Religious deconversion in adolescence and young adulthood: A literature review

Religious deconversion terminology

We use the term religiosity to mean religious behavior, affect, and cognition, which is synonymous with the term “religiousness.” Religiosity is a multidimensional construct (Hardy et al., 2019). First, the behavioral dimension includes public religious involvement (e.g. worship service attendance) and private religious practices (e.g. prayer and scripture study). Second, the affective dimension includes religious importance (i.e. salience), religious motivation (i.e. orientation or internalization), and religious identity (i.e. how central religion is to a person’s identity). Third, the cognitive dimension includes religious beliefs (e.g. belief in God, belief in life after death, and belief in sacred texts like the Bible) and religious knowledge (e.g. knowledge of the religious history, doctrine, and policies). These are continuous dimensions of religiosity, where people are relatively more or less religious. Another dimension of religiosity, often thought of as categorical, is religious identification. We define religious identification as the demographic religious category with which people most strongly identify. At the general level, people identify as religious or not religious. Those who identify as religious often have a specific religious affiliation, such as an overarching religious tradition (e.g. Christian) or even a particular religious denomination within that religious tradition (e.g. Protestant, or even more specifically Baptist, Methodist, or Lutheran). Those who are not religious, often called “unreligious,” “non-religious,” “religiously unaffiliated,” or simply “Nones,” typically fall into three categories: (1) atheists who do not identify with a religious affiliation and firmly believe there is no God, (2) agnostics who do not identify with a religious affiliation and are skeptical if there is a God, and (3) those who report “nothing in particular” when asked about their religion. However, there is likely more nuance to the category of Nones just as there is to religious people (e.g. Silver et al., 2014).

We define religious deconversion as any movement away from religion (Hardy & Longo, 2021). Religious deconversion is in many ways the opposite of religious conversion (Paloutzian et al., 2013). The study of religious conversion has a long history in psychology, by pioneers such as William James, G. Stanley Hall, and Edwin Starbuck (Rambo et al., 2014). Conversely, the study of religious deconversion has only emerged in recent decades (Streib, 2021). Religious conversion is often thought of as a discrete transition from identifying as non-religious to religious but may also involve becoming more religious (before and/or after identifying as religious; Paloutzian et al., 2013). This latter kind of religious conversion characterized by increased religiosity could be reflected in terms of the various continuous dimensions discussed above, such as religious behavior, cognition, and affect. Similarly, religious deconversion, as a mirror phenomenon, could involve a discrete transition from identifying as religious to non-religious, decreases in continuous dimensions of religiosity, or both (Hardy & Longo, 2021).

We distinguish the term religious deconversion from related terms, such as “religious disaffiliation” and “religious deidentification.” Often these three terms are used synonymously to mean the decision to no longer identify as religious, and thus become a None. However, recently Van Tongeren and DeWall (2021) differentiated religious disaffiliation as distancing oneself from a religious community from religious deidentification, which is no longer identifying with that religion, often with the former preceding the latter. We use the term religious deconversion as a general term capturing all the processes involved with becoming less religious on continuous dimensions of religiosity as well as categorical religious deidentification.

Religious deconversion trends

Despite religious growth in areas such as Africa and the Middle East, largely due to the spread of Islam, a lot of Western countries, such as the United States and many throughout Europe, are becoming less religious (Bullivant, 2019, 2022; Pew Research Center, 2017a). Currently, about a third of US adults do not identify with a religious affiliation (Public Religion Research Institute, 2023), which is roughly double the number from two decades ago and nearly triple from three decades ago (Pew Research Center, 2022). This means that currently the size of the Nones category in the United States is on par with the largest religious groups, such as Catholics and Evangelical Protestants (Bullivant, 2022; Burge, 2021). Indeed, despite small and steady increases in the American Muslim population, most religions in the United States are losing numbers (Pew Research Center, 2019). Part of this shift is due to religious deconversion, but over time an increasing percentage of the population is growing up without religion—raised by parents who are Nones (Bullivant, 2022; Hardy & Longo, 2021; Thiessen & Wilkins-Laflamme 2017). However, historical movements in religious identification should be considered in light of trends in numbers of births and deaths (Bullivant, 2019, 2022; Pew Research Center, 2017a; Sherkat, 2014).

Most data on trends in religious identification focus on adults in the United States, with limited available for youth or people in other countries. In the United States, one data source estimated that 32% of youth aged 13–17 years were Nones in 2019 (Pew Research Center, 2020), while another estimated that 38% of late adolescents and young adults aged 18–29 were Nones in 2022 (Public Religion Research Institute, 2023). And, as with adults, the percentage of the adolescents and young adults in the United States who are Nones has increased in recent decades. Only 10% of US high school seniors were Nones in the 1980s, but this doubled by the 2010s (Twenge et al., 2015), and as previously mentioned, may have tripled, or quadrupled since. Developmentally, adolescence and young adulthood are heightened times for religious deconversion (Hardy & Longo, 2021; Hardy & Nelson, 2022). As evidence of this, in the National Survey of Youth and Religion (NSYR), 16% of US youth were Nones in early adolescence (13–17 years old; 2002–2003), whereas only 5 years later 27% were Nones (18–23 years old; 2007–2008; C. Smith & Snell, 2009).

Globally, the regions of Asia-Pacific, Europe, and North America tend to have higher rates of Nones than Latin America-Caribbean, the Middle East-North Africa, and sub-Saharan Africa (Pew Research Center, 2015). Furthermore, throughout much of the world (Pew Research Center, 2018), particularly in the United States (Pew Research Center, 2020), younger generations are often less religious and more likely to leave religion than older generations. One data source suggests that in 60% of countries there are no generational differences in religious identification, while in about 40% of countries the younger generation (those 18–39 years old) is more likely to identify as None, and in rare cases the opposite is true (Pew Research Center, 2018). Specifically, the age gaps in 2018 go in descending order as follows: North America—Canada (51% of young adults were Nones; 23% of older adults were Nones), North America—United States (34% of younger adults were Nones; 17% of older adults were Nones), Europe (25% of younger adults were Nones; 15% of older adults were Nones), Latin America—Caribbean (13% of younger adults were Nones; 7% of older adults were Nones), Asia—Pacific (18% of younger adults were Nones; 14% of older adults were Nones), sub-Saharan African (2% in both age groups were Nones), and Middle East—North Africa (0% in both age groups were Nones).

