We all know that polls aren’t a 100% perfect predictor of an election. But it’s occurred to me more than once that whenever the polls are “wrong,” it always seems to be the Republican candidate that wins. It’s never the Democratic candidate who got more votes than expected.
I can think of several possible reasons, up to and including confirmation bias, but I thought I’d open up the topic to get other ideas. Am I right? If so, why?
It is quite rare indeed for a Democrat to outperform the polls, vs a Republican.
I would theorize:
If you go by what FiveThirtyEight says, this just isn’t true. They’ve noticed leans both directions before. Heck, 2020 overestimated the Republican chances in many places.
IIRC polls overestimated Republican performance in some recent special elections.
One of the theories is that the personality traits that lead people to be willing to participate in surveys have, fairly recently, become more predictive of partisanship.
…people who don’t trust people and don’t trust institutions are way less likely to answer phone surveys. Unsurprising! This has always been true. It just used to not matter.
It used to be that once you control for age and race and gender and education, that people who trusted their neighbors basically voted the same as people who didn’t trust their neighbors. But then, starting in 2016, suddenly that shifted.
As a dem I activily avoid taking part in polls due to the nasty nature of politics theses days. No phone, internet or in person offers of being polled are accepted. I think its mostly pubs and magas that like to talk about their issues to pollsters.
My WAG… there’s peer pressure in the PC/progressive crowd to vote Democrat. And when asked, that’s what they will say. But in the voting booth, they begin to think about issues that affect their livelihood, like taxes (“I’m paying too much”), inflation (“It’s too high”), and jobs. Republicans seem to do a better job when it comes to addressing economic issues. (Quite frankly, many Democrats seem clueless when it comes to economics.) And then later, when asked, they will claim they voted for Democrats.
I also think the “other-way-around” doesn’t occur that often.
I realize this isn’t the Pit…and I’m not attacking you, personally. I’d just like to say that many Republican voters seem clueless when it comes to observing, historically (including quite recently), which party was in charge (presidency, at least) when major economic downturns occurred; which party was in charge during the recovery from these downturns; and which party was in charge to unfairly get credit as the recovery continued (albeit on increasingly shaky foundations).
Taxes, okay, that’s a fair point.
Michael Moore, who, whatever you think of him, was one of the only (THE only?) non-Conservative(s) to predict a trump win in advance of the 2016 election.
This time he’s predicting a Democratic Midterm landslide in oppostiion to most predictions. Unfortunately, I think he’s operating from wishful thinking this time, but I hope he’s right and I’m wrong. A key quote from the article:
“I think that there is going to be such a landslide against the traitors, especially the 147 Republicans who just, hours after the insurrection, voted to not certify the elected president of the United States, Joe Biden. I think there are going to be so many people coming out to vote. I wanna thank the Supreme Court for reminding women they are in fact second-class citizens, and taking their rights away like this . . . I think there’s gonna be such a massive turnout of women.”
I mean, this is clearly false, but whatever.
If there is outperformance (and I don’t know of that’s still true), it could be because Republicans are embarrassed about their positions in front of strangers but not among their own?
Do they view it as a bad position themselves, but hold it because of ???
I think a lot of forecasting is to estimate hew each demographic shows up and votes. The pollsters know that Republicans always show up and vote and assume that fewer Democrats will show up. So this year, I look for Democrats to exceed expectations due to the “Roetivation” factor.
.I don’t think it’s fair to say Democrats don’t understand economics. Look at inflation- I think it’s due to supply chain issues caused by the pandemic. That’s why it’s worldwide. If it was Biden’s fault, why the hell do we see it all over the world? Republicans love to use it as a sledgehammer since their gullible voters think that every nickel of government spending causes inflation. The Fed thinks that since all they have is a hammer, then inflation MUST be a nail. Raising interest rates to cause unemployment to spike is not going to solve inflation, but the Fed being the Fed that’s all they know how to do.
Really? I’m hearing a lot of Republican campaign ads that claim inflation is caused by federal spending. Apparently, elected Republicans aren’t too familiar with economics either.
This is correct.
What has Democrats gun-shy about polls is the big misses in 2016 and (to a lesser extent) 2020 in upper-Midwest and former rust-belt states. If those misses repeat, Dems will lose the Senate and have a bad night in the House.
