FACT CHECK: Trump Is at 32% in Quinnipiac—Not in Every Poll

Introduction

On July 29, 2026, Quinnipiac measured President Trump’s approval at 32%. That is an accurate, pollster-specific result. It is not the number reported by every late-July survey in this dossier, which range from 32% to 46%. A correct number becomes false when its label is stripped away.

The correction is not complicated: identify the pollster, retain the field dates and margins where the source provides them, and compare like with like. CNN/SSRS, AP-NORC, Reuters/Ipsos, Rasmussen, Morning Consult, and RMG Research did not publish the same number. No single result is a consensus by itself.

Quinnipiac’s 32%

A July 29 result

Quinnipiac University reported 32% approval for President Trump on July 29, 2026, described in the dossier as a low for that pollster’s second-term measurement. The result is real within Quinnipiac’s method.

It cannot truthfully be restated as the number reported by every pollster. A fact check begins by keeping the pollster attached to its number.

A house-specific low

The 32% holds as a Quinnipiac result. The universal claim fails.

Its direct consequence is limited but concrete: quinnipiac’s 32% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to quinnipiac’s 32%, and it cannot be carried into a claim the source does not make. Keeping that boundary visible protects the comparison from false certainty. Quinnipiac owns its result, not every pollster’s.

CNN/SSRS at 34%

Fieldwork from July 23 to 27

The CNN/SSRS poll measured 34% approval among 1,225 adults from July 23 to 27, 2026, with a 3.2-point margin of error. These methodological facts set the scale of the estimate.

A survey is a measurement of a sampled population during defined field dates. It is not a timeless census, and its margin is part of the reported result.

A sample and margin disclosed

CNN/SSRS supplies a nearby low number with a disclosed sampling frame.

The next interpretive step is narrower: cnn/ssrs at 34% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to cnn/ssrs at 34%, and it cannot be carried into a claim the source does not make. Its limited reach is the safeguard against a misleading shortcut. A sample and margin belong to the number.

AP-NORC at 33%

The same field window

AP-NORC reported 33% approval among 1,165 adults during July 23–27, with a 3.7-point margin of error. The overlap in dates with CNN/SSRS does not make the two surveys identical instruments.

Different pollsters can interview different samples and weight responses differently while covering similar dates. That is why agreement near the low end is information, not a license to erase every higher measure.

A different sample and margin

AP-NORC reinforces the presence of a low cluster. It does not create unanimity.

For the public record, one practical rule follows: ap-norc at 33% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to ap-norc at 33%, and it cannot be carried into a claim the source does not make. No larger conclusion can be built without new evidence. Similar dates do not create identical surveys.

Reuters/Ipsos at 39%

July 24 to 26 fieldwork

Reuters/Ipsos measured 39% approval among 1,248 adults from July 24 to 26, 2026, reporting a 2.9-point margin of error. It falls within the same late-July period as the lower results.

The six-to-seven-point distance from Quinnipiac is not a typographical nuisance. It is part of the methodological spread the article must disclose.

A higher contemporary reading

Reuters/Ipsos does not cancel Quinnipiac. It proves the available measures were not one number.

This detail changes the reading in a specific way: reuters/ipsos at 39% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to reuters/ipsos at 39%, and it cannot be carried into a claim the source does not make. The distinction prevents a true detail from being misused. A higher poll is evidence, not an error.

The actual range

From 32% to 46%

The named surveys in this dossier span 32% to 46%. That interval is the simplest factual correction to any claim that one low poll represents a universal consensus.

Poll houses differ in sampling, weighting, and collection mode, according to the fact block’s stated limitation. The resulting spread belongs in the headline logic as much as the lowest figure does.

Why range beats a slogan

A range is less convenient than a talking point. It is more faithful to the evidence.

The operative limit appears at precisely this point: the actual range must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to the actual range, and it cannot be carried into a claim the source does not make. Accuracy depends on leaving that constraint in place. The range is the fact the slogan leaves out.

Rasmussen at 44%

A July 30 publication

Rasmussen Reports published 44% approval on July 30, 2026. This result sits above the 32%, 33%, 34%, and 39% figures reported by other organizations.

The presence of a 44% reading makes it false to say every cited poll found 32% or even that every cited poll was in the mid-thirties. It remains one house’s result, not a corrective master number.

A number above the low cluster

Rasmussen is part of the documented spread. It has to be named.

The evidence carries one clear consequence: rasmussen at 44% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to rasmussen at 44%, and it cannot be carried into a claim the source does not make. The source earns confidence by being read at this scale. Forty-four percent is part of the record.

Morning Consult at 45%

A July 26 reading

Morning Consult reported 45% approval on July 26, 2026. Its number was published in the same general late-July period as the surveys reporting lower support.

The comparison shows why a fact check must distinguish an individual measure from a cross-poll average or consensus, neither of which is supplied here as a final official figure.

