Polls on Same-Sex Marriage State Laws

[Alcon]

New Iowa numbers:

Do you think same-sex marriage should be legal or illegal?

Legal............................................................... 41%
Illegal .............................................................. 48%
Not sure ........................................................... 11%

Would you vote for or against a constitutional
amendment that says the following: “Marriage
between one man and one woman shall be the
only legal union valid or recognized in this
state"?

Would vote for it..............................................  50%
Would vote against it ......................................  43%
Not sure ..........................................................   8%

No gap?  Bizare.

[Alcon]

What exactly makes SD so gay friendly ?

Their 2006 amendment was really restrictive, banning civil unions (like the failing Arizona one) and only passed 52%-48%.

[Alcon]

Studies indicate that polls significantly overstate public support for 'gay marriage."

http://prospect.org/article/when-can-you-trust-polling-about-ballot-measures

http://politics.as.nyu.edu/docs/IO/4819/egan_marriage_polling.pdf

This is pretty well-known already, but must be a shallow comfort for you folks considering the obvious trend you're working against.

[Alcon]

Studies indicate that polls significantly overstate public support for 'gay marriage."

http://prospect.org/article/when-can-you-trust-polling-about-ballot-measures

http://politics.as.nyu.edu/docs/IO/4819/egan_marriage_polling.pdf

This is pretty well-known already, but must be a shallow comfort for you folks considering the obvious trend you're working against.

Hmm.

You didn't "know" or acknowledge it when I pointed this out previous.

I'm not sure that's English, but...have I disagreed in the past?  I don't respond to every post you make to indicate my agreement or disagreement.

Oh, and would you please point out where the electorate (as opposed to inaccurate surveys) shows such a "trend"?

Are you claiming that national and state polling that shows a long-term trend toward support of gay marriage increasing is..."inaccurate" because of the gap between it and Election Day results?  Yeah, I'm not sure you know what a "trend" is either.

[Alcon]

Studies indicate that polls significantly overstate public support for 'gay marriage."

http://prospect.org/article/when-can-you-trust-polling-about-ballot-measures

http://politics.as.nyu.edu/docs/IO/4819/egan_marriage_polling.pdf

Yes, of course. All the polling in the universe does not address or reflect the actual performance of homosexuality at the ballot box.

I don't understand how you guys think about this.  Are you insinuating that the underperformance of gay rights at the ballot box somehow negates the obvious trends we've seen in social science polling (and at the ballot box, really) on this issue over the past decade?

[Alcon]

Studies indicate that polls significantly overstate public support for 'gay marriage."

http://prospect.org/article/when-can-you-trust-polling-about-ballot-measures

http://politics.as.nyu.edu/docs/IO/4819/egan_marriage_polling.pdf

Yes, of course. All the polling in the universe does not address or reflect the actual performance of homosexuality at the ballot box.

I don't understand how you guys think about this.  Are you insinuating that the underperformance of gay rights at the ballot box somehow negates the obvious trends we've seen in social science polling (and at the ballot box, really) on this issue over the past decade?

I believe he's suggesting there exists a sort-of Bradley Effect on the issue.

I get that, but that doesn't mean the polls "don't address" the actual performance, nor does it suggest the trend seen in polling on the issue is false.  It seems like both posters I'm quoting are making additional claims besides a sort of "Bradley Effect" (maybe I'm wrong) but aren't being clear what they mean.

[Alcon]

Carl already pointed that polling on this issue is misleading.

1. Polling companies routinely "cook the books" either deliberately or through the way in which they word their questions.

The Gallup question has been remarkably consistent for years, as far as I know.

Beyond that, unsure what you mean by "cook the books."  Are you alleging the trend on the gay marriage issue is a mass fabrication?

2. Individuals often do not give accurate information to pollsters.

Again, I'm not quite sure what you're getting at.  There's certainly a phenomenon where gay rights issues modestly overpoll on average.  But are you suggesting that the trend is instead just a massive increase in misrepresentation of opinion?  That doesn't even square with election results.

3. There is a marked difference between the national polls on this issue, and individual State polls. If the national polls like the Gallup garbage are correct, then Pennsylvania shouldn't have a 20+ point gap against homosexual 'marriage'.

So either Gallup is wrong, PPP is wrong, or something else is going on.

And then there are results like Arizona.  Have you added this up?  I'd be interested in what you got.

4. Polling companies changed the wording of the way that they asked the question. This means that any polls before/after that change cannot be compared.

Ehh, you can compare dissimilar things.  It just introduces a degree of subjectivity.  And again...you think this explains the shift in the Gallup Poll?  I'm just not quite sure what specific claim you're making, sorry.

5. A recent national poll using different wording found 62% opposition to changing the definition of marriage. (This would be more in line with the 20 point gap found in Pennsylvania).

