by Jay Johansen | Oct 28, 2016
I'm writing this in 2016. Republicans are spending a lot of time this election season speculating about bias in the polls.
Quit obsessing about it!
In the last few days I've seen poll results ranging from the Republican ahead by 2 to the Democrat ahead by 12. Most of these polls claim to have a margin of error of 3 or 4%. Clearly there's a problem. As the polls contradict each other, logically at least some of them are just wrong.
Maybe they're wrong because the media and the pollsters are biased. Maybe they're wrong because of unintentional flaws in their methods. But either way, what difference does it make? What are you going to do differently because of these polls?
People often say, "The only poll that matters is the one on election day." Suppose you carefully studied the pollsters methods and were absolutely convinced that a candidate you hate was leading. Would that cause you to vote for him? Maybe there are people out there who will vote for a candidate just because they're ahead. That makes no sense to me, it would be a particularly silly form of succumbing to peer pressure. (I saw a T-shirt once that said, "Resist peer pressure. It's what all the cool kids are doing.")
It's been suggested that polls showing that your candidate is losing "suppress the vote", causing people to not bother to vote because they conclude it's hopeless. But by the same reasoning, polls indicating that your candidate is headed for an easy victory could lead people to not bother to vote because they conclude it doesn't matter: their candidate is going to win anyway, why go to the trouble of getting off the sofa?
I don't know if anyone ever done a study on any of these theories. (Of course, if a poll found that biased polls don't hurt anything, would we believe them, or would we suspect they were biased?) Maybe it's true that biased polls distort the vote. But even if so, for conservatives to complain to each other about biased polls isn't going to do a thing to solve the problem. You'd have to convince the pollsters to be honest. And if someone is willing to lie and cheat, it's unlikely that saying "I suspect you may be lying and cheating, please stop" is going to make him stop. He's just going to deny it. Unless you have actual proof that he's lying and cheating and can publicly expose him, you're not going to accomplish anything.
Sure, for the candidates' campaign committees, detailed polls can be useful. For example, if the polls show that you are doing poorly among black women, and you think that your policies would be good for black women, it might make sense to work harder to reach this group.
When I make a campaign contribution, I'm no billionaire, I don't have the money to give to every candidate I like. So I figure: if a candidate is way ahead, no point giving to him, he's going to win anyway. If a candidate is way behind, no point giving to him either, my modest contribution is not going to turn the tide. But if a candidate is in a very close race, if he's ahead by just 1 or 2 points or behind by just 1 or 2 points, maybe my contribution will actually make a difference. It he can take out just one more newspaper ad or put up another dozen yard signs, that might tip the balance. So I'll look at polls and give to candidates in close races.
If a third party candidate was doing very well in the polls, that could nullify the classic argument against voting third party: "He's not going to win. By voting for that third party candidate, you're just throwing your vote away." If, say, the Libertarian was just a few points behind the Democrat and the Republican, then it would make sense for someone with libertarian leanings to vote for the candidate who is his first choice rather than the candidate who is his second choice but more likely to win.
Besides those special cases, the polls don't matter. Or even if some of these "suppress the vote" theories are true, speculating about whether polls are biased or insisting that any poll that shows your candidate behind just must be biased isn't going to help anything. Quit obsessing about them. Don't waste your time studying polls. Just get out there and work for your candidate.
© 2016 by Jay Johansen
No comments yet.