Left Behinds

The anti-andrewsullivan.com. Or, the Robin Hood (Maid Marian?) of bright pink Blogger blogs.

Sunday, October 29, 2006

The Effect of Ballot Initiatives on Man-in-the-Moon Elections

Richard Kim, who is being interviewed tomorrow morning at 11:30 am on 99.5 FM, draws a lot on some polling analysis in this excelent article about the hype surrounding the right-wing reaction to the NJ marriage decision and eight ballot initiatives.

In the article, ABC's analyst Gary Langer writes that

My own analysis of the 2004 exit poll results, presented at Stanford University on Nov. 9, 2004, found no consistent boost in turnout by pro-Bush groups — conservatives, Republicans, and churchgoing white Protestants — in states with gay-marriage initiatives.

An effect, to be an effect, should be consistent.

Nationally, Bush's greatest gain in 2004 came not among the highly churched, but among infrequent churchgoers. That election was about terrorism. [EA]
No consistent boost in turnout between 2000 and 2004.

But if his argument is that the ballot initiatives didn't consistently increase Republican voter turnout between 2000 and 2004, my initial retort is that those same voters might have had other factors driving disproportionate turnout in 2000 (such as, say, inaugurating an evangelical Christian Bush scion and defeating the "liberal" Clinton legacy they'd been resentfully living under for a decade). Factors which were absent or less motivating in 2004.

There are so many factors affecting turnout that it's hard to say "A was associated with Z one year, and M was associated with Z the next year, therefore M did not significantly boost Z in the second year." Without a control, it's hard to say if those Bush supporters would have been motivated to turn out in such huge numbers in 2004 without the ballot initiatives.

I.e., perhaps Rove's secret weapon in 2000 was Clinton, and in 2004 it was ballot initiatives. That seems plausible to me, and it's not clear that Langer's analysis contradicts it.

Does that make sense?

I'm just processing this info now.


As I reread the original article, the methodology seems pretty clear. I was really tired when I wrote this post.


  • At 12:30 AM, Blogger Antid Oto said…

    No, he's not saying no boost in turnout between 2000 and 2004. He's saying no differential boost among Republican-identified groups in states with gay marriage ballot initiatives.

    According to Jackman, states with the initiatives had a 3.3 percent higher turnout than in 2000, "but there [was] no relationship between change in Bush support and change in turnout." Rather, he said, the initiatives helped mobilize voters across party lines.

    Langer from ABC agreed with Jackman, noting that opposition to gay unions is not an exclusive issue of the religious right. In fact, a majority or a significant number of Democrats, moderates and liberals backed the initiatives, he said. "In a national exit poll, only 25 percent of Americans support gay marriage," he said.

    In other words, everybody hates the gays, not just Republican voters. Or, to put in another way, homophobia is so universal it's not a reliable way to motivate only Republican-leaning voters.

  • At 12:54 AM, Blogger Solomon Grundy said…

    Right, I got that part. The ballot initiatives were associated with a slight increase in voter turnout between '00 and '04, but for both Reps and Dems.

    But that still seems silent on the question of how evangelical Christians would have voted if the ballot initiatives were not on the ballot. Maybe 3% of Dem voters were turned out by the initiatives but 10% of right wing Christians were turned out by them. How can you really say, considering that 2000 and 2004 had such different factors affecting turnout (Clinton, 9/11, etc.). Something strikes me as wrongheaded about drawing conclusions about the ballot initiatives based solely on changes in turnout between '00 and '04. But I'm tired and maybe I'm just being dumb.

    I guess the closest thing to a control group would be similar states without ballot initiatives.

  • At 2:04 AM, Blogger Antid Oto said…

    Exactly. I'm pretty sure they're comparing the change within subgroups in states with gay marriage ballot initiatives to the change within those same subgroups in states without those initiatives.

  • At 2:40 PM, Blogger Solomon Grundy said…

    Oh yeah, when I reread that in the clarity of daytime it seems pretty clear. I was just being late-night muddleheaded.


Post a Comment

<< Home

FREE hit counter and Internet traffic statistics from freestats.com