skip to main content

Religious Intensity Predicts Support for McCain

Hispanic Catholics and black Protestants provide notable exceptions

PRINCETON, NJ -- Americans who say religion is an important part of their daily lives support John McCain over Barack Obama for president, 50% to 40%, while their less religious counterparts support Obama over McCain, 55% to 36%.

These conclusions are based on an analysis of the relationship between presidential vote choice and responses to the question "Is religion an important part of your daily life?" among almost 95,000 registered voters interviewed as part of Gallup Poll Daily tracking, from March through June of this year. (About two-thirds of American registered voters respond in the affirmative when asked whether religion is important to them.)

The finding that the Republican candidate does better among more religious Americans is not a new one. White evangelical Christian voters, for example, have traditionally been among the most reliable Republican voters. The current analysis shows that the divide in vote preferences based on religiosity is not confined to white Protestants, but also occurs among non-Hispanic white Catholics. Indeed, the relationship between religiosity and vote choice is apparent among other groups that may not have been thought of in this regard traditionally, including in particular American Jews and, to a slight degree, those who identify with non-Christian and non-Jewish religions.

There are exceptions to the correlation between religiosity and likelihood to vote Republican. The strong support for Obama among Hispanic Catholics, black non-Catholic Christians, and those who do not have a specific religious identity appears to be little affected by self-reported importance of religion.

The relationship of religion to vote choice is most clearly seen in the large segment of Americans who are white and identify themselves as either Protestant or some other non-Catholic Christian religion. Among those in this group who say religion is important in their daily lives, McCain beats Obama by a 36-point margin, 63% to 27%. Among those in this group who say religion is not important, McCain and Obama are essentially tied, at 46% for McCain and 45% for Obama.

The same pattern, although not quite as dramatic, is seen among non-Hispanic white Catholics. Those in this group for whom religion is important prefer McCain to Obama by a 16-point margin. Non-Hispanic white Catholics for whom religion is not important give Obama a slight, two-point margin over McCain, 47% to 45%.

These patterns, so starkly evident among white Christians, essentially disappear among Catholics who are Hispanic, and non-Catholic Christians who are black. Cultural considerations deriving from ethnicity and race among these groups are likely the dominant factors influencing their vote decision. The differential impact of personal intensity of religion within these groups, it appears to follow, is much less important than it is among other groups examined in this analysis.

Only 39% of U.S. Jews report that religion is important in their daily lives, well below the overall national average. Among this smaller group of religious Jews, however, Obama and McCain break even, 45% to 45%. This compares to Obama's 68% to 26% lead among the majority of Jews for whom religion is not important.

There is a modest impact of the importance of religion among Americans who report identification with non-Christian, non-Jewish religions. Obama wins by a significant margin among these individuals regardless of whether they report religion to be important in their daily lives, but the Obama over McCain gap is larger among those for whom religion is not important (a 45-point margin versus a 30-point margin.)

Finally, there is the group of about 12% of Americans who say that they have no specific religious identity or religious affiliation. This group taken as a whole is strongly likely to support Obama. There is no substantive difference in the pattern of support for Obama over McCain among the small group of 11% of this voter segment who -- somewhat anomalously -- say religion is important in their daily lives.


It has been well established from an analysis of previous survey data that certain groups of highly religious Christians -- usually defined as "evangelicals" -- are reliably Republican in their presidential vote preferences. The current analysis expands the exploration of this relationship between religiosity and vote choice, taking advantage of a large sample of almost 95,000 interviews conducted as part of the Gallup Poll Daily tracking program from March through June.

The data confirm that among non-Hispanic white Christians -- Catholics and non-Catholics alike -- those who report that religion is important in their daily lives are significantly more likely to report voting for McCain over Obama than are those who say religion is not important.

Importantly, the analysis suggests that the presence of this relationship between religious intensity and presidential vote choice can be extended to include Jews and to some degree those who identify with a non-Christian, non-Jewish religion. In both of these groups, those who say religion is important are more likely to support McCain than are those who say religion is not important in their daily lives.

At the same time, the data show that for two voter groups, Hispanic Catholics and black non-Catholic Christians, there is little evidence of a relationship between the importance of religion and the vote. These two groups of voters appear to be strong Obama supporters regardless of whether they report being personally religious. And, among the small segment of Americans who do not report identification with any religion, support for Obama is also strong and relatively unaffected by the self-reported importance of religion.

Survey Methods

Results are based on telephone interviews with 94,872 registered voters, aged 18 and older, conducted March-June 2008. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of sampling error is ±1 percentage point for the entire sample. Sampling error for subgroups will vary.

Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only).

In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.

To provide feedback or suggestions about how to improve, please e-mail

Gallup World Headquarters, 901 F Street, Washington, D.C., 20001, U.S.A
+1 202.715.3030