the smarter average
2024 AVERAGES: HOME | NATIONAL VOTE | ELECTORAL COL. | HOUSE | SENATE | Battlegrounds: AZ - GA - NC - MI - PA - WI - NV
         
  Rumble   Facebook   X   YouTube   Truth
  PENNSYLVANIA
 
>>>LOOK AT MY PAST PROJECTIONS. THE MOST EXTENSIVE POLL AVERAGES, PERIOD.
POLITICAL REF'S AVG:
Trump +0.8
Harris 48.0, Trump 48.8
Sample Size: 16,377 LV
Theoretical Margin of Error: 0.8%
Sample Model: 2020 Exit Poll Turnout Model (Polls that did not use 2020 Exit Poll were adapted to it, justifying a margin or error).
EXIT POLL & 2024 TURNOUT ASSUMPTIONS
POLLSTERS:
Trump +0.7
Harris 48.1, Trump 48.8
If a poll is missing, it does not release cross tabs, the sample is too small and the topline is an outlier (less than 500), or averaged 3 or more off the result in the last two presidential elections.
POLL RESULT SUMMARY (SEE FULL RESULTS & EXPLANATIONS BELOW)
Poll Date Sample + Party Distribution + Response Bias Adjustments Result Margin Poll- ster Avg Ref's Avg
ATLAS INTEL 11/3-11/4 1840LV, Party: 39D/18I/44R. Gender: 45/54 M/F. Education: 47/52 BA/<BA. Race: 88/5/3/2/2 WBLAO. Age: 11/17/34/39 18-30-45-65+ Harris 48.4, Trump 49.1   x  
Political Ref's Take Poll Bias Score = 43 to Dems. Insignificant bias. Bias summary on lines 9-11. Included in my average. Harris 48.4, Trump 49.1 +0.7   x
EMERSON 10/30-11/2 1000LV, Party: 40D/18I/41R. Gender: 48/51 M/F. Education: 40/60 BA/<BA. Race: 80/12/5/2/2 WBLAO. Age: 13/16/13/21/38 18-29-30-40-50-60+ Harris 48, Trump 49   x  
Political Ref's Take Poll Bias Score = 2, no bias. Bias summary on lines 9-11. Included in my average. Harris 48, Trump 49 +1.0   x
NYT / SIENNA 10/29-11/2 1527LV, Party: 44D/12I/43R. Gender: 46/53 M/F. Education: 38/62 BA/<BA. Race: 70/7/5/1/2 WBLAO. Age: 13/22/33/30 18-30-45-65+. Harris 48, Trump 48   x  
Political Ref's Take Poll Bias Score = 42 toward Dems, insignificant bias. Bias summary on lines 911. Included in my average. Harris 48, Trump 48 TIED   x
TRAFALGAR GROUP 11/1-11/3 1089LV, 40D/20I/40R. Gender ID: 48/52. Race: 80/12/5/2/2 WBLAO. Age: 13/14/15/38/21 18-30-40-50-65+ Harris 46.7, Trump 48.4   x  
Political Ref's Take Poll Bias Score = 38 toward Dems, insignificant bias. Bias summary on lines 9-11. Included in my average. Harris 46.7, Trump 48.4 +1.7   x
INSIDER ADVANTAGE 11/1-11/2 800LV, All voter IDs inferred. Party: 40D/18I/42R. Gender: 47/53 M/F. Race: 83/9/6/2 WBLO. Age: 18/50/32 18-40-65+. Harris 48.2, Trump 49.2   x  
Political Ref's Take Poll Bias Score = -95 toward GOP, definite bias. Bias summary on lines 9-11. After adjustment, Harris 48.6 to Trump 48.7 Harris 48.6, Trump 48.7 +0.1   x
MUHLENBERG COLLEGE 10/27-10/30 469LV, Party: 45D/9I/46R. Gender: 49/50 M/F. Education: 38/62 BA/<BA. Race: 75/10/7/3/6 WBLAO. Age: 15/27/29/30 18-30-50-65+ Harris 49, Trump 47   x  
Political Ref's Take Poll Bias Score = 86 toward Dems, definite bias. Bias summary on lines 9-11. After adjustment, Harris 48.2 to Trump 47.6. Harris 48.2, Trump 47.6 +0.6   x
USA TODAY / SUFFOLK 10/27-10/30 500LV, Party: 45D/15I/40R. Gender: 47/52 M/F. Education: 39/61 BA+/NoBA. Race: 80/12/5/1/1 WBLAO. Age: 20/24/26/28 18-35-50-65+. Harris 48, Trump 48   x  
Political Ref's Take Poll Bias Score = 65 towards Dems, definite bias, will adjust. Far too many Dems and too few white ID, pro-Harris biases, not offset by too few college degrees, a trump bias. After adjustment, Harris gets 48.1 to Trump 48.9. Calculations here. Harris 47.8, Trump 48.9 +1.1   x
ECHELON INSIGHTS 10/27-10/30 600LV, Party: 35D/23I/39R. Gender: 47/53 M/F. Education: 38/62 BA+/NoBA. Race: 83/9/4/2/2 WBLAO. Age: 19/23/27/31 18-35-50-65+. Harris 46, Trump 51   x  
Political Ref's Take Poll Bias Score = -35, insignificant bias. Too many GOP and too few college degrees, pro-Trump biases, offset by disproportionately pro-Harris age groups. Included in my average. Harris 46, Trump 51 +5.0   x
MARIST 10/27-10/30 1400LV, Party: 40D/19I/41R. Gender: 50/50 M/F. Education: 39/61 BA+/NoBA. Race: 85/8/5/6 W/B/L/O. Age: 20/16/64 18-35-45+ Harris 50, Trump 48   x  
Political Ref's Take Poll Bias Score = -73 towards the GOP, definite bias, will adjust. Too few women/many men, too few college degrees, too few voters from Philly, too many white ID and disproportionately pro-Trump age groups. After adjusting, Harris 50.0 to Trump 47.8. Harris 50.0, Trump 47.8 +2.2   x
WASHINGTON POST 10/26-10/30 1204LV, Party: 43D/15I/42R. Gender: 47/52 M/F. Education: 39/61 BA+/NoBA. Race: 82/10/4/1 WBLAO. Age: 31/40/29 18-40-65+. Harris 48, Trump 47   x  
Political Ref's Take Poll Bias Score = 35, insignificant bias. Primary bias is in the age groups, pro-Harris, but it is insignificant. Included in my average. Harris 48, Trump 47 +1.0   x
RASMUSSEN REPORTS 10/25-10/28 849LV, Party: 40D/19I/41R. Gender: 48/52 M/F. Race: 81/11/8 WBO. Age: 14/16/17/28/24 18-30-40-50-65+. Region: 11/21/17/20/26/5 Phi/PhiSub/NE/Cent/W/ Elsewhere Harris 47, Trump 49   x  
Political Ref's Take Poll Bias Score: 35, insignificant bias. Primary bias is in the age groups, pro-Harris, but it is insignificant. Included in my average. Harris 47, Trump 49 +2.0   x
FOX NEWS 10/24-10/28 1057LV. All voter IDs are inferred. Party 40D/17I/43D. Gender: 47/53 M/F. Education ID: 40/60 BA+/NoBA. Race: not provided. Age: 34/66 18-45+ Harris 49, Trump 50   x  
Political Ref's Take Poll Bias Score = -45 to GOP, insignificant bias. Slightly too many GOP. Too few 18-45 and too many 45+. The biases were small. Included in my average. Harris 49, Trump 50 +1.0   x
QUINNIPIAC 10/24-10/28 2186LV, Voter category IDs inferred. Party: 40D/17I/43R. Gender: 47/53 M/F. Education ID: 47/53 BA+/NoBA. Race: 81/10/5/1/1 WBLAO. Age. 21/21/31/28 18-35-50-65+. Harris 47, Trump 49   x  
Political Ref's Take Poll Bias Score = 29 to Dems, insignificant bias. Too many GOP/few Dems, too many white ID, disproportionately pro-Trump age groups. These pro-Trump biases were offset by far too many college and post-grad degrees, a pro-Harris bias. Harris 47, Trump 49 +2.0   x
CBS NEWS 10/22-10/28 1273RV, All Category IDs inferred. Party: 40D/18I/42R. Gender: 47/53 M/F. Education: 53/47 BA+/NoBA. Race: 81/11/5/1/1 WBLAO. Age: 11/21/38/31 18-30-45-65+. Harris 49, Trump 49   x  
Political Ref's Take Poll Bias Score = 44 to Dems, insignificant bias. Disproportionately pro-Trump age groups offset by far too many college and post-grad degrees, a pro-Harris bias. Included in my average. Harris 49, Trump 49 TIED   X
FRANKLIN & MARSHALL 10/9-10/20 583LV, Party ID: 44D/15I/41R. Gender ID: 49/50 M/F. Education ID: 43/57 BA+/NoBA, Racial ID: 81/19 W/O. Age ID: 20/33/47 18-35-55+. Regional ID: 32/12/19/8/30 SE/NE/SW/NW/Cent. Harris 49, Trump 50   x  
