Louis Harris: the pollster who was right and why
In the shadow of the presidential sweepstakes hung another contest of no little interest: Which pollster would most accurately forecast the voters' choice?
The winner: Louis Harris.
"My neck was out by a mile," said the New York-based pollster in a celebration party in his office election night. "I was the only pollster to say that Ronald Reagan would win."
Other national pollsters, including Gallup, roper, and CBS/New York Times, reported only 0-3 percentage points separated Mr. Reagan and President Carter in the final days. But because their margins of error also were one to three points, the race was considered "too close to call." The Associated Press/NBC poll did not issue a last-minute prediction.
But Loui Harris, working late into the night before the election day, went on ABC News on the morning of Nov. 4 to boldly predict a REagan win by at least five points, with a possibility of a 9 percent lead over Carter. His margin of error was 1.5 percent. The final spread was roughly 10 percent. After a "lonely night" as the sole pollster with such a prediction, he says it was a sweet victory for the firm which had lost some credibility for miscalling the New York primary last spring.
The bigger-than-others Reagan win prediction was based on last-minute telephone surveys of 640 screened voters in the 14 big states with large electoral votes.
Still, the Harris poll predicted that Carter would win 140 electoral votes from Georgia, West Virginia, Minnesota, Wisconsin, Hawaii, Maine, Rhode Island, South Caroline, Tennessee, Arkansas, and Louisiana. Of these however, the Presidend failed to win Wisconsin, Maine, South Carolina, Tennessee, Arkansas, and Louisiana. States too close to call, Mr. Harris says, were Kentucky, Illinois, Massachusettes, New york, Oregon, Mississippi, and Alabama.
Why did Harris win the pollster contest? "We went to a larger sample. We were struck by how many people were not taken by either candidate. Anything could have happened. The 1976 race was close but not scary for pollsters," says David Neft, executive vice-president at the Harris firm.
And unlike other pollsters who saturated calls into selected "clusters" of voters around the nation, the harris poll was designed to select one voter in many areas and to keep phoning until voters were reached. The firm, located in Manhattan's Rockefeller Center, relies on many jobless actors and actresses as callers.
Failing to contact voters at home, says Harris, means a survey is skewed to views of people who tend to stay at home -- the relatively immobile and nonaffluent who tend to be more Democratic. By ensuring contact with the affluent nad mobile voter, the Harris poll was able to detect the conservative sweep that helped bring in Ronald Reagan.
Polling was done every day to keep up with a volatile voter shifts possibly reacting to the debate, the Iran hostage situation, and the John Anderson candidacy. (The poll showed Anderson the winner if people were asked to assume he could win.)
"I view the Reagan win as a protest against Carter who appeared not to get anything done and a vote against the federal government. Right-wing concervatives would be mistaken to see it as a mandate for Reagan's social policies," says Harris. The televised debate solidified the Reagan vote. "In hindsight, we could have wrapped up this election last Thursday," says Harris's vice-president Neft.