PPP/DailyKos weekly poll: Obama+5 nationally (likely voters)

[Tender Branson]

Daily Kos/SEIU Weekly State of the Nation Poll:

50-45 Obama

Public Policy Polling, 1200 likely voters, MoE ±2.8%, September 20, 2012 - September 23, 2012.

http://elections.dailykos.com/weeklypolling/2012/9/20

[MorningInAmerica]

Last week Obama led 50-46% (he had led by 50-44% margin since the week of the Republican convention). Party ID of the likely voters is D+7 (the same as 2008 turnout). So if the November electorate looks as Democratic as the 2008 electorate, then Romney trails by about 5 points according to this poll. McCain lost in the 2008 electorate by 7. Romney probably needs to trail in an Obama electorate by 3 or less in order to actually win in November (presuming a less Obama friendly electorate in November).

[Craigo]

Last week Obama led 50-44% (he had led by that margin since the week of the Republican convention). So this is the first movement in Romney's favor (if you can call it that, w/in MoE) on the Daily Kos tracking poll since the end of August. Not to mention that the party ID of the likely voters is D+7 (the same as 2008 turnout). So if the November electorate looks as Democratic as the 2008 electorate, then Romney trails by about 5 points according to this poll. McCain lost in the 2008 electorate by 7. Romney probably needs to trail in an Obama electorate by 3 or less in order to actually win in November (presuming a less Obama friendly electorate in November).

See rule #7. Or visit unskewedpolls.com, they'll tell you what you obviously want to hear.

Also, a one point shift in a poll with a margin of error of 2.8% is not "movement."

[MorningInAmerica]

Last week Obama led 50-44% (he had led by that margin since the week of the Republican convention). So this is the first movement in Romney's favor (if you can call it that, w/in MoE) on the Daily Kos tracking poll since the end of August. Not to mention that the party ID of the likely voters is D+7 (the same as 2008 turnout). So if the November electorate looks as Democratic as the 2008 electorate, then Romney trails by about 5 points according to this poll. McCain lost in the 2008 electorate by 7. Romney probably needs to trail in an Obama electorate by 3 or less in order to actually win in November (presuming a less Obama friendly electorate in November).

See rule #7. Or visit unskewedpolls.com, they'll tell you what you obviously want to hear.

Also, a one point shift in a poll with a margin of error of 2.8% is not "movement."

I see you missed this part, or chose not to read it:

You must be logged in to read this quote.

2nd, who said anything about this poll being skewed? Instead of projecting your own thoughts on my comments, read what I actually said. In a 2008 styled electorate (which is what a D+7 electorate is), Romney trails by 5. What exactly is there to argue about that? What exactly suggests the poll is skewed, based on that? Stop projecting.

Edited to note that last week, when obama led 50-46%, party ID was D+6. This week, Obama leads 50-45%, and party ID is D+7.

[Craigo]

You asked "if you can call it that." The answer is: No, you can not. Trying to cover one's ass with "It is irresponsible not to speculate" does not make the speculation any less foolish.

You have repeatedly demonstrated that you do not understand polling, particularly party ID.

1) It is not a demographic characteristic
2) It is a dependent variable of voter intent
3) The relationship is weak

There aren't "more" or "less" Democratic of Republican electorates, since party identification is an attitude, not a characteristic. Party ID breakdowns reflect how the sample intends to vote - most respondents decide what party they identify as by deciding what candidates they support, not the other way around. And even those two variables don't correlate very well.

There's a reason why professionals don't worry about party ID.

Bonus question: Which election had a 38D-35R party split?

A. 1984
B. 1988
C. 1992

[MorningInAmerica]

You asked "if you can call it that." The answer is: No, you can not. Trying to cover one's ass with "It is irresponsible not to speculate" does not make the speculation any less foolish.

You have repeatedly demonstrated that you do not understand polling, particularly party ID.

1) It is not a demographic characteristic
2) It is a dependent variable of voter intent
3) The relationship is weak

There aren't "more" or "less" Democratic of Republican electorates, since party identification is an attitude, not a characteristic. Party ID breakdowns reflect how the sample intends to vote - it does not drive it. And even those two variables don't correlate very well.