Religious deconversion is not just about deidentifying from religion, but also decreasing religious behavior, affect, and cognition. In terms of religious behavior, 34% of US teens 13–17 years old attend church at least weekly, whereas 26% never do (Pew Research Center, 2020). Regarding prayer, 11% pray several times a day, while 18% never do (Pew Research Center, 2020). In terms of religious affect, 24% of US teens 13–17 years old report religion being “very important” to them, while 18% said religion was “not at all important” (Pew Research Center, 2020). In terms of religious cognitions, about 85% of US teens 13–17 years old believe in God, with 40% being “absolutely certain,” and about 14% do not believe in God, with 4% being “absolutely certain” (Pew Research Center, 2020). These trends also manifest historically—in the 1970s, 40% of high school seniors attended religious worship services weekly, whereas only 30% did by the 2010s (Twenge et al., 2015).

Unfortunately, limited data exist on trends in religiosity outside of the United States. Among adults in other countries, religious worship service attendance and religious importance tend to be lower in Asia, Australia, Canada, and Europe, and higher in Africa, the Middle East, and South America, compared with the United States (Pew Research Center, 2018). Young adults are included in these trends for adults, and we might assume that these cultural differences in religiosity among adults are similar for adolescents. However, as noted above, adolescents and young adults are less religious than older adults in many regions of the world (Pew Research Center, 2018).

Only a few dozen studies have systematically examined changes in religiosity across adolescence and young adulthood. Religious attendance typically declines across adolescence and into young adulthood, at least in the United States (e.g. Dyer et al., 2020; Pearce & Denton, 2011). Attendance may then stabilize across young adulthood (e.g. Koenig et al., 2008), followed by an increase in middle adulthood (Hayward & Krause, 2013). We know less about changes in other dimensions of religiosity. The few studies examining prayer over time report declines across adolescence and into young adulthood, at least in the United States (e.g. Dyer et al., 2020). Dyer and colleagues also noted a dampening of this decline in young adulthood. Findings regarding changes in religious importance across adolescence and young adulthood are more mixed. Some studies report stability (e.g. Lopez et al., 2011), while others report declines (e.g. Dyer et al., 2020). Findings for religious beliefs and doubts are similar, with some reporting stability (beliefs, Pearce & Denton, 2011; doubts, Hardy, Hendricks et al., 2023), while others report declines in beliefs (Schwadel, 2017). In short, the developmental trend across adolescence and young adulthood, at least in the United States, is a movement toward religious deconversion.

Nevertheless, there is interindividual variability around developmental trajectories of religiosity. Three main studies used person-centered analyses to identify classes of people with different developmental trajectories (B. H. J. Lee et al., 2017; Petts, 2009; Upenieks, 2021). All three studies used nationally representative samples of US adolescents and young adults. Although the number of and characteristics of the classes vary across studies, they found that some people remain stable (often staying low, medium, or high in their religiosity), some people decrease in their religiosity (and such decreases can occur early, late, or gradually), some people increase in religiosity, and some nonlinearly decrease and then increase. Thus, while the normative developmental trend looks like religious deconversion, some people are converting, reconverting, or remaining stable.

There is also evidence for nonlinearity in religious development. Normative developmental declines in religiosity across adolescence, at least in Western cultures, may level off in young adulthood (Dyer et al., 2020; Koenig et al., 2008; Schwadel, 2017). Then, as people move into middle adulthood they may modestly increase in religiosity (Hayward & Krause, 2013), perhaps as they take on new adult roles such as parenting. This suggests a normative developmental “reconversion” of sorts. In terms of religious identification, there is evidence that a substantial minority of people change religious identification more than once in their life (Pew Research Center, 2009). Thus, not all Dones remain non-religious—some may reconvert to religion. Using data from the National Study of Youth and Religion, a nationally representative study of 3,290 US teens from 2002 to 2013, we estimated religious reconversion rates. From Wave 1 to Wave 2, 11% left religion, but 35% of those reconverted to religion by Wave 3, and 42% of those were still religious at W4. Thus, 29% of Dones reconvert and stay for at least a few years.

Little is known about longitudinal trends in religiosity in other countries. On one hand, declines in religious worship service attendance were found among Croatian (Cranney et al., 2018) and Canadian (Hardy, Pratt, et al., 2011) youth. On the other hand, Indonesian Muslim youth may either evidence stability (French et al., 2011) or increases (French et al., 2014) in Muslim-specific public and private religious practices. Perhaps other Western cultures follow similar trends as the United States, while Eastern cultures follow different trends.

In short, what we are seeing, at least in many Western cultures, is both a historical and a developmental phenomenon. Historically, religious people are becoming less religious and the proportion of the population that is non-religious is growing. Developmentally, religiosity typically declines across adolescence and into young adulthood, and adolescence and young adulthood are the most likely periods for religious deidentification. According to ecological systems theory, individual changes are embedded within historical changes (Bronfenbrenner, 1992). As such, although religious decline and exit during adolescence and young adulthood is not a new phenomenon, it is compounding over time. Each generation seems to leave religion at higher rates than the one before (Bullivant, 2019, 2022; Sherkat, 2014).

Processes of religious deconversion

Religious deconversion, like other change processes, involves individuals embedded within layers of interconnected social contexts embedded within historical time. This idea was initially proposed by Bronfenbrenner (1992) in his ecological systems theory and recently extended by Overton (2015) and Lerner et al. (2015) with relational developmental systems theory. Lerner emphasizes the active role of the individual in their own development, the dynamics of various processes within the individual (e.g. biological, cognitive, and emotional), and the bidirectional nature of relationships between individuals and contexts (e.g. family, peers, religious communities, schools, media, and culture). Within this framework, religious deconversion would be viewed as a complex and dynamic process.

Theory and research are beginning to elucidate processes of religious deconversion (for an earlier review, see Hardy & Longo, 2021). Nearly three decades ago, Barbour (1994) proposed a model of religious deconversion recently expanded by Streib (2014). According to Streib’s version, there are five components to deconversion that may roughly proceed in this order: (1) a loss of religious or spiritual experiences; (2) intellectual doubt of, denial of, or disagreement with religious beliefs or practices; (3) moral criticism, including the rejection of the way of life of members of one’s religious group; (4) emotional suffering, such as grief, guilt, despair, and loneliness; and (5) eventual disaffiliation from one’s religious community. These processes can lead to various religious deconversion trajectories distinguished by where the individual started and where they ended up in terms of religious identification and levels of religiosity (Streib et al., 2009). Streib and colleagues (2009) also showed how these processes are similar across the United States and the United Kingdom, despite potential descriptive differences.

