Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Declines in American Adults’ Religious Participation and Beliefs, 1972-2014:

Declines in American Adults’ Religious Participation and Beliefs, 1972-2014: Previous research found declines in Americans’ religious affiliation but few changes in religious beliefs and practices. By 2014, however, markedly fewer Americans participated in religious activities or embraced religious beliefs, with especially striking declines between 2006 and 2014 and among 18- to 29-year-olds in data from the nationally representative General Social Survey (N = 58,893, 1972-2014). In recent years, fewer Americans prayed, believed in God, took the Bible literally, attended religious services, identified as religious, affiliated with a religion, or had confidence in religious institutions. Only slightly more identified as spiritual since 1998, and then only those above age 30. Nearly a third of Millennials were secular not merely in religious affiliation but also in belief in God, religiosity, and religious service attendance, many more than Boomers and Generation X’ers at the same age. Eight times more 18- to 29-year-olds never prayed in 2014 versus the early 1980s. However, Americans have become slightly more likely to believe in an afterlife. In hierarchical linear modeling analyses, the decline in religious commitment was primarily due to time period rather than generation/birth cohort, with the decline in public religious practice larger (d = −.50) and beginning sooner (early 1990s) than the smaller (d = −.18) decline in private religious practice and belief (primarily after 2006). Differences in religious commitment due to gender, race, education, and region grew larger, suggesting a more religiously polarized nation. Keywords religion, cultural change, prayer, religious affiliation, generations Are Americans less religious than they used to be? In previ- these sentiments. Finke and Stark (1988, 2005) contended that ous research, the answer depended on how religious commit- the overall religiousness of the American public has remained ment was measured. Most studies agree that religious relatively constant as a whole, although fluctuations in affilia- affiliation has declined in the United States since the 1970s; tion and expression have occurred. Similarly, Berger’s (1999) for example, more Americans in recent years chose “none” work explored the overall constancy of American religious when asked to identify their religion (e.g., Hout & Fischer, affiliation over time, with a particular focus on how religious- 2002; Lim, MacGregor, & Putnam, 2010; Pew Research ness moved back to the forefront of political and economic Center, 2015). However, several recent studies have con- discourse in recent decades—sentiments also echoed in cluded that religious service attendance, belief in God, and many other seminal works (Berger, 2011; Berger, Davie, & prayer have not changed or have even increased in recent Fokas, 2008). years (e.g., Dougherty, Johnson, & Polson, 2007; Presser & Thus, at least up to the mid- to late-2000s, research sug- Chaves, 2007; Smith & Snell, 2009; P. Taylor, 2014; gests that Americans’ private religious practice and beliefs Wachholtz & Sambamoorthi, 2011). Based on data up to (e.g., those religious practices, disciplines, and beliefs that 2008, Chaves (2011) concluded that belief in God and fre- may be conducted alone or without explicit religious affilia- quency of prayer did not change in the General Social Survey tion) and religious service attendance remained unchanged (GSS) since the 1980s. Examining 18- to 24-year-olds in the GSS 1972-2006, Smith and Snell (2009) found only small San Diego State University, CA, USA changes in religious affiliation and service attendance, and Florida Atlantic University, Boca Raton, USA no changes in frequency of prayer and belief in God. They Case Western Reserve University, Cleveland, OH, USA concluded that emerging adults “have not since 1972 become Corresponding Author: dramatically less religious or more secular . . . if such a trend Jean M. Twenge, Department of Psychology, San Diego State University, is indeed perceptible, it would seem to be weak and slight” 5500 Campanile Drive, San Diego, CA 92182-4611, USA. (pp. 99-100). Other sociologists of religion have echoed Email: jtwenge@mail.sdsu.edu Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License (http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open even as more did not affiliate with a particular religious tradi- Millennial generation (born approximately 1980-1994) is tion. Another possibility is that religious belief has been less religious because they are young or because of genera- replaced by spirituality (Fuller, 2001; Saucier & Skrzypińska, tional or time period change. That is, are Millennials less 2006). In other words, the prevailing conclusion has been religious than Generation X (born 1961-1979) and Boomers that Americans have remained just as religious and/or spiri- (born 1943-1960) were