Bonus question: Which election had a 38D-35R party split?

A. 1984
B. 1988
C. 1992

Listen, there's no need to act like the jedi of poll interpretation. The answer to your question is 1988 according to the Washington post http://www.washingtonpost.com/wp-srv/politics/interactives/independents/data-party-identification.html, but comparing 1980s party identification (or even party registration) to today is pointless. We don't have a consistent measure of party ID/registration across the country, but the data we do have tells us there were a lot more people identifying as Democrats in the 1960s and 1970s than today. Even  in 1984, when Regan won by nearly 20 points, party ID in the exit polling was D+2.

Looking at the last several elections, you can certainly see a correlation between party ID in the exits and the results. In 2002, when Republicans had a great year, party ID was near even, or R+1. In 2004, when Republicans won their first MAJORITY of the popular vote since 1988, party ID was even. In 2006, when Dems had a great year, turnout was D+3. In 2008, when Dems had an even better year, it was D+7. In 2010, when Reps had a great year, party ID was back at parity again. So you can see, as party ID approaches parity, the results bode well for the GOP. As they approach D+7 (2008) and above, the results imply landslide status for the Democratic party. I don't see what is so wrong with discussing this.

Your contention that there are not "more" or "less" Democratic or Republican electorates is an opinion, not a fact. You can read an exit poll and see that 39% of voters identifed as Democrat in 2008, and 32% identified as Republican. Political and poll analysts frequently cite the MORE Democratic electorate of 2008, often pointing specifically to the party ID in that years exit polling to make their point.

I'll repeat that there is nothing in my post that demonstrates I know nothing about polling, ole wise professor. The party ID of the poll is D+5 (a fact). Obama leads by 5 (a fact). The party ID of the 2008 electorate based on exit polling was D+7. A fact. Nothing about pointing out those facts implies you know nothing about polls. Stop being the "poll analysis" police.

And since you're so wise, please do tell us what we are allowed to discuss when looking at poll internals? You know, just so we can know what we're allowed to talk about around here.

[Lief 🗽]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

[MorningInAmerica]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

I don't think it's "ridiculous" to anticipate similar Dem turnout to 2008, I just don't personally think it's likely. Reasonable people can disagree on this. But yes, there was a time where I would have found it ridiculous to assume 2008 turnout. I don't find it "ridiculous" anymore.

[Sbane]

You asked "if you can call it that." The answer is: No, you can not. Trying to cover one's ass with "It is irresponsible not to speculate" does not make the speculation any less foolish.

You have repeatedly demonstrated that you do not understand polling, particularly party ID.

1) It is not a demographic characteristic
2) It is a dependent variable of voter intent
3) The relationship is weak

There aren't "more" or "less" Democratic of Republican electorates, since party identification is an attitude, not a characteristic. Party ID breakdowns reflect how the sample intends to vote - it does not drive it. And even those two variables don't correlate very well.

Bonus question: Which election had a 38D-35R party split?

A. 1984
B. 1988
C. 1992

Listen, there's no need to act like the jedi of poll interpretation. The answer to your question is 1988 according to the Washington post http://www.washingtonpost.com/wp-srv/politics/interactives/independents/data-party-identification.html, but comparing 1980s party identification (or even party registration) to today is pointless. We don't have a consistent measure of party ID/registration across the country, but the data we do have tells us there were a lot more people identifying as Democrats in the 1960s and 1970s than today. Even  in 1984, when Regan won by nearly 20 points, party ID in the exit polling was D+2.

Looking at the last several elections, you can certainly see a correlation between party ID in the exits and the results. In 2002, when Republicans had a great year, party ID was near even, or R+1. In 2004, when Republicans won their first MAJORITY of the popular vote since 1988, party ID was even. In 2006, when Dems had a great year, turnout was D+3. In 2008, when Dems had an even better year, it was D+7. In 2010, when Reps had a great year, party ID was back at parity again. So you can see, as party ID approaches parity, the results bode well for the GOP. As they approach D+7 (2008) and above, the results imply landslide status for the Democratic party. I don't see what is so wrong with discussing this.