Given that religious deconversion seems to be a mirror phenomenon to religious conversion, some have taken models for religious conversion and flipped them to apply to religious deconversion. For instance, Marriott (2020) adapted Rambo’s (1995) stage model of religious conversion to become a seven-stage model for religious deconversion. First, the context includes characteristics of the individual as well as the various contexts in which they are embedded, particularly their family and religious community. For instance, often people who leave religion come from authoritarian homes and/or fundamentalist religions (Streib et al., 2009). Second, there is a faith crisis, which could entail a variety of religious and spiritual struggles, such as doubts about religious beliefs, difficulty integrating science and religion, trouble accepting church policies and procedures, negative interactions with other church members or church leaders, thought-provoking interactions with non-believers, or stressful life events. Third, there is truth seeking, where people look more deeply into the issues that drove their faith crisis. They may consult various sources such as family, friends, church leaders, social media, internet searches and websites, church publications, scriptures, and even God. Fourth, individuals attempt to retain their faith. Many skip this step, but some make a final effort to avoid deconversion, given the potential benefits of staying and perceived costs of leaving their religion. Fifth, there is a move to agnosticism. This is essentially a step in the direction of atheism, reflecting ambivalence toward one’s belief in God. Sixth, individuals become atheist. At this point, they have a high level of certainty regarding the non-existence of God. Seventh, they embrace and publicly express their new identity as atheists.

Recently, Van Tongeren and DeWall (2021) used Saroglou’s (2011) universal multidimensional model of religion as the basis for a model of religious deconversion. The four components of Saroglou’s model are belief, bonding, behaviors, and belonging. Thus, four processes that may ultimately lead to religious deidentification involve moving away from belief, bonding, behaviors, and belonging. First, shifting from belief to disbelief entails doubt in the existence of God, truthfulness of scripture, or validity of various religious doctrines and practices. Second, disengaging rather than bonding implies stopping rituals intended to facilitate connection to God, often in the form of ceasing public (e.g. worship service attendance) and private (e.g. prayer and scripture study) religious practices. Disengaging also involves negative emotions, such as feeling angry at, distant from, or abandoned by God. Third is discontinuance from both prescriptive and prospective behaviors—individuals stop living shared religious norms. Fourth is disaffiliation from community belonging. At this point they fully disconnect and cease attending worship services and social gatherings with members of their religion. They may also sever social ties with those members, which results in lost relationships with neighbors, friends, and family. These four processes may not occur in this specific sequence but depend upon the specific experiences that trigger the deconversion process. All four processes typically play a role in religious deidentification, when people decide to no longer identify with a religion.

Fisher (2017) proposed a model of religious deconversion focused on the role of doubt, at least among those in Western cultures. It often starts with questions, particularly those related to religion and science (e.g. evolution vs creation), social issues (e.g. religious intolerance toward sexual minorities), and the problem of suffering in the world (Altemeyer & Hunsberger, 1997). When these questions are not satisfactorily resolved, they can lead to doubt. This emergence of doubt is often triggered by negative life events or exposure to information critical of religious beliefs (Hunsberger et al., 1996), or when secular explanations seem more viable than religious explanations (J. M. Smith, 2011). Doubt is more likely to emerge and grow given certain individual and contextual features, such as affinity for scientific explanations (J. M. Smith, 2011), negative family relationships (Hunsberger et al., 2002), and negative interactions with other religious individuals (Krause & Ellison, 2009). At some point, doubt becomes sufficient that individuals feel compelled to make a change—they can leave religion, switch religions, or reconfigure how they live their current religion. If reconfiguration does not seem feasible or desirable, and switching religions will not seem to resolve doubts either, then religious deconversion progresses. Individuals move away from religion in terms of their beliefs, practices, and interactions with the religious community. At some point they may officially deidentify from religion, and then maybe even become oppositional toward religion.

As these religious deconversion models suggest, people typically move gradually away from religion before religious deidentification. In other words, deconversion is typically a long process rather than a rapid, discrete life event. In the United States, adolescent religious deidentification is preceded by decreases in dimensions of religiosity, specifically worship service attendance, prayer, scripture study, religious importance, and spirituality (Hardy, Hendricks, et al., 2023). Thus, usually teens gradually became less religious before deciding to no longer identify as religious. However, sudden and profound religious conversion and deconversion experiences can occur (Skalski & Hardy, in press). While noteworthy these experiences are uncommon.

The religious deconversion process is not finalized by religious deidentification. Recent studies have demonstrated a “religious residue effect” in various Western and Eastern cultures (e.g. the United States, the Netherlands, New Zealand, and China). Religious behavior, cognition, and the affect of people who leave religion (“Dones”) tend to fall somewhere between that of people who have never been religious and people who remain religious, and this seems to hold across various Western (e.g. the United States, the Netherlands, and New Zealand) and Eastern (e.g. China) cultures (Van Tongeren, DeWall, Chen, et al., 2021). In other words, the religiosity of individuals who deidentify from religion does not instantaneously match that of other Nones. Rather, their religiosity gradually decreases toward that of Nones, and this process continues after religious deidentification. This residual trend is also seen with prosocial behaviors (Van Tongeren, DeWall, Chen, et al., 2021), moral intuitions (Van Tongeren, DeWall, Hardy, et al., 2021), and values (Saroglou et al., 2020; Schwadel et al., 2021), where Dones similarly fall someone between religious people and other Nones. Furthermore, the more time passes since religious deidentification, the more Dones resemble other Nones (Van Tongeren, DeWall, Hardy, et al., 2021). However, a study of adults found that while some Dones still engage in religious practices and hold positive attitudes toward religion, others do not (McLaughlin et al., 2022).

Clearly, there is more than one religious deconversion path. The general deconversion processes outlined above manifest differently depending on individuals and contexts. Indeed, Streib and colleagues have identified six religious deconversion trajectories (Streib et al., 2009). These trajectories are largely characterized by where the individual is moving from and to, religiously speaking. First is secularizing exit, where people leave organized religion entirely and become Nones. Second is oppositional exit, where people move from a well-integrated religious group to one that exists on the periphery of mainstream society. Third is religious switching, where people shift from one well-recognized and integrated religion to another. Fourth is integrating exit, where people move from a minority or peripheral religion to one that is more accommodating of or integrated with broader society. Fifth is privatizing exit, where people terminate their formal membership with a religion but otherwise keep living their religion in private. Sixth is heretical exit, where people change their religious beliefs and behaviors without changing their religious identification. Only secularizing exit involves leaving religion entirely.

