A Trump approval rating of 34% has prompted comparisons with Harry Truman’s standing before the 1946 midterm elections. The comparison appears in published reporting, but it requires an important qualification: different surveys and voter populations do not produce one interchangeable approval number.
Reports describing the New York Times/Siena University poll put President Donald Trump’s approval among registered voters at 34%. Siena’s own October 21 summary highlights a different result: 36% approval and 59% disapproval among likely voters.
Those figures should not be combined or presented as contradictory without explaining whom each estimate describes. Registered voters and people classified as likely voters are different survey populations.
What the Trump approval rating comparison measures
A report carried by AOL described the 34% Times/Siena result as Trump’s lowest in that polling series. It attributed the comparison with Truman’s 33% in 1946 to The New York Times.
Silver Bulletin also reported the historical comparison, explicitly attributing it to the Times. That supports describing the claim as a published analysis, rather than treating it as a universal ranking independently established across every available poll.
The distinction between a poll and an average matters. Silver Bulletin explains that its approval average combines surveys and weights them using factors including reliability, sample size and recency. A single Times/Siena result is not that average.
Truman’s 1946 figure is documented
Gallup’s historical midterm table lists Truman’s approval at 33% for 1946. The American Presidency Project separately records 33% approval, 52% disapproval and 15% without an opinion in a September 13–18, 1946, survey.
That historical date matters. The number was measured before the election, not on election night. Describing both presidents as approaching midterms does not mean the surveys occurred the same number of days before voting.
Using the reported 34% Trump figure and Truman’s 33% produces a difference of one percentage point. It does not establish a statistically meaningful difference between surveys conducted eight decades apart with different methods.
The historical comparison concerns approval before a midterm election. It is not a claim that no president has ever recorded lower approval at another point in a presidency.
Siena’s accessible release shows 36% among likely voters
The Siena Research Institute’s October 21 release reports Trump’s job approval as 36% to 59%. Its summary also describes weak ratings on several policy issues and dissatisfaction with both major parties.
The release identifies a national sample of more than 3,000 voters. However, the full toplines and registered-voter crosstabs linked from the release could not be directly reviewed for this article.
The poll’s full methodology could not be independently verified. Exact fieldwork dates, the precise approval-question wording, subgroup sample sizes and applicable margins of error are therefore not asserted here.
For the Trump approval rating claim, the 34% registered-voter figure remains attributed to published reporting. The directly accessible Siena summary supports the separate 36% likely-voter result.
Trump disputes the unfavorable polling
Trump disputed the polls and claimed his approval was closer to 60%, The Independent reported on October 24.
A politician’s assessment of personal support is different from a documented survey result. Support among a president’s strongest party supporters also cannot be substituted for approval among the national population.
Low approval does not decide the midterms
Gallup’s historical analysis found that unpopular presidents’ parties generally experienced larger House losses. That is a historical relationship, not a guaranteed seat forecast for 2026.
The Trump approval rating comparison supplies context about the political environment. It does not determine turnout, individual candidate performance or which party will control Congress after November 3.
Hourly Times previously covered a separate September Reuters/Ipsos approval poll. That earlier survey should not be relabelled as the October Times/Siena poll or used as its historical benchmark.
