Generic ballot polls and Senate polls don't jive.

[triron]

Why is that?

[Codeine]

1. Senate races are concentrated

in red states this cycle.

[DemocratSinceBirth]

2. /Thread

[triron]

3. What about Nevada and some competitive 'red' states (like Missouri, Florida)?

I believe Florida is blue is actuality; Russians made it red.

[Awsi Dooger]

11. Hillary got 38% in Missouri

Hardly a competitive state. That changed a long time ago.

Florida is only perhaps one point red in neutral conditions. But Republicans do an extremely good job of motivating and maximizing in Florida. I think Gillum will win. Senate race was looking good until Scott benefitted from this hurricane.

Nevada is close to even with the nation in presidential years but has a tendency to tilt red in midterms, especially in midterms favoring Republicans. This cycle we should have a chance to take away the governors seat in an open race. Ousting an incumbent like Heller will be difficult but I am somewhat optimistic.

The senate map was extraordinarily difficult this cycle. Tennessee has changed dramatically. There was a chat on 538 yesterday that contained quite a bit of talk on how much more conservative Tennessee has become since 2006. That race was basically a pipe dream even though we had a candidate to candidate edge.

I do not agree that the generic doesn't match the senate polling. Simply too many hard line conservatives in many of those states we are trying to retain or take away. It's the reason I always look at those liberal/conservative percentages in each state. They are like rigid poles stuck in the ground and preventing too much flow. Anyone who ignores those percentages and takes polling at face value, especially early in a cycle, is destined to be fooled. Beto is a phenomenal candidate but Texas is not going to deport those 44% hard line conservatives in time for election day.

Arizona is the disappointing state because it is ready to flip this cycle and the opponent is hardly a great candidate. But apparently Sinema had plenty of unfortunate comments in her recent past and now the Republican opposition research crew has uncovered them and using them effectively against her. That is one reason I am opposed to nominating so many females. I have emphasized that time and again. Kanye West may be a noodle brain but what he described is similar to what I've heard from countless simplistic males. They are scared of females in position of power and use any excuse to vote the other way. Sinema is falling in the polls due to those comments partially because female nominees in tight races have lower margin for error and not nearly as much teflon.

[GulfCoast66]

15. Respectfully, You are incorrect.

Nelson is our only statewide Democratic office holder.

There are more registered Democrats in Florida. But most do not vote. Republicans do.

[PoliticAverse]

4. "Generic ballot polls" don't matter. Ignore them. n/t

[Loki Liesmith]

6. That's empirically false.

[PoliticAverse]

7. Tell me when you can vote for a "generic" ballot. n/t

[DemocratSinceBirth]

8. You are correct

If you’re interested in how the 2018 midterm elections will go, you may want to keep an eye on the generic congressional ballot, one of the best tools for predicting how many U.S. House seats each party will win.

https://fivethirtyeight.com/features/heres-the-best-tool-we-have-for-understanding-how-the-midterms-are-shaping-up/

There's also peer reviewed research that attests to the efficacy of the generic ballot:

Will Democrats Catch a Wave?
The Generic Ballot Model and the 2018 U.S. House Elections

Alan Abramowitz

This article presents a simple model that uses results of generic ballot polling to predict the outcome of the 2018 elections to the U.S. House of Representatives. The “generic ballot” refers to a question included in numerous national polls asking voters which party they favor in the House elections without providing the names of the candidates. When combined with two other variables—the party of the president and the number of seats held by each party prior to the election—this model produces very accurate predictions of seat swing in House midterm elections. The model currently predicts a strong likelihood of a Democratic takeover of the House.

For the 18 midterm elections since World War II, the results of the generic ballot test are highly correlated with the party division of the national popular vote for the House (r=.82). The generic ballot also can accurately predict seat swing in midterm elections when incorporated in a statistical forecasting model (Moore and Saad 1997; McGhee and Baldassare 2004; Abramowitz 2010; Abramowitz 2014). In this article, I use results of a simple three-variable model including the generic ballot to provide conditional forecasts of seat swing in the 2018 midterm election.


...

I wish i could link it. It's in a google doc.

[PoliticAverse]

13. If you are really interested in how the 2018 midterms will go you may want to look at the polling

for the individual races.

[DemocratSinceBirth]

16. Would you agree that 82% is high correlation ?

For the 18 midterm elections since World War II, the results of the generic ballot test are highly correlated with the party division of the national popular vote for the House (r=.82). The generic ballot also can accurately predict seat swing in midterm elections when incorporated in a statistical forecasting model (Moore and Saad 1997; McGhee and Baldassare 2004; Abramowitz 2010; Abramowitz 2014). In this article, I use results of a simple three-variable model including the generic ballot to provide conditional forecasts of seat swing in the 2018 midterm election.

