📉 🪦 Requiem for Gallup
The death knell of Gallup’s presidential approval polling and what it means for understanding American politics
🍺 What’s on tap 🚰
Today’s newsletter features:
Opening Bell: Must-read items about elections and politics.
The Frontrunner: A look at Gallup’s departure from presidential approval polling, including how it could affect political analysis and why it’s a big, if unsurprising, development in the history of polling.
Around the Corner: Upcoming elections we’re tracking at DDHQ.
🔔 Opening Bell 🐏
Must-read items about elections and politics.
Last week, Maine Sen. Susan Collins formally announced that she will seek a sixth term. Collins is the only Republican senator from a state that former Vice President Kamala Harris carried in 2024, making her one of the Democrats’ top targets in the 2026 midterms. If reelected, Collins would be on course to become the longest-serving senator in Maine’s history.
On Friday, the Virginia Supreme Court allowed the state’s redistricting referendum to proceed after a lower state court had ruled that the Democratic-backed effort ran afoul of state law. However, the state’s high court will consider the case, which leaves the door open to a ruling against the redistricting campaign even after the state holds a vote.
In Arizona’s GOP primary for governor, 2022 gubernatorial contender Karrin Taylor Robson announced her withdrawal from the race. After narrowly losing the 2022 Republican primary for governor, Taylor Robson quickly announced a 2026 campaign and earned President Donald Trump’s endorsement. But she struggled to win over more conservative, MAGA-aligned elements in her party. Taylor Robson’s campaign took an additional hit when Trump issued a co-endorsement of her main GOP rival, Rep. Andy Biggs, who is the favorite in the Republican nomination race.
☑️ Introducing DDHQ Votes ☑️
Decision Desk HQ is excited to announce that it will soon release Votes, our new election data portal. Dive into the world of election data with DDHQ Votes:
Up-to-date polling on the key issues and races
Track how prediction markets are moving
25 years of election results at your fingertips
Track election results as they come in
From local to federal, find every race
Coming Soon: DDHQ Forecast
Please join the waiting list for Votes now! Those on the waiting list will receive early access to Votes when it launches in the very near future.
📈 The Frontrunner 🥇
The end of the Gallup Poll as we know it
Quick summary:
Gallup announced that it would no longer track presidential approval. Quantitatively, removing Gallup’s monthly approval poll will not greatly alter the polling averages that track President Donald Trump’s standing with the public. Still, Gallup’s move could subtly produce a higher approval rating for Trump because the pollster had regularly released numbers on the lower end of Trump’s approval spectrum.
After decades of leading the political polling field, Gallup has shifted away from some of its more polarizing survey questions over the past dozen or so years. After struggling in the 2012 presidential election, the firm ceased polling elections ahead of the 2016 race. And after increasing its approval polling during President Barack Obama’s tenure, the pollster began dialing back its output, releasing approval data monthly at the time of its termination.
It’s the end of the world as we know it. Well, the end of a polling era, that is. Last Wednesday, Gallup announced it would cease measuring presidential approval ratings — ending an 88-year run of tracking how the American public views its chief executive.
In some ways, this was only the latest step in the slow demise of Gallup’s traditional political polling. The pollster stopped surveying presidential elections after the 2012 contest. Similarly, the firm had dialed back its approval polling, offering fewer data points about public attitudes in recent years than it had in the past. Still, it’s striking that the gold standard name in polling will no longer release numbers regarding the president’s standing, one of the most commonly polled questions in U.S. politics.
Gallup’s move will have some consequences for contemporary political analysis, as a long-running and well-regarded poll will no longer appear in polling averages. More broadly, the decision to stop surveying this important question will curtail a nine-decade long time series of approval data that has served as one of the cornerstones of understanding American politics. And, this latest step back from political polling could also portend further erosion in Gallup’s public-facing political information.
You won’t have Gallup to kick around anymore
The immediate effects of Gallup ending its presidential approval polling won’t be earth-shattering. This is one of the most-polled questions asked by pollsters, and one that most major firms regularly include in their surveys. Quantitatively then, removing Gallup’s monthly approval poll will not greatly alter the polling averages that track President Donald Trump’s standing with the public.