when they were 18- to 29-year-olds? tual in a private or personal sense, but less religious in a pub- This data may also provide an early look at iGen (born 1995- lic sense. This may be due to a more general disassociation 2012) and their religious attitudes. from large groups—for example, Americans have become Changes over time and generations in attitudes, values, significantly less confident in virtually all large institutions and personality traits are rooted in cultural change (Stewart & from government to medicine (Twenge, Campbell, & Carter, Healy, 1989; Twenge, 2014), with cultures and individuals 2014). Such an explanation would also be consistent with mutually influencing and constituting one another (Markus & many popular conceptions of religion as a socially organiz- Kitayama, 2010). One cultural change relevant for religious ing institution (e.g., Ysseldyk, Matheson, & Anisman, 2010) orientation is the rise in individualism, a cultural system plac- that transmits cultural values, mores, and rules (Graham & ing more emphasis on the self and less on social rules (e.g., Haidt, 2010). As societal norms have shifted away from Bellah, Madsen, Sullivan, Swidler, & Tipton, 1985; institutional identification to individualism, one would Fukuyama, 1999; Myers, 2000; Twenge, 2014). Several stud- expect commitment to religion, a ubiquitous social institu- ies have documented increases in focus on the self (Twenge, tion, to similarly decline. However, it is unclear whether Campbell, & Gentile, 2013; Twenge & Foster, 2010) and such decreases in external commitment would also be asso- declines in focus on institutions, empathy for others, and ciated with decreases in personal religious involvement or moral rules (Kesebir & Kesebir, 2012; Konrath, O’Brien, & practice. Hsing, 2011; Twenge et al., 2014). There are several reasons Despite popular conceptions that public religious involve- we would expect religion to decline with greater individual- ment has decreased while private expressions of religion and ism. First, religiosity implies some level of commitment to a spirituality have stayed about the same, stark distinctions larger group or organization. As Welzel (2013) suggests, the between religion and spirituality may be more theoretical trend in Western societies has been toward more freedom and than practical. Although religion and spirituality are known less commitment to groups. Second, belonging to a religious to be distinct constructs (i.e., religion comprises social and group may require assent with the group’s beliefs, opinions, ritualized aspects of personal belief, whereas spirituality and practices. This can create tension when differences in includes the search for meaning or transcendence in daily opinion arise between an individual and an organization (e.g., life; Pargament, 1999), these two constructs often overlap, Exline, Pargament, Grubbs, & Yali, 2014; Exline & Rose, and highly religious individuals often identify as being 2013). Third, religiosity usually involves some rule-follow- highly spiritual as well (for a review, see Hill & Pargament, ing and submission to authority (e.g., Graham & Haidt, 2010), 2003). Moreover, although some individuals certainly do another characteristic that goes against emancipation and identify as spiritual but not religious (e.g., Saucier & individualism. Fourth, religion often focuses on concerns out- Skrzypińska, 2006), a much larger proportion of individuals side of the self, such as helping others and serving God (e.g., identify as both religious and spiritual (Pargament, 1999), Pichon, Boccato, & Saroglou, 2007; Shariff & Norenzayan, and many people have difficulty substantively differentiating 2007). Thus, the increasing individualism of American cul- between the two on an individual level (Hill et al., 2000; ture may have produced decreased religiosity among more Zinnbauer et al., 1997). Therefore, as religious commitment recent time periods and generations. has decreased, one may also expect decreases in private reli- Based on previous research and cultural changes, we gious practice and individual spirituality. expect a decline in religious affiliation. We also predict In this article, we seek to examine whether Americans’ declines in religious service attendance; while religiously religious service attendance, religious practice, religious unaffiliated Americans may attend services for a time, they beliefs, religiosity, spirituality, confidence in religious insti- may become less likely to do so as they feel more disassoci- tutions, and religious affiliation have changed since the ated from religion. Most crucially, we predict declines in 1970s, with a particular focus on the years since 2006 and on more private expressions of religious belief and practice, 18- to 29-year-olds. We take the additional step of calculat- such as prayer, religiosity, and belief in God, with the declines ing effect sizes and performing statistical significance testing especially evident among young people. Belonging