And Sometimes Upsets Don't Happen

The party on the outside wanted back in after more than a decade of frustration. The party occupying the White House was the underdog, trailing badly in presidential polls. "I don't believe the polls," the incumbent president declared a few days before the election. "The people ask me where I get my figures. I get 'em from the best possible sources, from the people themselves."

On Election Day, pundits and polls had the challenger well ahead. "The nation goes to the polls again tomorrow," The Sun reported, "and unless usually reliable signs fail, the voters will elect the first Republican president in 16 years."


George Bush will tell you that of course it didn't happen that way. In an upset nearly unrivaled in presidential politics, Harry S Truman returned to the White House, beating Thomas E. Dewey by an astonishing 114 electoral votes. (The Sun, which had endorsed the Republican on its front page, was so apoplectic that it seemed hardly able to publish for a few days. Then it blamed it all on the unions.)

Could it happen to Mr. Bush? Could he pull a Truman, confounding pundits and polls again?


There are some parallels, similarities, ironies. In '92, an independent candidate, Ross Perot, may well affect the outcome, just as Henry Wallace and Strom Thurmond siphoned off more than 2 million votes between them in 1948. That year, California's Gov. Earl Warren, Dewey's running mate, complained in a speech at Baltimore's Lyric Theater about the "paternalism" of the Truman administration, and Truman called the Republicans "gluttons of privilege." There was even a mission to Moscow. Late in the campaign, Truman promised to visit the "Red" capital in an effort to smooth relations.

But the differences between 1948 and 1992 overwhelm the similarities. The incumbent had run as vice president in 1944 and had been thrust into the presidency upon the death of Franklin Roosevelt. The post-World War II economy was still booming.

And presidential campaigns hadn't become television spectacles. The candidates didn't start campaigning until mid-September. Truman traveled 23,000 miles, all by train, and made 271 speeches, most of them dutifully reported in lengthy ** newspaper articles and by radio networks. Dewey made only three campaign tours.

What a candidate said, the content of his views, was what counted, and the news was transmitted over telephone lines, not satellite dishes. Newspapers reprinted campaign speeches that would be considered banal today. How a candidate sounded was important, too; Truman and Dewey both used the four national radio networks in the weeks before the election. How a candidate looked was least important. Today, of course, the order is reversed.

Television was a nascent industry in 1948. By the late '60s it had become the dominant force in presidential politics, and by the '80s it had turned the quadrennial presidential campaigns into wondrous circuses replete with hour-by-hour polls, face-to-face debates followed by instant analysis of "who won" and what Johns Hopkins media critic Mark Crispin Miller calls the "suffocating barrage of information" about the candidates, their wives and children, their astrological signs and whatever else one might want to know. Sound bites, spin doctors and image-makers had consumed the political process.

In 1948, there was little interest in "character issues." If Dewey or Truman cheated on their wives or escaped military service in World War I, the public was not to know. The candidates didn't attack each other on the personal level. In fact, Warren angered his own party by praising some of Truman's policies.

And polling was quite obviously an inexact science 44 years ago. Newspaper stories that mentioned polls almost always coupled the references with those "usually reliable signs" or "reliable sources." Polling allegedly is a more exact science today. The pollsters even tell us how wrong they could be; it's called the "margin of error."

So the hugely favored Tom Dewey, who had run against Washington and its "brazen people who put politics above the public welfare," fell to ignominious defeat. Could it happen to Bill Clinton? Unless something happens in the next 10 days to change public opinion -- something that will be spotted instantly by the polls and monitored to death, so that we won't have to be vexed by not knowing the results before we vote -- it doesn't seem likely.