Quinnipiac: Americans Oppose Almost All of Trump's Proposals

[Rjjr77]

Sample skews heavily democrat. 1.36 dems for every 1 republican. Income over 100k is also oversampled.

[Rjjr77]

Sample skews heavily democrat. 1.36 dems for every 1 republican. Income over 100k is also oversampled.

A lot of those are opposed by double digit margins. Reducing the party gap wouldn't increase the popularity that much.

.36 is a pretty big party gap, it would absolutely affect the poll.

The whole sample seems out of whack, poll looks to be an outlier.

[Rjjr77]

And yet we have a President Trump, a Republucan Senate, and a Republican House. So either this poll is very wrong or Democrats are even more incompetent than I imagined.

Gerrymandering, the influence of money and opinion-shaping, a poorly educated public, and yes the Dems are probably more incompetent than we can imagine. (They lost to Trump! That's proof of some pretty big incompetence right there.)

Gerrymandering gets a lot of pull on here, but it really isn't that bad, one of the things that hurts democrats with district draws is the minority-majority requirements

[Rjjr77]

Sample skews heavily democrat. 1.36 dems for every 1 republican. Income over 100k is also oversampled.

A lot of those are opposed by double digit margins. Reducing the party gap wouldn't increase the popularity that much.

.36 is a pretty big party gap, it would absolutely affect the poll.

The whole sample seems out of whack, poll looks to be an outlier.

But are registered dems and registered reps actually balanced IRL?

No, which is why it shouldn't be asked in general, ideology should be used, but if you are going to ask it with Registered and Lean they should be equal. In an opinion poll they should be looking at ideological balance from the US in general.

Regardless Quinnipiac actually weights their polls, which mean if sampling margin is off( as it appears to be with what they provide) it can be waaaay off with how they weight. Quinnipiac is normally a solid poll, this set of polls they've been running on the subject seems off.

[Rjjr77]

.36 is a pretty big party gap, it would absolutely affect the poll.

The whole sample seems out of whack, poll looks to be an outlier.

I wasn't implying it would have no effect. Party identification usually skews more Democratic, and so my post assumed a smaller gap but still not even.

Gerrymandering gets a lot of pull on here, but it really isn't that bad, one of the things that hurts democrats with district draws is the minority-majority requirements

It's a number of things hurting Democrats. VRA districts, natural packing, incumbency (for now), and gerrymandering all play a part. Getting fair maps in the midwest/rustbelt and select other states could still give back to Democrats a meaningful number of seats, though. It is much easier to obtain a House majority if you are only 12 - 15 seats in the hole and not 24 - 30.

Quinnipiac weights, so if they sample is off that much on party ID it's going to have wild MOEs down the line, especially if they weight for race/age/income. Normally Quinnipiac has good sound polling but looking back I think they've made a sample error on a couple of similar polls in the same direction. They ought to stop weighting for demographics.