Retrospective research also captures individual differences in religious deconversion trajectories by directly asking people who have deconverted why they left religion. Numerous books have been published detailing such self-proclaimed reasons for leaving religion, at least among adolescents and young adults in the United States (for a review, see Hardy & Longo, 2021). First, those who leave religion feel unaccepted or out of place in their religious community. Second, they feel their religion is too intolerant and exclusive. Third, they experience their religious community as being overprotective and repressive. Fourth, they feel shame for not living their religion, particularly its prescriptions and proscriptions. Fifth, they view religion with intellectual skepticism and disbelief, particularly the truthfulness of scripture (e.g. the Bible). Sixth, they think their religion is anti-scientific and their religious community has no place for doubt or religious questioning. Seventh, they think their religion is irrelevant and shallow. Eighth, they feel religion is inconvenient or boring. Ninth, they have experienced a salient life change, a difficult life challenge, or a stressful life event. Tenth, they disagree with their religion or religious leaders regarding doctrinal, moral, or political issues, often about sexuality and reproduction. Thus, the experience of deconversion often involves certain common threads of behaviors, thoughts, emotions, and relationships. Although these typical reasons for deconversion are echoed across numerous sources, more systematic qualitative research is needed to map out reasons for deconversion and examine differences across populations.

Predictors of religious deconversion

As discussed earlier, religious deconversion is a change process involving individuals embedded within multiple layers of interconnected social contexts. Prior research has identified numerous individual predictors of deconversion. First, in terms of demographics, at least among US adolescents and young adults, males tend to be less religious than females (C. Smith & Denton, 2005) and may be more likely to decline in religiosity (B. H. J. Lee et al., 2017; Lopez et al., 2011) and to deidentify from religion (Hardy, Hatch, et al., 2022). Nevertheless, currently in the United States, male youth are only slightly more likely to be religiously unaffiliated than females (33% vs. 32%, respectively; Pew Research Center, 2020). This trend in gender differences of males generally being less religious than females has also been found in Europe (Robinson et al., 2019), and perhaps may hold in other regions as well. Regarding ethnicity, African American youth in the United States are more religious than those in other ethnic groups (C. Smith & Denton, 2005), and may be less likely to decline in religiosity (Petts, 2009). More generally, ethnic minority youth in the United States are less likely to be religiously unaffiliated than White, non-Hispanic youth (28% vs. 36%, respectively; Pew Research Center, 2020).

One demographic that may be of particular importance to religious deconversion is sexual orientation. Religious identities may conflict with sexual and gender identities due to religious proscriptions against non-heterosexual sexual identities and non-cisgender gender identities (Wolf & Platt, 2022). This identity conflict can cause severe psychological strain, even suicidality, unless successfully resolved through either prioritizing one or the other identity or effectively integrating them (Wolf & Platt, 2022). As such, religious adolescents and young adults who are LGBTQ (Lesbian, gay, bisexual, transgender, and queer) are at a heightened risk for religious deconversion. For example, US young adults who are sexual minorities are more likely to decrease their religious attendance and deidentify from religion over time across adolescence and young adulthood (Woodell & Schwadel, 2020). Specifically, in one dataset (Add Health) 25% of those with same-sex attraction deidentified from religion, compared with 14% for heterosexuals, while in a second dataset (National Study of Youth and Religion) these were 43% and 27%, respectively. However, LGBTQ concerns can also affect those who are not LGBTQ themselves. Often the reasons people give for leaving religion have to do with concerns regarding sexuality and gender, including perceived judgment toward the LGBTQ community (Hardy & Longo, 2021).

Second, in terms of personality, traits such as conscientiousness, agreeableness, and openness to experience tend to be positively associated with adolescent and young adult religiosity (for a meta-analysis, see Yonker et al., 2012). In other words, certain types of people may be more inclined toward religion than others, and thus less inclined to deconvert. One longitudinal study of adults in New Zealand found that changes in personality traits may precede rather than follow deconversion (Stronge et al., 2021). Specifically, honesty/humility increased while agreeableness decreased prior to deconversion. However, interestingly a longitudinal study of Australian adolescents found that religiosity predicted later personality, but not vice versa (Huuskes et al., 2013).

Moving beyond the big five personality traits, a cross-section study of Polish adolescents found that those higher in social anxiety were more likely to have reported recent declines in their religiosity, and this was mediated by their tendency toward social comparison (Zarzycka, Razmus, & Krok, 2023). Similarly, a longitudinal study of US teens found that those who were less social competent at the initial wave were more likely to leave religion 3 years later (Longo & Kim-Spoon, 2014). Two additional studies of Polish youth demonstrate the complex role of identity formation processes in religious deconversion. Those less sure about their identity were more likely to deconvert (Puchalska-Wasyl et al., 2022), while those who relied heavily on social norms in their identity exploration were less likely to deconvert (Zarzycka, Tomaka, et al., 2023). Also, a longitudinal study of Chinese adults found that, prior to their religious deidentification, those who left religion were lower on emotional stability and social-focused values (e.g. conformity, tradition, and benevolence), and higher on personal-focused values (e.g. self-direction, stimulation, hedonism, and achievement) than those who stayed religious (Hui et al., 2018).

Third, norm-relevant behaviors may also affect religiosity over time. Specifically, youth who violate prescriptions and proscriptions are more likely to subsequently decrease their religiosity (Regnerus & Uecker, 2006; Vasilenko & Lefkowitz, 2014). Such violations are also provided by deconverts as reasons for leaving religion (Hardy & Longo, 2021), although this has not been quantitatively demonstrated. Few studies have addressed norm-relevant behaviors as predictors of changes in religiosity and most have focused on US populations.

Fourth, some of the most salient predictors of religious deconversion are various dimensions of individual religiosity. For example, decreases in perceived religious importance across adolescence in the United States are associated with decreases in worship service attendance (Desmond et al., 2010). Furthermore, religious importance and worship service attendance are bidirectionally linked over time in the United States, with attendance more strongly predicting importance in early adolescence, and the inverse in late adolescence (Dyer et al., 2020). As noted earlier, decreases in several dimensions of religiosity (e.g. worship service attendance, prayer, scripture study, religious importance, and spirituality) across adolescence in the United States are predictive of religious deidentification in young adulthood (Hardy, Hendricks et al., 2023). Furthermore, various dimensions of individual religiosity in adolescence uniquely predict subsequent religious deidentification when controlling for each other (worship service attendance, religious youth group attendance, prayer, religious importance, and doubt), with youth lower on the first four dimensions and higher on doubt being more likely to deidentify (Hardy, Hatch, et al., 2022).

