It took only a few hours for media talking heads to come up with an explanation for why the tax-cutting, war-mongering ignoramus George W. Bush got re-elected on Tuesday. It couldn't have been, they agreed, that the electorate truly wanted four more years of Republican domestic and foreign policies. After all, didn't John Kerry win every substantive point on the issues that really matter? Wasn't Bush an accidental, illegitimate president in the first place? Another explanation had to be found.
Quickly scanning the exit polling, pundits spotted that 22 percent of voters had said "moral issues" were the most important in the race, and they broke heavily for Bush. A-ha! Now they had a talking point: evil mastermind Karl Rove had used the same-sex marriage issue to turn out so many religious conservatives that they overwhelmed everyone's likely-voter models and became a disproportionate share of the 2004 electorate.
The initially unspoken, but soon loudly proclaimed, implication of this fact was that Bush and the Republicans had bamboozled these voters, whose real economic interests lay with the Democrats. They didn't truly support Bush on substantive domestic and foreign-policy issues. They were just anti-gay bigots. As the spin got more frenzied Wednesday and Thursday, words like "jihadis" and "mullahs" got attached to these deluded and dangerous Bush voters, whom Democrats and sympathetic analysts described as something akin to a bizarre and perversely fascinating lost tribe just discovered in the rain forests of Borneo. By weekend, an inevitable backlash against the frenzy had set in.
The problem with all this is that, while comforting to many Kerry supporters and exhilarating for some social-conservative leaders, the notion that Bush won primarily because religious voters turned out for him does not seem to be backed up by any real evidence. Few reporters or commentators appear to have gone back to examine the 2000 exit polls, which would seem to be necessary if one wishes to assert a trend.
I did. I found that the percentage of voters sampled who said they attended church at least weekly was the same—42 percent—in both 2000 and 2004. The percentage never attending church was also the same, at 15 percent. The middle group, those attending occasionally, was, you guessed it, 42 percent each time. Interestingly, while Bush slightly improved his standing among frequent churchgoers, by about a point in 2004, his support grew by 3 to 4 points among those attending seldom or never.
Yep, it was the atheist vote that really put Bush over the top in 2004.
There could be other ways to salvage the myth of the Republican mullah-cracy. For example, one might argue that it is unfair to equate church-going with religiosity or cultural conservatism.
Another potential proof: More people identified themselves as conservatives in 2004 (34 percent) than in 2000 (29 percent). But there are all kinds of conservatives, including quite a few who are socially conservative and hawkish and in favor of privatizing Social Security. Sorry, this doesn't prove anything other than people are increasingly willing to label themselves as conservative rather than moderate.
OK, what about issue positions? In 2000, about 40 percent of voters in the exit poll said that abortion should be mostly or always illegal. In 2004, it was 42 percent. Not exactly a huge jump. And we don't know how many of those are single-issue voters on abortion. Both parties have significant minorities who disagree with the official party position: about a quarter of pro-lifers voted for Kerry, while around one-third of pro-choicers picked Bush. On same-sex marriage, the issue was not polled in 2000 so it is impossible to say with certainty how the two electorates compare, but it is unlikely that this year's voters were significantly more conservative on it. In fact, the public's position is more nuanced here than the insta-spin would have you believe. About as many favored civil unions but not official marriage (35 percent) as favored neither (37 percent), and Bush was preferred by both groups over Kerry.
Well, perhaps there was no national trend but it happened in selected states such as Ohio. Nope. In the 2000 exit poll for Ohio, the percentage of frequent churchgoers was higher (45 percent) than in 2004 (40 percent). Bush did win a larger majority of religious Ohio voters in 2004 than he did four years ago, but there were fewer of them proportionally. Besides, saying that the religious-vote affect mattered in a few key states changes the nature of the media spin, which has been trying to assert it as a sweeping national "explanation" for Bush's popular vote.
That leaves the initial assertion about 22 percent of voters citing moral issues as most important, higher than the share citing terrorism, Iraq, the economy, or other issues. When I looked more closely at this question, however, doubts immediately presented themselves. For one thing, the answers were broken out in ways that biased the analysis. While the poll did not attempt to distinguish the various moral issues that voters might be thinking about—abortion, marriage, wars for oil, etc.—it did list "taxes" and "the economy" separately, as well as "terrorism" and "Iraq." Of course, for many voters, these are not separate issues. You may disagree with them, but most voters sampled in the exit poll said that the war in Iraq was part of the overall war on terrorism. And many right-leaning voters see tax policy as inextricably linked with economic growth and job creation (at least a few freedom-loving folks even see tax cuts as a moral issue—imagine that!)
In short, the question is flawed and the answers easily misunderstood. Moreover, it doesn't compare well with the 2000 exit poll, in which "moral issues" was not listed as an option. On the other hand, you can track the impact of foreign policy over time. In 2000, only 12 percent said that "foreign affairs" was the most important issue in the presidential race, and they broke 54 percent to 40 percent for Bush over Gore. In 2004, a combined 34 percent identified foreign policy (either Iraq or the war on terrorism) as the most important, and they appear to have broken for Bush by 59 percent to 40 percent. Put it all together, and the increase in salience and small increase in Bush preference for foreign policy constitutes a gain of 13.5 percentage points in the Bush vote in 2004.
Obviously, he didn't win by that much. He lost ground on economic issues, because of the recession. But without his edge on war on terrorism, Bush would have lost. And that proposition—unlike the "it's all about gay marriage meme"—is testable and fits the available data. Voters worried about partial-birth abortion, same-sex marriage, and other cultural issues are obviously an important constituency within the current GOP majority, but they are no more responsible for Bush's national victory on Tuesday than voters motivated by other issues to re-elect the president.