Nine Polls in Two Weeks Put the President's Approval at a Second-Term Low

What the polls said

The last two weeks of July produced an unusually dense run of national approval polling, and it pointed one direction. A Quinnipiac University survey released July 29 put President Trump's job approval at 32 percent, the lowest reading in that pollster's series for him (Quinnipiac University, July 29, 2026). An Associated Press and NORC poll released July 30 had him at 33 percent (AP-NORC, July 30, 2026). A CNN survey conducted by SSRS from July 23 to July 27, among 1,225 adults, found 34 percent approval and 66 percent disapproval, matching the career low he first reached in January 2021 (CNN, SSRS, July 29, 2026).

The rest of the field filled in the range. Pew Research Center had him at 34 percent approval and 64 percent disapproval (Pew Research Center, July 23, 2026). CBS News and YouGov measured 39 percent and 61 percent (CBS News, YouGov, July 26, 2026). Emerson College put approval at 39 percent (Emerson College, July 23, 2026). A CNBC survey had 40 percent approval and 59 percent disapproval (CNBC, July 17, 2026), and a Washington Post and Ipsos poll landed at 37 percent and 61 percent (Washington Post, Ipsos, July 16, 2026). Eight polls, a spread from 32 to 40, every one of them under water.

The average is the number that matters

Any single poll can miss. The discipline is to read the average, and the average agreed. The Silver Bulletin polling tracker on July 30 showed 38.2 percent approval against 58.8 percent disapproval, a net of minus 20.6 points (Silver Bulletin, July 30, 2026). That net was the lowest of the president's second term, and it had moved there quickly, from minus 16.4 points on July 12 (Silver Bulletin, July 30, 2026).

A four-point slide in a polling average across eighteen days is a real move, not statistical noise. It is the kind of shift that shows up when a specific event pushes across party lines rather than hardening each side in place. This month there was such an event.

The visible driver

The clearest weight on the numbers was the American military action against Iran. Approval of that action fell from 44 percent in March to 39 percent by late July in CBS News and YouGov polling (CBS News, YouGov, July 26, 2026). Quinnipiac found 60 percent of voters disapproved of the conflict and 34 percent approved, the widest gap it had recorded on the question since the fighting began (Quinnipiac University, July 29, 2026). A war that a country is not sure it wanted is a heavy thing to carry into an approval number, and it carried.

That is the honest attribution. The decline is broad, but it is concentrated where the news was loudest, and the news this summer was a shooting war with a rising energy bill attached to it. Approval numbers are a moving average of how a month felt, and this month felt like that.

Read all the way down

Here is the part a one-day headline leaves out. A reading in the low 30s is a bad number, and it is also a number several presidents have gone well below and lived past. Harry Truman fell to 22 percent in February 1952 (Gallup). Richard Nixon sat at 24 percent the month he resigned (Gallup). George W. Bush reached 25 percent in October 2008 during the financial crisis (Gallup). President Trump's own first-term average of roughly 41 percent was the lowest full-term average in Gallup's records, and he still won the office again four years later (Gallup). The current 34 percent in the CNN survey does not match Truman or Nixon or Bush. It matches his own January 2021 mark (CNN, SSRS, July 29, 2026).

Approval is volatile by design. It rewards good months and punishes bad ones, and it says almost nothing about where a number will sit six months on. Treating a single week of polling as a verdict is the same error as treating one quarter of growth as an economy. It is a data point in a long series.

What US Political Rank measures instead

This site does not rank presidents by how they poll. It ranks them by what the record shows, which is a deliberate choice and the reason the site exists. When US Political Rank grades presidents by economic record, it uses full-term averages of real growth and inflation from the Bureau of Economic Analysis and the Bureau of Labor Statistics, not the mood of any month. Popularity and performance are different measurements, and history is full of presidents who were unpopular and effective, popular and idle, and every combination in between.

Polls belong to the reader who knows how to weigh them, which is a skill worth building before the midterm season floods the zone with them (see how to read a presidential approval poll). The eight numbers from late July are real, they agree, and they are low. They are also a snapshot of one hard summer, taken by pollsters whose own track records vary (see pollster accuracy rankings). Both of those statements are true, and holding them together is the whole job.