Political Ref's Take Poll Bias Score = 30 towards Dems, or insignificant bias. Included in my averages. Too many Dems and college and post-grad degrees, pro-Harris biases sufficiently offset by too few women/many men and slight pro-Trump regional and racial ID biases. Included. Harris 49, Trump 50 +1.0   x
OUTDATED OR DUPLICATE POLLSTER BELOW. NOT INCLUDED IN AVERAGE.
ATLAS INTEL 10/27-10/30 1738LV, Party: 40D/18I/43R. Gender: 46/53 M/F. Education: 49/51 BA+/NoBA. Race: 77/10/9/2/3 WBLAO. Age. 15/23/39/24 18-30-45-65+. Harris 47.4, Trump 48.5   x  
Political Ref's Take Poll Bias Score = 164 to Dems, severe bias, will adjust. Too many college and post-grads, non-white ID and pro-Harris age groups, not offset by slightly too many GOP. Must adjust. After adjusting, no change in the result. Again the pollster seems to have accounted for the bias in the raw numbers. Harris 47.4, Trump 48.4. Harris 47.4, Trump 48.4 +1.0   x
INSIDER ADVANTAGE 10/26-10/27 800LV, All voter IDs inferred. Party: 40D/19I/41R. Gender: 50/50 M/F. Race: 84/7/7/2 WBLO. Age: 27/48/26 18-40-65+. Harris 47.4, Trump 48.4   x  
Political Ref's Take Poll Bias Score = -96, definite bias, will adjust. Too many men, white ID and pro-Trump age group.

Harris 48.0, Trump 48.2

 +0.2   x
EMERSON 10/21-10/22 860LV, Party ID: 40D/19I/41R. Gender ID: 48/52. Education ID: 40/60. Racial ID: 80/12/5/2/2 W/B/L/A/O. Age ID: 21/21/30/27 18-35-50-65+ Harris 48, Trump 49   x  
Political Ref's Take Poll Bias Score = -22 towards GOP, or insignificant bias, included in my average. No biases large enough to mention. The poll balances. Included Harris 48, Trump 49 +1.0   x
NYT / SIENNA 10/7-10/10 857LV, Party ID: 37D/23I/35R. Gender ID: 47/52 M/F. Education ID: 39/61 BA+/NoBA. Racial ID: 69/8/4/2/3 W/B/L/A/O. Age ID: 13/21/33/29 18-30-45-65+. Harris 47, Trump 43 x    
Political Ref's Take Poll Bias Score 67 towards Dems, definite bias. Will adjust. There are too many Dems, which makes up most of the lead for Harris. Also far too few white ID, only 69% of the electorate instead of 81%. This is not going to happen. GOP outnumbered Dems in PA in 2020 despite months of Dem GOTV and dream conditions for Dems, and Republicans have closed the registration gap by over half since 2020, so no way this turnout manifests in PA this time, especially with Harris's relative unpopularity with unions and men. But when you inflate Dems by a net of 3 points, you get Harris only five behind with men and nearly tied with white voters. This poll is an outlier. After adjustment, Harris 48.4, Trump 47.9. Harris 48.4, Trump 47.9 +0.5   x
WALL STREET JOURNAL 9/28-10/8 600RV, Party ID: 47D/7I/46R. Gender ID: 48.5/51.5 M/F. Education ID: 41/59 BA+/NoBA. Racial ID: 79/9/5/2/5 W/B/L/A/O. Age ID: 23/23/24/27 18-35-50-65+. Harris 46, Trump 47 X    
Political Ref's Take Poll Bias Score 79 towards Dems, definite and excessive bias, will adjust. Slight too many Dems, slightly too many college and post-grad degrees and far too disproportionate pro-Harris age groups not sufficiently offset by too few women/many men. After adjustment, Harris gets 47.5 to Trump 49.2. See all assumptions and calculations here. Harris 47.5, Trump 49.2 +1.7   x
AMERICAN GREATNESS / TIPP 10/7-10/9 803LV, Party ID: 42D/17I/41R. Gender ID: 48/52 M/F. Education ID: 44/56 BA+/NoBA. Racial ID: 85/9/6 W/B/L. Age ID: 4/28/35/35 18-25-45-65+. Harris 48, Trump 49 x    