Would you mind linking to this?

[Alcon]

Pennsylvania is a swing state with a left lean. Hence, if the national polls indicated a 53%-44% split in favor of redefining marriage, Pennsylvania should be somewhat close to that number, not 20 points in the opposite direction.

When you consider the other PPP polls (OH, MI, MN, ect...) you don't get numbers IN SWING STATES which would support what the Gallup polls are showing.

When you add in the numbers in places like MARYLAND (about as far left as you can get) then something is wrong with the way Gallup is conducting the poll.

This is not just Gallup's result.  It's been echoed by Pew and others.  Gay marriage also relatively lags in the Midwest compared to New England and the West, so a simple adjustment of the Pennsylvania results may be insufficient.

Have you done a statistical analysis to try to demonstrate whether the state and national polls are incompatible?  It may seem that way facially, but by contrast, Arizona's -1 would suggest that a national +3 (or whatever) is more than plausible.  If all of this oddness adds up to the difference of only a few points, claiming it throws the trend towards gay rights support into question is an absurd conclusion.

Moreover, since multiple national polls echo this result, and because it's not like state and national polling have immensely different methodologies, I'm not exactly sure what you're suggesting explains the discrepancy you're sensing.

[Alcon]

Oh, and changing ONE WORD in a polling question can drastically change the results.

Hence, social "science" is not science at all.

That doesn't really follow.  There's a use of scientific methodology; there is just subjectivity involved in finding a representative sample and in finding the right question to gauge an abstract concept.  The fact that people respond differently to different phrasing shows that oftentimes people aren't even sure of their real opinion.  There's no perfect phrasing anyway -- After all, ballot measures themselves have varying language.

Even in hard science, complete control for variables oftentimes is impossible.  The presence of subjectivity in science is not alone enough to make it "unscientific"; it certainly isn't enough to make information inherently useless.

Self-reporting surveys are inherently flawed, especially on hot button issues in the face of propaganda and violence from those pushing an agenda on the issue. (White powder sent to mormon temples, assaults, death threats, boycotts, fines, ect...)

There may well be pressures that cause people to self-report inaccurately -- Social desirability, fear, anxiety about being challenge, plenty of things enter the equation.  You can establish a gap between the polling on the issue and the actual results.  That's actually relatively easy to do, even if you can never tell how big the gap really is.  But we agree with you that the gap exists, and it tends to overestimate gay rights support modestly.  Are you trying to make a further point?

[Alcon]

I always thought Minnesota was more progressive on these issues ...

The Midwest has some considerable cultural traditionalism going, for better or worse.

[Alcon]

Just write up a repeal of marriage licensing in your state and let churches take it over.  You can get married at the First Fabulous Church of His Lord in your area.

[Alcon]

Those colored bars are of...interesting relative lengths.

[Alcon]

LOL!!!

Push poll commissioned by homosexual activists. Even the far-left Post says, its the "rosiest" picture yet.

OF COURSE! It's a push poll.

Got to love that spin. Goes to show that PPP is nothing more than a propaganda outfit.

How did they word the push questions?

[Alcon]

According to that Mississippi poll, voters aged 18-29 are only at 21% support for same-sex marriage, but those 30-45 are much more supportive at 39%.  Interesting.

That's really weird, it should be the opposite.

The entire sample of voters 18-45 is 30% of n=502, so 18-29 and 30-45 combined are only 151 people.  That means that the sample of 18-29 is probably around n=50 at most, for a gigantic Margin of Error of +/-14%.

[Alcon]

They might not like SSM, but they would be willing to vote to legalize it.

That hypothesis doesn't make much sense, considering the first question asks if they want same-sex marriage to be allowed, not whether they approve of same-sex marriage.

There is considerable research indicating that the "real" numbers on same-sex marriage come only after you re-assign to the "No" column people who give inconsistent answers or indicate personal discomfort/theological opposition to same-sex marriage.  And that applies to the straight-up "do you support same-sex marriage rights?" question.  The second question in this Ohio poll is about as vague and feel-good as they come.  There's absolutely no reason to believe it's more accurate than the first question, based on either wording or past history with similar poll language.

I would peg same-sex marriage for about -5 points in Ohio based on those numbers...which isn't bad at all.  The +14 number is pretty ridiculous to accept, though.

[Alcon]

They might not like SSM, but they would be willing to vote to legalize it.

That hypothesis doesn't make much sense, considering the first question asks if they want same-sex marriage to be allowed, not whether they approve of same-sex marriage.

There is considerable research indicating that the "real" numbers on same-sex marriage come only after you re-assign to the "No" column people who give inconsistent answers or indicate personal discomfort/theological opposition to same-sex marriage.  And that applies to the straight-up "do you support same-sex marriage rights?" question.  The second question in this Ohio poll is about as vague and feel-good as they come.  There's absolutely no reason to believe it's more accurate than the first question, based on either wording or past history with similar poll language.

