A double standard for polls with incumbents below 50 (GOP House vs Dem Senate?)

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

When we see House polls with Republican incumbents below 50, this is usually taken as a bad sign for the Republicans.

However, we do not usually see this same reaction when we see polls with a Dem Senate incumbent below 50.

So, is there a double standard for how these House polls with Republican incumbents below 50 and how Senate polls with Democratic incumbents below 50 are treated?


Examples:

The highest that Claire McCaskill has polled recently is 47%.

The highest that Joe Donnelly has polled recently is 49% if you take the most favorable matchup for him (without the Libertarian) in the Marist poll. Most polls have had him substantially lower though, often in the low 40s.

The highest that Heidi Heitkamp has polled recently is 44%, in the FOX poll before the most recent FOX poll that had her down 12 and at only 41%.

Jon Tester and Joe Manchin are polling a bit better than this trio, but still often have been polling below 50%. For example, the highest that Manchin has polled since August is 48%, and that is in his own internal. He is only consistently leading Morrisey by a good margin because Morrisey is usually below 40 and the WV polls have large numbers of undecideds.



One possible argument for treating these Dem incumbents differently from the Republican House incumbents is that "it is a good Dem year and the undecideds will break to Dems." It may be true that it is a good Dem year (although I think this puts the horse before the cart to some extent), and that could plausibly make a significant difference.

However, these Dem Senate incumbents are all running in deep red states, whereas most of the districts where House Republicans are running are at least somewhat swingy/competitive districts. The voters in the deep red states with Dem Senate incumbents, and also the undecideds, are likely to be much more Republican leaning than the undecided voters in swing House districts are Dem leaning. So on that basis (partisan lean of undecideds), one might think that Republican House incumbents are better positioned to do comparatively well among undecideds than Democratic Senate incumbents are.

Personally, I would expect this latter effect (partisan lean of undecideds) and the previous effect (good democratic year) to offset each other to a significant degree in these red state Senate races. This means that undecideds will likely break more Republican in these Senate races than undecideds in the House races with Republican incumbents.



Anyway, is there a double standard here? Because if we saw polls with Republican House incumbents showing the same sorts of numbers that many Dem Senate incumbents are showing, we would in many cases be wondering how long until they get triaged.

Is this just an example of how Atlas is Dem biased in its analysis, or is there really a good basis, that we can be confident about, for this seeming difference in analysis?

[IceSpear]

The obvious answer is Dem hackery.

But there's also the fact that most of the time a Senate challenger is going to have far higher name recognition than a House challenger does because the races are much higher profile.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

The obvious answer is Dem hackery.

But there's also the fact that most of the time a Senate challenger is going to have far higher name recognition than a House challenger does because the races are much higher profile.

If it is just an issue of name recognition, then isn't it even worse for a Senate incumbent to be below 50? That would suggest that a House incumbent that is below 50 may be able to pick up some support simply by running some ads and increasing name recognition.

[America Needs a 13-6 Progressive SCOTUS]

Because undecideds tend to break heavily towards the party out of power.

[IceSpear]

The obvious answer is Dem hackery.

But there's also the fact that most of the time a Senate challenger is going to have far higher name recognition than a House challenger does because the races are much higher profile.

If it is just an issue of name recognition, then isn't it even worse for a Senate incumbent to be below 50? That would suggest that a House incumbent that is below 50 may be able to pick up some support simply by running some ads and increasing name recognition.

Most of the time, the more evenly matched the name recognition, the more representative the poll will be of reality. I'd rather be leading by 5 with my opponent being just as well known as me than leading by 10 with my opponent being a complete unknown, especially if they're a serious candidate that has the resources to eventually get their name recognition on par with mine. For instance, the name recognition gap between Claire McCaskill and Josh Hawley is far smaller than the name recognition gap between like, Peter Roskam and Sean Casten. In fact, in the former case there basically is no gap at all.

[Fmr. Gov. NickG]

The obvious answer is Dem hackery.

But there's also the fact that most of the time a Senate challenger is going to have far higher name recognition than a House challenger does because the races are much higher profile.

If it is just an issue of name recognition, then isn't it even worse for a Senate incumbent to be below 50? That would suggest that a House incumbent that is below 50 may be able to pick up some support simply by running some ads and increasing name recognition.

I think the point was that the challenger tends to have much lower name recognition in House races than Senate races, and thus more room to grow.

But I don't really think the 50% threshold is particularly magical in either case.  The margin is much, much more important.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

The obvious answer is Dem hackery.

But there's also the fact that most of the time a Senate challenger is going to have far higher name recognition than a House challenger does because the races are much higher profile.

If it is just an issue of name recognition, then isn't it even worse for a Senate incumbent to be below 50? That would suggest that a House incumbent that is below 50 may be able to pick up some support simply by running some ads and increasing name recognition.

You misread. Senate polls tend to be more accurate than House polls when incumbents have narrow leads because both Senate nominees already have high name rec. That generally isn’t the case for House challengers. That said, IN, MO, MT and WV still concern me a lot

Oh yeah, you are right. I read it as "incumbent," whereas IceSpear actually said "challenger."

[MT Treasurer]

Yeah, people like to point out (and for good reason!) that Heller seems to be stuck in the low to mid 40s, but it’s not as if someone like Joe Donnelly is very close to 50 right now either (I’m aware that he could win with 47%/48% or something like that because of the Libertarian, but I’m talking more about the high % of undecideds in general). It really all comes down to whether you think the environment will carry those red state Dems, and while there are legitimate reasons to believe that undecideds will break towards Democrats in states like IN or MO, it’s definitely not a sure thing by any means.

[Xing]

I would imagine that it's because this looks as though it'll be a good year for Democrats. I wouldn't be too excited about a poll showing a Republican up 48-42 though, for example. In the case of Heller, it's not just that he's under 50. It's that Nevada polls are consistently bad.

[Co-Chair Bagel23]

Yeah, this is why I think Morrisey edges out.