Ethnicity demographics showed that support for McCain rose the most among Hispanic evangelicals, where support increased by 14 percent to earn McCain 74 percent total. Support for Obama dropped from 19 percent to 12 percent in the same category. 

Among Caucasian evangelicals, the percentage of those voting for McCain rose from 81 percent to 88 percent. The African American evangelical vote for McCain dropped while support for Obama increased – a 2 percent shift from McCain in favor of Obama.

These numbers show McCain/Palin’s support among evangelicals to be even stronger than the support President George W. Bush enjoyed in 2004. The so-called religious right appears to have reemerged, with 81 percent of Caucasian evangelicals backing McCain compared to Bush’s 78 percent.

Daughter’s Pregnancy Earns Increased Support

This new-found support correlating with McCain’s VP announcement remained unaffected when news surfaced that Palin’s unwed daughter, 17-year-old Bristol, was pregnant. Instead, support actually rose again.

Of the 1,738 people surveyed, only 1 percent said they no longer support McCain as a result of Bristol’s pregnancy. Twenty-four percent of evangelical voters said the news actually strengthened their support of McCain/Palin, and 61 percent said their support for Palin was unaffected by the news.

One participant who voiced increased support for McCain/Palin said, “I believe it makes Sarah more real and in touch with normal people. Her reaction, love and support for her daughter and commitment to life have actually caused me to support the Republican party for this election.”

Other participants echoed that parents can teach their children, but children will ultimately make their own decisions. Many also expressed frustration at the media attention Bristol’s pregnancy has drawn.

Religion Still Matters

As both the Democratic and Republican national conventions proved, religion remains a critical question in the election.

In part one and two of the survey, 76 percents of participants said a candidate’s religion affects their vote.

By denomination, Baptists were most likely to say that a candidate’s religion mattered, with only 12 percent dismissing it as a voting issue. Catholics were the least likely to be influenced by a candidate’s religion, with 46 percent saying it did not influence them. The same percentage of Catholics, however, said it was a factor in their voting decision.

The motivation behind choosing a candidate also stems from religion. Almost 93 percent responded that voting is a Christian responsibility. Eighteen percent, however, said that abstaining from voting is an acceptable option when no candidate is preferred.

Abortion Isn’t the Top Issue

In spite of the importance of religion, pocketbook issues rose above social concerns in the minds’ of survey participants.

Asked to rank eight issues in order of importance, participants placed the economy of highest significance. Abortion ranked second. Concerns about the war in Iraq and defense issues trumped homosexuality and same-sex legislation, followed by health care, education, illegal immigration, and the environment.

The economy surfaced as the biggest concern and issue for all ages except those over age 65, who placed war and defense issues in the top spot.

“In a way, that’s not too shocking, given that so-called ‘pocketbook’ issues and matters of war-and-peace tend to always dominate the pecking order for voters,” said Dr. Paul Kengor, professor of political science and executive director of the Center for Vision & Values at Grove City College.

“However, you might think evangelicals would swim against the tide when it comes to the moral issues. This suggests that evangelicals are not all that different from voters generally, which is an interesting finding that ought to surprise the secular press,” he continued.

Editor’s Note:

This survey was sent by email to Crosswalk.com subscribers and was not an online poll. Participants included in this survey were chosen randomly and proportionally from a pool of subscribers with verified geographic data from all 50 states.  Based on the sampling method and sample size, the estimated margin of error is +/- 1.9%.