There was indeed a flood of evangelicals to the polls--but it now appears that the shift in the Catholic vote was just as important and, in crucial states, probably more so.
In addition, Bush also made gains among the moderately religious--and the secular--not just the heavy-duty religious voters who attend religious services weekly or more.
Bush's strong performance among Catholics, it turns out, was crucial to his victory. Bush won Catholics 52%-47% this time, while Al Gore carried them 50%-46% in 2000. If Kerry had done as well as Gore, he would have had about a million more votes nationwide. According to Gallup Polls, only one Democrat since 1952 (Walter Mondale in 1984) lost the Catholic vote by this large a margin.
The Catholic impact was starker in key states. In Ohio, Bush got 55% of the Catholic vote in 2004 compared to just under 50% of them in 2000. That means a shift of 172,000 votes into the Republican column. Bush won the state by just 136,000 votes this year.
In Florida, Catholics made up 26% of the electorate in 2000. This year, they made up 28%. In 2000, 54% of Catholics went for Bush; in 2004, 57% of them voted for him. The combination of those two factors meant a gain of 400,000 voters in the Sunshine State-about Bush's margin of victory.
Bush also did better among Hispanic Catholics, getting 42% of the vote in 2004 compared to 31% in 2000.
During the campaign, polls showed the Catholic vote shifting back and forth between the candidates. Kerry's standing improved after the third debate when he spoke about his faith. But President Bush's views on abortion and gay marriage are more in line with official church teachings, and the campaign made the Catholic vote a high priority.
Another surprising finding: Bush did not dramatically improve his standing among people who go to church weekly or more often.
There was a big increase in the portion of the electorate made up of weekly churchgoers in the South--and yet nationally the figure didn't budge, indicating that regular churchgoers actually played a smaller role outside the South.
Here again, the pattern was apparent in two decisive battleground states. In 2000, 43% of Ohio voters were people who attended church weekly or more often. In 2004, that percentage actually declined to 40%*--meaning regular churchgoers were, relatively, less important in the close outcome.
The same thing happened on a larger scale in Florida. In 2000, 41% of voter attended services weekly or more often. This year, the portion dropped to 35%.
Nationally, Bush did improve his standing among those who attend worship services monthly instead of weekly (his share of this vote rose from 46% to 50%.)* A possible explanation: contrary to the common stereotype, many religious people, including "born-again" Christians do not attend religious services weekly. It was with this group that GOP outreach efforts may have borne the most fruit.
Amusingly, one big improvement in Bush 's performance actually came from those who never go to church. He won 36% of this group compared to 32% last time.*
While it is certainly not the case that Bush rode to office on a wave of atheism and secularism, these patterns reveal the complexity of Bush coalition-it was not just the "religious right."
None of this is to suggest that white church-going evangelicals didn't play a significant role. They were probably particularly important in growing Bush's overall popular vote and in some close swing states. A good example is Iowa where where close to a third of the voters this time were white born again protestants.
Though changes in the wording of exit polls make it difficult to directly compare the evangelical vote this election and last. In addition, the evangelicals who did vote went for Bush by a greater margin-78% rather than 72% in 2000. The improved performance among those evangelicals who voted proved to be just as important as the turnout.
The combination of those two factors-the higher evangelical turnout and the greater margin-meant that Bush did beat his Karl Rove's much publicized target of drawing in four million evangelical voters.
Second, evangelical turnout was at least partly offset by increased turnout from pro-Kerry groups. Kerry got roughly two million more votes from 18-29-year-olds than Gore did in 2000. He received approximately 1.6 million more votes from African Americans than Gore did. Churchgoers voted in greater numbers-but so did secular voters, and, in fact, nearly everbody else.