bq. The 2004 U.S. presidential election was a wartime contest that entailed a great deal of discussion about the role that previous military service plays in elections for both candidates and the electorate. Using polling data throughout 2004, this article examines party identification, candidate affect, and vote choice preferences among veterans and nonveterans in the electorate. Despite widespread assumptions depicting the veteran population as deeply Republican, those with military experience in 2004 largely mirrored their nonveteran peers in terms of partisan identification, warmth toward candidates, ballot intentions, and vote choice. One important exception manifested after the “Swift Boat” advertisement in September, which impelled significant numbers of veterans who identify with the Democratic Party to express the intention to vote for George W. Bush.
That is from this paper by Jeremy Teigen. There will be much speculative discussion of the veteran vote during this campaign; this piece (and this one) provide some welcome data.
Teigen does find that, in 2004, McCain was perceived more positively by veterans (as was Colin Powell) than by non-veterans, but it’s an open question whether that will remain true in 2008. Teigen suggested to me:
bq. Perhaps veterans respect or admire political elites who share their military experience, but inside the pressure cooker of a presidential election, partisan affinities, issues, and electioneering outweigh the biographical effects. At the same time, John McCain’s military service as a Navy pilot and POW is much more well-known and less likely to be disparaged.
Ultimately, beware of bivariate relationships expressed as fact. Teigen notes:
bq. Republican tendencies among veterans disappear once one controls for the fact that veterans are overwhelmingly a group of older males.
For more on attitudinal variation among veterans, see this post by the artist formerly known as Mystery Pollster.
[Addendum: See also this paper by Ben Bishin and Matt Incantalupo, in which they find that veterans are no more likely to vote for either Republicans or veteran candidates, once other factors are taken into account.]