NEW YORK - A growth in early voting and tough economy for the media are forcing changes to the exit poll system that television networks and The Associated Press depend upon to deliver the story on Election Night, all with the pressure-filled backdrop of a tight presidential race.
The consortium formed by ABC, CBS, CNN, Fox News Channel, NBC and the AP is cutting back this year on in-person exit polls while upping the amount of telephone polling. This is to take into account more people voting before Nov. 6 and households that have abandoned land lines in favor of cell phones.
"It makes it trickier," said Joe Lenski, executive vice president of Edison Research, the company that oversees the election operation for the news organizations. "It means there are a lot of different pieces to keep track of."
On a perfect Election Night, Americans who are tracking results won't notice all the work being done behind the scenes. The Associated Press reports actual vote counts nationwide and news organizations use those numbers, plus the exit polls, results from precinct samples in some states and telephone polls of absentee voters to do their own race calls.
But things haven't always gone perfectly. The news organizations completely rebuilt their exit poll system after the 2000 embarrassment, when TV networks mistakenly called the race for George W. Bush when it wasn't decided until a month later (the AP mistakenly called Florida for Al Gore, retracted it but, unlike the networks, never called the overall race for George W. Bush). In 2004, early exit poll results overestimated the strength of Democrat John Kerry.
To save money this year, the consortium is doing bare bones exit polling in 19 states. Enough voters will be questioned in those states to help predict the outcome of races, but not enough to draw narrative conclusions about the vote — what issues mattered most to women voting for Mitt Romney, for instance, or how many Catholics voted for Barack Obama.
The affected states are: Alaska, Arkansas, Delaware, Georgia, Hawaii, Idaho, Kentucky, Louisiana, Nebraska, North Dakota, Oklahoma, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, West Virginia and Wyoming, along with the District of Columbia.
Each is considered a non-battleground state with polls showing a strong advantage for one of the presidential candidates. Some non-battleground states will get the full exit poll for other reasons, like Massachusetts and its hotly contested U.S. Senate race between Scott Brown and Elizabeth Warren.
"What we are doing is taking our resources and using them where the stories are," said Sheldon Gawiser, NBC's elections director and head of the steering committee for the AP-network consortium.
Spending figures were not made available. News organizations have had a tough few years financially, but the consortium noted that it is interviewing a total of 25,000 voters this year, up from 18,000 in 2008.
Because of early voting, there are no traditional exit polls in Oregon, Washington and Colorado. A phone poll is done prior to Election Day in those states, taking in a mixture of people who have and haven't voted. Others states have a mixture of telephone polling and exit interviews. California, North Carolina and Arizona are among the states where the percentage of telephone polls has grown because of more people voting early.
More people are interviewed on cell phones because it is the primary way to contact them. The consortium said cell phone interviews are twice as expensive as those on land lines because of manpower costs, in large part because it is harder to reach people and federal law requires the phone numbers to be manually dialed instead of done by computers.
In addition to the exit poll changes, the news organizations are taking steps to improve their ability to include actual vote counts in their decisions on when to call particular states as a winner for either candidate. This usually involves collecting sample precincts that reflect a state's demographics.
Even this is complicated by local customs. Some states report precinct results more quickly than others. New Mexico, for example, sets up polling places where anybody from a particular county can cast a ballot; while this makes voting easier, it makes projections based on precinct samples more difficult.
Television viewers may notice that networks are being slower than in the past to project winners in certain states, but the consortium believes people won't see a difference.
If the actual election is as close as the pre-election polls are suggesting, it will be a long night, anyway.
With all the factors increasing the difficulties and costs associated with exit polling, it's worth wondering whether a time will come that the news organizations abandon them in favor of the pre-election polling. The experts say that time is nowhere near.
"One of the great advantage of exit polls is you don't have to worry about who voted. You don't have all of these `likely voter' issues that you have now," said Lee Miringoff, a pollster at Marist College.
Gawiser noted how the minds of voters can change, even up until the last possible minute.
"It's a story we want to be able to tell on Election Night and we want to be able to tell it accurately and rapidly," he said. "I really don't think it's much different than any other story we tell."