The Accuracy of the Final National Polls

by John Sides on November 7, 2012 · 9 comments

in Campaigns and elections

Fordham University political scientist Costas Panagopoulos has compared the last pre-election national poll from each of the pollsters available in the Pollster data to the estimated national popular vote.  He writes:

For all the derision directed toward pre-election polling, the final poll estimates were not far off from the actual nationwide vote shares for the two candidates. On average, preelection polls from 28 public polling organizations projected a Democratic advantage of 1.1 percentage points on Election Day, which is only about 1 point away from the current estimate of a 2.2-point Obama margin in the national popular vote (Obama 50.3% versus Romney 48.1%).
Following the procedures proposed by Martin, Traugott and Kennedy to assess poll accuracy, I analyze poll estimates from these 28 polling organizations. Most (22) polls overestimated Romney support, although some only slightly, while 6 overestimated Obama support. Most (22) polls overestimated Romney support, while six (6) overestimated Obama strength, but none of the 28 national preelection polls I examined had a significant partisan bias.

I’ve plotted the data above a simpler quantity than what Panagopoulos calculated, which conveys the thrust of his analysis.  If the overall picture changes dramatically as the votes are counted, we will report back.  I also hope to have something on the state polls soon.

For more on this subject, see Mark Blumenthal, Harry Enten, and Simon Jackman.

(Errata: This post has been corrected to account for my misunderstanding of what Panagopoulos had done.  The data in my chart are not the basis of his analysis that I quoted above.)

{ 9 comments }

RB November 7, 2012 at 2:50 pm

It would be great if you could include the margins of error in these plots.

Brice D. L. Acree November 7, 2012 at 4:25 pm

Agree. I threw it together. I couldn’t grab UConn, and I’m off to class, but here: http://marginoferror.org/pollfinals.png

J November 7, 2012 at 3:16 pm

Also, similar scalings would be nice..though the final number as compared to the final outcome would be give people a better idea of accuracy isntead of margin!

PollyUSA November 7, 2012 at 3:23 pm

Fascinating thanks

m0nty November 7, 2012 at 11:48 pm

The final margin is likely to be over 3%, BTW. That will skew it even more.

Robert Poyourow November 8, 2012 at 2:34 pm

Using the final polling and the focus on the margins ignore several problems. First, not all pollsters are in the field near the end of the election campaign, leading to errors in movement over time. Second, exaggeration of the undecideds can also misrepresent the condition of the race. Last, if one ignores differences from the mean of the polls throughout the campaign, then there is little accountability for those pollsters whose functions are to gin up their patron’s base or manage contributors’ enthusiasm. Opportunists and partisan outfits can always adjust their last poll to save their standing and rescue their reputations.

mz November 10, 2012 at 2:11 am

Not quite so fast, sparky. As of now, Obama leads by 3% and counting, which means that the 4 most accurate polls are currently Pew, Hartford Courant, Angus Reid and ABC/WAPO, with Democracy Corps joining your previous champs as the runners up.

Shock of shocks, Rasmussen and Gallup are among the 4 worst (YAY, Gallup, woefully wrong once again, and good thing you had Sandy come along to give you a chance to adjust your numbers so you didn’t have a laughable 7 point Romney lead.) Let’s review, Gallup. In the last three national elections your final tallies have been off 4, 8 and 4. Likely Voter model my keister. Rasmussen we expect to have dubious and partisan results, but when are we going to wake up about the farce that Gallup has become. They may beat out Rasmussen as the worst pollster of the year this cycle, in terms of accuracy. Oh, and let’s not forget their final November 5th swing state poll calling it a 48-48 “tie”. Almost…Obama only won the 12 swing states they polled by about 5 points…

sue November 15, 2012 at 2:33 am

what he said!

El Pedro IV November 12, 2012 at 8:47 am

Dick Morris needed new scaling to fit on the page !

Comments on this entry are closed.

Previous post:

Next post: