Since the RNC and DNC ended a plethora of new polls have come out, nationally and in many swing states, showing Obama well ahead of Mitt Romney.  A few recent examples come to mind.  A new WashPo poll out of VA found Obama up eight points over Romney.  In Iowa and Nevada PPP (D) has found Obama leading by 5-10 points.  The latest batch of polls from CBS/NYT/Quinnipiac find Obama ahead by 10 in Ohio, 8 in Florida and 14 in Pennsylvania.  So is Obama really ahead by this much or does it have more to do with the polls themselves?

First off, before I address this question it does bear mentioning that I state that polling is an inexact science.  Especially when we consider we are still a month in a half from the election.  Most pollsters are largely guessing about what turnout will be.  Still……

As Weekly Standard analyst Jay Cost notes there is a distinct different in poll results depending on the sample.  In his article he creates some nice simple charts for us to use.  To paraphrase those charts and no big shocker to the informed, the polls that have larger Democratic samples, whether through weighting or random chance, show larger Obama leads.

Using data from the 2004 and 2008 election exit polls and comparing it to the partisan composition of certain polls provides some enlightenment.  In Florida, recent polls from PPP (D), Gravis and Fox all show a partisan composition greater than or equal to 2008 partisan composition turnout.  By contrast, Rasmussen, WashPo and Purple Poll all have samples closer to 2004 (less Democrats).  Now consider Ohio.  Partisan turnout in Ohio vastly favored Democrats (39D-31R) so only Gravis has a poll showing greater partisan turnout for Democrats than indicated by 2008 exit polls.  But WashPo and Fox expect turnout to be almost equal to 2008.  Only Purple Poll and Rasmussen expect it to be closer to 2004.

There is also something else to consider.  In the last few months both Gallup and Rasmussen have been criticized for their samples.  More specifically, on the racial composition of their samples.  Both Gallup and Rasmussen show a tight race nationally in their tracking polls but also in their swing state polls.  These results conflict with other pollsters results.

So in short we are left with a bimodal distribution of poll results.  One mode has Obama well ahead of Romney and the other shows a tight race.  Which one is right I cannot say but I would say the odds are better this election resembles 2004, or even 2000 turnout than 2008.

Keep in this mind it is now easy to see why Obama seems to have a such a strong lead.  His lead is essentially built on the fact that his base is expected to turnout out at 2008 or even greater levels while the GOP base drops from 2008.  Contradicting this view however are polls that show Republicans are more enthusiastic to vote, whether against Obama or for Romney.

Two other thoughts and to be fair Cost gave me the idea for them.  No incumbent president (at least in my lifetime), has ever won an election with less than 45% of the independent vote.  According to Gallup tracking and some national surveys Obama does not get this number.  Gallup daily tracking has his approval numbers among this group right at 45%.  So if Obama is not winning independents his lead must be built on pollsters expectations for partisan turnout favoring Democrats.

Lastly, this is something I have noticed before but did not pay much attention to.  President Obama is holding his coalition together better than Mitt Romney.  Considering all the hype the media has made about the Tea party and its far right conservatism it is revealing polls show Obama doing better among his base.

Historically it has been Republicans who do better among their base.   Consider Cost’s chart that shows Republicans doing better n the past.  If the historical average occurs this election that effectively gives Romney a leg up of about a 1-1.5% advantage.  But currently polls show Obama doing better and that is just extremely hard to believe.

Last point, the sampling models are also swinging Senate races in the Democrats direction.  Senate Republicans or candidates once thought to be competitive or favored are now either tied or trailing their Democratic counterparts.  For GOP campaigns this spells trouble in motivating their base and raising money.

So I cannot say one way or the other if the polls are skewed but they seem to reflect turnout models that are even better for Democrats than 2008.  Call me a betting man, but this election I don’t see that occurring.


Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s