Editor's note: Mark Mellman is a Democratic pollster and strategist who has guided hundreds of campaigns, from president to city council. Mellman is a member of the American Association of Political Consultants Hall of Fame and has taught at Yale University. The opinions expressed in this op-ed are Mellman's own. Find more opinion on CNN.
CNN —
We are in the midst of a historic first: As the trial of Donald Trump nears its final stages, a jury is expected to decide as early as next week whether the former president committed a crime as a result of his alleged attempts to conceal hush money paid to Stormy Daniels.
Some analysts point to surveys such as the ABC News/Ipsos poll that found a fifth of Trump's supporters would reconsider or withdraw their support if he were convicted of a felony, and argue that while a conviction is by no means a certainty, it could cost Trump the White House in November's presidential election.
Melman Group
Mark Melman
Despite our desire to believe, there is every reason to be skeptical: Trump will likely lose the presidential election again, but a conviction in this case is unlikely to have a significant impact on that outcome.
Pollsters can ask voters questions, and voters will generally answer accordingly, but sometimes the answers will be a meaningless jumble. Part of the science and art of polling is understanding what questions are more or less likely to lead to meaningful insights.
Two types of questions tend to fall into a vague category: questions that directly ask people about changing their views, and questions that ask individuals to transport themselves into an alternate reality and predict what their reaction would be in that new situation.
Asking people how their vote would change if Trump were convicted falls into both categories.
The highly partisan and totally unjust impeachment of then-President Bill Clinton in the late 1990s illustrates how research and the human brain work in cases like this.
A Washington Post/ABC News poll conducted in January 1998 found that 55% of voters thought Clinton should be impeached if he had lied in an affidavit about not having had an affair with Monica Lewinsky, while 40% were opposed to impeachment under such circumstances.
Months later, it was revealed that the president had lied, but only 30% favored impeachment, compared with 55% who opposed it (66% opposed).
Americans had argued that they would support impeachment if Clinton had lied about the relationship, but once it became clear that he hadn't been honest, they overwhelmingly opposed impeachment.
Similarly, when Washington Post pollsters asked whether Clinton should resign or oppose the prosecution in the Senate if the House voted in favor of impeachment, only 38% of Americans said she should oppose the prosecution, while 58% said she should resign.
Did voters maintain that view after Clinton was impeached by the House?
no.
Indeed, the numbers changed almost instantly: Within days of being impeached, the percentage of Americans who thought Clinton should resign rather than fight the prosecution had fallen to 42%.
The inherent difficulties with these polling questions also emerged after the 2020 election. At the time, a majority of Republicans believed the illusion that Trump had actually won the election. Many of them told my pollster they would change their views if the courts ruled against Trump. But despite their earlier protestations, their thinking remained largely unchanged even after Trump lost some 60 lawsuits challenging the integrity of the election.
Similarly, a March 2016 NBC/Wall Street Journal poll found that half of women who voted in the Republican primary said they couldn't imagine themselves voting for Trump in the general election. What was unimaginable in March, nearly 90% did in November.
Faith is enduring and stubborn.
People think they are open to new information, and sometimes they are, but in my 40 years studying public opinion, I've learned that most of the time, we hold fast to our preconceived ideas about important social and political issues.
Subscribe to our free weekly newsletter
Additionally, people are not very good at predicting their own reactions to altered realities: humans have transported themselves into parallel universes and have a hard time correctly predicting their own reactions in that alternate reality.
Could convicting Trump change the outcome of the vote? Of course.
The charges against Trump are more serious than those against Clinton, and no former president has ever been convicted of a crime.
Of course, some presidents have been seriously damaged by scandals: Richard Nixon's approval rating fell from 67% to 31% between January and August 1973, when the Watergate scandal came to public attention, and when he left office a year later, only 24% approved of his performance.
A similar fate is possible if Trump is convicted — though history suggests that's unlikely — but the point is that today's poll responses tell us almost nothing about how the public will ultimately respond in November.
We can only wait and see what happens after the jury delivers its verdict.
