Even people who aren’t political junkies are aware of polls. They have also for a couple of decades now been wary of what they hear and read about polls thanks to the mess in Florida during the 2000 election. Even since reputable polling organizations have worked to address changes in the landscape and improve methodology for getting a good representative sample on which to base their conclusions. In addition to national, well known polling outlets political Parties and even many candidates have their own “internal” polling operations. It seems for the GOP they have good reason to doubt what their own internal polling is telling them.
Newsweek is out with a story describing the situation. Many a candidate over the years has publicly dismissed “polls” because they said their own numbers reflected something different. Usually that were slinging bullsh*t or whistling past the graveyard. I can recall plenty of election night reports where correspondents report that a losing candidate pretty much knew what was coming, that despite the brave front they had put forward their own “internals” matched independent polling. However, in some cases there have been surprises, if not in who (or what ballot measure) won/lost but by how much. In a way, especially when you get down to Congressional district level and below (to state races) it makes sense. A good, professional pollster down in the weeds in theory should be able to both get a better sample of respondents. AND be able to trust they are getting truthful answers to their questions.
Now, all that seems to be coming into question. At least for Republicans. Democrat Marylin Lands’ surprisingly lopsided victory in a statehouse race in freaking ALABAMA is just the latest example that, as Newsweek puts it “Republicans have an internal polling problem:”
Republican Teddy Powell lost his bid for a seat in the Alabama House to Democratic candidate Marilyn Lands on Tuesday despite internal polling conducted in December which suggested he was on course for a comfortable victory.
Lands flipped the previously Republican district with 62.4 percent of the vote against 37.6 percent for Powell, a Madison City Council member. However, an internal poll conducted by Powell’s campaign in December put the Republican on 47.7 percent of the vote against 36.8 percent for Lands.
Prior to yesterday’s special election both campaigns had cited polling that showed they would prevail. Clearly someone was either lying or was wrong. Normally given how I feel about GOPers in general I’d assume Powell and his people were lying but now I think they were genuinely surprised. I think they actually did have internal polling that showed them winning. And therein lies the tale, or rather problem the GOP faces.
In retrospect perhaps yesterday’s unpleasant (for the GOP) surprise shouldn’t have been all that surprising. Ever since Trump was elected Democrats have usually won special elections, even in red states. Republicans expected to lose seats in the 2018 midterms because historically that’s what usually happens. Instead they got wiped out. If not for the extreme gerrymandering they’d done around the country after 2010 based on history Democrats would have wound up with an eighty or more seat majority – double what they actually got.
GOPers were stunned of course but viewed it as a “one-off.” Many were confident they’d win the WH and both chambers of Congress in 2020. They lost all three. Then of course we spent 2022 hearing all about the “Red Wave” that would banish Democrats to the wilderness. It was barely a trickle. And the GOP just can’t understand what’s been going on. Now, it seems we are learning what the problem is.
I can’t help but think the sees of this were planted back in 2012 when Mitt Romney and the old-guard GOP was shaken. His “Orca” operation had assured everyone on that side that Obama was vulnerable. Not so much as it turned out. And then came 2016 and the outlines of the phenomena began to reveal themselves – plenty of people out there, and in numbers large enough to affect polling results were willing, if not eager to LIE to pollsters! So many of us were stunned the day after the 2016 election. Not that it ever should have been close in the first place but the pollsters, all the big name ones got it wrong.
Yet one outfit I’d never heard of (I doubt many of you had either) named Trafalgar was one of the few who got it right:
The Trafalgar Group established a reputation after being one of the few pollsters to correctly predict Donald Trump would defeat Hillary Clinton in 2016 by heavily weighing “shy” supporters of the then-GOP candidate
It would seem since then they’ve become something of a go-to group for either polling or personnel by Republicans. Or their model of weighting for the folks who lie to pollsters has become an article of faith. It seems that despite so many instances of what happens when people actually cast votes proving them wrong, Republicans and their internal pollsters aren’t learning the lesson. Now, do I believe there are lots of American conservatives out there that will lie through their teeth to a pollster? You bet. Often it’s to “own the libs.” What maybe isn’t being accounted for is that perhaps, when talking to a Republican pollster they lie for a different reason. They don’t want to admit they no longer worship Trump, or agree with the GOP’s stance on issues like say abortion! Let’s face it, if some GOP polling outfit has their name many realize any apostacy might well haunt them for years to come.
It’s as good an explanation as I can think of for why the GOP keeps thinking they are going to win a race or a referendum “bigly” only to be left the next day wondering: What the hell happened? Out internals said we had this! Now that I think about it rather than admit independent polling firms that were right and that they’re out of touch with the real feelings of their votes their minds jump to “cheating” or “election fraud.” Then they say it out loud and their supporters think that HAS to be what happened!
