Emerson swing state polls: trump leads, MN tied

[Person Man]

It's insane how well Democrats are doing downballot.

[Florida Man for Crime]

Emerson has a 2 point GOP bias. Remember to take that into account.

Biases for pollsters are not necessarily the same from year to year. You should not just blindingly add +2 to a poll on that basis. It is true that 538 for example makes adjustments to polls, but they do so on the basis of a consistently applied methodology which on average is neutral. I don't see you going into all the other threads for polls where this would shift things the other direction and consistently reminding everyone to "take into account the Dem bias," and if you are not doing that, you cannot validly do so here.

Party ID for each state:

AZ: R+4
GA: R+5
MI: Even
MN: D+1
NV: D+3
PA: R+1
WI: R+3

All of these except for NV are too R-leaning.

Party ID is not a stable variable. You are doing literally the same thing as Republican Romney 2012 unskewers.

It can be sensible to look at things other things such as recalled vote in a previous election (2020) as a sanity check, but just because party ID is a bit off from what you might expect does not mean that a poll is wrong.

[Florida Man for Crime]

It's insane how well Democrats are doing downballot.

It's almost as though generic democrats have a stronger brand than Biden.

[dspNY]

Emerson has a 2 point GOP bias. Remember to take that into account.

Biases for pollsters are not necessarily the same from year to year. You should not just blindingly add +2 to a poll on that basis. It is true that 538 for example makes adjustments to polls, but they do so on the basis of a consistently applied methodology which on average is neutral. I don't see you going into all the other threads for polls where this would shift things the other direction and consistently reminding everyone to "take into account the Dem bias," and if you are not doing that, you cannot validly do so here.

Party ID for each state:

AZ: R+4
GA: R+5
MI: Even
MN: D+1
NV: D+3
PA: R+1
WI: R+3

All of these except for NV are too R-leaning.

Party ID is not a stable variable. You are doing literally the same thing as Republican Romney 2012 unskewers.

It can be sensible to look at things other things such as recalled vote in a previous election (2020) as a sanity check, but just because party ID is a bit off from what you might expect does not mean that a poll is wrong.

Emerson predicted Senators Masters, Oz and Laxalt last cycle. They also predicted Ron Johnson and Ted Budd winning by 5 points when they won by 1 and 3 respectively. Forgive me for thinking Emerson is crap. Emerson is systematically off as a pollster at the moment

[Florida Man for Crime]

Emerson predicted Senators Masters, Oz and Laxalt last cycle. They also predicted Ron Johnson and Ted Budd winning by 5 points when they won by 1 and 3 respectively. Forgive me for thinking Emerson is crap. Emerson is systematically off as a pollster at the moment

I agree they are not the greatest pollster.

However, pollsters generally look at their errors after a cycle and try to make improvements and changes to their methodology. It is probable that Emerson made some adjustments, so their bias is probably not the same now as it was in 2022 (even aside from the fact that polling Presidential years is a different matter from polling off-years in general due to the substantial differences involved in predicting turnout and the nature of the electorate).

So for this reason, you cannot simply project their past bias on to their current polls (unless you are doing it systematically and consistentlyas a way to smoothen random variation in polling averages caused by differences in pollster methodology, i.e. what 538 does, which is a very different thing).

[dspNY]

Other examples of Emerson’s systematic failure as a pollster from last cycle:

They had Bennet winning CO-Sen by 7 and he won by 14.
They had Blumenthal winning CT-Sen by 12 and he won by 15.
They had Duckworth winning IL-Sen by 10 and she won by 15.
They had Hassan winning NH-Sen by 4 and she won by 9.
They had Ryan losing OH-Sen by 9 and he lost by 6.
They had Murray winning WA-Sen by 9 and she won by 15.

In the high profile senate races that were close or expected to be close last cycle, the only one they got right was GA. Every other race they polled that was considered competitive was heavily biased to the GOP.

[Tekken_Guy]

This is the first poll I can recall in which Wisconsin is Trump's best state of the rust belt trio. Usually it's PA, and earlier in the year it was usually MI.

[Amenhotep Bakari-Sellers]

I don't believe these Emerson polls id Ds are ahead in S

[wbrocks67]

Honestly this just tells me that all the swing states will be close. These are at least more realistic and within MoE than some of the ridiculous Trump +9 results we were getting for some of these states.

Just like the whiplash of which state is better for Biden. It was PA, then MI, then WI, then NV was completely lost, now it's his best state, etc. It's all over the place.

I don't really put too much stock into the MN poll specifically since it has Klobuchar barely clearing double digits which is pretty unlikely for her.

[Spectator]

The Rosen numbers continue to blow my mind, yet all polls are showing this kind of split in NV.

[Eraserhead]

Emerson isn't exactly the best pollster but yeah these are mostly quite poor results indeed.

[wbrocks67]

Why Emerson insists on having Republicans lead among "party registration" in PA is beyond me. Party ID is one thing, but Party Registration is still D+4.

[Tekken_Guy]

Emerson isn't exactly the best pollster but yeah these are mostly quite poor results indeed.

They're not terrible for Biden given Emerson's R bias, since he's only down a couple of points.

[oldtimer]

This is the first poll I can recall in which Wisconsin is Trump's best state of the rust belt trio. Usually it's PA, and earlier in the year it was usually MI.

They alternate randomly because they are lumped together.

[7,052,770]

Those Senate results wow. Rosen in NV just surging.

Dems are up:
+4 Arizona
+4 Michigan
+12 Nevada
+6 Pennsylvania
+2 Wisconsin

I just can't see this split holding through November.

