If Harvard picked the president, half the men who have occupied the Oval Office in the past 128 years never would have made it to the White House.

According to records of Harvard straw polls over The Crimson’s 139-year history, the winner of the poll at Harvard—this year, Barack Obama by a wide margin—has won the real thing only 50 percent of the time.

The Crimson records date back to 1884, when James G. Blaine beat out Grover Cleveland by just 20 votes among Harvard men. Presidential polls at Harvard have been speculatively traced all the way to 1788, the very first presidential election in which George Washington essentially ran unopposed.

Over that time, the political views of the student body have shifted along with—and sometimes against—the tides of history. Harvard picked Republicans—the party of Calvin Coolidge and Theodore Roosevelt, Class of 1880, but also of losing contenders like Charles Evan Hughes—in every contest from 1884 to World War II, save 1912. In that year, as the Republican Party split its vote between incumbent President William Howard Taft and Roosevelt, who decided four years after his term ended that he had not had enough of the presidency, Democrat Woodrow Wilson managed to take the most Crimson votes. The same phenomenon occurred on a national level.

Even the Harvard pedigree of Franklin D. Roosevelt, Class of 1904 and a president of The Harvard Crimson, was insufficient to take the top of The Crimson’s poll. As the rest of the country voted overwhelmingly for Roosevelt, the student body picked Herbert Hoover—the man hated nationwide for supposedly bringing on the scourge of the Great Depression—then Alfred M. Landon and Wendell Willkie, all by significant margins. Clearly, economic woes dogged the Harvard voter far less than the ordinary election-goer.

At Harvard, the famously incorrect front-page headline “Dewey Defeats Truman” would indeed have been accurate, as twice as many Crimson voters picked Thomas E. Dewey as the real victor Harry S. Truman.

In 1952, the Harvard Young Republican Club charged The Crimson with “an irresponsible display of partisanship,” saying the votes had been manipulated through “insidious statistical maneuvering” when competing tallies emerged for the picks of the Law School and Adams House.

In 1960, a Democrat won the poll by an absolute majority for the first time: Harvard alumnus and youth favorite John F. Kennedy ’40. By 1972, Harvard had grown to be such a Democratic stronghold that Richard M. Nixon, who won 49 states and more than 60 percent of the popular vote in the national election, carried only 8.7 percent of Harvard students’ votes.

Harvard voters picked a third party candidate in a 1980 poll, with 40 percent choosing independent John B. Anderson, who won 6.6 percent of the nationwide popular vote. The Crimson’s editorial board veered still further from the mainstream, endorsing Barry Commoner, who took only .27 percent of the national vote.

Political preferences were not always consistent across the College. In 1996, for example, while Kristen Gore ’99 was living in the House that her father Al Gore ’69 inhabited long before he ran for vice president that year, 73.9 percent of her Dunster housemates chose Bill Clinton. Cabot was the least Democratic House at the time, with 57.7 percent of voters preferring Clinton over Bob Dole.

That poll drew on information from 430 respondents; since then, the polls that picked Democrats Gore, Kerry, and Obama were based on the participation of 351 to 471 students. This year’s landslide victory for Obama is based on the preferences of more than 1,500 students who indicated their choices in The Crimson’s online survey.

But in years of high poll participation or low, of conformity to national opinion or contrarianism, The Crimson has consistently endorsed voting, for one candidate or another. As a 1924 editorialist wrote, “If one wishes to criticize the government during the next four years (and who would give up that priceless privilege?) voting is the prerequisite.”

—Nicole M. Iacopetti contributed to the reporting of this story.