One place my thoughts turn during the holidays is to the men and women serving overseas. From there – my day job being what it is – it’s a short leap to the veteran vote. With the 2008 election dust now settled, how’d it look?
There are some surprises. While veterans clearly favored John McCain, it was by less of a margin than you might suppose for a candidate with a celebrated war record: McCain won veterans by 10 points, compared with George W. Bush’s 16-point margin in 2004.
Digging’s worthwhile, since it’s a pretty big group: Per the exit poll, veterans accounted for 15 percent of all voters, about as numerous as, say, blacks (13 percent) or single men (14 percent). And as I noted in a posting based on our pre-election polls last summer, veteran voters stand out demographically. Almost all of them – 89 percent – were men. More than a third were senior citizens, triple the level of seniors among non-veterans. They were more apt to be Southerners, including Southern whites.
Those fit expectations. But the biggest surprise is for anyone who assumed a major tilt toward the Republican Party among veterans. Thirty-five percent of veterans voting in 2008 identified themselves as Democrats, 34 percent as Republicans (the rest were independents). That’s a striking change from 2004: Republican allegiance among veterans dropped from 41 percent then to 34 percent this year. Democrats gained 4 points, independents 3. What had been a 10-point Republican advantage over Democrats among veterans vanished.
Veteran voters 2008 2004 ChangeDem 35% 31% +4 pointsRep 34 41 -7Ind 31 28 +3
An overall drop in Republican participation was one of the major takeaways of the election. What we see here is that it was even more pronounced among veterans (as above, down 7 points) than among non-veterans (down 4 points, from 37 percent Republican in 2004 to 33 percent in 2008). Veterans’ party allegiance is far from fixed in the firmament. (This fits, as it happens, with a survey of partisanship among active-duty military by Army Maj. Jason Dempsey that I wrote about in July.)
At the same time, ideology was more stable: Veterans this year were 6 points more apt than non-veterans to be conservatives, about the same as the 8-point gap in 2004. Forty-one percent of veteran voters this year were conservatives, identical to 2004. (Among white veterans only, conservatism peaked, at 46 percent.) Nineteen percent of veterans were liberals, compared with 24 percent of non-veterans.
Greater conservatism and/or lessened liberalism would push the veteran vote toward Republican candidates, party affiliation aside. That helps explain why McCain did a little better with Democratic veterans than with non-veteran Democrats – and more important, a lot better with independent veterans than with non-veteran independents. Another table:
Vote pref: Obama-McCain Veterans Non-veteransDem 83-16% 90-9%Rep 7-92 12-87Ind 41-56 55-40
In the end, veterans voted for McCain by 54-44 percent, compared with a 57-41 percent preference for Bush in 2004. Sadly, given the construction of the exit poll, we don’t have attitudinal questions to carry our evaluation beyond the demographic differences reported here. When we run a regression analysis, holding constant other factors (such as age, race, sex and partisanship), being a veteran is a significant predictor of vote preference in 2008 exit poll data, but was not in 2004. That’s similar to the inconsistent regression results we’ve seen previously in pre-election data.
As I suggested last summer, it seems to me that most coherent voter groups exist via political or religious affinity or demographic characteristics – such as independents, blacks, white evangelicals or young voters. With the exception of career officers, veterans seem to have less of a common denominator, aside from the fact of their service – which, for many of them, was a long time ago. That makes them a more elusive group politically than conventional wisdom might suggest.
I’ll close with a few details:
-Like sex, age is a sharp difference: Thirty-five percent of veterans who voted were 65 and over, triple the share of seniors among non-veterans, 12 percent. And just 5 percent were under 30, compared with 22 percent of non-veterans. That mattered, since older voters were McCain’s best group; young ones, his worst.
-Veteran voters were 9 points more likely than non-veterans to be Southerners, 40 percent vs. 31 percent. Among whites only, 34 percent of veterans were Southerners, vs. 28 percent of non-veterans.
-White veterans voted 63-36 percent for McCain, compared with 53-46 percent among white non-veterans. Interestingly, this difference disappeared in the South, where white veterans voted almost exactly like all Southern whites, 68-31 percent for McCain; it was in other regions where white veterans were better for McCain than white non-veterans. Among nonwhites, meanwhile, McCain got 28 percent support from veterans, compared with 16 percent from non-veterans.
-Among veterans who voted for Bush in 2004, 12 percent defected to Obama. That's less than Obama's share of non-veteran Bush voters, 19 percent. At the same time, among veterans who did not vote in 2004, or who voted for a third-party candidate, 73 percent backed Obama this year (as did 69 percent of non-vets in this group).
-A last nugget: Veteran voters were 20 points more apt than non-vets to be gun owners.