Still, Gallup’s exit could subtly produce a higher approval rating for Trump. During his second term, Trump has regularly polled worse in Gallup’s surveys than his average approval in five trackers: Decision Desk HQ, FiftyPlusOne, The New York Times, RealClearPolitics, and Silver Bulletin. In Gallup’s polling in 2025, with just one exception, Trump’s net approval was lower than his average net approval in each of these trackers on the same date. The only exception was November 2025, when Gallup’s net rating equaled the -13 net average in FiftyPlusOne’s tracker.
Now, some of these trackers adjust a pollster’s results to account for “house effects” when calculating an average. That is, the tracker’s model accounts for the tendency of some firms to consistently produce more favorable or unfavorable approval numbers — not because of a particular bias (in most cases), but more due to aspects of a pollster’s methodological choices. Considering Gallup tended to find Trump doing worse than most other pollsters, a house effects adjustment would generally increase Gallup’s raw approval numbers before affecting an average, at least slightly. Even then, Gallup’s adjusted approval would remain on the low side. Take Silver Bulletin’s calculations: In 10 of 12 surveys conducted in 2025, Silver Bulletin’s adjusted approval figures for Gallup still fell below Silver Bulletin’s overall average.
Gallup’s absence will also remove a pollster whose polls often held somewhat more influence in many averages. For instance, Gallup is one of about 20 pollsters marked as “select” by The New York Times that has released approval data during Trump’s second term. The Times identifies a pollster as “select” if it meets two of three criteria: accuracy in recent elections, membership in a professional polling organization, and/or use of probability-based sampling in its polling. The Times’s tracker gives surveys from select firms somewhat more weight in their average.
As a side effect, this also means that Gallup’s polls will no longer influence election forecasts. Historically, presidential approval is a fairly predictive indicator of the overall electoral environment, particularly in a midterm election. For that reason, average presidential approval is often an input in election forecasting models as a “fundamental;” that is, one of the underlying factors that helps tilt the overall electoral landscape in one direction or the other before even going through the circumstances surrounding each race on the ballot.
Not with a bang but a whimper
While Gallup’s departure perhaps registers as a 5.0 earthquake from a quantitative standpoint, historically it amounts to a much larger convulsion. Dr. George Gallup, the firm’s founder, helped pioneer public opinion research. In the 1936 presidential election, Gallup’s scientific survey famously proved more accurate than the Literary Digest’s straw poll, which foresaw President Franklin Roosevelt losing reelection (he won in a landslide).1
In 1937, Gallup began asking respondents if they were “for or against” Roosevelt, the firm’s initial attempt to gauge approval and disapproval for the president.
In 1938, the pollster tweaked the question to more directly ask if respondents approved or disapproved of the president.
And by 1941, the wording looked much like how pollsters have asked it for nearly nine decades: Do you “approve or disapprove” of how the president is handling his job?
Gallup established a deep well of polling on presidential approval, elections, issues, and most everything under the sun.2 In the age of the internet, the Gallup Presidential Job Approval Center has served as an incredibly helpful portal for comparing public attitudes toward presidents over time. Having built up such a lengthy record, it is hard not to feel sad (as an analyst) that a nine-decade time series of public opinion data is coming to an end.
Still, this news is not all that stunning when we put it in the context of Gallup’s shift away from political and electoral polling — and the financial realities of the survey business.
The first major hit came in the 2012 presidential election. In the final weeks of the campaign, Gallup regularly found Republican Mitt Romney ahead of President Barack Obama, sometimes by sizable margins. This looked bad for Gallup when Obama went on to win 51%-47%, though Gallup’s final survey had Romney only nominally ahead 49%-48% among likely voters and actually had Obama up 49%-46% among registered voters.
All together, though, Gallup had just about the worst showing of any major pollster in that election. In response, Gallup reviewed what happened and tested various approaches that would enable it to better identify likely voters — one of the main problems the firm had run into in the Obama-Romney matchup. But by the time 2015 rolled around, Gallup decided it would not survey the presidential race, making the 2016 contest the first since 1932 to not feature a Gallup horserace poll.