Your contention that there are not "more" or "less" Democratic or Republican electorates is an opinion, not a fact. You can read an exit poll and see that 39% of voters identifed as Democrat in 2008, and 32% identified as Republican. Political and poll analysts frequently cite the MORE Democratic electorate of 2008, often pointing specifically to the party ID in that years exit polling to make their point.

I'll repeat that there is nothing in my post that demonstrates I know nothing about polling, ole wise professor. The party ID of the poll is D+5 (a fact). Obama leads by 5 (a fact). The party ID of the 2008 electorate based on exit polling was D+7. A fact. Nothing about pointing out those facts implies you know nothing about polls. Stop being the "poll analysis" police.

And since you're so wise, please do tell us what we are allowed to discuss when looking at poll internals? You know, just so we can know what we're allowed to talk about around here.

Your analysis itself proves that party ID changes based on who people are voting for. So if Obama is up by about 5-6 points, it's very likely the people who turn out will identify as Democrats by about the same margin as 2008.

[J. J.]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

Ah, where do you get that?

I know that, at this point, there were both stronger registration and campaign activities for Obama in Phila in 2008.  I've been surprised how quiet the voter registration office was last week.

PPP is in line with Gallup.

[Umengus]

a more realist sample (36-33-31 for exemple) and it's Obama by 2,5 %.

[ChrisFromNJ]

You asked "if you can call it that." The answer is: No, you can not. Trying to cover one's ass with "It is irresponsible not to speculate" does not make the speculation any less foolish.

You have repeatedly demonstrated that you do not understand polling, particularly party ID.

1) It is not a demographic characteristic
2) It is a dependent variable of voter intent
3) The relationship is weak

There aren't "more" or "less" Democratic of Republican electorates, since party identification is an attitude, not a characteristic. Party ID breakdowns reflect how the sample intends to vote - it does not drive it. And even those two variables don't correlate very well.

Bonus question: Which election had a 38D-35R party split?

A. 1984
B. 1988
C. 1992

Listen, there's no need to act like the jedi of poll interpretation. The answer to your question is 1988 according to the Washington post http://www.washingtonpost.com/wp-srv/politics/interactives/independents/data-party-identification.html, but comparing 1980s party identification (or even party registration) to today is pointless. We don't have a consistent measure of party ID/registration across the country, but the data we do have tells us there were a lot more people identifying as Democrats in the 1960s and 1970s than today. Even  in 1984, when Regan won by nearly 20 points, party ID in the exit polling was D+2.

Looking at the last several elections, you can certainly see a correlation between party ID in the exits and the results. In 2002, when Republicans had a great year, party ID was near even, or R+1. In 2004, when Republicans won their first MAJORITY of the popular vote since 1988, party ID was even. In 2006, when Dems had a great year, turnout was D+3. In 2008, when Dems had an even better year, it was D+7. In 2010, when Reps had a great year, party ID was back at parity again. So you can see, as party ID approaches parity, the results bode well for the GOP. As they approach D+7 (2008) and above, the results imply landslide status for the Democratic party. I don't see what is so wrong with discussing this.

Your contention that there are not "more" or "less" Democratic or Republican electorates is an opinion, not a fact. You can read an exit poll and see that 39% of voters identifed as Democrat in 2008, and 32% identified as Republican. Political and poll analysts frequently cite the MORE Democratic electorate of 2008, often pointing specifically to the party ID in that years exit polling to make their point.

I'll repeat that there is nothing in my post that demonstrates I know nothing about polling, ole wise professor. The party ID of the poll is D+5 (a fact). Obama leads by 5 (a fact). The party ID of the 2008 electorate based on exit polling was D+7. A fact. Nothing about pointing out those facts implies you know nothing about polls. Stop being the "poll analysis" police.

And since you're so wise, please do tell us what we are allowed to discuss when looking at poll internals? You know, just so we can know what we're allowed to talk about around here.

