Political polls are paid for by three main groups: news organizations and universities, which publish polls to inform the public; campaigns and party committees, which commission private polls for strategy; and advocacy groups, super PACs and businesses, which use polls to make a case. Who pays shapes which questions get asked, how they’re worded, and whether you ever see the results.
The sponsor doesn’t automatically make a poll good or bad. But knowing who paid, and who actually did the fieldwork, is the first step in deciding how much weight a poll deserves.
News organizations and universities: the public polls #
Most polls you see in the news were paid for so they could be published.
News outlets #
Major outlets pay for polls because exclusive data drives coverage and audience. High-quality polling is expensive, so outlets often partner with a university or a professional firm:
- The New York Times and Siena College.
- CNN and SSRS.
- Fox News, whose poll is run jointly by a Democratic firm, Beacon Research, and a Republican firm, Shaw & Company.
Universities #
University polling centers, such as the Marquette Law School Poll, Quinnipiac University and the University of New Hampshire’s Survey Center, fund polls from their budgets, endowments and grants. They usually publish full question wording and detailed crosstabs, and they’re a big share of the independent polling in some states.
The money is fragile. Monmouth University closed its well-known polling institute on July 1, 2025, citing the resources required to run it. When a respected academic pollster shuts down, fewer independent polls get done, and cheaper polls fill the gap.
Campaigns and parties: internal polls #
Campaigns, the national party committees and their House and Senate campaign arms spend heavily on private polling. These polls are strategic tools, not public information.
- Benchmark polls at the start of a campaign test name recognition, issues and messages. See tracking polls versus benchmark polls.
- Brushfire polls in mid-campaign check how an ad, a debate or a controversy landed. We explain them in what a brushfire poll is.
- Tracking polls late in a race interview a small sample every night and roll the results into a multi-day average to catch movement quickly.
Why leaked internal polls deserve skepticism #
When a campaign releases an internal poll, it’s for a reason: to reassure donors, attract party money, or suggest momentum. Campaigns don’t release polls showing them losing badly. That selection alone makes released internal polls lean toward their sponsors, even if every individual poll was done honestly.
Advocacy groups, super PACs and businesses #
Unions, trade associations, issue groups, think tanks and super PACs commission polls too. Some are careful research; others are designed to produce a headline.
The main risk is question wording. A sponsor can get very different results from “Do you support investing in clean energy jobs?” and “Do you support new taxes to fund energy subsidies?” A legitimate issue poll still uses scientific sampling, but read the exact question before you accept the headline.
Push polls aren’t polls #
A push poll is persuasion disguised as a survey: a call that asks something like “Would you be less likely to vote for Candidate X if you knew…” followed by a damaging claim. The goal is to spread the message, not collect data, and there’s usually no real sample or published result. We cover how to recognize one in what a push poll is.
Why is polling so expensive? #
- People don’t answer. Pew Research Center’s phone surveys got a 6% response rate in 2018, down from about 9% a few years earlier. Every completed interview can take dozens of dials. Our piece on falling response rates explains why.
- Cell phones are dialed by hand. Federal robocall law restricts automated dialing of cell phones, so most pollsters have live interviewers dial them manually to stay safe.
- Weighting takes expertise. Adjusting a raw sample to match the electorate by age, education, race, region and past vote requires statisticians and good benchmark data.
- Better online methods cost more too. Probability-based online panels, recruited by mail from random addresses, are more expensive than opt-in web panels, which are cheap but less reliable.
Does the sponsor bias the result? #
It can, in a few ways: by choosing which polls to release, by wording questions to favor a side, or by hiring a pollster whose methods tend to lean one way. Consistent differences between firms are called house effects, explained in what a pollster house effect is.
It doesn’t have to. Many partisan-affiliated firms publish accurate public polls, and some news-sponsored polls miss. Judge the method, not just the logo.
How to check who paid for a poll #
- Find two names: the sponsor and the pollster. The sponsor paid; the pollster did the fieldwork. A poll sponsored by an advocacy group but conducted by an established independent firm can still be sound.
- Look for full disclosure. Credible releases include sample size, population, field dates, mode, weighting and exact question wording. Missing basics are a warning sign.
- Check transparency commitments. Many pollsters participate in the American Association for Public Opinion Research’s Transparency Initiative, which requires disclosing methods and sponsors.
- Read the question order. Questions asked before the horse-race question can prime respondents.
- Compare with the average. A sponsored poll far from other recent polls deserves extra scrutiny.
Election Tracker makes the first two checks quick. Every poll card lists the pollster, sample size, population (likely voters, registered voters or adults) and field dates, and the pollster’s release opens in an in-app browser so you can find the sponsor line and question wording. The app’s 30-day race averages count repeated polls from the same pollster only once, at their strongest sample, so one heavily funded firm polling every week can’t dominate a race.
Frequently asked questions #
Do tax dollars pay for political polls? #
Not for horse-race or approval polling. Federal agencies fund large surveys such as the Census Bureau’s, which don’t ask about candidates. Public universities may fund polling centers, and some academic research on voting receives government grants, but candidate polling isn’t a government program.
How much does a political poll cost? #
It varies widely with method, sample size and length. Live-interviewer phone polls of likely voters are among the most expensive per interview, and cheap online opt-in surveys among the least. Because costs have risen, some outlets poll less often or split costs with partners.
Can a campaign pay a pollster to fake numbers? #
Reputable pollsters won’t fabricate data; their business depends on credibility. Campaigns can legally choose which polls to release, and they can influence question design, which is why released internal polls deserve skepticism.
How do I find out who funded a poll? #
Look at the top or bottom of the poll release or the methodology statement, which should name the sponsor. News stories should say who paid. If no one will say who funded a poll, treat that as a red flag.
Are partisan pollsters included in polling averages? #
Usually, yes. Most averages include polls from partisan-affiliated firms and handle bias by averaging across many pollsters or adjusting for house effects. Some averages leave out polls sponsored directly by campaigns.