Yet Gallup did continue to invest in its presidential approval polling. Throughout Obama’s presidency and the first year of Trump’s first term, Gallup released daily updates for presidential approval. But in 2018, it reduced its output to a weekly approval figure — still quite often, to be sure. The downshift happened at least partly due to methodological realities. The pollster had used phone surveys to produce its approval polling, but it began shifting to mail surveys, which made a daily figure impossible. By 2019, Gallup had returned to periodic surveys rather than the daily or weekly approval surveys it used for a time.
But during President Joe Biden’s tenure, Gallup shifted to releasing its approval polling on a monthly basis, a practice it continued into the first year of Trump’s second term in office. While Gallup’s polls usually had relatively large sample sizes — about 1,000 or more respondents — they were conducted over an unusually long period of around two to three weeks. This gave the firm more time to acquire a larger sample without costing as much as a three- or four-day survey of the same size. The long survey period was not necessarily a huge negative when it came to measuring presidential approval, which is often quite stable. But it did make Gallup’s approval polling somewhat less responsive to major events that could affect the president’s standing in a short period of time.
In the wake of last week’s announcement, observers have wondered if Gallup’s routinely poor numbers for Trump may have contributed to its decision to drop the approval question. If that played into Gallup’s choice, it would not be the first company to make a business decision aimed at avoiding Trump’s potential wrath. As Chris Stirewalt noted at The Hill last week, Trump has filed lawsuits against pollsters before, such as one aimed at Ann Selzer and the Des Moines Register over their final survey of Iowa in the 2024 presidential race.3
The firm explained its decision as “an evolution in how Gallup focuses its public research and thought leadership.” Gallup pointed out (not incorrectly) that approval ratings are “widely produced, aggregated, and interpreted,” so the pollster felt it no longer could “make its most distinctive contribution” by polling that question. Regardless of what went into the decision, polling is more costly and difficult these days, which may have made it easier to forsake asking a polarizing question that, from the firm’s statement, it viewed as having limited upside for differentiation.
Nonetheless, having cut election and presidential approval polling, it’s easy to wonder what might be next on Gallup’s chopping block. For instance, media outlets and analysts often look to the firm’s long-running data on party identification as an indicator of how the political winds are shifting. Will the firm continue releasing that data? And what of the public-facing presidential approval data? Will that portal cease to exist in the near future because it is no longer adding new data?
In his own response to the Gallup news, Charles Franklin, director of the Marquette Law School Poll, pointed out with chagrin that there just is not much of a financial interest in providing public opinion data to the public these days. Gallup’s withdrawal from the approval scene is just the latest sign. That’s a big change from when George Gallup started polling: Back then, newspapers subscribed to Gallup’s polls and distributed the results to readers across the country. But not anymore; news media is increasingly fragmented and short on money to pay for surveys; as Franklin noted, it’s easier to just report on someone else’s poll.
📆 Around the Corner 📌
Notable upcoming elections:
March 10, 2026
GA-14 Special Election (runoff likely)
April 7, 2026
Wisconsin Supreme Court General Election
GA-14 Special Election Runoff (if necessary)
April 16, 2026
NJ-11 Special Election
Check out our 2026 Primary Primer for more information about the regular primaries happening in 2026!
To be clear, Gallup was not alone in polling the White House race, nor was his survey closest to the final mark. Gallup’s final 1936 survey, conducted in late October, found Roosevelt garnering 54% of the vote. By comparison, Elmo Roper — namesake of the Roper Center, an invaluable repository of historical polling data — found Roosevelt receiving about 60%. Roosevelt ended up winning about 61% of the national popular vote.
Just go to a university library and find copies of the annual book series for the Gallup Poll — they are tomes.
A poorly-performing poll is not indicative of malice or election interference, as Trump claimed in the suit. As it was Selzer’s last time polling Iowa for an election, this proved to be an ignominious endnote for her career. Now she’ll be remembered for a terrible survey, undermining her former reputation as one of the best pollsters in the country.