November 8, 2012

Study: Which Pollsters Most Accurately Predicted Election

By Rob Kall

In an exclusive interview, Costas Panagopoulos, Director, of Fordham's Center for Electoral Politics and Democracy discussed the result of a study that assessed the accuracy of 28 pollsters who predicted the outcome of the Presidential election.

::::::::

Pollsters not only predicted winners and losers. The pollsters themselves WERE winners and losers. A study by Fordham U. political scientist Costas Panagopoulos ranked 28 pollsters based on how well they predicted the actual outcome of the popular vote. 

In an exclusive interview, Costas Panagopoulos, Director, of Fordham's Center for  Electoral Politics and Democracy told Opednews.com that "Polls were pretty close to each other for the most part, were  pretty close to projecting what actually happened on election day."

Fordham's study found that the top pollsters, in terms of accuracy, were Ipsos/Reuters, YouGov, PPP,  Daily Kos/SEIU/PPP and Angus-Reid* 

Panagopoulos's study reports that none of the polls showed partisan biases, stating, "Most (22) polls overestimated Romney support, while six (6) overestimated Obama strength (indicated with a * below), but  none of the 28 national pre-election polls I examined had a significant partisan bias.  

Markos Moulitsas, founder and head of Dailykos, which used PPP, which was initially ranked first, before the election returns were updated, observed,

"PPP is a robo-pollster that doesn't call cell phones, which was supposedly a cardinal sin--particularly when their numbers weren't looking so hot for Obama post-first debate. But there's a reason we've worked with them the past year--because their track record is the best in the biz.

One last point--YouGov and Ipsos/Reuters were both internet polls. YouGov has now been pretty good two elections in a row. With cell phones becoming a bigger and bigger issue every year, it seems clear that the internet is the future of polling. I'm glad someone is figuring it out.

But let's be clear, you have to go down to number six on the list to get to someone who called cell phones. And Gallup called 50 percent cell phones and they were a laughingstock this cycle."

I asked professor Panagopoulos about his take on how his findings found that the top ranking polls did not include mobile phones. He replied,"They were not so far off to claim that we should not use mobile phones," but added, "The top polls were not based on contacting mobile phone users. They were automated or internet based surveys." And he concluded, referring to the robo-pollster and internet-based polls, "Some of these approaches may not be as  problematic as some of the critics believe."

Actually, Gallup, ranked among the lowest in both this 2012 poll AND a similar poll Panagopoulos did after the 2008 election did use mobile phones, at least in this poll. 

I asked "Are there are any methodologies that your findings suggest should be altered?" 

He replied, "I would look at the survey organizations -- I also did a similar study after the 2008 election and there are organizations like Gallup that come at the bottom of the list in both cycle. I think we ought to think about how those polls are being conducted and how they can be improved."

I also asked, "Any ideas on how orgs like Gallup should change their methodologies?" He replied, "Not at this time."

Panagopoulos did change the ranking, later in the day after he originally posted the study, when final counts continued to come, changing the margin of Obama's win from 1.7 to 2.2 percent. That moved PPP, which had been in first place in his first ranking, lower down, with Ipsos Reuters and YouGov ranking at the top. 

The study stated, "For all the derision directed toward pre-election polling, the final poll estimates were not 

far off from the actual nationwide voteshares for the two candidates. On average, preelection polls from 28 public polling organizations projected a Democratic advantage of 

1.07 percentage points on Election Day, which is only about 1.13 percentage points away 

from the current estimate of a 2.2-point Obama margin in the national popular vote 

(Obama 50.3% versus Romney 48.1%)."

The Full Ranked List of Pollsters From the Study (Dr. Panagopoulos says these could change as the final numbers of the popular vote count changes. Early articles by some major publications based their listings of rankings on an initial report that has since been updated, with revised rankings.)

"The following list ranks the 28 organizations by the predictive accuracy of their final, 

national pre-election estimates (as reported on pollster.com). "   

1. Ipsos/Reuters 

2. YouGov 

3. PPP (D) 

3. Daily Kos/SEIU/PPP 

4. Angus-Reid* 

5. ABC/WP* 

6. Pew Research* 

6. Hartford Courant/UConn* 

7. Purple Strategies 

8. NBC/WSJ 

8. CBS/NYT 

8. YouGov/Economist 

9. UPI/CVOTER 

10. IBD/TIPP 

11. Democracy Corps (D)* 

12. CNN/ORC 

12. Monmouth/SurveyUSA 

12. Politico/GWU/Battleground 

12. FOX News 

12. Washington Times/JZ A

2. Newsmax/JZ Analytics 

12. American Research Group 

12. Gravis Marketing 

13. National Journal* 

14. Rasmussen 

14. Gallup 

15. NPR 

16. AP/GfK

Bob Cesca observes, "For Gallup and Rasmussen this is just embarrassing. And Gallup seemingly knew this was how things were going to turn out because they ceased polling last Monday after having Mitt Romney up by as much as 6 points nationally among likely voters."

Panagopolous also did a similar study of the 2008 Pollsters. Here's the list from four years ago. 

1. Rasmussen (11/1-3)** 

1. Pew (10/29-11/1)** 

2. YouGov/Polimetrix (10/18-11/1) 

3. Harris Interactive (10/20-27) 

4. GWU (Lake/Tarrance) (11/2-3)* 

5. Diageo/Hotline (10/31-11/2)* 

5. ARG (10/25-27)* 

6. CNN (10/30-11/1) 

6. Ipsos/McClatchy (10/30-11/1) 

7. DailyKos.com (D)/Research 2000 (11/1-3) 

8. AP/Yahoo/KN (10/17-27) 

9. Democracy Corps (D) (10/30-11/2) 

10. FOX (11/1-2) 

11. Economist/YouGov (10/25-27) 

12. IBD/TIPP (11/1-3) 

13. NBC/WSJ (11/1-2) 

14. ABC/Post (10/30-11/2) 

15. Marist College (11/3) 

16. CBS (10/31-11/2) 

17. Gallup (10/31-11/2) 

18. Reuters/ C-SPAN/ Zogby (10/31-11/3) 

19. CBS/Times (10/25-29) 

20. Newsweek (10/22-23)

Rob Kall is executive editor, publisher and website architect of OpEdNews.com, Host of the Rob Kall Bottom Up Radio Show (WNJC 1360 AM), and publisher of Storycon.org, President of Futurehealth, Inc, and an inventor . He is also published regularly on the Huffingtonpost.com

Listen to over 200 of Rob's Podcast interviews here.

Rob is, with Opednews.com the first media winner of the Pillar Award for supporting Whistleblowers and the first amendment.

With his experience as architect and founder of a technorati top 100 blog, he is also a new media / social media consultant and trainer for corporations, non-profits, entrepreneurs and authors.

Rob is a frequent Speaker on the bottom-up revolution, politics, The art, science and power of story, heroes and the hero's journey and Positive Psychology. He is a campaign consultant specializing in tapping the power of stories for issue positioning, stump speeches and debates, and optimizing tapping the power of new media. Watch me speaking on Bottom up economics at the Occupy G8 Economic Summit, here.

See more Rob Kall articles here and, older ones, here.