Analysis · · by Eren Menteşe
The week in polls: November 2–8, 2026
A quiet week after the midterms: Trump approval rose to 37.5%, the 2028 primary averages held, and one poll moved Vance vs. Ocasio-Cortez to Vance +1.5.
The week of the midterms was busy at the ballot box and quiet in the polling averages. Four national surveys that finished fieldwork between November 2 and November 8 gave six results in the races we track. Only one average moved by more than half a point.
Trump approval: 37.5%
The Trump approval average rose 0.3 points to 37.5% approve. Three new polls ended this week. The Economist/YouGov had 36% approving and 62% disapproving, Reuters/Ipsos 36% and 63%, and John Zogby Strategies, polling likely voters on the eve of the election, 43% and 55%. The approval average uses a 14-day half-life, so it reacts faster than the others. Even so, three polls on either side of the average left it almost where it was.
The 2028 primaries: no change
McLaughlin & Associates published a post-election survey that ended November 5. It had Kamala Harris at 28%, Gavin Newsom at 16% and Pete Buttigieg at 9% among Democratic primary voters, and JD Vance at 38%, Marco Rubio at 17% and Donald Trump Jr. at 14% among Republicans. Both subsamples were small, at 471 and 427 likely voters.
Neither average moved. The Democratic primary average still has Harris ahead at 23.1%, and the Republican primary average still has Vance at 38.7%. Each average weighs more than 60 polls with a 60-day half-life, so one survey rarely changes them by a tenth of a point. Primary polling this far out mostly measures name recognition. No major candidate has declared yet; see who is running for president in 2028 for what each has said.
Matchups: one poll moves Vance vs. Ocasio-Cortez
The same McLaughlin survey tested a hypothetical general election between Vance and Alexandria Ocasio-Cortez and found Vance ahead 49% to 39%. That poll moved the Vance vs. Ocasio-Cortez average from Ocasio-Cortez +0.5 to Vance +1.5. McLaughlin is a Republican firm, and this matchup has few polls behind it, so a single survey can move it by two points. Treat the change as a data point, not a trend. There were no new polls of Vance vs. Newsom (Newsom +3.4) or Vance vs. Buttigieg (Buttigieg +2.8).
The numbers
| Average | Now | A week earlier | Change |
|---|---|---|---|
| Trump approval | Approve 37.5% | Approve 37.2% | +0.3 |
| Dem primary | Harris 23.1% | Harris 23.1% | no change |
| GOP primary | Vance 38.7% | Vance 38.7% | no change |
| Vance vs. Newsom | Newsom +3.4 | Newsom +3.4 | no change |
| Vance vs. Ocasio-Cortez | Vance +1.5 | Ocasio-Cortez +0.5 | 2.0 toward Vance |
| Vance vs. Buttigieg | Buttigieg +2.8 | Buttigieg +2.8 | no change |
Every poll that finished fieldwork this week, newest first, with a link to the pollster's release:
| Poll | Ended | Race | Result |
|---|---|---|---|
| The Economist / YouGov, 3,166 RV | Nov 6 | Trump approval | Approve 36%, Disapprove 62% |
| McLaughlin & Associates, 1,000 LV | Nov 5 | Vance vs. Ocasio-Cortez | Vance 49%, Ocasio-Cortez 39% |
| McLaughlin & Associates, 427 LV | Nov 5 | GOP primary | Vance 38%, Rubio 17%, Trump Jr. 14% |
| McLaughlin & Associates, 471 LV | Nov 5 | Dem primary | Harris 28%, Newsom 16%, Buttigieg 9% |
| Reuters / Ipsos, 1,357 RV | Nov 2 | Trump approval | Approve 36%, Disapprove 63% |
| John Zogby Strategies, 1,001 LV | Nov 2 | Trump approval | Approve 43%, Disapprove 55% |
LV means likely voters and RV registered voters. Averages are as of November 8 and a week earlier, computed with the ElectionsUSA method.
What to watch
The national House popular vote is still being counted. How far the final polls missed it is tracked, with every count labelled preliminary until states certify, in polls vs. results. On the 2028 side, the next milestone is the first formal campaign launch. In recent cycles that came within weeks to months of the midterms; see when 2028 candidates are likely to announce.
Numbers on this page are computed from the open dataset each time the site is rebuilt. Method: methodology. Corrections: corrections log.