Good Friday Morning! Especially to my readers down around the Gulf Coast. Invest 98L is kicking around in the Caribbean Ocean as we head into the weekend and will likely start heading into the Gulf of Mexico sometime next week. This is likely our first true major hurricane event of the season. It’s been a tranquil season thus far.
Hurricane Fiona struck Puerto Rico, which was still recovering from past hurricanes. But all things considered, Fiona could have been far worse. Bermuda is taking the brunt of it now, but the mainland has stayed safe. There was some spectacular footage of a drone that sailed through Fiona’s category four core — it’s astounding.
I hope everyone stays safe and monitors local weather. This week, I will talk about why the polls are wrong about the midterms and what needs to happen to make them right—links to follow.
Where you can find me this week
Please subscribe, rate, and review my podcast on iTunes, Spotify, or Google Play — the reviews help listeners, and readers like you find me in the algorithms. Make sure to sign up for the Conservative Institute’s daily newsletter and become a subscriber at The Dispatch, where I’m a contributor.
[9/19/2022] Gas prices are lower, but inflation marches on – Conservative Institute
[9/23/2022] A Post-Putin Russia is likely nearing – US needs a plan – Conservative Institute
The polls are wrong.
The polls are wrong, again. We’re nearing the end of September, October is right around the corner, and I can tell you without a doubt that the polls are wrong on the midterms. It’s not some conspiratorial take; it’s just a fact of how pollsters are conducting polls right now.
To understand this, here is a quick primer on a vital aspect of political polls. There are three basic types of polls: Registered Voter (RV), Likely Voter (LV), and Adults (A). As we get closer to an election, the most accurate type of poll you can run consists of likely voters (LV).
We know, for instance, that when elections get near, a poll consisting of registered voters only will miss those registering at the last minute and includes people who won’t vote. A poll of adults only doesn’t give you an idea of who will vote. That’s why the closer you get to an election, the more pollsters typically start running likely voter polls.
The methodology of how a polling outfit weights a likely voter poll is eternally up for debate. But most elections analysts and pollsters consider likely voter polls the most accurate of the bunch. Campaigns run likely voter polls for their internal purposes. If you want accuracy, you switch to that methodology.
How do I know that the polls are wrong right now? Because pollsters are still relying on RV polls. At the time I’m writing this, the RealClearPolitics average of polls of Biden’s approval numbers includes eleven of the most recent polls. Only two are likely voter polls, while the others are a mix of RV and adults-only polls.
The two LV polls in that average are both conservative-leaning: Rasmussen and Trafalgar. Rasmussen has Biden at 45% approval, and Tralfagar has him at 39% approval. The other polls are interesting, but they’re not telling us the whole story. Biden could be improving, sinking, or plateauing. We don’t know right now.
In the generic ballot polls, it’s the same story. Everyone is running RV polls, while Rasmussen and Tralfagar are running LV polls. The two LV polls have Republicans with a lead in the generic ballot, while the RV polls have Democrats in front. The problem with RV polls is that they tend to miss when undecided voters break for one party over the other.
Every election cycle has a group of undecided voters. We always see in the campaign’s closing weeks that those voters start breaking in one direction or another. To capture that, you must begin setting a baseline of where the electorate sits before watching for movement down the stretch. Pollsters establishing a baseline with RV polls are telling an inaccurate story.
This isn’t just me pointing this out, Nathanial Rakich of FiveThirtyEight is saying the same: “It’s Sept. 22. There is a month and a half until the midterms. If you’re a pollster and you’re still not using likely-voter screens, please reconsider.”
Logan Dobson, a pollster attached to the Republican Senate, made a similar point: “There are only 2 LV polls in the RCP average at the moment — both of them show a Republican lead on the generic ballot. If the GOP has a good night, many people will be justifiably confused why so many media pollsters kept putting out Registered Voter polls deep into September.”
Electoral fundamentals point to a good night for Republicans. Pollsters sticking to the RV polls this late are doing a disservice to everyone. They’re not measuring the actual political environment — and this is true whether Democrats are gaining ground or Republicans are standing with their gains.
Aside from polling, the primaries point to a GOP edge in the fall. David Byler at the Washington Post did a writeup of how primary turnout can point to advantages for the general election:
According to pollster John Couvillon, 52 percent of 2022 primary voters cast ballots in GOP races, while 48 percent voted in Democratic races. That’s a good sign for Republicans. High primary turnout signals enthusiasm for the general election — and the party with the stronger primary turnout typically does better in the midterms.
