FOR RELEASE NOVEMBER 9, 2006

Which Polls Had it Right?

<p class="MsoNormal">The final tally from the mid-term election is in—you know the poll in which people actually vote—and the media polls can now be graded. With each election, there are more media outlets, and more polls. Usually, the polls begin to converge as election day nears. This year, the polls varied widely. How did they stack up against the actual vote?</p>

BY Pew Research Center: Journalism & Media staff

FOR MEDIA OR OTHER INQUIRIES:

Communications Department
202.419.4372
www.pewresearch.org

RECOMMENDED CITATION

Pew Research Center, November 2006, "Which Polls Had it Right?"

About Pew Research Center

Pew Research Center is a nonpartisan, nonadvocacy fact tank that informs the public about the issues, attitudes and trends shaping the world. It does not take policy positions. The Center conducts public opinion polling, demographic research, computational social science research and other data-driven research. It studies politics and policy; news habits and media; the internet and technology; religion; race and ethnicity; international affairs; social, demographic and economic trends; science; research methodology and data science; and immigration and migration. Pew Research Center is a subsidiary of The Pew Charitable Trusts, its primary funder.

© Pew Research Center 2026

Table of contents

Which Polls Had it Right?

<p class="MsoNormal">The final tally from the mid-term election is in—you know the poll in which people actually vote—and the media polls can now be graded. With each election, there are more media outlets, and more polls. Usually, the polls begin to converge as election day nears. This year, the polls varied widely. How did they stack up against the actual vote?</p>

One of dominant phenomena in media coverage of politics is the proliferation of polling. As technology has exploded, the number of news outlets has expanded, and with it, so has the number of pre-election polls.

The polls can become a powerful lens through which reporters, pundits and editors shape coverage. Sometimes, stories can seem as if they were designed more to illustrate what the polls are showing than they are to explain what the candidates did or said.

How did the media polls do this year?

The final vote tally shows that Democrats won the popular vote for Congress by a margin of about 7 percentage points. An estimated 52% of citizens who cast ballots for the House voted for a Democrat, while 45% voted a Republican and 3% went for a third party candidate or didn’t vote for a House candidate.

Normally, as election day gets closer, the pre-election “generic” or national vote survey results tend to converge as undecided voters make up their minds and people settle on their final choices. This year, the opposite happened. In the last weekend before election day, the surveys got further apart, not just sowing confusion among pundits and creating anxiety among the pollsters, but probably creating public doubts about the validity of the polls. CNN’s survey showed a 20 point advantage for Democrats. The Pew Research Survey for the People and the Press had the narrowest range, 4 points.

So who was right? The polls that came closest were USA Today/Gallup poll at 7 points, and the ABC News/Washington Post poll, at 6. The Pew poll was next closest at 4. CNN, which formerly used Gallup, was furthest off at 20, followed by Newsweek (16) and Time (15).

The average of all the polls the final weekend came in at 11.5%, off by just under 5 points.

And no news organization or polling operation missed the overarching trend since they all predicted a victory for the Democrats.

As for the media exit poll, the one that asks voters as they leave polling places who they voted for and that proved so unreliable in the last three election cycles, there were no similar embarrassments on election night. Maybe that’s because news organizations exercised caution before using exit polls to call close races.

The verdict then on the media polls? This year, perhaps people will have to grade on a curve.