|
2016 US
Presidential Election Pollster Ratings
The following table lists the
accuracy of pollsters in the November 8, 2016 US Presidential election using the measure of polling
accuracy proposed by Martin, Traugott, and Kennedy.
An accuracy measure of 0.00
indicates that the odds ratio of the survey results matches the odds
ratio of the actual results (the closer to 0.00, the better).
| 2016
Presidential Election |
|
Clinton |
Trump |
Odds
Ratio |
Accuracy |
|
| Rasmussen |
45% |
43% |
1.00 |
0.00 |
| UPI/CVOTER |
49% |
46% |
1.02 |
0.02 |
| CBS/NY Times |
45% |
42% |
1.02 |
0.02 |
| Politico/Morning
Consult |
45% |
42% |
1.02 |
0.02 |
| McClatchy/Marist |
44% |
43% |
0.98 |
-0.02 |
| Bloomberg/Selzer |
44% |
41% |
1.03 |
0.03 |
| Reuters/Ipsos |
42% |
39% |
1.03 |
0.03 |
| Fox News |
48% |
44% |
1.04 |
0.04 |
| Boston Herald/Franklin Pierce/RKM |
48% |
44% |
1.04 |
0.04 |
| Angus Reid |
48% |
44% |
1.04 |
0.04 |
| ABC/Washington Post |
47% |
43% |
1.05 |
0.04 |
| Gravis Marketing |
47% |
43% |
1.05 |
0.04* |
| Economist/YouGov |
45% |
41% |
1.05 |
0.05 |
| CBS News |
45% |
41% |
1.05 |
0.05 |
| Insights West |
45% |
41% |
1.05 |
0.05 |
| NBC/WSJ |
44% |
40% |
1.05 |
0.05 |
| Times-Picayune/Lucid |
45% |
40% |
1.08 |
0.07 |
| Monmouth |
50% |
44% |
1.09 |
0.08 |
| NBC/SurveyMonkey |
47% |
41% |
1.10 |
0.09* |
| IBD/TIPP |
43% |
45% |
0.91 |
-0.09 |
| Pew |
46% |
40% |
1.10 |
0.10 |
| LA Times/USC |
44% |
47% |
0.89 |
-0.12* |
| USA Today/Suffolk |
49% |
39% |
1.20 |
0.18* |
| CNBC/Hart/POS |
43% |
33% |
1.25 |
0.22* |
| Saint Leo University |
45% |
34% |
1.27 |
0.24* |
|
| Actual |
48.18% |
46.09% |
|
|
*The accuracy measures for the Gravis Marketing
poll (sample size of 16,639), the USA Today/Suffolk poll (sample size
of 1,000), the Saint Leo University poll (sample size of 1,050), the
NBC/SurveyMonkey poll (sample size of 70,194), the CNBC/Hart/Public
Opinion Strategies poll (sample size of 804), and the LA Times/USC
poll (sample size of 2,972) fall outside the 95% confidence intervals.
The accuracy measures for the other 19 polls fall within the
95% confidence intervals.
The average absolute accuracy for all polls in
2016 is 0.070. The median absolute accuracy for all polls in 2016 is
0.049. The average absolute accuracy for the 19 polls in 2016
where the accuracy measures fall within the respective 95% confidence
intervals is 0.045.
The average absolute accuracy for all polls in the
2012
presidential election was 0.055. Three polls in 2012 had accuracy measures fall outside the 95% confidence intervals (YouGov, Rasmussen, and Gallup). The average absolute accuracy for all polls in the 2008 presidential election was 0.037. None of the polls in 2008 had accuracy measures outiside of the 95% confidence intervals.
|