to a reli- to quantify the size of the changes. We draw from the nation- gion and more privately believing in its tenets are tradition- ally representative GSS of U.S. adults conducted 1972-2014. ally linked (e.g., Park et al., 2013; Smith, Denton, Faris, & Because this survey draws from a multiage sample above 42 Regnerus, 2002); as more Americans are unaffiliated with years, it can isolate the effects of age from those of time religion, a greater proportion may become not just unaffili- period and generation. Thus, unlike some surveys conducted ated but secular in their beliefs and practices. These declines over a shorter period of time (e.g., 7 years: Pew Research may be especially evident in recent years and among 18- to Center, 2015), this data set can determine, for example, if the 29-year-olds, given the generational and cultural trends Twenge et al. 3 toward emphasizing social rules less and individual freedom sizes; (b) to examine whether these changes are due to genera- more (known as “Generation Me”: Twenge, 2014; or “eman- tion or time period; and (c) to examine whether the trends dif- cipative values”: Welzel, 2013). Moving away from social fer by gender, race, education, or U.S. region. institutions and community engagement would likely detract from one of the key facets of religion as a whole—that is, Method community involvement and social value transmission. A secondary question is whether changes in religious ori- Sample entation over time are caused by time period or generational We drew from the GSS, 1972-2014, a nationally representa- (cohort) effects. If successive generations are less religious tive survey of U.S. residents over 18. Depending on the item, (forming their religious orientation while young and not ns range between 12,862 and 58,893. As suggested by the changing), any decline would be due to generation. If people GSS administrators, we weight the analyses by the weight of all ages have become less religious during certain times, variable WTSSALL to make the sample nationally represen- any decline would be due to time period. New hierarchical tative of individuals rather than households and correct for linear modeling techniques (called APC or age-period- other sampling biases. However, these weighted analyses cohort analyses) attempt to separate the effects of age, gen- differ only very slightly from unweighted analyses. Also as eration, and time period (Yang, 2008; Yang & Land, 2013). suggested by the administrators, we excluded the black Some have argued that these techniques do not resolve the oversamples collected in 1982 and 1987. identification problem that has long plagued simultaneous analysis of age, period, and cohort effects (e.g., Bell & Jones, 2013, 2014); however, these criticisms appear to Items largely rest on untenable assumptions that are not consistent with basic APC models (Reither et al., 2015). In addition, We identified and analyzed all items on respondents’ own APC techniques have become widely used. For example, religious orientation asked in at least six administrations of Schwadel (2011) performed an APC analysis on some of the the GSS. They were as follows: GSS religion variables up to 2006. However, at that time the data included only a handful of Millennials, a generation 1. Religious preference: “What is your religious prefer- purported to be less religious; by 2014, however, Millennials ence? Is it Protestant, Catholic, Jewish, some other were the entirety of 20- to 29-year-olds. In addition, further religion, or no religion?” We analyzed the percentage time period change may have occurred in the 8 years of data of respondents who chose “no religion.” Asked available since 2006. Thus, we perform APC analyses to 1972-2014. examine whether shifts in Americans’ religious orientation 2. Strength of religious affiliation: “Would you call are due to generational or time period effects. yourself a strong (Christian, Jew, etc.) or not a very In addition, we examine possible moderators of change strong (Christian, Jew, etc.)?” Response choices were over time in religious orientation. Trends may differ among “not very strong,” “somewhat strong,” and “strong.” men and women, Blacks and Whites, education levels, and Asked 1974-2014. U.S. regions, as these groups differ in their levels of religios- 3. Religious service attendance: “How often do you ity and cultural focus (Blaine & Crocker, 1995; Piff, Kraus, attend religious services?” Response choices were Cote, Cheng, & Keltner, 2010; Plaut, Markus, & Lachman, “never,” “less than once a year,” “about once a year,” 2002; R. Taylor, Chatters, Jayakody, & Levin, 1996; Vandello “about once or twice a year,” “Several times a year,” & Cohen, 1999). We theorize that the decline in religious “about once a month,” “2-3 times a month,” “nearly orientation will be larger among demographic groups and every week,” and “every week.” Asked 1972-2014. regions with higher social power and more individualism, 4. Belief in the afterlife: “Do you believe there is a life