issues, district boundaries and voters who support candidates from different political parties.

Koning pointed to New Jersey’s 7th Congressional District as an example of how voting patterns can differ between elections. President Donald Trump carried the district in 2024, while Democrat Mikie Sherrill won the district during the following year’s gubernatorial election. She said state and district-level polling is especially valuable for understanding competitive races, but the declining availability of independent public polling at those levels makes it more difficult for voters to assess local contests.

Why competitive House races are difficult to poll

New Jersey’s 7th Congressional District is among the closely watched House races heading into the midterm elections. Koning said congressional districts present additional challenges for polling organizations because they cover smaller geographic areas, resulting in smaller samples and potentially larger margins of error.

District boundaries can also cross county and municipal lines, requiring pollsters to use voter registration records to identify eligible respondents. In closely divided districts, even modest changes in voter turnout can significantly affect election results.

Polling individual congressional districts is also expensive, largely because fewer people respond to surveys than in previous years. As a result, independent public polls are often limited to the most competitive races, leaving voters with relatively few surveys to compare.

Koning identified Siena, including its polling work with The New York Times, and Emerson College as organizations that regularly survey individual congressional districts. The limited availability of independent polling can create opportunities for campaigns and advocacy organizations to release their own surveys.

While such polls may use legitimate research methods, Koning cautioned that campaigns and advocacy groups typically publicize results that portray their preferred candidates favorably.

What voters should watch in the final days

With Election Day approaching, Koning said voters should pay attention to changes across multiple polls rather than focusing on a single survey. Undecided voters may make their decisions during the final days of a campaign, although political polarization means many voters have already settled on their preferred candidates.

In closely contested elections, a relatively small number of undecided voters could influence the outcome. Changes in turnout among different groups can also produce results that differ from earlier polling, while early voting and mail-in ballots create additional complications because some voters have already cast their ballots before the final surveys are released.

Koning recommends examining polling trends, the percentage of undecided voters and the enthusiasm of each candidate’s supporters rather than treating any individual poll as a definitive indication of the election’s outcome.

“Remember, polls aren’t meant to be crystal balls. They are simply taking a pulse in a moment in time that very well may change based on who does or does not turn out by Election Day.”

How to determine whether a poll is reliable

Koning also urged voters to carefully examine the organizations conducting election surveys, particularly following a recent controversy involving Wedgewood Polls. According to Koning, Wedgewood Polls admitted that most of its surveys were fabricated, with some results already appearing in news coverage and polling averages before the issue was discovered.

She said voters should determine whether a polling organization has identifiable staff members, a documented track record and publicly available research methods. The source of a poll’s funding is another important consideration because campaign-sponsored and advocacy-sponsored surveys are not necessarily inaccurate, but voters should understand who commissioned the research and why the results were released.

Koning also emphasized the importance of knowing whether a poll surveyed all adults, registered voters or likely voters, since each group represents a different population. Other factors include when the survey was conducted, how respondents were contacted, the sample size, the margin of error and the exact wording of questions.

“A pollster should always publish the exact question text. If you don’t know anything about the wording, you don’t know anything about the resulting public opinion.”

For additional guidance, Koning recommended resources from the American Association for Public Opinion Research, the American Association of Survey Research Organizations, the Roper Center at Cornell University and the Pew Research Center. She said understanding how polls are conducted and recognizing their limitations can help voters interpret election coverage more accurately as the midterm elections approach.