Instead of taking a hard look in the mirror, and deciding to flush old assumptions and take a hard look at their advisors including pollsters they keep hoping against pretty clear evidence next time will be different. Still, the actual evidence in the form of electoral results that Trafalgar and/or their model of weighting for voters who don’t want “outsiders” (non GOPers) to know what they think or how they’ll vote seems clear:
The Republicans underperformed during the 2022 midterm elections, only narrowly taking control of the House of Representatives and actually losing a seat in the Senate to increase Democratic control.
According to a Newsweek analysis, polling conducted by The Trafalgar Group, a conservative-leaning pollster, correctly predicted the winner within the poll’s final margin of error in just five of the two dozen polls they conducted.
I don’t know about you but if I were a candidate or Party organization whether county, state or national that’s not a track record I’d have much faith in. I would be looking and hard at making a change. (Here’s an ugly thought but since we are talking about conservatives plausible – Trafalgar and their acolytes might know they are blowing smoke but they are getting rich telling GOPers what they want to hear) I’m sure for the MAGAs, the hard core types nothing gets through to them. However, there are plenty of hard core conservatives with at least one foot in actual “Reality Land.”
They know two things. The first is that they can’t be sure GOP internal polling is worth anything at all and that the truth is closer to what reputable independent firms are saying. Which is mostly bad news for the GOP. The second is that it’s likely the GOP is in for a major crash in November AND that MAGA still won’t get the memo and agree it’s time to adjust. As a result some are getting out, and now we see some key ones leaving NOW instead of waiting until the end of their current terms! Don’t be surprised to see quite a few more old timer GOPers announce they are done after November. That’s a given. The interesting question is whether they walk right away or serve out their terms.
I’m guessing there’s a lot of anguished soul searching in hard-core but sane conservatives these days. The linked article cites a NYT analysis that said:
An analysis by The New York Times concluded that Republicans in 2022 were misled by “right-leaning pollsters using opaque methodology, in some cases relying on financial support from hyperpartisan groups and benefiting from vociferous cheerleading by Mr. Trump.”
For older GOPers packing it in makes a lot of sense these days when they see reputable polling on what issues are likely to matter most this fall and that overall the lay of the land favors Democrats. Some of the younger ones might feel the same way. If they can “get out clean” they can always wait in the wings until Trump/MAGA makes what’s left of the GOP implode and the conservative movement finally accepts it’s time to start over from scratch.
If they weren’t conservatives I’d actually feel sorry for some of these people. Still, I can’t help but be amused that for a growing number of them all the way down to statehouse legislature level some are realizing they can’t trust their own freaking internal polling! And worse, the “lamestream” (as they see it) stuff looks very, very bad for them. If paranoia increases the GOP infighting and brings about the final collapse quicker then maybe this will be the catalyst.
Another aspect of GOP polling as I see it, is they slant their questions. I was contacted by an Austin phone number that was doing a poll. Since I had written letters to my reps (all repubs) I was pretty sure it was a GOP poll. I thought I could get my views through, but the questions were worded in such a way that I agreed to things that are their talking points, without any other options. So now I don’t answer any numbers I don’t recognize.
Interesting that you mentioned slanted questions, etc., and numbers you don’t recognize. When I wrote my comment below I had not yet read yours, but I said more or less the same thing in a different way. I hope you read it.
Yes I did just now read it and it is well written and well reasoned. My main point is when the republicans are conducting polls in this manor, no wonder they feel misled.
I don’t think lying to pollsters is the only problem. I think there is a bias introduced because very few people actually participate in the polls. By that I don’t refer to the small number of people polled, but to the number who simply refuse to participate. That introduces the possibility of bias if it turns out that Republican voters are more eager to be polled than Democratic voters. There are many people who simply do not want to pick up the phone for any unknown caller, and far easier to know the phone number of the caller. Even if you don’t have caller ID, it is easy to tell whether you are being called through an automatic dialing program, and hang up immediately. I do that all the time. Consequently I never get polled. The main reason I don’t like to be polled, by the way, is that they usually formulate questions in such a way that it’s impossible to answer accurately when all they will take iis a yes or no answer. Another obvious reason is that people don’t like to give out personal information, eve if it’s just about their preferences. They know who you are and you don’t really know who they are. Maybe the pollers are limited to land lines, which would also introduce some kind of bias. I don’t claim to have the statistics on this, obviously you’d have to use a different methodology than telephone calls to test the hypothesis that Trump voters may be more willing to participate in telephone polls than Biden voters.
several years back there was an issue with a pollster for John Boehner’s number 2 guy (don’t remember his name) who said he would easily win his primary and provided I believe it was three polls over a few months showing that to be the case. no one noticed that the numbers were the same as polls they had done for him in previous years. he got crushed. this polling firm did a lot of work for Republicans and I don’t think they ever got dropped. obviously, not one of the reputable ones. they apparently did this a lot, just digging out old polls and not bothering to do the work.