Political Ref's Take Poll Bias Score -34, insignificant bias, included in my average. Slightly too many Dems, too many college and post-grad degrees, offset by slightly too few women/many men, too many white ID and disproportionately pro-Trump age groups. The sample balances out sufficiently. See calculations here. Harris 48, Trump 49 +1.0   x
ATLAS INTEL 10/27-10/30 1738LV, Party: 40D/18I/43R. Gender: 46/53 M/F. Education: 49/51 BA+/NoBA. Race: 77/10/9/2/3 WBLAO. Age. 15/23/39/24 18-30-45-65+. Harris 47.4, Trump 48.5   x  
Political Ref's Take Poll Bias Score = 164 to Dems, severe bias, will adjust. Too many college and post-grads, non-white ID and pro-Harris age groups, not offset by slightly too many GOP. Must adjust. After adjusting, no change in the result. Again the pollster seems to have accounted for the bias in the raw numbers. Harris 47.4, Trump 48.4. Harris 47.4, Trump 48.4 +1.0   x
WASHINGTON POST / GMU 9/30-10/15 707LV, Party ID: 35D/31I/34R. Gender ID: 48/51 M/F. Education ID: 40/60 BA+/NoBA. Racial ID: 83/10/4/1/3 W/B/L/A/O. Age ID: 14/17/14/25/29 18-30-40-50-65+. Harris 49, Trump 47 x    
Political Ref's Take Poll Bias Score -33, insignificant bias, included in the average. Too few Other ID and too few women. The biases are small. Included in the average. Harris 49, Trump 47 +2.0   x
RASMUSSEN REPORTS 10/9-10/13 1072LV, Party ID: 43D/15I/42R. Gender ID: 48/52. Racial ID: 81/11/8 W/B/O. Age ID: 15/16/17/28/24 18-30-40-50-65+. Regional ID: 10/21/17/20/26/6 Philly/PhillySubs/NE/ Central/West/Misc. Harris 47, Trump 50 x    
Political Ref's Take Poll Bias Score 9, no bias, included in my average. Too many Dems and slightly too few voters from Western and Central PA, pro-Harris biases offset by slightly too few women/many men, slightly too few other Racial ID, and slightly too few 18-29, pro-Trump biases. The sample balances well, included in my average. Harris 47, Trump 50 +3.0   x
INSIDER ADVANTAGE 10/7-10/8 800LV Harris 47, Trump 49 x    
Political Ref's Take Awaiting release of the cross-tabs. Until then, excluded. Excluded    
QUINNIPIAC UNIV 10/3-10/7 1412LV, Party ID: 36D/23I/34R. Gender ID: 48/52 M/F. Inferred Education ID: 38/62 BA+/NoBA. Inferred Racial ID: 81/12 W/B. There is no reported data related to Latinos or Asians. Inferred Age ID: 18/18/40/25. 18-35-50-65+. This age data is invalid, but necessary to make these numbers work. Harris 49, Trump 47 x    
Political Ref's Take Poll Bias Score 64, definite and excessive bias, will attempt to adjust for it. Quinnipiac polls often do not add up in the internals and this one is particularly inconsistent. The vote ratios simply don't add up unless 40% of the sample is between 50-64. The education ID is always incomplete in these polls, which represents a serious defect. After working through five voter ID adjustments, I verified that a Harris two-point lead was justified by the sample. Quinnipiac must report the full percentage of college grads as well as the Asian and Latino vote to be adequately transparent. Harris 47, Trump 45 +2.0   x
ATLAS INTEL 10/12-10/17 2048LV, Party ID: 41D/21I/39R. Gender ID: 49/50. Education ID: 51/49 BA+/NoBA. Racial ID: 76/11/7/3/3 W/B/L/A/O. Age ID: 14/26/38/23 18-30-45-65+. Region ID: 55/21/24 Sub/Urb/Rur. Harris 47, Trump 50 x    
