How non-atheists have viewed atheists.

Public opinion surveys have often shown a gap between broad support for freedom of religion and acceptance of atheists as neighbors, civic participants or family members. The figures below are historical records, not measurements of present-day opinion.

Tolerance and the meaning of atheism

People may support religious diversity in principle while holding less favorable views of people who do not believe in a deity. That distinction matters because atheism is not a single organization or creed: it can describe a lack of belief in gods, or a considered conclusion that no gods exist. Individual atheists may otherwise differ widely in ethics, politics, culture and personal identity.

Older surveys also used inconsistent labels. Some asked about “atheists,” while others grouped together people with no organized religious affiliation, including agnostics, secular humanists and people who simply preferred not to identify with a tradition. Those categories should not be treated as interchangeable.

Selected historical findings

Polls reported from the 1980s

Accounts of surveys from the 1980s described a sharp difference between support for general religious freedom and support for unrestricted criticism of religion by atheists. In the same period, many respondents were reluctant to support the use of publicly funded civic venues for atheist advocacy. These results suggest that “freedom of religion” was sometimes understood more readily as protection for religious communities than as equal protection for nonbelief.

The historical context is important. A question about the right to criticize religion measures attitudes toward speech and offense, not simply whether respondents accept atheists as citizens. It is therefore unsafe to collapse these separate issues into one measure of tolerance.

Barna Research findings from 1995

A 1995 Barna Research survey asked American adults whether different religious or nonreligious positions had a positive or negative effect on U.S. society. The archived account reported substantially more negative views of atheism than of several named religious traditions, with especially negative responses among respondents identifying as born-again Christians.

The result may reflect unfamiliarity and the belief, common in that era, that shared religious commitment was necessary for moral behavior. It does not establish that respondents had direct experience with atheists, nor does it show how they would respond to every form of nonbelief.

For comparison, the site provides background on Islam, Buddhism, Judaism and Hinduism, as well as the basic definition of atheism.

ARIS research in 2001 and 2008

The American Religious Identification Survey reported growth in the share of U.S. adults who did not identify with an organized religion: from about 8% in 1990 to 14% in 2001, and approximately 15% in 2008. This broad “no affiliation” category included atheists, agnostics, secular people and others who did not select a religious identity.

Growth in non-affiliation does not necessarily mean that all respondents had adopted the label atheist. It does, however, show why older assumptions about a uniformly religious public became less reliable over time. Related perspectives are discussed in the pages on agnosticism and humanism.

University of Minnesota study reported in 2006

A national telephone study conducted by University of Minnesota researchers found that atheists were often placed near the bottom of a list of groups respondents considered compatible with their vision of American society. Respondents were also especially hesitant about the possibility of a child marrying an atheist.

The findings were interpreted as evidence that many Americans associated trustworthiness and moral conduct with religious belief. That interpretation describes a social assumption revealed by the survey; it does not demonstrate that atheists lack ethical commitments. For a broader comparison, see the collection on morality and ethics.

Why older polls need caution

• Definitions varied. “Atheist,” “agnostic,” “unaffiliated” and “nonbeliever” were frequently used differently across studies.
• Disclosure could be difficult. Respondents may have avoided identifying with a stigmatized position, particularly in telephone interviews.
• Question order mattered. Questions mentioning religion, morality or social trust could influence answers that followed.
• Samples were limited. A national survey can describe its sample without representing every region, generation or community equally.
• Historical results age. Social attitudes and terminology change, so figures from 1980–2008 should not be presented as current polling.

A broader view of religious freedom

Equal freedom of conscience includes the right to practice a religion, change religions, combine traditions, remain uncertain or hold no religious belief. It also includes the right to express criticism within the limits established by general law. Protecting atheists does not require agreement with atheism; it requires applying civic rights without making religious belief a condition of trust or belonging.

Modern readers can use these historical surveys as evidence of changing social boundaries, not as a scorecard for entire religious communities. A careful approach distinguishes prejudice from disagreement and treats every percentage as a product of its method.

References and further reading

The original essay cited Wendy Kaminer’s The Last Taboo, Barna Research material, the 2006 University of Minnesota study and the American Religious Identification Survey. Links and institutional pages change over time; readers should consult current primary sources before using historical figures in research.

For related topics, explore religious freedom, religious tolerance, religiously motivated hostility and the glossary.