Indeed, doubt could play a potentially critical role in religious deconversion. Doubt is a part of a broader category of what Exline and colleagues (2022) call religious and spiritual struggles, which are experiences of conflict, tension, or distress pertaining to religion and spirituality. They outline six domains of religious and spiritual struggles: divine struggles (e.g. anger toward God, fear of divine punishment, or fear of being abandoned by God), demonic struggles (e.g. anxiety about being attached or tempted by the devil or evil spirits), interpersonal struggles (e.g. conflicts with or negative emotions toward religious people or institutions), moral struggles (e.g. difficulty following religious prescriptions and proscriptions and the associated negative affect such as guilt), ultimate meaning struggles (e.g. concerns about a lack of meaning in life), and doubt-related struggles (e.g. questions or doubts about religious beliefs). In a study of US college students, Exline and colleagues (2022) compared religious and spiritual struggles across individuals with various trajectories of religiousness. Those who had consistently been religious or non-religious had fewer religious and spiritual struggles than those who deconverted, whether they officially deidentified from religion or not. In addition, Canadian adolescents higher in religious doubt were more likely to decrease in worship service attendance (Hunsberger et al., 2002). Furthermore, all religious deconversion models outlined above include doubt as fundamental to the process (Fisher, 2017; Marriott, 2020; Streib, 2014; Van Tongeren & DeWall, 2021), and individuals report doubt as being one of the key reasons why they left religion (Hardy & Longo, 2021).

Prior studies have also identified numerous contextual predictors of religious deconversion. The family is the most salient and proximal context for religious development (for a review, see Hardy et al., 2019). Parent or family religiosity is connected to adolescent and young adult religiosity (Hardy, White, et al., 2011), and predictive of religiosity change trajectories across these life phases (Petts, 2009; Regnerus & Uecker, 2006) in the United States. For example, youth from less religious families in the United States tend to have low and stable or low and declining religious trajectories rather than high and stable or even high but declining trajectories (Goodman & Dyer, 2020; Petts, 2009). In terms of deidentification, individuals from less religious families or who have family members who have left religion are more likely to leave religion themselves (sample of Canadian college students, Altemeyer & Hunsberger, 1997; sample of US families, Bengtson et al., 2008). More specifically, lower parent worship service attendance, lower parent religious importance, and fewer family religious conversations in adolescence predict greater likelihood of later religious deidentification in the United States (Hardy, Hatch, et al., 2022). In addition, in this study the link between family religiosity and adolescent deidentification was mediated by adolescents’ individual religiosity. Specifically, in less religious families teens are less religious, and thus more likely to eventually deidentify from religion. Family religiosity also played a more salient role in deidentification than peer religiosity and religious community supportiveness. This link between parent or family religiosity and religious deconversion seems to hold across European (e.g. Poland; Łysiak et al., 2020) and Asian (e.g. Indonesia; French et al., 2013) cultures.

Family dynamics, such as parenting style and parent–child relationship quality, also influence adolescent religious deconversion (for a review, see Hardy et al., 2019). For example, US youth who are less satisfied with their family relationships tend to decrease more in religiosity through adolescence, even accounting for parent religiosity (Regnerus et al., 2004). Furthermore, poorer quality family relationships predict US adolescents leaving religion (Longo & Kim-Spoon, 2014). In Indonesia, family religiosity and family dynamics interact in that family religiosity more strongly predicts adolescent religiosity when parents use an authoritative parenting style characterized by warmth, structure, and autonomy support (French et al., 2013). Thus, adolescent religious deconversion is more likely in religious families where parents use other parenting styles, such as authoritarian (where they may pressure youth too much regarding religion), permissive (where they pressure youth too little regarding religion), or neglectful (where there is the absence of a parent–child relationship). In fact, religious parents who use authoritarian parenting practices risk child religious embitterment (Bornstein et al., 2017). While most research on family and religion involves US families, there is evidence for the importance of parenting and family relationships to religious development and deconversion in European (e.g. Poland; Łysiak et al., 2020) and Asian (e.g. Indonesia; French et al., 2013) cultures.

Surprisingly little research exists on family religious socialization. In other words, there is a fair amount of work on how family religiosity and family dynamics relate to adolescent religious development, but very little specifically on how parents deliberately go about nurturing the religious development of their children. As notable exceptions, Christian Smith and colleagues recently published two books articulating their theoretical and empirical work on the issue (C. Smith & Adamczyk, 2021; C. Smith et al., 2019). One major theme in their research is that parents are undoubtedly the most important source of religious socialization. In addition, they suggest the most effective mode of religious socialization is modeling—or as the saying goes, “practice what you preach.” Similarly, if the family is living their religion, youth often internalize their own faith as a matter of course, seemingly unconsciously. Nevertheless, parents can aid in strengthening religious development by providing youth opportunities to practice their religion, such as public and private forms of worship, as well as serving others. As such, effective religious socialization is more about practicing and less about preaching—that is, didactic teaching should be lower on the list of strategies. Discussions can be productive if they are youth-led. Parents can share experiences and ideas if they minimize preaching. Parents should be patient and expect gradual religious development in youth. Along the way, parents can progressively release responsibility to youth for their own faith development. Parents need to balance their desire for successful faith transmission with respect for their children’s autonomy (Barrow et al., 2021). In short, parents can help prevent adolescent religious deconversion if they are thoughtful and diligent about religious socialization but avoid being overbearing.

Peer relationships are another important context for adolescent religious deconversion. US youth with fewer religious friends who talk less frequently with their friends about religion are less likely to be religious themselves (Pearce & Denton, 2011). Furthermore, US adolescents with less religious friends are also more likely to decrease in religiosity across adolescence (Regnerus & Uecker, 2006). Even more, among Muslim youth in Indonesia, changes in individual and peer religiosity are linked over time, possibly due to processes of influence and selection (French et al., 2014). Analyses of friendship networks in the United States (Cheadle & Schwadel, 2012) and Indonesia (Shen et al., 2022) suggest the priority of influence processes over selection processes in longitudinal links between individual and peer religiosity. Peers also play a significant role in religious identification shifts because adolescents tend to consult their peers more than their parents when contemplating such shifts. Retrospective accounts from Canadian adults reveal that discussions with peers during adolescence can lead to any kind of shift, including conversion, deconversion, or switching religions (Altemeyer & Hunsberger, 1997). Furthermore, US youth who have fewer religious friends, fewer friends attending religious youth groups, and fewer religious conversations with friends are more likely to subsequently deidentify from religion (Hardy, Hatch, et al., 2022). In fact, as US college students deconvert they are drawn to form social networks with others who are deconverting or have deconverted, particularly in contexts where most people are religious (Sepulvado et al., 2015).

