Skip Navigation
I Voted sticker
Justin Sullivan/Getty
Analysis

How Polls Can Help or Hurt Democracy

Attacks on legitimate polls and the rise of fake polls make it hard for voters to know what to believe.

  • Caitlin Durkovich
  • Zarine Kharazian
August 12, 2026
I Voted sticker
Justin Sullivan/Getty
August 12, 2026

A poll is an important democratic tool that does more than just measure who is ahead in an election. It serves as a gauge for public opinion, allowing candidates and officials to understand what voters think and value. At the same time, polls can shape voters’ perception of candidates, electoral participation, and confidence in election results. As recent years have demonstrated, this makes polling a prime target for those looking to influence the electorate and shape public opinion.  

Polling has become harder and more expensive because the public is increasingly difficult to reach and less willing to participate. Traditional polling methods relied heavily on landline telephone surveys when people were more likely to answer unknown calls and remain on the line for interviews. In recent years, the widespread use of mobile phones, caller ID, and spam filtering has dramatically reduced response rates. Many people do not answer calls from unfamiliar numbers, while others decline to participate even when reached. As a result, polling organizations must contact far more individuals and adapt their interviewing protocols to obtain a statistically meaningful sample that reflects the broader population.

Notable misses in the 2016 and 2020 presidential elections highlighted some of the challenges with polling, particularly with respect to underestimating support among certain voter groups. Those incorrect forecasts, combined with an increasingly polarized political environment, have made polling an attractive target for manipulation and efforts to undermine public trust.

Today, skepticism about poll credibility is challenging how voters decide what polls to believe. Two distinct but reinforcing threats are contributing to the problem. First, legitimate polls are often attacked or dismissed when results are unfavorable and inconvenient. Second, misleading or outright fabricated polls can distort perceptions of public opinion and potentially influence election outcomes. Understanding both issues is essential to protecting voters’ ability to make informed decisions.

Attacks on Legitimate Polls

A legitimate poll is a scientific survey designed to measure what people think, believe, or intend to do. It follows established standards for sampling, survey design and methodology, and verifiability. When done correctly, polls provide a trustworthy snapshot of what the public thinks about important issues and give people a reliable voice in the democratic process by helping leaders understand their views.

There are dozens of independent organizations in the United States that regularly conduct public opinion polling. Economist/YouGov, CBS/YouGov, ABC/Washington Post, NBC/Wall Street Journal, Reuters/Ipsos, Quinnipiac University, and the Associated Press are among the organizations that produce most of the widely cited national political polls used by campaigns, media outlets, analysts, and academic researchers. Gallup — the pioneer of modern polling — has stepped back from regular U.S. horse-race electoral polling, but it continues to conduct issue polling, as do other independent and university-based polling organizations, such as the Pew Research Center, Siena College, and AP-NORC.

These organizations use rigorous scientific methods to ensure the people they survey reflect the broader population. For example, the Pew Research Center recruits participants using probability-based sampling so that most U.S. adults have a known chance of selection. It weights results to reflect national demographics and fully discloses its methodology.

Attacking the credibility of legitimate polls has become a recurring political tactic to erode the public’s trust in the results. Politicians and political parties have taken to declaring polls whose results they dislike to be biased or part of a conspiracy. For example, in January, President Trump threatened to sue The New York Times over polling results indicating that his approval rating had fallen. The Times defended the poll’s methodology and the company’s track record. 

The Rise of Fake Polls

A fake poll manufactures or manipulates data to create a misleading impression of voter sentiment. Unlike legitimate polling, its purpose is not to measure public opinion but to influence and change voter behavior. Fake polls are designed to target voters who remain undecided or weakly committed. Research shows that these voters often look to polling as a source of information about candidate viability, momentum, and broader public sentiment. By exaggerating a candidate’s strength or suggesting an opponent has little chance of winning, manipulated polling can create a bandwagon effect, discourage turnout, and alter perceptions of a race during a moment many voters are deciding how to cast their ballots. 

Misleading polls have become more visible in recent election cycles. Candidates from both parties have at times released polls showing themselves leading or locked in tight races and have not disclosed the underlying methodology, even when those results diverged significantly from more established or independently validated polling. The problem has grown more acute in recent years.

During the 2022 midterm elections, a wave of partisan and methodologically weak polls distorted polling averages and contributed to misleading perceptions of several races. In some cases, polling appeared designed to inflate perceived support, shape media narratives, and preemptively cast doubt on potential election outcomes. Starting in 2024, fabricated polls have been used to inflate a party’s apparent strength and to potentially help lay the groundwork for challenging the legitimacy of an inconvenient result. In the 2026 cycle, the Maine Senate race illustrates a trend of campaigns commissioning internal polls and selectively releasing favorable findings to bolster perceptions of a candidate’s strength and create the appearance of campaign momentum.

Manipulation of voter sentiment can occur in other ways as well. A fabricated CNN graphic showed Democratic candidate Kamala Harris ahead of Donald Trump in early voting by 6 percentage points in Texas. While CNN never aired the graphic, and it was quickly debunked as manipulated media, it still circulated online, leading some social media users to falsely conclude the election was being rigged in favor of Democrats.

Manipulated polling is not just a domestic problem. In Hungary’s 2026 parliamentary election, concerns about manipulated polling and information operations intensified in the final days of the campaign. Observers noted a widening gap between independent polling organizations, many of which showed the Tisza opposition party leading, and government-aligned pollsters that projected strength for incumbent Viktor Orbán’s Fidesz party. Some analysts argued that the divergence was difficult to explain on methodological grounds alone. 

Verifying Poll Legitimacy 

Attacks on legitimate polls and the spread of fabricated ones gain steam, in part, because most voters lack the tools to evaluate them. These questions can help: 

  • Who conducted the poll? Is this an established organization or an unfamiliar entity? Stick to established organizations with a history of accuracy and transparency.
  • How was the poll conducted? Is the methodology publicly available? Does it use representative sampling?
  • Why was the poll released now? Does the sponsor have a stake in the outcome? Was it released at a moment designed to shape perceptions?

Polling is an important tool for understanding public opinion, but it is becoming harder for voters to distinguish credible polling from attempts to manufacture consensus. Helping voters understand how polling works and how to distinguish legitimate polls from fake ones strengthens their judgment and makes democracy more resilient against manipulation.