As I read somewhere the other day, Obama isn't leading in the polls because the Democratic sample is high, the Democratic sample is high in the polls because Obama is leading.

That kind of ignores the 2010 elections, doesn't it? While there are differences in presidential-year turnout, you don't get the 2010 results without the electorate shifting somewhat, just as in 2006, which foretold that 2008 would be different from 2004.

As I read somewhere the other day, Obama isn't leading in the polls because the Democratic sample is high, the Democratic sample is high in the polls because Obama is leading.

Obama won independents big in 2008. If the D advantage is only +1 from 2008, as in today's CNN poll, then it's quite odd that Obama could now be losing among independents but still be winning by 3 points. The only way the math works out is if there's a large number of GOP defectors, which seems highly unlikely.

That kind of ignores the 2010 elections, doesn't it? While there are differences in presidential-year turnout, you don't get the 2010 results without the electorate shifting somewhat, just as in 2006, which foretold that 2008 would be different from 2004.

Midterms always, always skew R. In addition, Republican congresscritters ran campaigns specifically targeted at seniors. The efficacy of those tactics has waned after they have taken certain votes.

Party ID is fluid. It's not really that odd at all. All it would take would be more Republicans identifying as Independent on a survey.

I assume this is Joe? Regardless, let's talk apart this logic:

In 2010, the wave had crested and was draining back to status quo, which led to House victories for the GOP. As polls predicted.

Okay. You guys win. Obama is clearly behind and the polling organizations are ignoring all of their incentives to get the election right and are instead in the tank for Obama.

Your logic is so crystal clear I cannot refute it. This is among the worst of the conversations on this lengthy thread.

According to today's CNN poll, the Dems are only +1 from their 2008 advantage, while Obama is losing among independents. How can Obama lose 5-plus points among independents while gaining only one point from Dems and still maintain a 3-point lead? Are you really claiming that a 5-point chunk of voters identifying as Republicans are defecting to Obama?

There is powerful prima facie evidence (lots of current polling data) to support the idea that party affiliation is in fact different since 2010.

Okay. You guys win. Obama is clearly behind and the polling organizations are ignoring all of their incentives to get the election right and are instead in the tank for Obama.

According to today's CNN poll, the Dems are only +1 from their 2008 advantage, while Obama is losing among independents. How can Obama lose 5-plus points among independents while gaining only one point from Dems and still maintain a 3-point lead? Are you really claiming that a 5-point chunk of voters identifying as Republicans are defecting to Obama?

Regardless of what the correct answer is, what "incentives to get the election right"? Pollsters who work for candidates have a great incentive to get the election right; their efforts affect campaign strategy, and a candidate isn't going to pay someone who gives him bad info. But the only thing media polls do is provide fodder for punditry, and there's no penalty for pundits who get things wrong.

I haven't seen any logic from you except that a bunch of polls say X, so X is probably right.

According to today's CNN poll, the Dems are only +1 from their 2008 advantage, while Obama is losing among independents. How can Obama lose 5-plus points among independents while gaining only one point from Dems and still maintain a 3-point lead? Are you really claiming that a 5-point chunk of voters identifying as Republicans are defecting to Obama?

But the only thing media polls do is provide fodder for punditry, and there's no penalty for pundits who get things wrong.

Joe, it's not clear to me exactly what you're even getting at here. Lose 5 points among independents since....2008?

No, I'm claiming that it's quite likely a number of voters identifying as Independents were identifying as Republicans before. Or that a lot of Republicans died in the last two years. Or other things. I don't really care. Looking at one poll is silly. The broad polling average is pretty clear.

I do not have a strong idea if self-deportation was meant as a "dog whistle" or not, however, isn't a dog whistle one in which you expect/want the people on your side - the ones you want to vote for you to hear?

Exit polls in 2010 showed a 38/38 GOP/Dem split.

But the only thing media polls do is provide fodder for punditry, and there's no penalty for pundits who get things wrong.

If you honestly believe that Dems will have more than the D+7 advantage they enjoyed in 2008, then you're certainly entitled to that opinion. I happen to disagree.

I once read an intriguing counterfactual exercise in which Ford starts to say this, then stops brain farting and catches himself, and completes the answer instead in normal political platitude. Instead of this becoming one of the all-time debate bloopers and hurting Ford in the polls, he muddles through and wins instead of loses the close election of 1976.

A highly plausible scenario.

OK, then how would history have changed? The exercise I read kind of put the pedal to the metal and imagined all sorts of interesting consequences. Instead of becoming the nominee against the weak incumbent Carter in 1980, Ronald Reagan is forced to be a good Republican soldier and support a generally popular incumbent Gerald Ford in his re-election bid in 1980.