Political Ref's Take Poll Bias Score 212 toward Dems, severe bias, must adjust. One pro-Trump bias, too few women, overwhelmed by large pro-Harris biases including too many Dems, far too many college and post-grad degrees, too few white/too many Latino and other IDs and too many 18-44. After adjusting these biases, Harris gets 46.5 to Trump 50.7. See the calculations here. Harris 46.5, Trump 50.7 +3.2   x
TRAFALGAR GROUP 10/17-10/19 1084LV, Party ID: 40D/21I/39R. Gender ID: 48/52 M/F. Racial ID: 80/12/5/2/2 W/B/L/A/O. Age ID: 13/14/15/38/21 18-30-40-50-65+. Harris 43, Trump 46 x    
Political Ref's Take Poll Bias Score 13, no bias, included in my average. No areas of significant bias. Harris 43, Trump 46 +3.0   x
EMERSON 10/5-10/8 1000LV, Party ID: 40D/19I/41R. Gender ID: 48/52 M/F. Education ID: 40/60 BA+/NoBA. Age ID: 13/16/13/21/38 18-30-40-50-60+. Racial ID: 80/12/5/2/2 W/B/L/A/O. Harris 48, Trump 49 x    
Political Ref's Take Poll Bias Score -2, no bias, included in my average. There are no significant areas of bias in this poll. Included. Harris 48, Trump 49 +1.0   x
Trafalgar Group 9/26-9/29 1090LV, Party ID: 40D/20I/39R. Gender ID: 48/52 M/F. Education ID: Unavailable. Age ID: 13/14/15/38/21 18-30-40-50-65+. Racial ID: 80/12/5/2/2 W/B/L/A/Other. Harris 45.3, Trump 47.5 x    
Political Ref's Take Poll Bias Score -50 to the GOP, or insignificant bias, included in my average. There were slightly too few Republicans and slightly too few white ID, which was balanced but too many Dem leaning age groups. The biases balance well, included. Assumptions and calculations here. Harris 45.3, Trump 47.5     x
ATLAS INTEL 9/20-9/25 1775LV, Party ID: 46D/17I/37R. Gender ID: 49/50 M/F. Education ID: 47/53 BA+/NoBA. Racial ID: 81/9/6/2 W/B/L/A. Age ID: 17/21/39/23 18-30-45-65+. Harris 48, Trump 51 x    
Political Ref's Take Poll Bias Score 155 to Dems, severe bias, all biased categories will be adjusted and averaged with the topline result. Far too few Republicans and far too many college and post-grads not balanced by slightly too GOP age groups and slightly too few women. All assumptions and calculation are here. After adjustment, Harris gets 47.4 to Trump 51.7. Harris 47.4, Trump 51.7     x
FOX NEWS 9/20-9/24 775LV, Party ID: 45D/11I/44R. Gender ID: 51/48 M/F. Education ID: 28/72 BA+/NoBA. Inferred Age ID: 25/75 <45/>45. Inferred racial ID: 81/11/5/1 W/B/L/A. Harris 49, Trump 49      
Political Ref's Take Poll Bias Score -288 to GOP, severe bias, all biased categories will be adjusted and averaged with topline result. Too many men, non-college degree holders, too few under 45s, not sufficiently balanced by too many Democrats. The vote margins didn't differ enough to make a big difference, all assumptions and calculations can be seen here. Adjusted, Harris gets 48.0 to Trump 47.5. Harris 48.0, Trump 47.5     x
FABRIZIO / ANZALONE 9/17-9/24 600LV. Inferred Party ID: 41D/21I/38R. Inferred Gender ID: 47/53 M/F. Inferred Education ID: 38/62 BA+/NoBA. Inferred Racial ID: 70/17/9/2 W/B/L/A. Inferred Age ID: 22/23/29/26 18-35-50-65+. Harris 50, Trump 47 x    
Political Ref's Take Poll Bias Score 156 towards Dems, severe bias, will adjust. I had to infer all voter ID ratios, but vote ratios were provided so I was able to do that with a good degree of confidence. Too many Democrats and too few Republicans were polled, far too few white ID and too many black and Latino ID were included, and finally slightly too many 18-49 and and slightly too few 50+ were included. There was one pro-GOP bias, slightly too few college and post-grad degrees were included, but not nearly enough to offset all the pro-Dem biases. After adjusting all biased IDs and applying the vote ratios provided by the pollster, Harris gets 47.4 to Trump 47.7. See calculations here. Harris 47.4, Trump 47.7     x
