Religious “Nones,” Networks & Neighborhoods
“I’m interested in stories of religious de-conversion,” my friend Daryl reflected. Daryl and I were sitting outside sharing a beer our economist friend David. The topic had turned to a discussion of our academic career and interests. Daryl is a leading psychologist of religion and an all-around good guy with a naturally curious disposition.
“I am interested in what drives someone to leave their religious tradition.”
You can understand his interest if you begin to look at to what is happening in the larger US religious landscape.
Let me set the context.
By definition, I am an older Millennial. This means that, according to recent research out of the PEW Foundation, I am a part of a cohort of individuals that is de-affiliating from their religious upbringing at an unprecedented rate. Amongst older Millennials like myself, 34% now now best identified as “Nones,” or people without a religious affiliation. The number for younger millennial is a slight tick higher at 36%. The contrast and trend becomes apparent when you compare these numbers to the 11% for the Silent Generation (born between 1928–1945), 17% for Baby Boomers (1946–1964) and 23% for Generation X (1965–1980).
In a preliminary read of the data, New York Times Op-Ed writer Ross Douthat cautions against an overly negative interpretation of the data (negative from the perspective of someone who wants affiliation high), suggesting that the key trend is decline in affiliation, but not necessary participation. Douthat then goes on to suggest what he thinks is driving the trend — the increasing age of first marriage. Historically, people fall away from the church when single, and then come back in part when they get married, and more often when they have kids. With the increase in age of marriage, people are staying away longer and longer, meaning either (for a positive take on religious affiliation) the nones will eventually become “somes” or (for the negative take), the nones will get used to non-affiliation and thus not come back if they eventually have kids. Given that age of marriage is tied in part to where people live — New York City Millennials are getting married later than those in my neck of the woods in Grand Rapids, Michigan — Douthat argues that we need to pay attention to where and how these Millennials live.
But let’s take a slightly different angle, also tied to geography. As I mentioned earlier, the other friend at the brewery that day was our friend David. David is a development economist who studies neighborhood effects. Ultimately, David is interested in things like the way you grow up affects different life outcomes, and the role of policy changes like increased public transit and transportation vouchers in mitigating these effects..
As David reminded educated us, one of the biggest findings in this area over the last few years is a changing consensus on the role of geography on human outcomes. For a significant period of time, the key study in the field was build around an experiment in the mid 1990s called “Movement to Opportunity.” This policy involved giving 4,600 families living in public housing the chance to enter a lottery which would move them to a better neighborhood. The hope in this work was simple. Give people a better place to live, and then watch them live better. But as economist Justin Wolfers writes in the New York Times, “Early findings had been disappointing, with no effect on the employment and earnings of parents, some positive effects seen on their physical and mental health, and few notable effects on their children.”
The problem of these early interpretations was that the analysis grouped all lottery winners together. As a result, it did not distinguish the age at which the exposure to good neighborhoods took place. When switching to this more nuanced lens, the findings look significantly different. Here is Wolfers again:
The children who moved when they were young enjoyed much greater economic success than similarly aged children who had not won the lottery. And the children who moved when they were older experienced no gains or perhaps worse outcomes, probably the result of a disruptive move, paired with few benefits from spending only a short time in a better neighborhood.
In other words, neighborhood matters, but so does the ‘when.’
So, how might this framework and understanding of place apply to the rise of the “Nones” As a network theorist, I think of neighborhoods a bit differently. Specifically, while the data on county poverty and religiosity is interesting, I don’t think it is fine grained enough. I tend to think that neighborhoods matter in so much as they give us access to the people that form our social world. In the case of religious belief, this means they shape the people I know, who i interact with, and what that makes plausible to believe. The reason I like networks over neighborhood is that the latter doesn’t lead directly to the former (though it obviously plays a large role). Let’s put it directly. You and I live in the same neighborhood of about 1000 people. Ours is a highly religious neighborhood. When we look at the census data, it indicates that 900 of these people are highly religious, and affiliate specifically with the Catholic Church. Within our neighborhood, each of us has 50 friends. But 40 of my friends are the non-Catholics, with 10 retaining some weak affiliation to the church. In contrast, your friendship more closely mirrors the neighborhood with 45 of the people you know spending a significant part of their time in the church. If networks shape beliefs, should we expect each of us to find the same beliefs plausible? Of course not.
So how might the network matter? To get to Daryl’s question, we have to understand what might cause someone to de-convert. In general, I can see people leaving a faith as a result of three of the following things.
- Decreasing plausibility / ‘believability’ of the worldview or tradition
- Decreasing belief in the uniqueness of those truth claims coming out of the tradition
- Decreasing stigma of de-conversion
Each of these mechanisms is a bit distinct. For example, I might find the beliefs on my own tradition (mainline protestant christianity) less believable (#1), I might continue to find them believable and valuable, but not think of them as delivering unique value over and above other traditions (#2). Furthermore, even if I fail to believe in the value (#1) and uniqueness (#2) of my tradition’s worldview, I might still affiliate because I am afraid of what people might think if I do ‘leave the faith.’ I would say in general, you have to have either #1 or #2, and it helps to have #3 to reduce friction. Ok, so to flesh it out, let’s consider three hypothetical people and three different networks (seen below).
In each of the three examples (left, middle, and right) the larger box in the center with the number is the focal person. The eight boxes around them are people in their network. For the sake of simplicity, we assume each person has three friends. If the colors are the same, they are aligned with the same belief system. If the colors are distinct, they represent different belief systems. The more transparent the color, the less ‘central’ the belief is to that person’s identity. In the example above, person one exists in a relatively homogenous network, with the majority of the friends having the same belief as a central part of their identity. Person two is a in a network that has two competing systems of belief, with the majority of their friends reflecting a strong centrality to their identity. Person three is a part of a more fragmented community with people of four different faith systems represented in their 8 friends.
For these three, the question becomes how the network shapes what they find believable. Put in language of de-conversion above, how does this network shape the ‘believability’ of their worldview, their sense of its ‘uniqueness,’ and their sense of the ‘social stigma’ of de-conversion. My gut says person one finds their worldview most naturally believable, but also has the largest stigma of conversion. Person two on the other hand might find their system unique in the sense that uniqueness might be increased with a clear sense of ‘the other,’ but might decrease when exposed to a bunch of other views (person three). Finally, I would think the third person might have a harder time with believability and uniqueness given that their friendship group reflects four different world-views, all held with varying levels of centrality.
So here is where things get tricky. As the economics of place research suggests, it’s not just what one’s neighborhood (or network) looks like, but also when they are exposed to different groups. Today, given the growing popularity of urban centers, more and more people are 1) growing up in places that look like person #3's network, and/or 2) migrating to places like #3 when they move to the city post college to work. The question thus becomes, what becomes of the likelihood of de-conversions and their respective narratives (about believability, about uniqueness, about decreasing stigma) as people’s network shifts, or as people shifts to the city in general. In the visual below, I identify a few different potential trajectories. In example 1, a person in a homogenous community moves to a city upon graduation and is immediately set into a more pluralistic world (this is the story for many of my students as they graduate and go into the workforce). Situation 2 is an example of someone who finds the tension of multiple communities tricky to manage and so they work towards crafting a more homogenous group of friends, no matter what neighborhood they enter (I actually see some of this behavior of friends who live in a larger city, but whose friendship group is relatively homogenous in regards to religious belief). The third example is someone who is born into a pluralistic world, and stays within it the rest of their life.
So, what might the story of de-conversion look for each of these three hypotheticals? What might it look for other combinations? David and Daryl, the ball is in your court…