“Trump dismisses low poll ratings as ‘fake numbers’ in Time interview”, 1 October 2026 (Photo via The Guardian — the original report)
As it ran on the front
The number is good. I want to lead with that, because the number is good and I count maybe three people in this building willing to say it out loud: new claims at 197,000, the lowest since mid-July. And the report says why. Job security. That is not my phrase, that is the report's phrase. A worker who is not worried about the job does not file.
Now, I have seen the polling. September 21st, Reuters and Ipsos, approval at 32 percent — Reuters called that a record low for his entire political life, and I would push back on the methodology there, except that two days later The Hill's business letter had rural support chipping away on the economy, and on the 24th Emerson had disapproval at the highest of the second term, and when three outfits with three different samples land in the same place, that is not methodology, that is — that is a sampling artifact. Did I say that? That doesn't sound like me. Three readings are not a trend. Three readings are three readings.…
…(cont) What I would ask you to hold onto is the gap. The report describes workers who feel secure. The surveys describe workers handing him his worst economic marks yet. That is the problem.
On the litigation. The president sued an Iowa pollster and the Des Moines Register over the 2024 poll, and on September 24th a state judge dismissed that suit. We are reviewing options. The theory was never about one poll; the theory was that a poll is a product and a product can be defective, and the court did not rule that the number was — the court dismissed it. Fine.
And then today, Time. Asked about the low ratings, he called them fake numbers. So we have two explanations in one news cycle for one approval rating: job security, which accounts for the good number, and fake numbers, which accounts for the bad one, and those cannot both be — they are complementary. They are complementary. One is a measurement and one is a characterization, and I would never confuse the two, except that I just did, and I am walking that back.
In the same interview he deflected on the length of the Iran war, which he blamed on the media reporting his remarks, and he declined to rule out declaring a state of emergency before the midterms. Emergency authority is a tool every president has.
🍺
“A government report tells you that you feel secure in your job, and the same workers it was written to calm just handed him 32 percent — Reuters says the worst number of his life — so the number that flatters him is data and the number that doesn't is fake. He sued a pollster over a poll he didn't like and a judge threw it out last week. Now he won't rule out a state of emergency before the midterms, and the only instrument left that can tell him no is the one he's already calling fake.”
“Look, the claims number is real, the claims number is the lowest since mid-July, and I don't apologize for leading with the thing that is working. Polls are a snapshot, they are modeled, they are weighted, and anyone who says a snapshot is a verdict is selling you — well, he does say the bad ones are fake, which is a stronger claim than mine, and I would not have phrased it that way. Did I say that? That doesn't sound like me. We stand behind both numbers.”
President also doesn’t rule out declaring state of emergency before midterms and deflects responsibility for Iran war US politics live – latest updates Donald Trump deflected responsibility for misleading Americans about the length of the Iran war, dismissed low poll ratings as “fake numbers” and refused to rule out the possibility of…
This page is a back-issue: the story as it ran, receipts as they were. The current edition is at the front. The byline is a pen name for a column drafted by a machine and checked by the editor: how this is made.