I think what you just showed us is that election results drive Party ID, not the other way around. As of right now, we do not know what the results of the 2012 election will be. However, we have numerous pollsters making projections that are very fluid, and not sticking to a static party ID that projects conventional thinking onto the electorate. I think if you say that "there is no way the electorate is going to be just as Democratic as it was in 2008", you are projecting your own feelings into the race. Based on live interviews in the field, pollsters may be seeing something different.

[Lief 🗽]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

Ah, where do you get that?

I know that, at this point, there were both stronger registration and campaign activities for Obama in Phila in 2008.  I've been surprised how quiet the voter registration office was last week.

PPP is in line with Gallup.

I get it from the numerous polls that show as much, a considerably better source than your anecdotal evidence from a solid Obama state.

[mondale84]

a more realist sample (36-33-31 for exemple) and it's Obama by 2,5 %.

Stop posting junk numbers you just pull out of your ass. You're worse than Ben Romney.

[King]

I wonder how far Romney supporters are going to take this. Will they question the exit polls if D+6 is reported?  That's where we get the idea that 2008 was D+7 so they must be accurate.

How many pollsters (none who weigh by a preset partisan ID) are going to sample 1000 random people before we get the D+3 likely voter figure that is supposedly the truth?

[J. J.]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

Ah, where do you get that?

I know that, at this point, there were both stronger registration and campaign activities for Obama in Phila in 2008.  I've been surprised how quiet the voter registration office was last week.

PPP is in line with Gallup.

I get it from the numerous polls that show as much, a considerably better source than your anecdotal evidence from a solid Obama state.

Well, the polls are not showing it.  I'm not seeing it on the ground. 

[King]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

Ah, where do you get that?

I know that, at this point, there were both stronger registration and campaign activities for Obama in Phila in 2008.  I've been surprised how quiet the voter registration office was last week.

PPP is in line with Gallup.

I get it from the numerous polls that show as much, a considerably better source than your anecdotal evidence from a solid Obama state.

Well, the polls are not showing it.  I'm not seeing it on the ground.  

Obama supporters are going to door to door in my area signing people up for voting by mail.  I never heard a word from Romney, even before he "pulled out of New Mexico."  I at least got mailings from McCain.

Also, what poll is there to show "Obama organization?"  If it's enthusiasm, I have news for you:



R+2 in June is now D+9 in September.

[Fmr President & Senator Polnut]

This is why the right is freaking out about the polling at the moment. They were relying on the economy to keep Obama's voters disinterested and disenchanted. What's clear is that since the DNC, the Democratic base is enthusiastic and has at worst closed the enthusiasm gap the GOP had been talking about all year as being the foundation of their victory.

Having Romney/Ryan as the opposition certainly hasn't hurt...

They simply cannot believe that the base is enthusiastic, let alone anywhere near 2008 levels... the fervor is gone, but that doesn't mean the energy has too.

[J. J.]

This is why the right is freaking out about the polling at the moment. They were relying on the economy to keep Obama's voters disinterested and disenchanted. What's clear is that since the DNC, the Democratic base is enthusiastic and has at worst closed the enthusiasm gap the GOP had been talking about all year as being the foundation of their victory.

They simply cannot believe that the base is enthusiastic, let alone anywhere near 2008 levels... the fervor is gone, but that doesn't mean the energy has too.

Because no one seeing it after the convention bounce.  Fundraising hasn't been there.  It's not showing up in strongly approved numbers.  Democratic registration seems like it is down (it is in FL and NC, where I've checked.)

[Fmr President & Senator Polnut]

This is why the right is freaking out about the polling at the moment. They were relying on the economy to keep Obama's voters disinterested and disenchanted. What's clear is that since the DNC, the Democratic base is enthusiastic and has at worst closed the enthusiasm gap the GOP had been talking about all year as being the foundation of their victory.

They simply cannot believe that the base is enthusiastic, let alone anywhere near 2008 levels... the fervor is gone, but that doesn't mean the energy has too.

Because no one seeing it after the convention bounce.  Fundraising hasn't been there.  It's not showing up in strongly approved numbers.  Democratic registration seems like it is down (it is in FL and NC, where I've checked.)

Well we haven't seen September fundraising yet, there's a very clear reason why registration in FL is down.

