Gallup: highly religious people more liberal (small "l")

October 8, 2008

Worldwide, Highly Religious More Likely to Help Others
Pattern holds throughout the world and across major religions
by Brett Pelham and Steve Crabtree

WASHINGTON, D.C. -- Gallup Polls conducted in more than 140 countries worldwide between 2006 and 2008 show that those whose responses identify them as highly religious are more likely than less religious respondents to report that they have engaged in each of three "helping behaviors" in the past month. In all four major global regions, for example, highly religious people are more likely than those who report being less religious to report having donated money to a charity in that time.



The pattern is similar when Gallup asked respondents whether they had volunteered their time to an organization in the month prior to being surveyed. Though the overall numbers are lower here in all regions except Africa, highly religious respondents are again more likely to say "yes" than those who are less religious.



One question these findings raise is the degree to which highly religious people reserve their charitable activities for members of their own religious communities. After all, many religions encourage -- or even require -- members to donate their time or money to their local faith-based organizations. Are highly religious people also more likely than those who are less religious to say they've helped a stranger in the past month? The answer is yes -- though the differences are smaller in this case.



The "religion effects" we see in these questions are consistent not only across the major global regions, but also consistent across the world's largest faith traditions, including Christianity, Islam, Hinduism, Buddhism, and Judaism. Among respondents who identified with each of these major religions, those who fall into the highly religious category are more likely than those who are less religious to say they've engaged in all three helping behaviors, with differences for helping a stranger ranging from 7 percentage points among Buddhists to 15 points among Jews.

Bottom Line

We cannot conclusively attribute these helping behaviors to the direct influence of religiosity. It's possible, for example, that these differences occur because inherently helpful people may be inherently attracted to religion.

But it does make intuitive sense that religious people around the globe are likely to engage in helping behaviors. After all, selflessness is a principle tenet of many religious traditions: One of the five pillars of Islam is zakat or almsgiving, for example, similar to the long-standing practice of tithing among Christians. Important principles of Buddhism include Sila, which requires people to treat others as they would prefer to be treated themselves, and Dāna, which loosely translates as "generosity."

However, the tendency of highly religious people around the world to say they donate their time and money does seem more impressive when you consider that, from a global perspective, those people are consistently poorer than those who are less religious. Among those highly religious respondents worldwide who reported their annual incomes to Gallup, the average figure (converted into international dollars) was about $10,000. Among those less religious respondents who reported their incomes, the average was about $17,500. Seen in this light, the data presented here offer compelling evidence of the role of religious dedication in helping to encourage supportive, community-oriented behaviors in areas where they may be most needed.

Survey Methods

Results are based on telephone and face-to-face interviews conducted between 2006 and 2008 with at least 2,000 adults in most countries. Because most analyses focus on large numbers of participants (e.g., more than 40,000 highly religious people in Africa), confidence intervals were only a fraction of a point (i.e., less than ±1 percentage point). 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.