SUSQUEHANNA 9/16-9/22 700LV, Party ID: 45D/12I/43R. Gender ID: 48/52 M/F. Age ID: 10/22/16/21/32 18-30-45-55-65+. Education ID: 61/39 BA+/NoBA. Racial ID: 78/10/5/3 W/B/L/A. Harris 46, Trump 46 x    
Political Ref's Take Poll Bias Score of 325 to Dems, or severe bias. Each biased category will be corrected with actual vote ratios, or estimates based on averages of other polls where voter ratios aren't provided, and averaged with the pollster topline. This poll had too many Democrats, far too many college and post-grad degree holders and too few white ID. These biases were not offset by too few women. All assumptions and calculations can be seen here. After correction, Harris gets 46.8 to Trump 48.1, corrected bias score is 35 to Dems. Harris 46.8, Trump 48.1     x
MUHLENBERG COLLEGE 9/16-9/19 450LV, Party ID: 45D/9I/44R. Gender ID: 48/51 M/F. Education: 39/61 BA+/NoBA. Racial ID: 74/10/6/3 W/B/L/A. Age ID: Harris 48, Trump 48 x    
Political Ref's Take Poll Bias Score 65, not severe but definite and excessive bias, will adjust each biased category and average with original topline result. Slightly too Democratic, far too few white ID and too many 18-29 and 65+, groups both leaning Dem in this poll. Dem leaning biases not sufficiently offset by slightly too many four-year degrees, too few women and too many men. After correcting each biased category, Harris gets 46.7 to Trump 48.8. See all calculations here. Harris 46.7, Trump 48.8     x
RASMUSSEN REPORTS 9/19-9/22 1202LV, Party ID: 43D/14/43R. *Education ID: 53/47 BA+/No BA. Gender ID: 48/52 M/F. Racial ID: 81/11/8 W/B/Other. Age ID: 15/33/28/24 18>30>50>65+. Harris 48, Trump 48 x    
Political Ref's Take Poll bias score is 199 towards Dems, severe bias, will adjust. While the bias score is high, the adjustment needs not be massive to correct for it because BA+ voted 50-47 Harris while non-BA voted 50-46 Trump. The model is very sensitive to bias, so a high score just means you need to look very closely at the turnout assumption. Almost all the bias comes from including far too many four-year college degrees and graduate school grads. Will adjust with education ID. Change to 40/60 BA+/No BA, Harris receives 48 to Trump 49. Harris 48, Trump 49     x
EMERSON 9/27-9/28 1000LV, Party ID: 40D/18I/41R. Gender ID: 48/52 M/F. Education ID: 40/60. Racial ID: 80/12/5/2 W/B/L/A. Age ID: 13/29/21/38 18-30-50-60+ Harris 48, Trump 48 x    
Political Ref's Take Poll Bias Score 3, no bias, included in my average. Harris 48, Trump 48     x
THE HILL / EMERSON 9/15- 9/18 880LV, Party ID: 40D/18I/41R. Gender ID: 48/51 M/F. Education ID: 40/60 BA+/No BA. Racial ID: 80/12/5/2/2 W/B/L/A/Other. Inferred Age ID: 13/29/39/17 18>30>45>65+ Harris 47, Trump 48 x    
Political Ref's Take Poll bias score -3, no bias, included. Harris 47, Trump 48     x
MARIST 9/12- 9/17 1476LV, Party ID: 41D/21I/38R. Gender ID: 49/51 M/F. Age ID: 29/71 18>35+, 42/58 18>45+. Education ID: 37/63 BA+/No BA. Racial ID: 78/10/6/6 W/B/L/O. Harris 49, Trump 49 x    
