As we have seen, polls and surveys typically show that adolescents and college students are "less religious" than are older persons. However, the disengagement from religion that is more common among adolescents and younger adults is often characterized as a temporary phenomenon. As noted in Chapter 9, some "dropping out" may simply represent youthful exploration of alternative ideas and religions (e.g., alternative philosophies or belief systems, sects, cults) for a relatively short period in young people's lives. Gordon Allport suggested that after youthful disaffection with traditional religious values, many people return to religion by the time they are in their 30s.²⁶⁴ They may have children of their own and be concerned that their offspring should have some religious upbringing, or they may more generally have lost their rebellious tendencies and be settling down.

Similarly, Bibby has argued that disengagement from religion is typically temporary, and that many people who rejected religion in their teens eventually return to institutionalized religion, even if primarily to avail themselves of such "rites of passage" as marriage, funerals, and/or some religious instruction for their children.²⁶⁵ Bibby relies on Canadian evidence showing that the percentage of people claiming "no religion" is consistently highest among younger adults (18-34). Furthermore, a cohort analysis by Bibby suggests that almost half of the 16% of people aged 18-34 who claimed to have "no religion" in 1975 were reabsorbed into the religious realm, since in 1990 just 9% of the 35- to 54-year-olds (i.e., apparently many of the same people in the 1975 statistics) claimed "no religion." But of course we do not know how many of the 9% were the same people who claimed "no religion" 15 years earlier; individuals could not be followed longitudinally, and we do not know, for example, how many people might also have become apostates in the interim. But this evidence certainly suggests (albeit indirectly) that some people do return to religion after claiming "no religion" when they were younger.

Not all research findings are consistent with this "return to religion" tendency. In a study of rural Pennsylvania young people, Willits and Crider asked them questions about religion when they were high school students in 1970, and again in 1981 when they were about 27 years old.²⁶⁶ Data analyses were reported only for the 331 respondents who were married by 1981. The researchers concluded that these people were in fact less frequent church attenders at 27 than they had been in their middle teens. However, this study involved a relatively short-term follow-up, and it could be argued that the timing of the surveys (at ages 16 and 27, on average) might account for the unique findings. For example, a shift away from religion might well have occurred soon after the age of 16. Another follow-up when these people are in their 30s or 40s might be more informative.

An extensive longitudinal study of a U.S. national probability sample suggests that most religious "dropping out" probably occurs after age 16. Wilson and Sherkat followed the religious identification and other trends of more than 1,000 people from 1965, when they were seniors in high school, to 1973 and again to 1983.²⁶⁷ In the third wave of their study, they managed to retain more than two-thirds of the original 1,562 participants. They focused their attention on those who reported a religious preference in 1965, but then reported no preference in 1973. For these "dropouts," they found few differences between those who retained their apostate status in 1983 and those who had returned to religion. The returnees did report closer relationships with their parents in high school than did the continuing apostates. Furthermore, there was a tendency for early marriage and forming a family to be related to returning to religion, though this relationship was found only for men. In general, women were less likely to become apostates then were men, but women apostates were also less likely to return to the fold than were men. Wolson and Sherkat speculated that the men are more likely to be religiously affected by transitions to marriage and parenthood; they added, "Given the cultural understanding that the religious role is primarily allocated to women in the family, dropping out of the church is a stronger statement for women to make than for men, especially in a society where denominational affiliation of some kind is normative."²⁶⁸

Wilson and Sherkat's finding that marriage and parenthood are important factors in "returning to the fold" has been replicated elsewhere.²⁶⁹ This is consistent with our conclusion that parental religious socialization effects tend to weaken, and that other factors become more important, as people move on through the life cycle and begin to live independent adult lives themselves. However, we should not be too quick to conclude that marriage and parenthood contribute to greater religiousness.

Ploch and Hastings methodically analyzed General Social Survey data from the United States from 1972 to 1991, and found a general trend toward increased religiousness with increasing age.²⁷⁰ There was no indication in these correlational data that either marriage or childbearing was associated with an increase in church attendance. According to Ploch and Hastings, researchers who have concluded that family formation is positively related to church attendance may have confused a long-term trend toward an age-related increase in religiousness with short-term events such as marriage and childbearing. Recently, Stolsenberg and colleages have continued this debate, finding that "family life cycle" attitudes and events (marriage, cohabitation, parenthood, divorce, etc.) do affect religion, though they may interact with age in complex ways. That is, their results "suggest that an aging hypothesis and the Family Life Cycle hypothesis are not incompatible."²⁷¹ This issue is a complicated one. However, we should be careful not to assume that church attendance and membership are ideal, accurate indicators of personal religiousness.

These studies described above do agree that some people apparently become lifelong apostates. Also, for example, the "no religion" category seems to be either holding fairly steady (at about 7% in the United States²⁷²) or increasing (in Canada²⁷³). Thus, in spite of the fact that some apostates do return to religion later in their lives,²⁷⁴ apparently many people do remain "apostates for life."