POLLS: Wrong Again

This time the pollsters were rich in experience. They had pondered & pondered their failure to find enough Democrats in 1948; they were not making that mistake again. So when their figures repeatedly showed Dwight Eisenhower running in front, the more experienced pollsters went into learned loops to explain why such figures were not to be trusted. Almost all of them stressed the "undecided" vote. George Gallup's final poll showed:

Eisenhower 47%

Stevenson 40%

Undecided 13%

Once upon a time, Gallup would have ignored those undecideds; if he had done that this year, he would have come within 1% of Eisenhower's actual margin. But Gallup "allocated" the undecideds 2-to-1 and 3-to-1 for the Democrats. That kind of pattern this time, he told his readers, would take Eisenhower and Stevenson into 50-50 country. Some of Gallup's other calculations brought Stevenson out ahead. Pollster Elmo Roper dwelt on "basic conflicts" in voters' minds, refused to indicate how the conflicts would be decided.

Lacking the learned background of Gallup, Roper et at the New York Daily News wasted no time on the no-opinion crowd, flatly gave Eisenhower 52.1% in New York State (his indicated lead there: about 52.5%) and predicted that he would carry the state.

In many ways the best poll was one that did not approach the man in the street: Columnist David Lawrence polled the editors of daily newspapers in every state. Their verdict:, Eisenhower to win with 357 electoral votes.

But the "scientific" pollsters, who were so famously wrong in 1948, were even more wrong (in a different way) in 1952. This year they were right and did not have the courage to believe themselves.