This week, many columnists and bloggers took note of the apparent drop in Republican party identification on many of the recent national surveys, especially after both Gallup and the Pew Research Center released helpful summaries with charts showing their tracking of long term trends (those links are well worth following for more detail). Although the Democratic advantage has definitely grown over the last two to four years, the changes over the last few months look more like a modest rise in independence than a shift benefiting either party.
Nate Silver also took up the issue yesterday with a chart featuring party identification data from six national polls and a LOESS regression trend line similar to what we feature in our standard charts. That post goaded us to moving up a party ID chart on our own to-do list.
Our new flash chart, embedded below, appears on this page and will update with each new national poll that features party ID results. Nate's chart had data from six pollsters -- we add six more, most notably the 20 or so since Labor Day conducted by Gallup:
The most striking feature of the chart is not the drop in Republicans, but rather the increase in independent identification. The decline in Republican ID is a nearly parallel decline for the Democrats. That trend is not surprising, as partisan identification often increases slightly during the last few months of an election year and fade afterward. However, note that the Pew Research report labels the magnitude of increase among independents as "noteworthy" as it appears much greater than what they observed in 2005.
As for our chart, keep in mind that you can use all of the interactive features to explore the "house effects" in evidence in the party ID results reported by various pollsters. Point to any dot on the chart to see a "tool-tip" box with the name of the Pollster and information on the specific release. Click on it to "connect the dots" of polls released by that pollster. Use the "filter" tool to see how the trend line would change if you drop certain polls from the chart.
One thing that will become evident quite quickly is the wide variation in the degree of independence measured by various pollsters. For a few polls (Fox News, Diageo/Hotline) the greater apparent partisanship is likely related to their interviewing only registered voters rather than adults. Much of the poll-to-poll variation results from subtle differences in the text and the extent to which pollsters train their interviewers to push for an answer. Generally, the more interviewers push for an answer (even before asking independents how they "lean") the more partisans they get.
And one last thought: One very gratifying aspect of the chart is the number of pollsters that now routinely release their party identification results. Four or five years ago, it was not so.