Typically one does expect a dog whistle to only be "heard" by those in the know, but immigration in general has two groups that are very tuned into this sort of thing. The Sheriff Joe's of the world and the immigrant community both keep track and both are tuned into the language used.

The reason it is a dog whistle is shown on this thread, it sounds so innocent, but in reality it is all about making like as horrible, as miserable as possible. Oddly enough this process of making life hell for illegals does collateral damage to their friends and families, many of whom are citizens.

No, I'm claiming that it's quite likely a number of voters identifying as Independents were identifying as Republicans before.

Zogby's fallen off the map after spectacularly failing in 2008.

I love this logic. The view of polls that they should be weighted by party ID is just orthogonal to reality. Polls take a random sample and then weight by factors, but almost always by demographics. You would do better ignoring the party affiliation numbers. But you won't, and everyone needs hope.

IIRC at one point during the the 1960 campaign, William F. Buckley** suggested a GOP draft of Eisenhower for Vice President, presumably to help keep the Republicans in office by utilizing Ike's popularity.

The idea that the best way to count the numbers of party affiliated by taking multiple random samples where they ask "what's your party affiliation?" of the electorate, by multiple companies and comparing their results seems like a pretty logical way to go about it.

This is actually quite simple. When the public id is leaning Dem, then left-leaning "independents" will identify as Democrats while right leaning independents will identify as "independents." That's how you see a +margin on the Dem side without seeing any real movement from self-identifying "independents." The indies who are moving out of the column into the D column are being replaced by the group from the right moving out of the R column into the I column.

The polls are showing a +margin for Dem IDs because more people are comfortable ID'ing as a Dem than a Rep these days, much to Joe's consternation.

So, by your logic, if you call 1,000 people and 900 of them happen to be registered Dems, there are no problems with the sample. Bizarre.

Zogby was one of the most-cited polling outfits not very long ago, but is now something of an afterthought. The idea that nearly all of the polling companies would suddenly, simultaneously and deliberately sabotage their own businesses and risk joining Zogby on the sidelines is one of the most nitwitted claims of the campaign.

If I call 1,000 people that, considered by demographic, are a fair sampling of the country, and 900 of them are registered Dems, I conclude that there are more Dems than Reps. (Repubs? Whatever.) I don't take my polling data and replace it with my previous assumptions about parity.

And to amplify this a bit, if I have 7 or 8 different companies do this, and the results come back roughly similar, I can actually begin to quantify the difference between the number of party affiliated with greater accuracy.

Also, I'm not so sure that pollsters have any incentive to go rogue, so to speak. It seems like the upside for being an outlier and being right is much less than the downside from being an outlier and being wrong.

Make their lives as miserable as possible how? By asking employers to verify Social Security numbers against a federal database?

So, by your logic, if you call 1,000 people and 900 of them happen to be registered Dems, there are no problems with the sample. Bizarre.

That is quite a straw man you have up there.

First of all people who identify as Democrats is very different than party registration. Party identification is fairly fluid and often changes based on who the person is planning on voting for in the upcoming election. As stated upthread, Obama doing better means more people will self identify as "Democrat", this has zero to do with actual party registration (for some bizarre reason this distinction keeps getting lost when you write - repeat after me identifying to a pollster is not the same as registering with a political party).

Political party registration is fairly stable, with long term trend lines (in general) moving towards more independents. However for most of those that leave from the left, they still vote left and most who leave from the right to become independents still vote right.

Of course if we follow what you believe (as stated) every pollster takes a random sample. We know they weigh by demographics (the vast majority at any rate) and then release their figures. They (the majority) don't weigh at all by self identification of party. But still they are all getting way more D than you think is right.

Poll after poll, survey after survey. Randomly. They keep getting too many D. Why do you think that is? They are not weighing to get that proportion. Is it random chance? Have the demographics turned that heavily against Team Red? Are they all conspiring against Romney?

Why do you think they keep coming up with more D than R? My reason, the one the actual profession pollsters say, I gave above. What is your expert opinion?

Isn't it obvious? The Republicans are all at work while the lazy Dems sit at home and talk to pollsters.

So party ID is "fairly fluid" while party registration is "fairly stable," and most people vote for the same party in every election. I sense a conflict or two in there.

Wouldn't inaccurate polls cause the outcomes to vary rather than generally agree?

One thing to remember is the demographics of the nation continue to change, with groups favoring the D side growing faster than the R side. Of course they vote less proportionally and skew younger (which vote less as well), but as they age the migrate to a higher voting cohort. But in raw numbers the demographics will look unfavorable to the GOP going forward unless the demographic identity politics changes

Of course, as people get older, they vote more often and vote more conservatively.

Can't we just skip ahead to what extent President re-elect Obama was the beneficiary of massive ACORN voter fraud and librul media bias?

I'm so bored with the biased polls discussion...

Wednesday’s presidential debate promises sharp contrasts. One candidate wants to repeal Obamacare, one candidate invented it. One opposed the auto industry bailout, one takes credit for it. One doubts the scientific consensus about climate change, one believes in it. One wants to “voucherize” Medicare, one wants to save it. One dismisses nearly half of Americans as a bunch of moochers, and one claims to champion the struggling middle class.

It promises to be an epic clash: Mitt Romney vs. Mitt Romney. Oh, and President Obama will be there, too.

I'm so bored with the biased polls discussion...

It's what losers masturbate over.

Sigh... it's like people don't even care about all the great porn the internet has to offer.

Sigh... it's like people don't even care about all the great porn the internet has to offer.

I have read this is not true, but it has been established wisdom for a long time

No, I'm claiming that it's quite likely a number of voters identifying as Independents were identifying as Republicans before.

World's largest circle jerk

Nice to have company.

I have read this is not true, but it has been established wisdom for a long time



I understand that party affiliation doesn't change, but voting does. Essentially, once you pick your razor you use it for life, but behind the curtain you're not the same voter.

World's largest circle jerk

Nice to have company.

I bet that pole is screwed too!

Can you say that on the internet?

I understand that party affiliation doesn't change, but voting does. Essentially, once you pick your razor you use it for life, but behind the curtain you're not the same voter.

Politically, I love her, since she uses blunt language to describe exactly what the Republican agenda has been, which is to strip consumer protection regulations in order to cement the power of corporations over individual citizens. Few spokespeople have been as articulate as she has in defining the choice between the two parties, and the proof of her effectiveness is the degree to which Republican hatred has been directed against her. If I were in Massachusetts this year, she would definitely get my vote.

Politically, I love her, since she uses blunt language to describe exactly what the Republican agenda has been, which is to strip consumer protection regulations in order to cement the power of corporations over individual citizens. Few spokespeople have been as articulate as she has in defining the choice between the two parties, and the proof of her effectiveness is the degree to which Republican hatred has been directed against her. If I were in Massachusetts this year, she would definitely get my vote.

It is always open season on media bias, though I think that even more boring than poll bias. Poll bias at least there is the election and you can compare numbers and see if there was bias or not.

but other than that there's nothing there . . .

Is that really the Republican agenda?

Sigh... it's like people don't even care about all the great porn the internet has to offer.

And yet, those same polls show Romney leading among both Republicans and independents.

Obama won independents big in 2008. If the D advantage is only +1 from 2008, as in today's CNN poll, then it's quite odd that Obama could now be losing among independents but still be winning by 3 points. The only way the math works out is if there's a large number of GOP defectors, which seems highly unlikely.

At this point, the only interesting question to me is how the fever swamps will spin the loss, and how long it will take them to gin up enough "outrage" at whatever it is they focus on to attempt another coup-de-impeachment (assuming they hold the House.)

3: You area ware that the 1996 electorate looked much morelike the 1992 electorate than the 1994 electorate?

Is that really the Republican agenda? I mean, I can think of lots of unpleasant things the contemporary Republican party has accomplished, and "strip consumer protection" is not really on the list.

(Is it even technically possible to impeach a second term president for something that happened in his first term?

Of course, as people get older, they vote more often and vote more conservatively.

I have read this is not true, but it has been established wisdom for a long time.

It doesn't have anything to do with what I think is "right"; it's about the reported numbers in prior elections. The Dems had a Dem+7 advantage in 2008 and they won big. I find it hard to believe the Dems will have a bigger party ID advantage in 2012 but win by a smaller margin than in 2008. If the theory is that people identify with the winner, such a result would be entirely counterintuitive.

You do realize of course, this election is now gonna go down as a polling fiasco like 1948, Romney's gonna win, the pollsters will all have egg on their faces, and JoeK will be insufferable.

I can see this as an alternative description of her from a conservative's POV, but since her evidence of "bad character" pretty much rests on that questionable ancestry claim, that's not enough for me to shove her aside, given the alternative.

2004 was brutal and I admit I bought some (though not all) of the spin.

I just made my first political donation of the 2012 cycle - and I didn't so much donate TO Scott Brown as AGAINST the utterly odious Elizabeth Warren. I don't get why Warren has popularity on the left; she seems to me to be a really nasty personality and totally unlike an Obama or even a Bernie Sanders or Barney Frank type where you might disagree with the ideology but you respect the person as fundamentally decent and well-meaning.

and JoeK will be insufferable.

At least I think Liberals can see the humor in that now...

Scott Brown is a bad person too, in fact an overwhelming % of our elected pols are people most normal people would not want to eat lunch with