The upper end of the list

Morning Consult moves the ceiling higher. It does not rewrite the floor.

Here the record demands a separate inference: morning consult at 45% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to morning consult at 45%, and it cannot be carried into a claim the source does not make. The broader claim must wait for a record that actually supports it. The ceiling does not erase the floor.

RMG Research at 46%

The highest listed result

RMG Research measured 46% approval on July 26, 2026, the highest figure enumerated in the assigned fact block. Its presence fixes the upper edge of the documented range.

The article does not choose 46% as the “true” answer any more than it chooses 32%. Each poll is attributed to its own house and method.

A house result, not a verdict

The high number makes the low number less universal, not less real.

The information has a disciplined use: rmg research at 46% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to rmg research at 46%, and it cannot be carried into a claim the source does not make. Its proper weight lies in what it says, not in what rhetoric adds. A house result is not a national verdict.

Priority issues in CNN/SSRS

73% say focus is insufficient

In the CNN/SSRS poll, 73% of respondents said Trump was not focused enough on priority issues. That is a stated perception about attention to issues, not the same question as overall job approval.

Conflating the two measures would create a false causal story. The survey reports both, but neither alone proves why a respondent answered the other way.

A different question from approval

Issue focus and approval can travel together or apart. The questionnaire treats them as distinct.

What this establishes is deliberately bounded: priority issues in cnn/ssrs must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to priority issues in cnn/ssrs, and it cannot be carried into a claim the source does not make. The evidence remains strongest when it is not overextended. Approval and issue focus are different questions.

Republican approval at 78%

A low for that subgroup measure

CNN/SSRS put approval among Republican voters at 78%, described as a low for that measure. A partisan subgroup is not the national adult sample and cannot be substituted for it.

The internal-party number matters because it shows the poll’s composition beneath the national estimate. It does not independently forecast an election or define every Republican voter.

National and subgroup figures differ

A subgroup adds texture. It does not replace the national measure.

The factual result calls for one careful conclusion: republican approval at 78% must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to republican approval at 78%, and it cannot be carried into a claim the source does not make. That restraint is the practical meaning of a source-based account. A subgroup is not the country.

Republican enthusiasm

19% versus 38% in spring

Only 19% of Republicans said they were enthusiastic about the rest of Trump’s term in CNN/SSRS, compared with 38% in spring 2026. The fact block calls this a change in reported enthusiasm, not a seat forecast.

Enthusiasm is a different survey construct from approval. It can show an attitude shift in one poll without authorizing predictions about turnout, campaigns, or final results.

Mood is not an electoral forecast

Mood can be measured. The future cannot be manufactured from it.

The immediate implication is not expansive: republican enthusiasm must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to republican enthusiasm, and it cannot be carried into a claim the source does not make. The public record improves when the inference stops there. Enthusiasm measures mood, not a finished election.

Field dates before August 4

Immediately preceding context

All field dates listed in the fact block fall before the strict August 4–7, 2026 reporting window. They are immediate prior context, not interviews conducted during that week.

Polling freshness depends on when respondents were contacted, not merely when an article repeated the result. That time distinction is small on a calendar and large in a fast political story.

Publication date is not fieldwork

The interviews came first. The later coverage must say so.

This part of the record imposes a defined constraint: field dates before august 4 must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to field dates before august 4, and it cannot be carried into a claim the source does not make. Nothing in the assigned material authorizes a wider verdict. Fieldwork dates prevent false freshness.

The fact-check conclusion

One accurate figure used inaccurately

The statement Trump is at 32% in Quinnipiac, when properly dated and attributed, is accurate. The statement becomes false if it says or implies that all pollsters put him there.

This correction does not rehabilitate or attack any political figure. It restores the missing source names, field dates, margins where available, and the documented range.

A claim corrected without denial

One correct number can become a false generalization when its label is removed.

The final lesson from this evidence is restrained: the fact-check conclusion must be read on its own terms before anyone turns it into a broader conclusion. Its evidentiary boundary is specific to the fact-check conclusion, and it cannot be carried into a claim the source does not make. This is how fact stays separate from speculation. The fact check corrects the generalization, not the data.

Conclusion

Quinnipiac’s 32% is a dated and attributable measurement. The error comes when the result is made to speak for all polling. The available record contains a real methodological spread, not a single settled public number.

The honest public summary is a range with sources: 32% to 46% across the named late-July measures, with known field dates, samples, and margins where provided. That is not a weaker claim. It is the verified one. The 32% figure stands. The universal claim does not.

Sources

Primary sources

• CNN — Poll of Polls — accessed in the dossier
• CNN — Poll reporting and methodology — July 29, 2026
• CNN — Poll of Polls reference — accessed in the dossier

Secondary sources

• USA Today — CNN/SSRS polling report — July 29, 2026
• New York Times — Approval polling context — August 1, 2026
• The Guardian — Approval polling report — July 30, 2026
• Fortune — Poll comparison — August 2, 2026