As a party's share of the generic ballot rises there is a concomitant rise in their percentage of the two party congressional vote.Also, we don't have individual polls for the seventy or so races deemed competitive which will determine the control of the House, much less polls for all 435 House seats.

[PoliticAverse]

17. Who cares what the correlation with the "party division of the national popular vote for the house"

is? I want to know which party will have the most seats. What was the "party division of the national popular vote" of
the last Presidential election? Who is President? It's less than one month to the midterm elections. Look at the individual
races if you actually want to know who will be controlling the House and by how many seats.

[DemocratSinceBirth]

19. We don't have individual polls for all the races that will determine the control of the House.

The peer reviewed research indicates that the generic ballot is a proxy for the success or lack thereof of a party. Given the effects of gerrymandering and that Democratic voters are more clustered into densely populated areas a 3 point lead in the generic ballot is the inflection point for Democratic control of the House. For every point above that Democratic chances increase.


I will let you have the last word. Please enjoy the rest of your weekend.

[Awsi Dooger]

28. I am only paying attention to the generic polling average and 538 House percentage

As DemocratSinceBirth pointed out, the generic ballot is easily the greatest correlation to House outcome and number of seats to swing.

The polling on specific House races is infrequent and unsophisticated and comparatively not reliable. It is more like primary polling, which can be all over the place.

I can't imagine looking at House district polling and trying to pretend it has equal or greater weight to generic polling average nationally. Not even close.

Bottom line there will be dozens of tight House races. Where they fall nobody knows. But the generic House indications and actual generic outcome are far more likely to point the way than polling on individual races.

Senate is another matter. As Nate Silver pointed out a couple of months ago, we probably needed an 11-12 point House generic edge to yank the senate along. And since the polling in the Tennessee senate race was probably overly kind to Bredesen to begin with, and other senate seats have slipped further away, even 11-12 might not be enough.

[NurseJackie]

12. Thank you! Nobody votes for generic candidates!

[Generic Brad]

18. Which is exactly why I will never run for office

'Nuff said.

[NurseJackie]

27. Ha!! :-D

[beachbum bob]

5. Never place much faith in the generic ballot polls, only shows trends in time

across general locations. Of course I rather see democrats up 12 than down 12,

[JCMach1]

9. Remember the Competitive Senate Seats tend to be in Red States- Kavanaugh effect was

greater there for the Red side... Polls should tighten if the Dem. candidates are in the Race going forward. Texas, for example, is even more highly determined by who shows at the polls... so much so that your can almost ignore 'likely' voter polls... or just use them to follow the trend...

[DavidDvorkin]

10. For the same reason that someone can lose the popular vote but win the Electoral College

The generic ballot is a nationwide number, but in voting, the voters for one side or the other are concentrated in various states.

[DemocratSinceBirth]

23. Throw in the fact Kavamess made blue and purple areas bluer and red areas redder.

[DeminPennswoods]

14. That's why you hear political pros say that

Dems need to be 10+ in the generic ballot to take the House and 12+ to take the Senate. That factors in the gerrymandering and such.

[DemocratSinceBirth]

21. Three points is the point where the Dems have a 50% chance of retaking the House.

Our chances increase for every point over that. We're in real trouble if we need a 10% generic lead to take back the House. We only had an 8% lead in 06 and we have an 8.4 point lead the generic ballot as of now:

https://projects.fivethirtyeight.com/congress-generic-ballot-polls/

[karynnj]

20. The generic ballot is national, Senate races by state

Think 2016 -- Hillary Clinton polled ahead up to the election - not a polling error - she DID get about 3% more votes. The problem this year is we won many marginal seats in 2006 or 2006 and 2012, that are up this year.

[DemocratSinceBirth]

22. Every thing went our way in 2006. We won every close race but Ford-Corker.

[karynnj]

24. Exactly -- and we did very well in 2012

This is why a disproportunate number of seats up are currently held by us. That makes retaining all of them - let alone winning two more - a tough challange.

On the other hand, the next two cycles are people who won in 2014 or 2016. Those will be tough for the Republicans.

[DemocratSinceBirth]

25. A good night for us is to win the House and mitigate our losses in the Senate.

Hopefully we can limit those losses to one or two.

We are also set to pick up 500 state legislative seats and six to twelve governors.

I hope these forecasts are right.

[uponit7771]

29. +1, matter of fact they should aggregate the state levels vs giving generic national numbers

[Cicada]

26. The generic ballot asks about House of Representatives, not senate