The religious community is another influential context for religious development. When US adolescents and young adults are not engaged in their religious community or have negative perceptions of it, they tend to be less religious (Pearce & Denton, 2011). Similarly, Orthodox Jewish youth in Israel reported moving away from religion because they felt less accepted in their religious community (Itzhaki et al., 2020). In addition, if youth in the United States do not feel their religious community is welcoming with leaders who are open to discussion and truly living their religion, they are less likely to be religious in young adulthood (C. Smith & Snell, 2009). Furthermore, when youth in the United States feel their religious community is not engaging and open to dialogue about difficult issues, they are more likely to deidentify from religion (Hardy, Hatch, et al., 2022). Finally, in a study of Polish adolescents, those who perceived greater hypocrisy among those of their religious community were more likely to exhibit thoughts, emotions, and behaviors related to religious deconversion (Zarzycka, Tomaka, et al., 2023).

Another important developmental context of religious deconversion is education. Given that most educational institutions in Western countries are secular and often liberal, the popular assumption is that more educated people are less religious (Burge, 2021). Indeed, while this might have been true in the past, today the issue is much more complex. The link between education and religion seems to be conditional upon religious affiliation. For example, in the United States, members of the Church of Jesus Christ of Latter-day Saints (i.e. “Mormons”) who are more educated tend to be more religious, while among people of the Jewish faith, those who are more educated tend to be less religious (Pew Research Center, 2017b). Another data source shows that declines in religiosity are particularly likely to follow higher education for mainline Protestants (Schwadel, 2017). The family context also matters, as college education among young adults in the United States predicts less religious attendance for those who grew up with high-attending parents but predicts more religious attendance for those with low-attending parents (Schwadel, 2017). The link between education and religion also depends on the dimension of religiosity, as being in college at all predicts declines in prayer, while staying to obtain a bachelor’s degree predicted declines in religious belief and certainty (Schwadel, 2017). The link between education and religious deconversion outside of the United States is unclear, but the dynamics are likely different in countries with a higher percentage of faith-based or religious schools, such as some areas of South American, Africa, and Europe. Presumably the tension between religion and education is less when the two are intentionally integrated (Anthony, 2018).

An increasingly important but little understood context of religious deconversion is social media. This communication medium exposes people to new ideas and information that can validate existing questions and doubts or provoke new ones. Social media can also connect people with a community of others in faith crises or who have already left religion, give people a chance to explore secular identities, and provide a platform for sharing their religious questions and experiences (Bullivant, 2019; Hafen & Hafen, 2018). However, few studies have examined this issue. A qualitative study of religious deconversion narratives posted online by individuals of various ages showed that social media provided opportunities for people to construct and try out secular identities, explore doubts, and process their emotions and experiences throughout the deconversion process (Starr et al., 2019). Also, social media use during young adulthood provides exposure to a breadth of ideas, which is linked to more liberal views about religion such as endorsing beliefs and practices from multiple religious traditions (McClure, 2016). One consideration is that posts on social media tend toward the extremes (Burnett et al., 2022), suggesting that much of what is posted about religion online is fanatical against or toward religion. This polarized information about religion poses a challenge for those using social media as a primary source of information and connection in their faith journey.

Outcomes of religious deconversion

Even less research exists regarding outcomes. Often developmental processes involve bidirectionality (Lerner et al., 2015), so predictors of religious deconversion could also be outcomes. Furthermore, given prior work demonstrating potential adaptive outcomes of being religious (Hardy et al., 2019), logically we would anticipate that people leaving religion would be losing some of the potential benefits of religion. Unfortunately, there is limited research directly examining outcomes of religious deconversion. One longitudinal study of US teens and young adults found that those who were religious across adolescence and young adulthood were higher than Nones on binding moral foundations (i.e. ingroup loyalty, respect for authority, and concerns for purity), while those who deconverted seemed to progressively approach the level of Nones the longer they were away from religion (Van Tongeren, DeWall, Hardy, et al., 2021). In line with this, a cross-sectional study of adults found that religious people were the most prosocial, followed by Dones and then followed by Nones (Van Tongeren, DeWall, Chen, et al., 2021). In short, research on the religious residue effect suggests that when people deconvert from religion they gradually move away from the values and behaviors of religious people.

Other studies have noted possible outcomes of religious deconversion relevant to individual and relational well-being. On one hand, deconversion is often challenging for individuals and families. Indeed, one cross-sectional study of US adults suggest that deconversion could lead to losses in positive affect and interpersonal connection (Krause & Pargament, 2017). Furthermore, a cross-sectional study of Polish teens found religious deconversion predictive of lower life satisfaction and family support (Zarzycka et al., 2022). As additional evidence for potential individual challenges of deconversion, in a nationally representative study of US adults, deconverts had lower health and well-being than people who were consistently religious or not—although directionality is uncertain (Fenelon & Danielsen, 2016). As additional evidence for potential relational challenges of deconversion, a longitudinal study of US families found that discrepancies between the religious identification and degree of religiosity of parents and their young adult children is predictive of lower relationship cohesion (Hwang et al., 2018).

These findings are all buttressed by qualitative research on religious deconversion involving young adult deconverts from the Church of Jesus Christ of Latter-day Saints in the United States (Jindra & Lee, 2023), adult deconverts from Evangelical Protestant Christianity in the United Kingdom (K. A. Lee & Gubi, 2019), young adult and adult deconverts from Christianity in the United States (Marriott, 2020), and adult deconverts from the Jehovah’s Witnesses in Italy (Testoni et al., 2019). Individuals who deconvert often experience a range of negative emotions: regret, shame, and embarrassment regarding their prior religious life; anger against organized religion and religious people, including family; loneliness, isolation, and sometimes ostracism; depression and loss over losing their faith and faith community; and a sense of losing their identity and worldview, including their primary source of values, meaning, and purpose (Jindra & Lee, 2023; K. A. Lee & Gubi, 2019; Marriott, 2020; Testoni et al., 2019). Families where an individual has deconverted often experience relationship strain, such as awkwardness, conflict, and even estrangement (Jindra & Lee, 2023; Knight et al., 2019; Testoni et al., 2019).