Political Ref's Take Poll bias score -24 to GOP, insignificant bias, included. Too many Dems and too few white ID offset by too few women, four-year degrees and too many under 45. Included. Harris 49, Trump 49     x
INSIDER ADVANTAGE 9/14- 9/15 800LV, Party ID: 36D/28I/36R. Gender ID: 48/52 M/F. Racial ID: 82/9/4 W/B/L. Inferred Age ID: 8/10/40/41 18>30>45>65+. Harris 48, Trump 50 x    
Political Ref's Take Poll bias score -133 to GOP, severe bias, will be adjusted. Far too many people over 45, slightly too many white ID and slightly too few women, not offset enough by too many Dems in sample. Will adjust using Party and Gender ID. With Party ID correction only, it's Harris 48.7, Trump 49.9. With age ID correction alone, it's Harris 50.3 to Trump 48.5. Averaging Party ID and Age ID together, Harris 49.5 to Trump 49.2. Poll bias after correction -38 to GOP, insignificant bias. Harris 49.5, Trump 49.2     x
WASHINGTON POST 9/12-9/16 1003LV, Party ID: 44D/15I/42R. Gender ID: 47/52 M/F. Age ID: 30/41/30 18>40>65+. Racial ID: 81/10/4/2 W/B/L/A. Education ID: 40/60 BA+/No BA Harris 48, Trump 48 x    
Political Ref's Take Bias score of 53 toward Dems. Slightly too many Dems and under 30. But I adjusted and the result stayed the same, so the bias was insignificant. Included. Harris 48, Trump 48     x
QUINNIPIAC UNIV 9/12-9/16 905LV, Party ID: 36D/25I/33R. Inferred Gender ID: 46/54 M/F. Inferred Education ID: 40/59 Degree/No Degree. Inferred Racial ID: 86/14 White/Black, no other races are reported. Harris 51, Trump 45 x    
Political Ref's Take *See Row Below Harris 50.0, Trump 47.1     x
*This poll received a bias score of 72, reflecting a definite and excessive bias, although not severe, mostly from a Party ID bias. There are also slightly too many women and people under 35 or over 65, but the party adjustment will suffice. Applying the party vote D 95-4 Harris, I 48-43 Harris, R 91-7 Trump to more likely Party ID of 40D/19I/41R. Harris gets 50 to Trump 47.1.
USA Today / Suffolk 9/11-9/14 500LV, Party ID: 37D/24I/34R; Gender 51/49 F/M; Education ID 59/41 Degree/No Degree; Racial Demo 69/12/4/12 White/Black/Asian/ Latino Harris 49, Trump 46 x    
Political Ref's Take *See Row Below for full explanation Harris 49.2, Trump 50.4     x
*This poll is rated SEVERELY BIASED (Score 120) towards Democrats. Click here for a full breakdown. Too many Dems, College Degrees and Racial Demo is substantially less white than 2020. There is a slight GOP bias on gender, too many men and too few women. Adjusting to 2020 exit poll levels for party ID. Gender, racial and education adjustments are impossible or imprudent because too little information is provided about how they voted. The racial bias of the poll also throws off these categories. These categories will be spoiled for accuracy when racial demos are so inaccurate. When you have 12% less whites than the 2020 vote suggests you should have, these other categories are impossible to adjust. Party ID provides the most reliable option. Apply the partially inferred (see bias calculator section "Party ID Calculator When Hidden by Pollster) party vote of D 95-5 Harris, I 52.5-47.5 Harris and R 96-3 Trump to more likely turnout of 40D/19I/41R, Harris gets 49.2 to Trump 50.4. I arrived at I 52.5-47.5 by allocating the reported 43-38 split for Harris to a value of 1 (0.525+0.475=1), eliminating unlikely voters.
