The role of independents in the 2006 elections has been overstated

On Saturday the New York Times ran an op-ed from a Duke professor named David Rohde which ran hard against the early conventional wisdom coming from the elections:

..."THE midterm elections have been widely viewed as a sudden change of direction, with Democrats seizing the wheel from Republicans. While that may be true, the big electoral news — news that has gone largely unnoticed — is this: After decades of weakness, after sideswipes from independent candidates, the two major parties are back. Indeed, they are more potent and influential than at any time in the past century."

There are really two pieces to this argument.  The first is the role of the Parties themselves in relation to other organizations and leaders in the political firmament.  I will not tackle that argument today, though I do agree with him, and a lot of it has to do with the way the internet allows people to have a much more intimate and direct relationship with their parties.  What I do want to write about is a related trend, the increased partisanship of the electorate. 

In the last two elections, 2004 and 2006, 74% of all voters identified themselves as a partisan, either a Republican or Democrat.  Only a quarter, 26 percent, identified themselves as independent.  These ratios did not change from 2004 to 2006. 

In 2004 37% of the electorate described themselves as Democrat, the same Republican.  In 2006 Democrats picked up a point as a share of the electorate, Republicans lost a point, leaving it 38D/26I/36R.  Remarkably stable ratios given that the vote changed from 51/48 R to 52/46 D.  

The early storyline then is that the shift from 2004 to 2006 came about from how independents swung.  They did swing 17 points, from 48R/49D to 39R/57D.  But a far greater shift happened inside the two parties, where there was an 8 point shift within the Democratic electorate, and a 4 point shift inside the Republican electorate, or a total of a 12 point shift. The Democratic vote went from 89/11 to 93/7, and the Republican vote 93/6 to 91/8.   

While less in percentage terms this 12 point shift happened in what is 3/4 quarters of the electorate, and this 18 point shift happened in what is 1/4 of the electorate.   So this means a far greater number of votes shifted in the last two years between and among the parties than shifted with independents - meaning that Democrats owe their victory much more to gains with Democratic and Republican partisans than they do to the gains they made with independent voters. 

This reduced role for independents was evident even in 2004.  John Kerry did what every Democrat was told was necessary to do win the Presidency - he won independents - and yet he still lost the election.  Why? Because the Rove machine pushed the percentage of the electorate that was Republican to an all time high, 37%, equalling the Democratic share, and they kept 93% of these Republicans.  Kerry while winning independents, only won 89% of Democrats.  This difference - between Rove's 93 and Kerry's 89 within their own parties - cost Kerry the election. 

Tim Kaine, writing about his impressive win in Virginia in 2005 in the DLC's magazine recently, described his "Democrats first" strategy, one that seems very much in touch with the notion that winning elections in this more partisan era starts first with expanding and holding one's own partisans:

"Just as Warner had done in 2001, I had to accomplish three things to win in a red state. First, I had to find and energize Democratic voters. Second, I had to share my story with the voters. Third, I had to reach out to independent and Republican voters in a strategic way. And that's exactly what we did."

I'm in no way suggesting that winning with independents was not an important part of how Democrats won in 2006.  Of course the big swing with independents was impressive and critical.  But with so few people considering themselves independents these days, we have to be careful not to overstate their impact.  It is clear from the exit poll data of the last two years that what has been far more important in determining the outcomes of the two elections is what has happened within and between the two parties, which is today about three-quarters of all voters. 

Additionally, Democrats should not discount the power of what Rove, Mehlman and his team did and have left behind. Despite their epic collapse this year, the Republicans only lost a single point of market share as a percentage of the electorate, and today almost 40 percent more Americans consider themselves Republicans than independents, an historically very high number.  As Mehlman said after the 2004 elections, they spent a great deal of money persuading Republicans to vote and to vote Bush.  Their "Republicans first" strategy was actually very successful in many ways, as this investment they made in creating more Republican voters has changed the nature of the American electorate, dimished the influence of traditonal independents, and has indeed made more Republicans than there used to be. 

The problem they had this year wasn't these voters becoming independents and fleeing the party, which one would have expected.  They may have voted Democrat this time, but in this year's exits they still consider themselves Republicans.  To repeat, independents did not gain a single point as a share of the electorate despite the tremendous collapse of the Republican brand.  

Update: Blogger James Hupp has challenged my math in a new post.   He was right about one thing, I had the change in independents to be at 18% when it was 17%.  I changed that above (thanks James).  But in reviewing his calculations, his raw numbers still indicate what I wrote above: it still seems that a 12 point net shift in 74% of the electorate is greater than a 17 point net shift in 26% of the electorate (add the two point net shift for Democrats and away from Republicans)  Hey, if I am wrong with my numbers here I will chuck it.  In reviewing his math this morning am still not convinced (and James of course I had these numbers weighted - a little unfair there.  If we can talk of a 17 point shift with indies you can talk 12 with partisans).  Open to your thoughts.