Last week, the polling firm Gallup announced it would no longer survey presidential-approval ratings, a decision that prompted speculation about potential political pressure. President Trump’s approval numbers have been declining, and in December, Gallup’s poll registered him at 36 percent—lower than the RealClearPolitics average of 42 percent. President Trump has previously taken punitive action against those publishing unfavorable polls, including a lawsuit against The Des Moines Register and its pollster, Ann Selzer.
Reports indicate other companies targeted by the president have also yielded to pressure, with CBS and ABC paying settlements to Trump’s presidential-library fund after being sued. On Monday, Stephen Colbert reported that CBS scrapped an interview with a Texas Senate candidate due to threats from Brendan Carr, the chair of the Federal Communications Commission. This context raised questions about whether Gallup’s decision was a similar concession.
Gallup’s stated reason for ending the polls—changes in its business strategy—reflects the economic realities facing polling companies. These firms face rising costs, declining returns, and increased competition. Polling relied on a sense of civic duty, with citizens accepting calls as part of their responsibilities, but that willingness has diminished as trust in institutions declines.
The Legacy of George Gallup
The foundations of modern polling were laid by Iowa-born George H. Gallup during the 1936 presidential campaign. Prior to Gallup, The Literary Digest dominated election prediction, successfully forecasting results from 1916 to 1932 by mailing postcards to a large sample of Americans. In 1936, The Literary Digest surveyed 10 million people, predicting a win for Alf Landon, but Gallup’s survey of 50,000 people correctly predicted a Roosevelt victory.
Gallup’s success stemmed from a more representative sample. The Literary Digest relied on lists from sources like automobile registrations and telephone exchanges, disproportionately reaching more affluent, Republican voters. Gallup’s smaller sample was more representative of the broader electorate.
Over time, polling techniques improved, but challenges emerged. As caller ID became widespread, response rates declined. In the late 1990s, 28 percent of those contacted by Gallup participated; by 2017, that figure dropped to 7 percent, and currently stands at 5 percent. In other words Gallup now needs to craft approximately 20,000 calls to obtain a 1,000-person sample, significantly increasing costs.
The declining response rates raise concerns about the representativeness of samples. Pollsters attempt to correct for this, such as Pew Research Center oversampling those who do not volunteer to make the sample more representative. However, these adjustments also add to the cost.
Gallup, unlike some other polling organizations, is a profit-seeking company. It ended its long partnership with CNN in 2006 and has since operated its presidential-approval research hoping to attract more business. However, the proliferation of low-quality polls in the internet age has diminished the value of reliable polling.
The U.S. Polling industry has faced a series of setbacks in the past decade, reminiscent of the 1948 election when Gallup incorrectly predicted Thomas Dewey would defeat President Truman. In 2016, polls consistently showed Hillary Clinton leading Donald Trump by an average of 3.2 percentage points, but Trump ultimately won the election.
The 2016 polls failed to accurately measure the support for Trump among less educated, less affluent, and more alienated voters. A similar underestimation of Trump occurred in 2024. The core issue appears to be the changing nature of the public itself, and the difficulty of measuring opinions within a polarized society.
The Gallup poll, once considered a reliable voice of the people, now joins other once-ubiquitous institutions that have struggled to adapt to a more divided and suspicious America.
Frequently Asked Questions
Why did Gallup stop surveying presidential approval ratings?
Gallup cited changes in its business strategy as the reason for ending presidential-approval polling.
What happened with the 1936 election and The Literary Digest?
The Literary Digest predicted Alf Landon would win the 1936 election, but Gallup correctly predicted Franklin D. Roosevelt’s victory. This was because The Literary Digest’s sample was unrepresentative of the electorate.
What challenges do pollsters face today?
Pollsters face challenges including declining response rates, rising costs, and difficulty obtaining representative samples due to changing demographics and a decline in civic engagement.
As polling methods continue to evolve, will they be able to accurately reflect the views of a changing electorate?
