Summary: President Trump is attempting to counter falling approval ratings by promoting favorable statistics and using sympathetic channels to shape public perception. A White House email survey claiming that 91% of respondents saw lower gas prices is an example of data that is likely non-representative and potentially biased. Observers warn that the selective release of information by allies and officials risks turning federal institutions into instruments of partisan messaging; new claims about election fraud and other allegations should be evaluated for provenance and methodology.
Trump Turns to 'Real Numbers' — and Federal Power — to Counter Slipping Polls
Trump’s got a very Trumpian plan to fight off bad poll numbers
President Donald Trump is increasingly casting his falling approval ratings as a problem of measurement rather than leadership, promoting favorable statistics and relying on sympathetic channels to shape public perception. As his second term faces a weakening economy, persistent price pressures and contentious immigration policies, the White House is emphasizing numbers that portray the administration more positively.
What Trump Is Claiming
At a December White House event, Trump said,
"When the real numbers start coming out, and the real pollsters start doing the polls, I think you’re going to see some really fantastic numbers."He has long dismissed unfavorable polls as biased while spotlighting positive surveys as authoritative.
The 91% Claim And Why It’s Problematic
One example came on his social platform, where the White House shared an image reporting that 91% of respondents had noticed lower gasoline prices since Trump returned to the White House. The source listed for that figure was a "White House email survey." That kind of self-administered poll is an unreliable gauge of broad public opinion for several reasons: the sample is likely biased toward supporters, the question as reported appears to presuppose a drop in prices, and it addresses only a narrow slice of consumer experience.
Partisanship Shapes Perceptions
Partisan views can dramatically affect how people answer questions about policy. After the 2017 tax cuts, for example, most Americans reported by March 2018 that they had not seen a clear increase in their take-home pay — while roughly two-thirds of Republican respondents said they had noticed a benefit. That contrast illustrates how survey wording and respondent composition can skew results.
One of President Donald Trump’s recent Truth Social posts.Via Truth Social
Pattern Of Selective Releases
The White House’s reliance on friendly numbers fits a broader pattern: the administration has on several occasions released selective or partisan material to support its narratives. In previous episodes tied to the Russia inquiry, Attorney General William Barr appointed a special investigator, John Durham, to examine the probe’s origins; Durham’s work did not substantively overturn the prevailing account. More recently, the administration has used official channels to publicize selective information about investigations and election-related claims.
Why This Matters
Federal institutions carry a veneer of legitimacy that can amplify otherwise dubious claims. Many Americans instinctively give greater weight to information presented by government offices or officials; that trust can be abused if data and documents are selectively released to serve political ends. Observers worry the approach risks turning public institutions into tools for partisan messaging rather than neutral sources of information.
What To Watch
Trump has promised "real numbers" and pledged that additional materials alleging election fraud or other misconduct would be made public. Readers should scrutinize the provenance, sampling methods and context for any forthcoming claims — especially those presented through partisan or government channels.
Bottom line: Promoted statistics can be persuasive, but without transparent methodology and representative sampling they do not resolve the larger questions about policy performance or public opinion.