We don’t yet know whether this pattern will repeat in November. But a thorough examination of the data shows that Republicans do have a primary turnout advantage — even considering Dobbs and the other complexities of this election cycle.
The GOP’s four-to-five-point turnout advantage can be measured in several ways.
Measured by total votes, the GOP is the strongest it has been in five midterm elections.
In 2010 and 2014 — the last two midterm cycles in which Republicans enjoyed a turnout advantage — voter interest was low and Republicans won by out-mobilizing Democrats. This year appears to be different: Voter interest is high among all voters and Republicans have a primary turnout edge.
If you’re looking at those numbers, it’d be the equivalent to a GOP +4 – 5 environment. If this were a generic ballot, that would be close to the average between the Tralfagar (GOP +6) and Rasmussen (GOP +1) polls (average of GOP +3.5).
That’s not to say any of this is a given. Everything can change as we lead up to election day. But when we know, without question, that the polls feeding into the averages are poor, that has to make you question the averages—garbage in, garbage out.
In that kind of environment, the House is a lock for Republicans, and control of the Senate is a strong possibility. According to Sean Trende’s simple average, a GOP +4 environment is where you’d expect a low end of Republicans gaining three seats in the Senate and a high end of five seats. That’s more than enough for control of the chamber.
It’s a 50-50 Senate, and Republicans only need to gain one seat. I’ve seen a lot of talk about how Mitch McConnell is spending money on seats for the Senate. His strategy isn’t complicated: they’re hammering down the easiest path to a 51+ majority in the Senate. Republicans want control of both chambers. They’re not overextending resources into more challenging races and ensuring they walk away with the majority.
Odds are Republicans will do exactly that by winning a seat in Nevada (Sean Trende explains why Nevada looks good this cycle) and holding all their other seats. Georgia and Ohio are relatively easy seats to maintain too. Pennsylvania and Arizona are more brutal races but also have a ton of money flowing into them. If McConnell and Republicans have things at a toss-up or slightly GOP lean come November, they’re anticipating a typical midterm shift to help push others over the edge.
History says this is a sound strategy. The House is a near lock for Republicans, and the question is just how large the majority is — the GOP will seek to expand that majority in 2024. Republicans need 51 seats to control the chamber, bringing Biden’s agenda to a screeching halt.
Again, everything I’m describing here is electoral politics. We’re all guesstimating until votes start rolling in and getting counted. But we’re getting close to that moment of truth. The elections are much closer than you think, and we need good data to measure what’s happening underneath the surface.
When pollsters start shifting to LV polls, we’ll likely see a shift in sentiment. Biden’s approval numbers will be critical, as will the generic ballot numbers. Until now, it was fine using RV and Adult screens in polls. We can afford to be a little fuzzy about what’s happening.
But not now as we head into November. That’s what I can say for sure that the polls are wrong, and everyone knows it. The question is, when does this finally start changing? Hopefully, this next week. Though, I’m shocked I’m asking for this transition heading into the end of September. Usually, I don’t even have to factor this in by now. LV polls are just presumed.
So as you’re watching the elections get near and news anchors start talking about polls more, ask yourself: what kind of poll did they run? If they ran an LV poll, it’s likely more accurate than anything else. If they’re still running an RV poll this late, they’re purposely putting out lousy data. This is a fact every pollster knows.
We need good data to know what’s happening. We don’t have it. The polls are wrong.
Links of the week
The Woke and the Restless – Matthew Continetti, National Review
Why Progressives Undermine Civilization: From electricity and psychiatry to criminal justice and meritocracy, center-Left political leaders are dismantling the institutions our forefathers created. Why is that? – Michael Shellenberger
A Great Copper Squeeze Is Coming for the Global Economy: The recent downturn for copper prices only stands to worsen a coming deficit as the slump discourages new investments for the metal used in EVs and power grids. – Bloomberg
Do Biden’s words even matter anymore? As his staff once again rushes in to clarify, it isn’t clear who’s running the country – Stephen L. Miller, American Spectator
Housing starts data raises 5th recession red flag – Housing Wire
Meta Quietly Reduces Staff in Cost-Cutting Push: Facebook parent is looking to reduce costs by at least 10%, people familiar with the plans said, while Google has required some employees to apply for new jobs – WSJ
Biden Just Doesn’t Get It on Inflation – Jim Geraghty, National Review
Dems set the standard on illegal immigration – Katie Pavlich, The Hill
In the dollar we trust – Jeffrey Frankel, Project Syndicate
I’m still here – John Wall, The Player’s Tribune
Twitter Thread(s) of the week
Satire of the week
Thanks for reading!