Fortunately, such negative outcomes of religious deconversion can be mitigated over time as individuals reconstruct their identity and worldview (Nica, 2020) and social networks (Nica, 2019) as non-religious people (Jindra & Lee, 2023; Testoni et al., 2019). Families may also learn to resolve religious discrepancies so relationships are sustained and healed (Colaner et al., 2014). These processes are increasingly technologically mediated, as individuals can more easily connect with others going through similar experiences (Sepulvado et al., 2015). Also, perhaps there will be more resources and social support for deconverts over time, as the number of Nones and Dones increases (Bullivant, 2022), and more research is done on the deconversion process.

On the other hand, deconversion can lead to positive outcomes, at least for some individuals in some situations. These findings stem largely from qualitative research on religious deconversion. Individuals who deconvert may experience more authenticity and less cognitive dissonance, more happiness and peace, freedom from the boundaries of religious doctrines and norms, release from pressures to live according to those doctrines and norms, flexibility to decide their own values and lifestyle, and diminished shame and guilt regarding what was once deemed “sin” (Jindra & Lee, 2023; K. A. Lee & Gubi, 2019; Marriott, 2020; Testoni et al., 2019). In addition, deconverts often feel greater compassion and acceptance of others with diverse ideologies and lifestyles (Marriott, 2020). This accords with prior work suggesting that religious youth can be more prejudiced toward certain outgroups, such as non-heterosexuals (Shepperd et al., 2021).

The extent to which deconversion is a positive or negative experience varies depending on the characteristics of individuals and contexts. A longitudinal study of Chinese adults found that some experienced decreases in well-being following deconversion while others experienced increases, with those who started off lower in life satisfaction experiencing poorer outcomes (Hui et al., 2018). In addition, outcomes of deconversion can vary depending on the type of religion from which one is exiting. For example, deconversion from covenantal religions such as Latter-day Saints, Amish, and Orthodox Jews may be more challenging because social, emotional, and even physical costs may be higher compared with leaving mainstream religions (Engelman et al., 2020).

Future directions

The study of religious deconversion has been around for decades but has only recently gained momentum in the field. This growing interest in religious deconversion is likely due to recognition of the sharp increase in rates of religious deidentification in Western societies. This trend has caught the attention of religious organizations, funding agencies, and social scientists. Nevertheless, given the nascent state of research on religious deconversion, there remains much to be discovered. Here we outline 10 directions for future research on religious deconversion.

First, qualitative and quantitative research is needed regarding self-perceived reasons for leaving religion. Prior work has identified common themes in reported reasons for leaving religion, but this has been largely informal and unsystematic (for review, see Hardy & Longo, 2021). While individuals themselves likely do not have a complete understanding of their own religious deconversion process, their perspective is nonetheless critical.

Second, more research is needed identifying prospective predictors of religious deconversion. Given that religious deconversion is a process that unfolds over time, it needs to be examined using longitudinal methods to be fully understood. Longitudinal studies also add to what we learn from retrospective self-reports, as researchers can examine elements of the deconversion process outside of individual awareness. Two recent studies provide examples of such longitudinal research. One study found that decreases in various dimensions of religiosity tend to precede religious deidentification (Hardy, Hendricks, et al., 2023). The other found that various aspects of individual religiosity, family religiosity, peer religiosity, and religious community supportiveness were predictive of later religious deidentification, and that individual religiosity also mediated relations between the developmental contexts (family, peers, and religious communities) and religious deidentification (Hardy, Hatch, et al., 2022).

Third, more research is needed exploring the potentially important role of religious and spiritual struggles, questioning, and doubt, or what is termed a “faith crisis.” Indeed, as discussed earlier, a faith crisis is often an important catalyst for religious deconversion (Marriott, 2020). However, much remains to be learned about the role of faith crises in the deconversion process. For instance, in identity formation research an identity crisis, or at least some degree of identity exploration, is critical to successful identity formation (for a review, see Meeus, 2023). Without identity exploration either youth remain without a firm sense of identity (i.e. “identity diffusion”) or unquestionably take on an easily accessibly identity (i.e. “identity foreclosure). Either is less adaptive than creating or discovering an identity after a period of exploration (i.e. “identity achievement”). If a similar dynamic is at work in religious development, then a faith crisis could potentially lead to a deepening of faith. Indeed, perhaps some level of religious questioning is necessary for successful religious development—that is, for religious commitment to be freely chosen and internalized (McLaren, 2021). This may likely be the case, given the intertwined nature of identity and religious development (Hardy, Pratt et al., 2011).

Fourth, more research is needed for examining the role of sexual and gender identity in religious deconversion. Several studies, both qualitative (Dahl & Galliher, 2012) and quantitative (Lefevor et al., 2017), have revealed challenges and opportunities of LGBTQ religious adolescents and young adults. As noted earlier, one of the main struggles seems to be integrating religious identity with sexual and gender identity, given religious norms (Wolf & Platt, 2022). We know that due to this and other factors, LGBTQ individuals are more likely to deconvert from religion (Woodell & Schwadel, 2020). But we do not know all the processes involved.

Fifth, research is needed examining the role of the integration of science and religion in the religious deconversion process. Some of the frequently reported reasons for leaving religion relate to potential difficulties with integrating science and religion (Hardy & Longo, 2021). For example, some youths are skeptical of religious doctrine and scripture due to a perceived lack of supporting scientific evidence. In addition, other youth see their religion as antiscientific and unaccepting of doubt. In both cases, these youth struggle integrating science and religion, so they ultimately leave religion. Little is known about the process of integrating science and religion in adolescence, although work is emerging on the issue among young adults, particularly in higher education contexts (Uecker & Longest, 2017). Belief in science and belief in religion are in tension (O’Brien & Noy, 2021), so people want to reject one or the other, or integrate them.

Sixth, research is needed exploring the role of cognitive complexity, or intolerance for ambiguity or uncertainty, in religious deconversion. Several sources seem to converge on suggesting that religious development progresses through three general phases (although the specific number of phases depends on the source): simplicity, complexity, and then some form of resolution beyond complexity (Hafen & Hafen, 2018; McLaren, 2021; Streib et al., 2009). Generally, people seem to start off with simplistic views of religion, aligned with early stages of cognitive development. However, often during adolescence or young adulthood, as cognitive development progresses toward abstract thinking, they confront the complexities of reality and the inadequacy of their simplistic views of the world, including their religious beliefs. Many people who leave religion do so at this stage. However, if they persist through ambiguity and uncertainty, they may come to accept or resolve such complexities. Yet, little empirical research has examined the role of cognitive complexity and tolerance for ambiguity or uncertainty in processes of religious deconversion, and findings have been mixed. On one hand, one study found that people who are less tolerant of uncertainty have more negative beliefs toward God (Aryeh et al., 2021), while another study reported that people who internalized their religion had greater cognitive complexity (Neyrinck et al., 2006). Yet, some studies have found religious involvement to be inversely related to cognitive flexibility (Zmigrod et al., 2018) and belief in God to be inversely related to tolerance for uncertainty (Kossowska & Sekerdej, 2015).