But there have been plenty of misses in the other direction, especially this year post-Dobbs, and pollsters have been very aggressive about trying to fix their problems from 2016 and 2020. And of course 2018 was pretty much spot-on, as I remember it. So it’s possible the “errors” are largely about Trump.
The more worrying sign from polling (for me at least) is that while immediately post-Dobbs the partisan preference question swung to +1 or so for Democrats, it’s now back to +1 or so for the GOP. So whatever bounce Dobbs and a reasonably good summer for Biden gave to the Democrats seems to be gone. So if we’re back to “baseline mid-term” that’s bad news for holding the Senate, and keeping the losses low in the House is the best-case outcome there.
Polls are a snapshot of who voters say they will vote for. It’s often weeks or days before they vote, (although some may have voted early). Lots can happen between the polling and the time they vote, assuming they ever do, which skews the polling accuracy.
While we are bombarded with political ads for months before the election, many people don’t decide who to vote for until they step into the booth. People often change their minds after taking a poll or decide voting is just too much of a hassle and opt-out. Pollsters have no way of knowing that.
Pollsters sample whoever they can get to talk to them. I have voted all my life. No pollster has ever asked me who I was voting for, which is typical for many people. I live in a rural town in a low-population state. Pollsters seem to focus on metropolitan areas since that’s where most voters are.
Republicans are much more reliable than Democratic voters, presumably because they are older and many are retired. It seems that Democrats are terrible at “getting out the vote.” I have never been contacted by the Democratic Party to see if I was going to vote. I always vote absentee since I would rather sit on my couch and casually vote than have to do it after standing in line for an hour in a frigid high school gym.
No, I think Republicans view themselves as good, but are aware enough to know that most of society regards them as bad. They view themselves as a slightly-outnumbered, but righteous, minority.
Moore was boots-on-the-ground in Michigan for the Democratic primary, and he saw how much enthusiasm there was for Sanders and how little there was for Clinton. iirc once the primary was decided he was predicting a Trump win. And of course she lost that state and the election. (I think it’s reasonable to predict that if you believe she would lose MI, she would have lost WI and PA as well.)
I’m not sure how accurate he’ll be this time around. At the very least, it seems like the pendulum in MI has swung toward a reaction against the conservative SCOTUS, so he might be right there. Although he thinks it’s a reaction to the election stealing/jan 6th stuff, and I don’t think that really holds water with people other than the media class and liberals who were going to vote blue anyways.
Before we go trying to figure out the reason behind a phenomenon, we really ought to verify that the phenomenon even exists. Yes, Trump outperformed his polls, and that was the biggest election this country has, so it sticks in folks’ minds. And it’s not too hard to find other examples of Republicans outperforming polls. But there were also Democrats who outperformed their polls, and about as many of them.
Which is of course exactly what we’d expect, from competent pollsters. If there really were some effect that caused Republicans specifically to outperform their polls, and that effect were detectable and statistically significant, then pollsters would adjust their methods and calculations to compensate, until there was no longer a statistically-significant difference.
Many polls ask the general public, without checking for actual voters (or they accept people’s statement that they are registered voters (and people lie)). And the whole population leans more Democratic than just actual voters. So it’s more likely that the people get polled but don’t actually vote are Democratic voters. So that means Republicans are more likely to exceed polling numbers.
This. Pollsters, for the most part, have a better understanding of polling than most posters in this thread are giving them credit for. They aren’t helplessly making the same mistake year after year and powerless to do anything about it.
But the bias tends to shift unpredictably from election to election. Polls had a pro-Republican bias in 2012, for example. They’ve also had a pro-Republican bias in elections so far in 2017 and 2018. We strongly encourage readers to remember that polling error can occur in both directions and that it’s almost impossible to predict which direction in advance.12 There have been cases, such as in last year’s U.K. general election, in which pollsters overcompensated for past errors and introduced new ones that caused polls to miss in the opposite direction. Over the long run, U.S. election polls have had very little overall bias toward either Democrats or Republicans.
My bold.
There’s no need to to twist yourself into knots to explain something that doesn’t exist. And there’s really no need to do so by starting with a foundation of clearly false statements,
e.g.
Your wish is granted. 
(well, sort of. Not a landslide, but better Dem results than expected — in some places, much better — except in a few states like New York and Florida).