I would peg same-sex marriage for about -5 points in Ohio based on those numbers...which isn't bad at all.  The +14 number is pretty ridiculous to accept, though.

Also as I pointed out earlier, while the press release mentions the 47-48 result first, there is nothing to indicate it was asked before the question that gave the 52-38 result. If that second question was asked first, some of the respondents likely were not thinking about SSM when they gave their answer.  Indeed, thinking about again, that really is the only way to explain such disparate numbers on the two questions.  Some people who answered that question were thinking only in terms of its effects upon opposite-sex marriage, since they were thinking that is the only true form of marriage.

https://freedomohio.com/docs/polling-details.pdf

Unless I systematically misunderstand how PPP reports their results, I'm quite sure this is what they did:

1. Asked Q5 ("support or oppose allowing same-sex couples to get married?")

2. Asked Q6 ("support or oppose a constitutional amendment [etc. etc.]?")

3. Of those who opposed the constitutional amendment to allow same-sex marriage, they asked Q7, which strengthens the Q6 language about religious exemptions.

It's hard for me to believe that many people didn't pick up the context clue, but even if only 1-in-4 didn't, that could make for a big difference.  Otherwise, it's possible that some people support removing the constitutional amendment, and didn't understand that it would result in same-sex marriage's immediate legality.  I think straining interpretation here is unnecessary, though: we have seen Q5 a million times, and we already know it generally overpolls, as I described above.  If Q6 is polling better than Q5, and we know Q5 overpolls, I don't see much use in thinking too hard about Q6.

It looks like, in polls that ask both legality and on a specific referendum, the "legal" margin is usually 2-5 percentage points better than the specific referendum ask, which in turn tends to be 2-5 percentage points better than the real referendum result.  If a few more polls show a "legal" average around -1, it's pretty certain that Ohio would fail a referendum by about 5-10 points (maybe more, considering some voters are probably more wary of amendments than referenda.)

[Alcon]

Taking a look at their demographics, I see they are severely underpolling non-whites with them being only 9% of the sample, yet they make up 18% of Pennsylvania's population according to the 2010 census.  It also appears they are underpolling Protestants.  On the other hand, their sample is significantly skewed towards older respondents.

18% of Pennsylvania's registered voters, voting-eligible population, voting-age population, or population?  And was this a poll of likely 2014 voters, all voters, adults, or what?  Depending on the answers to those two questions, we could easily be within the MoE here.  And even if this sub-sample misses the MoE, that's a sign that they're not weighting well, not necessarily that their sampling is bad.  It takes more than one MoE miss to justify assuming bad sampling.

(They may suck -- uni polls often do -- I'm just explaining that we need soldier evidence than just this.)

[Alcon]

I think the ruling says that Kentucky (like Oregon) recognizes same-sex marriages from other jursidictions, not that same-sex marriage is legal there.

[Alcon]

I think the ruling says that Kentucky (like Oregon) recognizes same-sex marriages from other jursidictions, not that same-sex marriage is legal there.

Not according to this.

I don't know why you interpret "legally recognized" to mean "issues same-sex marriages" instead of "recognizes out-of-state same-sex marriages."  If anything, "legally recognized" sounds more like the latter to me.  Plus, every source I've seen indicates this recognizes pre-existing marriages only.

[Alcon]

Here's where YouGov most recently had Wyoming on approval and disapproval on same-sex marriage (SSM):

WY    33    50    -17

I don't know where the 53-39 poll came from. If such is the pattern, then it would seem that soon after SSM is tolerated by law, the public follows. Maybe people find that it does not hurt them.  

PPP could poll North Carolina. This could compel me to update the approval rating for SSM in at least one state.

Why does Wyoming get this reputation of being all libertarian and whatnot... like a gaggle of cowboys that don't want the government involved in anything?  They are clearly just a bunch of nasty hicks like the rest of Republamerica.  

I think there was a time where it was like that, but the energy boom of the late 2000s brought in a lot of white trash. I think as the value of Hydrocarbons decreases, that these people will move out and things may or may not turn to normal in the next few years. Though even then, I don't think they are anti-Government as they are "we don't really care about feminism or civil rights or religion for that matter, we just want to get rich and enjoy the outdoors".

Wyoming's population hasn't really changed enough to cause any kind of seismic shift in their politics.  Also, I doubt that young energy industry workers are substantially more socially conservative than native Wyoming citizens.  They probably vote at fairly poor rates, too.

The interior west may have a "leave me alone" approach to politics, but that often takes the form of "don't force social liberalism on me and my family."  I guess that's libertarian in a way, but in a socially right-wing way.