Making the Case Against Polls Over-Sampling Democrats — Today’s Must Read

Time after time we see polls with funny results invariably in favor of President Obama and get our panties in a bunch that the results have no value while Obama advocates masquerading as journalists trumpet the results as confirming the inevitability of Obama winning this election .

But what is our reason for holding on to this belief when poll after poll says we are losing, even if ever-so-slightly? Jay Cost smartly made the quantitative case. Today I am going to make the substantive case.

First is the set-up: Barack Obama’s 2008 performance.  As we have outlined countless times, in November 2008 the Democrat turnout machine had a banner day when actual voters at the polling booths out numbered Republicans by 7 percentage points (note other exits have it as high as 8-points).  The representative electorate self-identified as the following: Democrats: 39%, Republicans: 32%, Independents: 29% or D +7. Very impressive following 2004 when the split between the parties was dead even (37 to 37).

In 2008, ahead of the election, Rasmussen Reports, who the Left loves to decry as partisan, had party affiliation at Dems +7, Dem: 41, Rep 34, Ind 25 — just as it was on election day. Others showed similar but greater Democrat advantages than what showed up on election day. For example, the generic ballot according to Pew Research showed Democrats with a +10 point advantage 38 to 28 . The generic ballot from Gallup had Democrats +12, 53 to 41.

There were many factors that gave Democrats such an advantage: historic viable candidate (Obama), war fatigue, financial meltdown, divisive incumbent, and inept opponent (McCain — good biography, bad candidate) all factored into a strong Democrat performance in 2008. Today none of those advantages are there for Obama and many of those are now largely working against Obama: 8%+ unemployment, sub-2% GDP, 23 million unemployed, Arab Spring blowing up, historic vote is yesterday’s news.

All of this adds up to a steep change in his own base’s enthusiasm which means Obama’s huge advantages at the voting booth in 2008 will be greatly diminished in 2012, the only question is by how much.  This is why President Obama has spent his entire campaign and Convention on hot-button cultural issues to fire up his base (Quite honestly can anyone outline his 2nd term agenda?).  His coalition for victory in 2008 relied heavily on voters who typically do not show up on election day like they did in 2008. The problem today is that survey after survey comes to the same conclusion:”Democratic Voting Enthusiasm Down Sharply From 2004, 2008.” This is why I title some blog posts: “What happens if Obama’s voters don’t show up?” The simple truth is that the pendulum of party affiliation has swung back from the strong showing in 2008 to a more balanced electorate in 2012.  This is what you see in the weekly generic tracking poll from Rasmussen Reports where either party has a 1-2 point advantage  depending on the week.  President Obama got a bump from his Convention so a 2-point advantage for Republicans shifted to a 2-point advantage for Democrats last week (it’s first lead since January).  The overall trend is unchanged — a very closely divided electorate.

This closely divided electorate is in no way shape or form represented in the relentless drumbeat of polls we get today from NBC, CNN, Fox News, Quinnipiac, ABC, The Washington Post, The Wall Street Journal, Marist, etc.

But what happens when you re-weight polls to reflect the last election when all the current data tells you voter preference has sharply changed?  You get stories like this one in 2010:

NEWSWEEK Poll: Democrats May Not Be Headed for Midterm Bloodbath

Obama’s approval continues to slide, but Bush’s legacy still haunts the GOP.

As Democrats prepare for considerable losses in the November elections, there’s reason to believe the party in power may not be headed for the bloodbath it might expect. According to a new NEWSWEEK Poll, President Obama’s approval rating—47 percent—indicates that the party is better off this year than Republicans were in 2006, when the GOP lost 30 House seats, and than the Democrats were in 1994, when they lost 52 House seats. Obama’s approval has fallen 1 percentage point since the last NEWSWEEK survey in June, but the White House has gained ground on several specific issues, specifically his handling of the economy, which has risen to 40 percent (from 38 percent) over the past two months.

Newsweek found a  45-45 tie  for Congressional preference ahead of the 2010 mid-terms. But when you look at the party vote breakdown, Democrats are winning 90% of the Democratic vote and Republicans are winning 94% of the Republican vote BUT Independents favor Republicans by 12 points: 45 to 33 – sound familiar?  It’s impossible to have a tied race when both sides lock up their bases and one side is winning Independents decisively.  So Newsweek re-weighted this poll to reflect the 2008 party identification for Democrats with a 7-point advantage to create a tied race ahead of what became a historic bloodbath at the Federal and State level the likes of which Republicans had never seen before.

This is exactly what is happening today.

In yesterday’s NBC/WSJ/Marist poll Romney led with Independents in both Ohio and Virginia but was losing decisively in both polls. We see this in nearly every state poll.  In the ABC/Washington Post national poll on September 11 Romney led Independents by 11-points but by oversampling Democrats they made the race virtually tied. In the CNN/ORC September 10 poll Romney led among Independents by 14-points but was losing the top-line result by 6-points. These are not surveys of the electorate, they are advocacy pieces pushing an agenda.

Te reality is very simple.  Major blocs of Obama voters are less enthusiastic and increasingly unlikely to show up at the polls in anywhere near the fashion they did in 2008.  This is true for Hispanics and the youth vote while the African-American vote has largely returned to historic voting patterns (90% support for Democrats). At the same time Republicans are greatly enthusiastic about voting in no small part due to the divisiveness of President Obama. The generic Congressional ballots and party affiliation surveys reflect a near 50/50 divided electorate between the parties more reminiscent of 2004 than 2008. Today’s polling, however, for whatever their stated reasons may be continue to report polls more representative of the 2008 electorate than anyone can reasonably argue will be the case. All the “micro-targeting” in the world won’t make up for a 5-6 percentage point shift in the electorate.

In 2008 you had a greatly fired up Democrat Party show up at the polls in droves while Republicans were dejected over a disappointing candidate and horrible confluence of events in the Fall of 2008.  Today Barack Obama has to own the failed economy, the dissatisfaction with his signature achievement Obamacare and meet a rejuvenated Republican Party.  The electorate in November 2012 will look dramatically different than current polling indicates which will leave more than a few “independent” journalists wiping away tears with printouts of these worthless polls.

2 Comments

  1. bobdevo
    Posted September 14, 2012 at 2:57 pm | Permalink | Reply

    It’s pretty clear from all the polling numbers that absent some earth-shattering event, this election is over. Romney’s pathetic showing at the convention and his ill-advised whining about Libya have destroyed any possible credibility with moderates or independents. Obama has numer paths to 270 electoral none, at present, Romney has no viable route based on poll numbers.

    • Posted September 14, 2012 at 4:05 pm | Permalink | Reply

      Thanks for visiting the blog and not reading any word I wrote. If Obama loses this election and you find yourself crying in your cereal about “how could he lose when ever poll said he was winning?”, don’t say you weren’t warned.

One Trackback

  1. […] went to great lengths to debunk the over-sampling of Democrats in polling used across polling outfits.  As a brief reminder, they use the prior election as their […]

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: