About Them Chickens

Untitled document

Paul Geary at The New Editor has a must read on polls, polling, election analysis, political punditry and pretty much everything else you need to know about the 2006 midterm elections. All the coverage you are seeing lately predicting the end of the world for the Republicans may be, to put it nicely, misguided.

For all the criticism we hurl toward 43rd St., The New York Times is still a first-class source for information and a must read. I became addicted to the Times real estate section 10 years ago while living in the tri-state area, during a stint at Nielsen Media Research in Fairfield County, Connecticut.

The interactive election guide online is a fantastic tool for the political junkie and the sort-of-interested alike, and probably the most comprehensive source for election information.

Having said that (you knew this was coming), the analysis part of the tool is, well, interesting.

The House is considered to be in play by most people, and the current analysis shows 57 seats in play — 19 "leaning Dem," 22 "leaning Rep," and 16 "toss ups." All 16 "toss ups" are in districts currently held by Republicans. All 22 "leaning Republican" are in current Republican districts.

In other words, there are zero Democratic seats that are even leaning Republican. This is less an "analysis" than it is a Democratic best-case scenario.

Consider a few facts: Incumbents historically win more than 90% of the time in House races. An incumbent who is not involved in scandal and who has a money advantage over a challenger wins north of 95% of the time. And most Americans don't pay attention to elections, especially off-years, until the last two weeks of the race.

And consider that there is a consistent skew toward Democrats in polls, which we've chronicled. This inability to get an appropriate random sample of voters has made pollsters unable to correctly predict close races. Remember not too long ago when their accuracy was alarming to the degree that we were worried they suppressed turnout on the West Coast? That is because fewer races were close then. Their clairvoyance was illusory.

Let's have a look at what "leaning" and "toss up" mean to the analysts at the Times.

Here are two examples of "toss up" races:

In New York-26, incumbent Republican Thomas Reynolds is running against Democrat Jack Davis. Reynolds had more than $3 million on hand vs. $63,000 for Davis as of the last FEC filing. That means Reynolds can afford to get out the vote and advertise his unredacted, unedited, and unfiltered message to the voters.

The Western New York district is historically Republican, having voted for George Bush by 12 points over John Kerry. Reynolds beat Davis by 12 points in 2004.

The downsides, according to pundits, are his stewardship of the National Republican Congressional Committee, as if everyday voters care about that. Reynolds admitted he knew about the Foley e-mails, and his top aide resigned. A poll shows Davis up by five points, but that was taken at the height of the coverage of Reynolds' relationship to the scandal. Also, the report doesn't say how "likely" voters were determined, or whether the Republican vs. Democratic breakdown of those polled was reflective of the district. Too often, it's not.

Does knowing about the Foley e-mails erase 12 points and a $3 million fundraising lead in an historically Republican district? History tells us otherwise.

This really is worth taking the time to read. Because there are some things that Geary points out that indicate just how much cheerleading is going on. And how much misleading "analysis" is being reported at straight fact, when history does not back those analyses up.

This entry was posted in Politics. Bookmark the permalink.