The number of individuals in the United States who do not identify as being part of any religion has grown dramatically in recent years, and “the nones” are now larger than any single religious group. According to the General Social Survey, religiously unaffiliated people represented only about 5% of the U.S. population in the 1970s. This percentage began to increase in the 1990s and is around 30% today.

At first glance, some might assume this means nearly 1 in 3 Americans are atheists, but that’s far from true. Indeed, only about 4% of U.S. adults identify as an atheist.

As sociologists who study religion in the U.S., we wanted to find out more about the gap between these percentages and why some individuals identify as an atheist while other unaffiliated individuals do not.

Many shades of ‘none’

The religiously unaffiliated are a diverse group. Some still attend services, say that they are at least somewhat religious, and express some level of belief in God – although they tend to do these things at a lower rate than individuals who do identify with a religion.

There is even diversity in how religiously unaffiliated individuals identify themselves. When asked their religion on surveys, unaffiliated responses include “agnostic,” “no religion,” “nothing in particular,” “none” and so on.

Only about 17% of religiously unaffiliated people explicitly identify as “atheist” on surveys. For the most part, atheists more actively reject religion and religious concepts than other religiously unaffiliated individuals.

Our recent research examines two questions related to atheism. First, what makes an individual more or less likely to identify as an atheist? Second, what makes someone more or less likely to adopt an atheistic worldview over time?

Beyond belief – and disbelief

Consider the first question: Who’s likely to identify as an atheist. To answer that, we also need to think about what atheism means in the first place.

Not all religious traditions emphasize belief in a deity. In the U.S. context, however, particularly within traditions such as Christianity, atheism is often equated with saying that someone does not believe in God. Yet in one of our surveys we found that among U.S. adults who say “I do not believe in God,” only about half will select “atheist” when asked their religious identity.

In other words, rejecting a belief in God is by no means a sufficient condition for identifying as an atheist. So why do some individuals who do not believe in God identify as an atheist while others do not?

Our study found that there are a number of other social forces associated with the likelihood of an individual identifying as an atheist, above and beyond their disbelief in God – particularly stigma.

Many Americans eye atheists with suspicion and distaste. Notably, some social science surveys in the U.S. include questions asking about how much tolerance people have for atheists alongside questions about tolerance of racists and communists.

This stigma means that being an atheist comes with potential social costs, especially in certain communities. We see this dynamic play out in our data.

Political conservatives, for instance, are less likely to identify as an atheist even if they do not believe in God. Just under 39% of individuals identifying as “extremely conservative” who say they do not believe in God identify as an atheist. This compares with 72% of individuals identifying as “extremely liberal” who say they do not believe in God.

We argue that this likely is a function of greater negative views of atheists in politically conservative circles.

Adopting atheism

Stating that one does not believe in God, however, is the strongest predictor of identifying as an atheist. This leads to our second research question: What factors make someone more or less likely to lose their belief over time?

In a second survey-based study, from a different representative sample of nearly 10,000 U.S. adults, we found that about 6% of individuals who stated that they had some level of belief in God at age 16 moved to saying “I do not believe in God” as an adult.

Who falls into this group is not random.

Our analysis finds, perhaps unsurprisingly, that the stronger an individual’s belief in God was at age 16, the less likely they are to have adopted an atheistic worldview as an adult. For instance, fewer than 2% of individuals who said that “I knew God really existed and I had no doubts about it” as a teenager adopted an atheistic worldview later on. This compares with over 20% of those who said that “I didn’t know whether there was a God and I didn’t believe there was any way to find out” when they were 16.

However, our analysis reveals that several other factors make one more or less likely to adopt an atheistic worldview.

Regardless of how strong their teenage belief was, for instance, Black, Asian and Hispanic Americans were less likely to later identify as an atheist than white individuals. All else being equal, the odds of individuals in these groups adopting an atheistic worldview was about 50% to 75% less than the odds for white individuals. In part, this could be a product of groups that already face stigma related to their race or ethnicity being less able or willing to take on the additional social costs of being an atheist.

On the other hand, we find that adults with more income – regardless of how strong their belief was at 16 – are more likely to adopt the stance that they do not believe in God. Each increase from one income level to another on an 11-point scale increases the odds of adopting an atheistic worldview by about 5%.

This could be a function of income providing a buffer against any stigma associated with holding an atheistic worldview. Having a higher income, for instance, may give an individual the resources needed to avoid social circles and situations where being an atheist might be treated negatively.

However, there may be another explanation. Some social scientists have suggested that both wealth and faith can provide existential security – the confidence that you are not going to face tragedy at any moment – and therefore a higher income reduces the need to believe in supernatural forces in the first place.

Such findings are a powerful reminder that our beliefs, behaviors and identities are not entirely our own, but often shaped by situations and cultures in which we find ourselves.

Christopher P. Scheitle receives funding from the National Science Foundation and the John Templeton Foundation. The research presented here was supported in part by the Explaining Atheism project at Queen's University Belfast.

Katie Corcoran receives or has received funding from the National Science Foundation, the John Templeton Foundation, the Patient-Centered Outcomes Research Institute, Presbyterian Health Foundation, Society for the Scientific Study of Religion, International Research Network for the Study of Belief and Science, and West Virginia University Humanities Center.