Seventh, research is needed for exploring the possible role of certain virtues in religious deconversion. One potentially salient virtue is intellectual humility. Some scholars have suggested that people need intellectual humility to stay religious—that is, they need to acknowledge the limitations of human reasoning, especially their own (Marriott & Wicks, 2022; Mason, 2015). Indeed, studies have shown that intellectual humility opens people up to the influence of God (Krause & Hayward, 2014), and helps them resolve perceived conflicts between their own beliefs and those of others (Davis et al., 2017). As noted above, people who are intellectually humble are more tolerant of ambiguity, which may factor in religious deconversion (Leary et al., 2017). Another virtue potentially relevant to religious deconversion is patience. For people to stay religious they need the patience to wait for answers to questions (Hafen & Hafen, 2018; Mason, 2015). Thus, patience is tied to tolerating ambiguity and uncertainty. Indeed, more intrinsically religious people are better able to find meaning and optimism in life, and thus be more patient in difficult situations (Schnitker et al., 2020). A third virtue of potential importance to religious deconversion is forgiveness. For people to stay religious they need the ability to forgive others, such as church leaders and other congregation members (Givens & Givens, 2014; Mason, 2015). Interpersonal struggles can be dynamically connected to one’s relationship with God (Choe et al., 2016), and trigger faith crises and deconversion (Exline et al., 2022).

Eighth, research is needed to identify and understand outcomes of religious deconversion, particularly relational outcomes. Few outcomes are likely as impactful as those involving the family. Often researchers investigate the role of relationships in religious development, but rarely investigate relational outcomes of religious development. Incongruence between parent and child religiosity can cause conflict (Hwang et al., 2018; Stokes & Regnerus, 2009). Such religious incongruence seems more likely now than ever due to historical trends away from religion and adolescence being a heightened time for deconversion. Furthermore, adolescent religious deconversion leads to subsequent losses to parent–child relationship quality, while teens switching religions or converting to religion generally does not (Hendricks et al., 2023). Perhaps the move to non-religiousness is seen as less favorable by parents than other religious transitions.

Ninth, research is needed examining religious reconversion—that is, the process of Dones (people who were once religious but then left religion) returning to religion. While few people who leave religion ever return (Uecker et al., 2016), with increasing rates of religious Nones, the potential pool of religious reconverts is growing. With a growing population of potential reconverts, the importance of studying processes of religious reconversion increases. Unfortunately, to our knowledge, no significant research exists on processes of reconversion. A few exceptions suggest the importance of family relationships for adolescents and young adults returning to religion (Uecker et al., 2016; Wilson & Sherkat, 1994).

Tenth, more research is needed for examining moderators of processes of religious deconversion. Of particular interest are studies looking at the moderating role of religious affiliation and culture. Pathways away from religion depend on broader social contexts involving cultural shifts and divergent values (Bullivant, 2019; Streib et al., 2009). Religious deconversion trajectories can be classified based on the religious context people are leaving and the religious (or non-religious) context to which they are moving (Streib et al., 2009).

Conclusion

The purpose of this article was to review theory and research on religious deconversion during adolescence and young adulthood. We defined relevant terminology, presented statistical trends, reviewed work on processes, predictors, and outcomes of religious deconversion, and outlined recommendations for future research. Religious deconversion involves movement away from religion and is a mirror phenomenon to religious conversion (Marriott, 2020). This movement away from religion may include becoming less religious or shifting to being non-religious (Hardy & Longo, 2021). More specifically, becoming less religious may appear as diminished religious beliefs, disengagement from religious behaviors, discontinuance of the religious lifestyle, and disaffiliating from one’s religious community (Van Tongeren & DeWall, 2021). Shifting to being non-religious involves changing one’s religious identification to something like atheist, agnostic, or “nothing in particular.” Becoming less religious often precedes the shift to being non-religious (Hardy, Hatch, et al., 2022), yet the decline in religiosity seems to continue after religious deidentification (Van Tongeren, DeWall, Hardy, et al., 2021).

The process of religious deconversion is influenced by characteristics of the individual (e.g. demographics, personality, and religiosity) within interconnected layers of social contexts (e.g. family, peers, and religious communities; for a prior review, see Hardy & Longo, 2021). While religious deconversion, like all developmental phenomena, is complex, dynamic, and idiosyncratic, there are general patterns (Marriott, 2020; Streib, 2014). Often it starts with a faith crisis inducing an event or series of events, such as stressful life events, the presence of negative religious or spiritual experiences, or the absence of positive religious or spiritual experiences. Such experiences can spark questioning and doubt. What once seemed simple becomes complex. This searching can lead to the deepening of faith, or to criticism of the faith tradition and community, depending on how people resolve the complexity. If the latter, it comes with emotional suffering. Hence why many attempt to retain their faith. If this fails, they further solidify their disbelief, disengage from their religious behaviors, discontinue their religious lifestyle, disaffiliate from their religious community, and perhaps, ultimately deidentify from that religion altogether. They may take on a new identity of being agnostic or atheist.

Obviously not everyone who leaves religion follows this narrative, but many do. The process of deconversion seems similar across cultures, with some nuances unique to each cultural context (Streib et al., 2009). Furthermore, trajectories of religious deconversion can be distinguished by where the individual starts and where they end up, particularly in terms of the level of religiosity and religious identification (Streib et al., 2009). Thus, religious deconversion processes also vary somewhat by religious affiliation (Bullivant, 2022). Typically, religious deconversion seems maladaptive in terms of its impact on well-being, but this depends on the individual and context and the trajectory followed.

Religious deconversion is an emerging area of scholarship that is becoming more relevant to the field and society over time. It is an area that is fascinating in terms of theory, research, and practice. Much remains to be learned on the topic of religious deconversion, providing ample opportunity for future study. We hope our review provides a solid foundation for the field to build on and inspires much more work on religious deconversion, particularly during adolescence and young adulthood, when the most shifting occurs.