Strongly approve might suggest something, that's a fair point, but as I said, enthusiasm doesn't have to be dramatic enthusiasm, as I said, the fervor is gone. There are segments of Obama's base who have been disappointed by him, so they approve of him, but aren't head-over-heels in love with him like they were.

There are plenty of ways to prove the GOP no longer has the clear edge on enthusiasm they had 4 months ago. Of course, having a terrible GOP nominee helps. The Romney strategy was to hope people were so angry, they'd go for an alternative without really thinking about it.

[Lief 🗽]

Uh, we are seeing an enthusiasm gap in fundraising. Obama is out raising Romney with small donors 5 to 1 ($271 million to $58 million), similar to his ratio over McCain in 2008.

[Kalimantan]

The idea that it's somehow ridiculous to predict Democratic turnout similar to 2008 may have held some truth during the summer, but it doesn't anymore. Democratic enthusiasm is back up to 2008 levels and this election is looking more and more like a repeat of four years ago.

Ah, where do you get that?

I know that, at this point, there were both stronger registration and campaign activities for Obama in Phila in 2008.  I've been surprised how quiet the voter registration office was last week.

PPP is in line with Gallup.

I get it from the numerous polls that show as much, a considerably better source than your anecdotal evidence from a solid Obama state.

Well, the polls are not showing it.  I'm not seeing it on the ground. 

I'm sure the poster JJ has a rule about anecdotal evidence (yard signs etc.) not being important. 2nd rule maybe? We could ask him?

[J. J.]

This is why the right is freaking out about the polling at the moment. They were relying on the economy to keep Obama's voters disinterested and disenchanted. What's clear is that since the DNC, the Democratic base is enthusiastic and has at worst closed the enthusiasm gap the GOP had been talking about all year as being the foundation of their victory.

They simply cannot believe that the base is enthusiastic, let alone anywhere near 2008 levels... the fervor is gone, but that doesn't mean the energy has too.

Because no one seeing it after the convention bounce.  Fundraising hasn't been there.  It's not showing up in strongly approved numbers.  Democratic registration seems like it is down (it is in FL and NC, where I've checked.)

Well we haven't seen September fundraising yet, there's a very clear reason why registration in FL is down.

I was looking at it in NC as well, but in regard to early voting.  As of 9/22, R gained about 6,000 voters since 2008; Ind. gained about 12,000.  D lost 88,000.  The electorate looks less Democratic.

You must be logged in to read this quote.

I will agree that enthusiasm is down, overall, but the premise is that D/R turnout will be what it was, or very close to what it was, in 2008, is one I'm buying.

[pbrower2a]

This is why the right is freaking out about the polling at the moment. They were relying on the economy to keep Obama's voters disinterested and disenchanted. What's clear is that since the DNC, the Democratic base is enthusiastic and has at worst closed the enthusiasm gap the GOP had been talking about all year as being the foundation of their victory.

They simply cannot believe that the base is enthusiastic, let alone anywhere near 2008 levels... the fervor is gone, but that doesn't mean the energy has too.

Because no one seeing it after the convention bounce.  Fundraising hasn't been there.  It's not showing up in strongly approved numbers.  Democratic registration seems like it is down (it is in FL and NC, where I've checked.)

Democratic registration could be down where people must register as Republicans to vote in contested R primaries and can participate only in uncontested Democratic primaries if they remain registered Democrats. Many people register in the most interesting primaries; with the re-nomination of Barack Obama as a foregone conclusion this year, Democratic participation in the Democratic Presidential primary might be a wasted effort. But getting to choose between Gingrich, Huntsman, Bachmann, Santorum, Perry, and Romney? That implies plenty of choices.

In Michigan primaries one gets a shared ballot that allows some secrecy... but one can't vote for the Republican nominee for President but also the Democratic candidate for any down-ballot candidates (like county sheriff, let alone US Representative).

[Antonio the Sixth]

Uh, we are seeing an enthusiasm gap in fundraising. Obama is out raising Romney with small donors 5 to 1 ($271 million to $58 million), similar to his ratio over McCain in 2008.

Let's be fair: Romney isn't raising much money with small donors because he doesn't need to.