[wbrocks67]

Biden at that magic 31% of white voters in GA...

[Florida Man for Crime]

Why Emerson insists on having Republicans lead among "party registration" in PA is beyond me. Party ID is one thing, but Party Registration is still D+4.

Well, if you look at the poll results, you can see that they ask the question "What is your Party Registration?"

This implies that they do not already know the party registration of the person they are talking to. And that in turn implies that they are not polling from a voter file list, but instead are using some form of random digit dialing.

If their sample is weighted sensibly, this most likely means that there are some people in PA who are in reality registered as Democrats, but do not realize that they are registered as Democrats and think they are Republicans.

That's not necessarily such a good thing for Democrats, because it implies a significant number of traditional Dems who are now (in their own minds) solidly Republicans/Trump supporters outside of the Philly metro.

Of course, there is also a possibility it could mean their sample is weighted poorly or that they are estimating that turnout will be relatively favorable for Republicans. But most likely it is primarily people misremembering their registration and assuming that they must have registered as Republicans, unless Emerson is doing an unusually bad job at sampling.

[Fancyarcher]

I don't really put too much stock into the MN poll specifically since it has Klobuchar barely clearing double digits which is pretty unlikely for her.

I mean it's not unrealistically for Klobuchar crossover support to dim considerably. I expect her out to outrun Biden by quite a bit, but the GOP opponent crossing 40% isn't out the of the question.

It's Emerson anyway, so I don't take it that seriously.

[wbrocks67]

I don't really put too much stock into the MN poll specifically since it has Klobuchar barely clearing double digits which is pretty unlikely for her.

I mean it's not unrealistically for Klobuchar crossover support to dim considerably. I expect her out to outrun Biden by quite a bit, but the GOP opponent crossing 40% isn't out the of the question.

It's Emerson anyway, so I don't take it that seriously.

I think Klob's crossover will definitiely fade, but also her opponent is a complete whack job this year too, so I actually wouldn't be surprised if she still won by like 15%.

But at the end of the day, this is a make or break cycle for Emerson. IF they botch this one again, they'll have 2022 and 2024 in a row of bad cycles.

[GAinDC]

Biden at that magic 31% of white voters in GA...

How was Biden polling with white voters back in 2020? I don't recall it being this consistently high. (30% white support for a Dem in the Deep South is a BIG deal)

[wbrocks67]

Biden at that magic 31% of white voters in GA...

How was Biden polling with white voters back in 2020? I don't recall it being this consistently high. (30% white support for a Dem in the Deep South is a BIG deal)

Emerson's last poll in 2020 had Trump +38 among whites, 67-29, which was actually nearly on point (68-31) to the real results.

It's the same old story as most other polls, Biden holding up remarkably well among white voters but lesser so among black voters. Interesting thing looking at the GA 2020 poll is that Biden is basically the same (84-12 then, 82-18 now), but their small slice of Hispanic voters is about 50/50 right now compared to 80-18 for Biden.

[wbrocks67]

Biden at that magic 31% of white voters in GA...

How was Biden polling with white voters back in 2020? I don't recall it being this consistently high. (30% white support for a Dem in the Deep South is a BIG deal)

Emerson's last poll in 2020 had Trump +38 among whites, 67-29, which was actually nearly on point (68-31) to the real results.

It's the same old story as most other polls, Biden holding up remarkably well among white voters but lesser so among black voters. Interesting thing looking at the GA 2020 poll is that Biden is basically the same (84-12 then, 82-18 now), but their small slice of Hispanic voters is about 50/50 right now compared to 80-18 for Biden.

Comparing the two just tells me that these small sub groups can make big changes. Biden is a small crosstab error away from being tied here. The difference between Biden 84-12 among blacks to 82-18 can make a big difference in a place like GA.

[TodayJunior]

Press X to doubt. Somebody is lying….why? Bc when Trump is actually on the ballot, every state with a senate race has voted the EXACT same way as President, with the lone exception being Maine 2020, which goes with their independent-minded streak here and there.

I just don’t see how we can reconcile the presidential numbers with the senate numbers in EVERY single swing state. Are there THAT many people who are going to split ticket this year? I just don’t buy it when we have strong history of what I mentioned in the first paragraph.

“Ya know, I think Trump would be good for the country bc I was better off and I support his agenda, but you know what, I really REALLY LIKE *insert Dem candidate* who’s gonna oppose everything trumps trying to do.”

Make it make sense for me.

[Donald Trump’s Toupée]

Very good news for president Trump!

[wbrocks67]

Press X to doubt. Somebody is lying….why? Bc when Trump is actually on the ballot, every state with a senate race has voted the EXACT same way as President, with the lone exception being Maine 2020, which goes with their independent-minded streak here and there.

I just don’t see how we can reconcile the presidential numbers with the senate numbers in EVERY single swing state. Are there THAT many people who are going to split ticket this year? I just don’t buy it when we have strong history of what I mentioned in the first paragraph.

“Ya know, I think Trump would be good for the country bc I was better off and I support his agenda, but you know what, I really REALLY LIKE *insert Dem candidate* who’s gonna oppose everything trumps trying to do.”

Make it make sense for me.

Yeah, I imagine there will be a few ticket splitters in swing states as there always is, but I also don't imagine many Dem-leaning voters willing to vote for their Dem senate candidate but then Trump for president, given Trump's views are in diametric opposition to most of the swing state Dem candidates. Joe Biden is not *that* disliked among Dem-leaning voters for that to happen. You'd have to also be totally against his policies... which are the same as Dem senate candidates.