CBS NEWS / YOUGOV 9/3-9/6 1085LV, Inferred Party ID: 41D/19I/40R. They assume Dem's lead in turnout despite R's improving in registration and R's turning out more than D's in 2020. Harris 50, Trump 50 x    
Political Ref's Take *See Row Below Harris 49, Trump 51     x
Party Vote: D's 92-8 Harris, I's 50-49 Trump, R's 93-7 Trump. Education Vote: White No Degree 62-38 Trump, White Degree 56-44 Harris. Assuming 2020 repeat of 40D/19I/41R, reallocating the party vote above, Harris receives 49 to Trump 51. Can't reallocate using education because YouGov doesn't report non-white education vote. Just looking at the white education vote, however, it would very likely increase Trump's lead if they provided the full education vote. Trump +24 among non-college whites.
Trafalgar Group 8/28-8/30 1082LV, 40D/20I/39R Harris 44.9, Trump 47.0 x    
Political Ref's Take The D+1 is wrong I think, but it's a small variation from the R+1 I assume. The racial makeup was 80% white, 12% black, 5% Hispanic, 1.5% Asian, 1.8% other. These are good racial turnout assumptions. No adjustment necessary. Harris 44.9, Trump 47.0     x
INSIDER ADVANTAGE 9/14-9/15 800LV, Party ID: 36D/28I/36D. Racial ID: 82/9/1/4 White/Black/ Asian/Latino. Gender: 52/48 Female/Male. Harris 48.4, Trump 49.8 x    
Political Ref's Take There were only slight GOP biases on gender and racial demos and a slight Dem bias on party ID, totaling a bias score of -20, meaning a slight GOP bias. Falls well within acceptable bias as explained on the poll bias calculator spreadsheet, no adjustment necessary. Included. Harris 48.4, Trump 49.8     x
THE HILL / EMERSON 8/25-8/28 950LV, Party turnout: D30/I36/R34, Education: 36% college degree, 64% no college degree Harris 48, Trump 48 x    
Political Ref's Take The R+4 is a bit high, but again it's a small variation and this is a likely voter poll. Education percentages are close. Included. Harris 48, Trump 48     x
CNN 8/23-8/29 789LV, 37D/26I/37R, Education level not provided, but can be inferred from the cross-tabs. College degree 45%, Non-college degree 55%. Harris 47, Trump 47 x    
Political Ref's Take Education level: Non-college grads voted 55-39 Trump, College grads voted 59-36 Harris. Allocating this vote ratio to the more likely education turnout, Harris 47 and Trump 47.4 Harris 47.0, Trump 47.4     x
FULL POLL RESULTS AND EXPLANATIONS OF ADJUSTMENTS
Poll Date Sample + Party Distribution Result Pollster Avg   Ref's Avg
CURRENT TURNOUT ASSUMPTION: 40D/19I/41R. This is the 2020 turnout, which is a heavy Dem turnout. Republicans are registering more new voters than Democrats over the last four years, a trend that has continued recently. Using 2020 exit polls is an assumption that should slightly favor Dems, a conservative approach. Gender: 47/53 Female/Male. Racial ID: 81/11/5/1 White/Black/Latino/Asian. Education ID: 60/40 No Degree/Degree.
 
2020 EXIT POLL: 40D/19I/41R without leaners (PDF); Demo - White 81%, Black 11%, Hispanic/Latino 5%, Asian 1%; Age - 18-29(13%), 30-44(23%), 45-64(37%), 65<(28%); Education - 40% College degree, 60% no college degree.
 

To see results of 2020 Pennsylvania polls below, click here

Best 2020 Pennsylvania Poll: Rasmussen Reports

Polls that got Pennsylvania right: Rasmussen Reports, Monmouth Univ, Frederick Polls, Muhlenberg College, RMG Research, Franklin & Marshall

Polls that got Pennsylvania wrong: Susquehanna Polling, AtlasIntel, Insider Advantage,

 
         
  Rumble   Facebook   X   YouTube   Truth