including Whites, men, those with a college education, and after death?” Response choices were “yes” and “no.” living in the Northwest and West, and lower or nonexistent Asked 1973-2014. among groups with lower social power and less individual- 5. Believing the Bible is literal: “Which of these state- ism, including Blacks, women, those without a college edu- ments comes closest to describing your feelings cation, and the Midwest and South (e.g., Piff et al., 2010; about the Bible?” Response choices were “The Bible Vandello & Cohen, 1999). Groups with relatively high social is an ancient book of fables, legends, history, and power might not see themselves as having a significant need moral precepts recorded by men”; “The Bible is the for religion or God, so these groups might pioneer the move- inspired word of God but not everything in it should ment toward less religiosity. be taken literally, word for word”; and “The Bible is Thus, we have three goals in this article: (a) to perform a the actual word of God and is to be taken literally, comprehensive examination of American adults’ religious ori- word for word.” Asked 1984-2014. entation from 1972 through 2014, with a particular emphasis 6. Frequency of praying: “About how often do you on 2006-2014 and 18- to 29-year-olds, and including effect pray?” Response choices were “never,” “less than 4 SAGE Open once a week,” “once a week,” “several times a week,” or by individual year. Given our focus on both overall change “once a day,” and “several times a day.” Asked since the 1970s and change since 2006, we separated the data 1983-2014. into 5-year intervals from 1972-2004 and reported data by 7. Belief in God: “Please look at this card and tell me individual year from 2006 to 2014. We report the effect sizes which of the statements comes closest to expressing (d, or difference in terms of standard deviations) and p values what you believe about God.” Response choices were for t tests comparing 1972-1974 with 2014 and 2006 and “I don’t believe in God”; “I don’t know whether there 2014. We also include two figures with all of the year-by- is a God and I don’t believe there is any way to find year data for some variables. We report both continuous vari- out”; “I don’t believe in a personal God, but do ables (e.g., the 0-8 scale for religious service attendance) and believe in a Higher Power of some kind”; “I find dichotomous variables (e.g., the percentage who never attend myself believing in God some of the time, but not at religious services). We use the tables for means and report others”; “While I have doubts, I feel that I do believe percentage changes in the text. in God”; and “I know God really exists and I have no For the APC models, we estimated random coefficient doubts about it.” Asked 1988-2014. models allowing intercepts to vary across time periods 8. Confidence in religious institutions: “I am going to (years) and generations (cohorts). Thus, effectively, an inter- name some institutions in this country. As far as the cept (mean religious orientation) score is calculated (using people running these institutions are concerned, empirical Bayes) for each cohort and each survey year. In would you say that you have a great deal of confi- addition, a fixed intercept (grand mean) is estimated along dence, quite a lot of confidence, only some confi- with a fixed regression coefficient for age and age squared. dence, or very little in them?” One of the items is This model has three variance components: One for variabil- “organized religion.” Response choices were “hardly ity in intercepts due to cohorts (τ ), one for variability in u0 any confidence at all,” “only some confidence,” or “a intercepts due to period (τ ), and a residual term containing v0 great deal of confidence.” We excluded “don’t know” unmodeled variance within cohorts and periods. Variance in and “refused” responses. Asked 1973-2014. the intercepts across time periods and cohorts indicates 9. Identification as a religious person: “To what extent period and cohort differences, respectively (Yang & Land, do you consider yourself a religious person?” 2013). Thus, the technique allows for a separation of the Response choices of “not religious at all,” “slightly effects of generation/cohort, time period, and age. Weighting religious,” “moderately religious,” and “very reli- could not be used for the mixed-effects analyses because gious.” Asked 1998 and 2006-2014. proper probability weighting for variance component estima- 10. Identification as a spiritual person: “To what extent tion requires taking into account pairwise selection probabil- do you consider yourself a spiritual person?” ities, which is not possible in current statistical software. Response choices of “not spiritual at all,” “slightly In describing the trends in the text and tables, we will spiritual,” “moderately spiritual,” and “very spiri- sometimes use common labels for the generations such as the tual.” Asked 1998 and 2006-2014. G.I. or “Greatest” generation (born 1900-1924), