Tuesday, October 12, 2010

What to look for in poll results between now and Election Day

image At left is a chart showing the results of the Rasmussen Reports™ generic congressional ballot poll from January 2009 through October 12, 2010.

That’s an 8-point lead in the last monthly generic poll we’ll see from Rasmussen.  Gallup’s generic poll is showing a similar gap.

Traditionally, political pundits will tell you that the general public usually doesn’t start paying attention to elections until after Labor Day.  That is usually the case—but interest in this midterm election was high all summer.  So it’s likely that more of the general public have been paying attention than usual.  The second thing to keep in mind is that while the general public earnestly begins paying attention around Labor Day, solid opinions are usually developed by Columbus Day.  This puts the focus on the so-called "undecideds.”  The “late-breakers.”  These are the people that candidates and the parties will spend tens of millions trying to influence over the next three weeks.

History indicates that while spending all of that money is a necessity, when both sides are well funded, it has rarely shown to be effective in changing the electoral picture that is forming this week.  Here are five rules of thumb to keep in mind as you read all of the breathless news stories, blog posts and tweets.  Stories will abound regarding races tightening, leads evaporating and candidates fading.

  • In the first week of October, any candidate with greater than 50% among surveys of likely voters tend to end up winning the election.
  • In the first week of October, an incumbent with less than 50% among surveys of likely voters tends to be unseated.
  • Races showing support levels that are outside the surveys’ margins of error tend to tighten up between Columbus Day and the week before Election Day, only to break back to levels consistent with the early October estimates.
  • Late-breaking, undecided voters tend to break for challengers in tight races.
  • Late-breaking, undecided voters tend to break for the leader in tight races also.

There will undoubtedly be races that don’t conform to these rules—all politics is local, observed Tip O’Neill. Elections can turn on last minute gaffes, local voter turnout and voter base enthusiasm.  A leading challenger in this election is a strong bet to win the election, and this is why so many experts are calling this a “wave” election.

Ed Morrissey at Hot Air has a nice discussion of likely voter turnout models and what that portends for Democrats seeking to hold onto the House of Representatives and the Senate.  Conclusion: It’s not good for the incumbents:

Note that the composition of likely voters will not differ widely from the overall composition of the electorate.  Democrats have an advantage between 1-3 points in party affiliation in the general population, but that reverses to a two-point GOP advantage among those likely to vote in these midterms.  If that 3-5 point swing was the only difference in the electorate, Democrats wouldn’t face a wave of the magnitude we’re seeing.  The difference is that their radical agenda and profligate spending has thoroughly discredited them not with their base but with everybody else.

Related Posts:


Anonymous said...

Great! I was familiar with the incumbents-below-50% rule but not with the others. But, all of the rules make sense if you think about them.

Post a Comment

You must have a Google Account to post a comment.

WARNING: Posting on this blog is a privilege. You have no First Amendment rights here. I am the sole, supreme and benevolent dictator. This blog commenting system also has a patented Dumbass Detector. Don't set it off.

Note: Only a member of this blog may post a comment.