2005-12-21T12:45:11Z 2005-12-21T12:45:11Z 2005-12-21T12:45:11Z

<p>Barry Burden <a href="http://www.iq.harvard.edu/blog/pb/2005/12/looking_for_variance.html" title="Political Behavior Blog: Looking for variance">notes that party identification explains too much variance in vote choice these days</a>:</p> <blockquote> <p>The old Michigan triad of partisanship, issues, and candidate evaluations as an explanation for vote choices is proving less useful in recent days. The main reason is that party identification and the vote are practically one and the same. In the 2000 and 2004 <span class="caps">NES</span> data, better than 90% of partisans voted for the presidential candidate of their party. In 2004 only 40 respondents (7% of partisans) voted against their stated party identification.</p> </blockquote> <p>He sets out a few intriguing directions for future research on party identification.</p> http://blog.lordsutch.com/archives/3137 http://blog.lordsutch.com/atom.cgi/entryid=3137