Home > General > Poll Insider Method Objection, And Happy Memorial Weekend

Poll Insider Method Objection, And Happy Memorial Weekend

This will be my last update for the weekend but wanted to lay something out there. I have received a number of emails and comments questioning my using of 2010 figures to recalculate PPP polling data. One such comment came from Brandon on my North Carolina post that read like this:  “Why are you assuming that 2012 turnout will be like 2010? Midterm electorates are COMPLETELY different than presidential elections. Even as a Republican, I know that’s a HORRIBLE assumption that you’re making.

The comment that the electorate for President and Midterm elections are completely different is simply not true. Turnout models actually go in heavy trends over a number of election cycles, usually two, and not based on the type of election it is. For instance, Republicans fared well in turnout in both the 2004 Presidential elections as well as in the 2002 midterm election. The mood was considerably Republican in those 2 election cycles, and for that matter in the 2000 election cycle as well, making the trend 3 cycles long. The mood switched dramatically in off-year 2006 and continued into the Presidential election of 2008. So in reality, the 2002 midterm and 2004 Presidential election were very similar in turnout that favored Republicans. The 2006 midterm and the 2008 Presidential election were very similar in turnout favoring Democrats. The reality is the electorate in 2002 and 2004 were very similar, and the electorate in 2006 and 2008 were very similar. The mood switched big-time after 2008 and, as trends would have it, the swing went back to Republicans in 2010. As Obama’s ratings have remained in the low-mid 40s, and he is double-digit underwater on his handling of everything, there is little reason to assume that the electoarte would suddenly convert back to 2006/2008 instead of 2010, which is the typical trend. The Presidential electorate in 2008 was considerably more similar to the midterm electorate of 2006 than it was to the Presidential election of 2004, which completely disproves the point that argument makes.

Regardless, I always point out how poor the Public Policy Polling samples are by quoting BOTH 2008 and 2010 turnout figures, to show that PPP oversamples Democrats by 20% when even compared to 2008 figures. I put all of my analysis out for everyone to see, and to agree or disagree with. For instance, in my reanalysis of North Carolina, PPP had a Democrat sample that was 48%. In the unlikely to be repeated 2008, super-heavy-Democrat turnout year, turnout among Democrats was only 42%. In the Republican Rebound in 2010, Democrat turnout tumbled to 37%. Either way, PPP oversampled Democrats by, at best, 6% and at worst 11%. Need more proof? When analyzing North Carolina, Pennsylvania, and Ohio we find that the electorate almost completely reverted back to the 2004, pro-Republican electorate. In 2004 North Carolina, Democrat turnout was 39%. Similar stories for PPP turnout in other Purple states such as Ohio and Pennsylvania. Take a look:

PPP   Democrat Sample:                                    48% in NC/ 45% in OH/51% in PA

2004 Presidential Democrat Turnout:       39% in NC/ 35% in Ohio/ 41% in PA

2006 Midterm  Democrat Turnout:            NA in NC/ 40% in Ohio/ 43% in PA

2008 Presidential Democrat Turnout:       42% in NC/ 39% in OH/44% in PA

2010 Midterm Democrat Turnout:              37% in NC/ 36% in OH/ 40% in PA

As you can see, In the two strong years, Democrat turnout was nearly identical (and note that one was a midterm and one was a Pres election). In 2010, as Democrat enthusiasm has waned and Bush Derangement Syndrome has dissipated, the turnout figures reverted back to those of the early 2000’s.

Most states I’ve looked t show similar results. PPP heavily oversamples Democrats compared to any of 2004, 2006, 2008 OR 2010.  I’ve considered running numbers of both 2008 AND 2010, but quite frankly I do not have the time. If I ran the numbers using 2008 figures, yes Obama would favor slightly better than the figures I use, but still far worse than whatever poll that PPP comes out with on any random day. My guess is that the turnout for both D’s and R’s will be somewhere between 2008 and 2010. And while I could average the two out (so, I could uses a NC turnout model of 39.5% theoretically) I prefer to use the most current trend and data that’s available, and want to avoid as much interference in the figures as possible.

So why do I use 2010 numbers? For unbiased purposes, I use whatever data is most current, whether that be 2010, 2008, or 2006.

I would assume that turnout is likely to be more similar to 2010 for the following reasons:

1) This would follow the typical 2 year trends seen in the past. My prediction, of course, is not that the Republican will win big. Obviously Obama won quite handily in 2008. Most of the poll numbers I show reveal that a landslide is not in the cards the next time around, however, and we will probably return to the closer elections of 2000 and 2004.

2) Republicans/Tea Party candidates are eager to throw Obama out. Now that he has a record, they are more eager. If anything, Obama as an enthusiasm deficit compared to 2008.

3) Hopey-changey was a one-time thing. Other than black voters, Obama is down in excitement levels among all groups, especially the 18-35 crowd that he needs big time once again. Enthusiasm in this segment has been decimated.

4) No more Blame Bush boogey-man.

5) Republican self-ID has surpassed Democrat self-ID and has remained that way for awhile. In 2008 Dems were way ahead of Republicans.

6) Republicans still ahead on congressional generic ballots, just as they were in 2010. Democrats were way ahead on the generic ballots in 2006 and 2008, years where they won big.

7) 4 more years of… what? Obama made a lot of promises, but other than shoving an unpopular healthcare bill down voters throats, what can Obama point to as an accomplishment? Soaring deficits? High unemployment that was never suppose to pass 8%, yet has remained well above that for over 2 years? Escalating crisis in the housing market? Stagnant economy? Higher taxes? Crony capitalism? He’s been unable to keep his promises to liberals of quick withdrawal in Iraq, has escalated the war in Afghanistan, and started a new one (and bumbled it) in Libya. Oh, and he never closed that wretched Gitmo.

So, the one thing I do know, is that PPP oversamples Democrats badly. Really, really badly. Compare 2008, compare 2010… doesn’t matter. In North Carolina, PPP says Obama beats Palin by 12. If I recalculate the numbers to 2010 turnout levels, which is a reasonable expectation, it is a 2 point race. If I recalculate them to 2008, it’s probably a 5 point race. Either way, it’s not the blowout PPP portrays (and the media willingly swallows) and it’s nowhere near what Obama did in 2008.

Advertisements
Categories: General
  1. May 28, 2011 at 12:07 pm

    THANK YOU !! Even though we know that polls are skewed by the Left, and we guard against it, at times they can be demoralizing. I have a small, private, Sarah Palin supporters site (180 members) that I’ll be posting this at with full credit, of course.

    Everything that you say reinforces what I constantly preach…Obama will be a one term President. I know you haven’t stated that…I am.

    BTW, we would be honored to invite you to join the site if you’d be interested.

  2. serfer62
    May 28, 2011 at 1:30 pm

    It was immediately apparent that OHbama was a one term TOTUS and he ekpt reenforcing that…the MFM can’t cover all the poop

  3. ayamo
    May 28, 2011 at 2:25 pm

    Very interesting to read about the “two year turnout” thing. Thanks. 🙂

  4. Franklin
    May 28, 2011 at 6:31 pm

    Not too long ago Rasmussen brought out his presidential poll and PPP, ABC/Washington Post and Marist came out with polls about the same time frame. They were within the margin of error for Romney, Gingrich, and Trump. Huckabee’s numbers were on the outer edges of some of the polls. Yet Palin’s numbers were completely outside the margin of error. Rasmussen had Palin -10 to Obama yet the other polls had her down 18-22 points. Given Rasmussen uses the largest sample of likely voters, I tend to favor Rasmussen. Makes me wonder if there isn’t some push polling going on with these other polls.

  5. Brandon
    May 29, 2011 at 11:55 am

    Thank you for this detailed post!

  6. Ryan.Iowa
    May 29, 2011 at 8:12 pm

    I’ve always maintained that the 2012 election will follow the shifts in the electorate that began to set in during the summer of 2009, and was ratified by the off-year elections in 2009 in New Jersey and Virginia, and the Massachusetts Senate special election and the mid-term elections of 2010. The trend lines are clearly visible, anyone who denies it is being very intellectually dishonest.

  1. June 13, 2011 at 8:42 am

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: