If it's in the News, it's in our Polls. Public opinion polling since 2003.


Do Independent Voters Matter?

A Commentary By Todd Eberly

Todd Eberly, a political science professor at St. Mary's College of Maryland, recently wrote a report for the centrist group Third Way about independent voters. Crystal Ball senior columnist Alan Abramowitz addressed that report in a recent article, and Eberly asked to respond. His point of view differs from Abramowitz's and the inclinations of the Crystal Ball staff, but in order to present readers a full view, we agreed to give Eberly a chance to share his views.

-- The Editors

Are independent voters a myth? That is certainly the conclusion of many who study political science. Research has demonstrated that, when pressed, independent voters often reveal significant partisan preferences: They lean Democratic or lean Republican. When leaners are reclassified and grouped among their partisan peers the share of pure independents in the electorate falls -- by some accounts -- to less than 10% of the electorate.

If the true number of independent voters is less than 10% of the electorate, then independent voters are of little concern. In an age of narrow victory margins in the national popular vote for the presidency and control of the House of Representatives, winning a majority of that 10% can be crucial, but appeals to a party's partisans would be a more important focus.

But what if the number of independent voters is greater than 10%, or even greater than 20%? Suddenly, winning a majority of independent voters becomes more important. In a recent report written for the centrist Democratic organization Third Way, I examined whether or not leaners are indeed independent. For my research, I used the 2000-2004 panel study conducted by the American National Election Studies (ANES). I selected the panel study for a simple reason: It's one of the few studies available that tracked the same group of voters across multiple elections. That's important. Most studies of voting and partisanship capture only a snapshot of a point in time and allow researchers to measure partisanship only during a given election cycle.

Such snapshots would be fine if partisanship were permanent and not subject to change. That is very much the view of partisanship taken by those who consider independent voters to be a myth. In my research for Third Way, I compared the partisan voting loyalty of Democrats and Republicans by looking at their partisan vote choice across three House elections (2000, 2002 and 2004) compared to their strength of partisanship in 2000. Survey respondents were classified as being strong, weak or independent partisans (leaners). I found that weak and independent partisans are less loyal to party in the short term and especially across time. While roughly 90% of strong partisans voted the party line in 2000, approximately a quarter of weak and independent partisans crossed party lines that year. In 2002 and 2004, strong and weak partisans held steady at roughly 90% and 75% loyalty, but independent partisans were more volatile -- especially independent Democrats. In 2002, 46% of those who identified as an independent Democrat in 2000 voted Republican. The share was 38% in 2004. I also found that independent partisans were far more likely to switch their partisan identification over time -- so 2000's independent Democrat could well be 2004's independent Republican. That's something a non-panel series could not account for.

The study suggested that during a given election period independent partisans are as loyal to party as their weak partisan peers, but that loyalty wanes over time. To me, a voter who switches his or her partisan vote choice from one election cycle to the next is not a loyal partisan -- rather, that voter is an independent voter. My findings have been criticized largely based on my selection of the 2000-2004 data series. Some contend that the events of Sept. 11 and the subsequent War on Terror made that time period unique and therefore unrepresentative. Unfortunately there is no other comparable data set exploring the same respondents across multiple elections. In a recent post challenging the findings contained in the Third Way report, Alan Abramowitz examined the 2008-2009 panel study and compared the partisan loyalty and partisan vote choice of respondents in the 2008 presidential election. Abramowitz came to the same conclusion as did I in my Third Way report: that independent partisans behave much like their more partisan peers in a given election.

Unfortunately, the 2008-2009 panel survey does not allow one to follow partisanship or partisan loyalty across multiple elections. As such it is not a useful data source for the study of partisan loyalty and the presence of independent voters. Additionally, I researched partisan loyalty by examining House elections, because it allows one to study multiple elections across a relatively short timeframe.

To address concerns about the 2000-2004 data, I conducted additional analyses with that data source and with the 1992-1997 panel survey by ANES. Having already demonstrated that leaners are less loyal to party over time, I wanted to focus on estimating the number of true independent voters in the electorate. Based on my study for Third Way, I placed the number at approximately 25% of the electorate, which is a number also endorsed by Linda Killian in her book, The Swing Vote.

I compared the 1994 and 1996 as well as the 2002 and 2004 partisan vote choice to the choice made in 1992 and 2000 respectively. In other words, what share of the folks who voted Democratic in 1992 voted Republican in 1994 or 1996? What share of folks voting Republican in 2000 voted Democrat in 2002 and 2004?

As I am interested in two-party vote shares, I limited my study to only those who voted for one of the two parties in each of the elections covered. Of those who voted for a Democrat in the 1992 House elections, 25% opted to vote Republican in 1994 and 24% opted to vote Republican in 1996. Among Republicans, 12% voted for a Democrat in 1994 and 21% voted for a Democrat in 1996. Based on the two-party vote shares in each election, nearly 19% of those voting in 1992 and 1994 changed their partisan vote choice. The overall share was closer to 23% between 1992 and 1996.

When looking at the more recent era, of those who voted for a Democrat in the 2000 House elections, 16% voted Republican in 2002 and 21% voted Republican in 2004. Among Republicans, 11% voted Democrat in 2002 and 21% voted Democrat in 2004.

Both panel series show that partisan loyalty declines over time, and that Democratic voters are less loyal than Republican voters. Consistent with the findings of the Third Way report, both panel series show that partisan loyalty is weakest among weak, but especially among independent, partisans (most defections came from independent partisans). Much is made of surveys by Gallup and Pew that suggest that a plurality of voters are independents -- perhaps as much as 40%. This is simply incorrect. But so too are arguments that independent voters make up less than 10% of the electorate. The stability of a partisan coalition is dependent upon sustained loyalty across elections, but roughly 20% of the voting electorate are not loyal partisans (and that share would grow if I expanded my study to include folks who opted to not vote -- as non-voters could hardly be considered loyal partisans). In an era of closely matched political parties and relatively narrow two-party vote shares, winning and maintaining the support of that 20% is crucial.

Larry J. Sabato is the director of the Center for Politics at the University of Virginia.

See Other Commentary by Larry Sabato

Rasmussen Reports is a media company specializing in the collection, publication and distribution of public opinion information.

We conduct public opinion polls on a variety of topics to inform our audience on events in the news and other topics of interest. To ensure editorial control and independence, we pay for the polls ourselves and generate revenue through the sale of subscriptions, sponsorships, and advertising. Nightly polling on politics, business and lifestyle topics provides the content to update the Rasmussen Reports web site many times each day. If it's in the news, it's in our polls. Additionally, the data drives a daily update newsletter and various media outlets across the country.

Some information, including the Rasmussen Reports daily Presidential Tracking Poll and commentaries are available for free to the general public. Subscriptions are available for $4.95 a month or 34.95 a year that provide subscribers with exclusive access to more than 20 stories per week on upcoming elections, consumer confidence, and issues that affect us all. For those who are really into the numbers, Platinum Members can review demographic crosstabs and a full history of our data.

To learn more about our methodology, click here.