Skip navigation.
 
 SEARCH
NCPP.ORG
THE WEB
Welcome 

Presidential Poll Performance 2000

A statement by the National Council on Public Polls' Polling Review Board

January 3, 2001

PRESIDENTIAL POLL PERFORMANCE 2000

The accuracy of the election projections based on the pre-election polls of 2000 was surpassed only by the polls of 1976 and 1960, according to a study release today by the National Council on Public Polls. This year's final polls had an average error of 1.1 percentage points on the estimates for George W. Bush and Al Gore. The error on the third place finisher, Ralph Nader, was 1.3 percentage points.

These results were based on the work of 10 polling organizations that used traditional methods for conducting their polls.

Poll Bush/Gore Error Nader Error
Harris Poll 0.0% 2.0%
CBS News 0.5% 1.0%
IBD/CSM/Tipp 1.0% 1.0%
ICR/Politics Now 1.0% 4.0%
Gallup/CNN/USA Today 1.0% 1.0%
Pew Research 1.0% 1.0%
Zogby/Reuters 1.0% 2.0%
ABC News/Wash Post 1.5% 0.0%
NBC News/WSJ 1.5% 0.0%
Battleground 2.5% 1.0%
Average 1.1% 1.3%

 

The 2000 election was a tie between Bush and Gore and was the closest election since the Kennedy-Nixon election in 1960. Democrat Gore had a slight edge in the CBS and Zogby polls, while seven of the other polls leaned to Republican Bush. The Harris poll had it tied. Four years ago, all 9 polls erred in favor of overstating Democratic Clinton. Challenger Nader was overstated by 7 of the 10 polls this year. Two got the Nader vote correct. All other polls overstated Nader's vote. Third party candidates typically get less support in the election than they do in the final pre-election polls.

Two other organizations used methods that previously had not been used. Harris Interactive conducted its polls on the Internet among a panel of e-mail users and forecast a tie. Rasmussen's Portrait of America poll was off by 4.5 percentage points on each of the top two candidates. Rasmussen had its interviews conducted by a computer playing a recorded voice with no live interviewer intervening.

Alternative Method Polls Bush/Gore Error Nader Error
Harris Interactive 0.0% 1.0%
Rasmussen 4.5%  

 

The ten traditional polls used random samples of telephone households and live interviewers to obtain vote intentions from likely voters. Screening questions that were unique to each poll identified likely voters. More detailed methods statements are available from the individual polling organization.

 

2000-Preliminary Gore Bush Nader Undecided Other
Election Results 48% 48% 3%   1%
Zogby 48% 46% 5% 0% 1%
CBS 45% 44% 4% 5% 2%
Harris (Phone) 47% 47% 5% 0% 1%
Gallup/CNN/USAToday 46% 48% 4% 0% 2%
Pew Research 47% 49% 4% 0% 0%
IBD/CSM/TIPP 46% 48% 4% 0% 2%
ICR/Politics Now 44% 46% 7% 1% 2%
NBC/WSJ 44% 47% 3% 4% 2%
ABC/WashPost 45% 48% 3% 3% 1%
Battleground 45% 50% 4% 0% 1%
           
Alternative Methods          
Harris Interactive 47% 47% 4% 0% 2%
Rasmussen 49% 40% 4%    

 

For this election, 2 of the 10 polls overstated Gore's vote while 7 overstated Bush. In the 1996 election, 8 out of 9 polls overstated Democrat Clinton. One poll each year neither under- or over- stated the winners' percentage.

Printer Friendly Version


Presidential Poll Performance 2000 Error Calculator