So were the preliminary polls wrong? Or did poll respondents simply not tell the truth?
In 1982, Tom Bradley had significant lead over George Deukmejian in all early polls for the California governor’s race. Bradley, the African-American Democrat and long-time mayor of Los Angeles, wound up unexpectedly losing to his white Republican opponent, though.
Why? Psychologists refer to this as a “social desirability bias” – white voters wanted to appear politically correct, many theorize, and didn’t personally reveal to pollsters they didn’t intend to vote for a black candidate. In the privacy of the voting booth, though, their votes went to Deukmejian, instead.
Thus, voters said one thing before the election, but voted a different way on Election Day.
On Election Day that year, though, Duke came in second with 32 percent of the vote, forcing a runoff against Edwin Edwards. Apparently, many voters didn’t want to admit to pollsters that they supported the openly-racist Republican. (Duke lost the final election, thank goodness.)
And that may well be an explanation to Donald Trump’s win on November 8. The Republican candidate was openly xenophobic – openly racist – openly sexist. He refused to reveal tax records. He filed multiple bankruptcies. And he’ll appear in court in three weeks for claims of fraud against his so-called Trump “University.” Who would want to openly admit they support such a discriminatory, scam-ridden failure?
But even though he lost the popular vote (so far in ballot counts still being conducted), Trump pulled ahead in the electoral count. And apparently because those who were too ashamed to publicly state support for him told pollsters one thing, but voted differently in the privacy of the voting booth.
While it was certainly not the only factor leading to Trump’s victory, this “Bradley Effect” may have had a prominent role.