Report suggests changes in exit poll methodology
WASHINGTON (CNN) -- Exit polls overstated John Kerry's share of the vote on November 2, both nationally and in many states, because more Kerry supporters participated in the survey than Bush voters, according to an internal review of the exit-polling process released Wednesday.
The report said it is difficult to pinpoint precisely why, in general, Kerry voters were more likely to participate in the exit poll than were Bush voters. "There were certainly motivational factors that are impossible to quantify," the report said.
Problems with the numbers first surfaced on Election Day, when exit polls showed Kerry with a 3-point lead nationally and an edge in some key battleground states. Those exit poll results were leaked and became widely known through the Internet.
CNN did not air those inaccurate results or post them on its Web site, and CNN's projections of winners on election night were accurate.
Nationwide, Bush got about 3.5 million more votes than Kerry.
The discrepancies stemmed from problems in interviewing voters at the 1,480 randomly chosen precincts where exit pollsters were stationed, not from how those precincts were selected or the way the data were processed, according to the 75-page report.
The report recommends a number of steps to deal with the problem, including better training for interviewers, as well as continued research aimed at boosting participation in the polls.
The report was issued by Mitofsky International and Edison Media Research, the polling firms that conducted the polls on behalf of the so-called National Election Pool, a consortium of six national media organizations (AP, ABC, CBS, CNN, Fox and NBC).
To prevent leaks in future elections, the news organizations have agreed not to access the data until 6 p.m. ET.
The report found that the exit polls offered no evidence of widespread fraud.
"Exit polls do not support the allegations of fraud due to rigging of voting equipment. Our analysis of the difference between the vote count and the exit poll at each polling location in our sample has found no systematic differences for precincts using touch screen and optical scan voting equipment," the report found.
The new report shows that exit polls overstated Kerry's support in 26 states, while estimates overstated Bush's support in four states. The problem is not new -- in every presidential election since 1988, exit polls have overstated support for Democrats nationally -- but the discrepancy in 2004 was more pronounced than in previous years.
The report identified several factors that may have contributed to the discrepancy, including:
- Distance restrictions from polling places imposed upon the interviewers by election officials at the state and local level.
- Weather conditions, which lowered completion rates at certain polling locations.
- Multiple precincts voting at the same location as the precinct in the exit poll sample.
- Interviewer characteristics, such as age, which were more often related to the errors last year than in past elections.
The pollsters said they plan to further investigate the recruiting and training procedures, the interviewing rate calculations, the length and design of the questionnaire, as well as characteristics of both the interviewers and the precincts chosen to be surveyed.
"Even with these improvements, differences in response rates between Democratic and Republican voters may still occur in future elections," the report reads. "However, we believe that these steps will help to minimize the discrepancies."
In addition to the information included in this report, exit poll data from this election are being archived at the Roper Center at the University of Connecticut and at the Institute for Social Research at the University of Michigan and will be available there for review and further analysis. A description of the methodology of the exit polls is posted at www.exit-poll.net.
From 1990 to 2002, exit polls were conducted by Voter News Service (VNS), whose exit polls in 2000 led to the networks' decisions to declare Al Gore the winner in Florida. In 2002, VNS was unable to deliver any exit poll data to the networks, resulting in the decision to disband it.
Read more. . . (off site)