Generic ballot
Pollster ratings →Asked which party they’d back for Congress, voters favor Democrats by 8.2 points in the average of 26 recent polls (adjusted for house effects and likely voters).
13 generic-ballot polls ended in the last 30 days; the latest ended Oct 12. 9 of them were entered from the pollsters’ own releases because our main poll feed doesn’t carry them.
| Pollster | End date | Sample | Result | Adjusted |
|---|---|---|---|---|
| YouGov | Oct 12 | 1,403 RV | D+7.0 | D+7.2 |
| YouGov | Oct 5 | 1,381 RV | D+9.0 | D+9.2 |
| Pew Research Center | Oct 4 | 3,917 RV | D+8.0 | D+8.1 |
| YouGov - CBS News | Oct 2 | 1,572 LV | D+9.0 | D+8.9 |
| YouGov | Sep 28 | 1,429 RV | D+8.0 | D+8.2 |
| HarrisX/Harris | Sep 28 | — LV | D+2.0 | D+4.7 |
| Quinnipiac University | Sep 27 | 1,032 RV | D+12.0 | D+9.7 |
| Emerson College | Sep 22 | 1,000 LV | D+11.0 | D+11.4 |
| The Argument/Verasight | Sep 22 | — LV | D+10.0 | D+9.2 |
| YouGov | Sep 21 | 1,401 RV | D+9.0 | D+9.2 |
| Verasight | Sep 21 | — RV | D+8.0 | D+6.2 |
| Echelon Insights | Sep 21 | 1,002 LV | D+8.0 | D+9.0 |
| Ipsos | Sep 20 | 979 RV | D+8.0 | D+9.8 |
| CNN/SSRS | Sep 17 | 867 RV | D+8.0 | D+7.5 |
| Marist College | Sep 15 | 1,280 RV | D+12.0 | D+10.2 |
| YouGov | Sep 14 | 1,461 RV | D+5.0 | D+5.2 |
| Beacon Research/Shaw & Co. Research | Sep 14 | 1,211 RV | D+7.0 | D+6.2 |
| YouGov | Sep 8 | 1,015 LV | D+11.0 | D+11.2 |
| Quinnipiac University | Sep 6 | 970 RV | D+11.0 | D+8.7 |
| Morning Consult | Sep 6 | 2,203 RV | D+4.0 | D+5.8 |
| YouGov | Aug 31 | 1,592 A | D+5.0 | D+4.3 |
| Ipsos | Aug 31 | 924 RV | D+5.0 | D+6.8 |
| YouGov | Aug 24 | 1,534 A | D+6.0 | D+5.3 |
| YouGov | Aug 24 | 1,377 RV | D+6.0 | D+6.2 |
| Ipsos | Aug 24 | 1,215 A | D+6.0 | D+6.9 |
Presidential approval
President Trump’s net approval (approve minus disapprove) averages -24.6. Unpopular presidents usually cost their party seats in midterms; the model uses this as a weak prior on the national environment.
12 approval polls ended in the last 30 days; the latest ended Oct 12. 10 of them were entered from the pollsters’ own releases because our main poll feed doesn’t carry them.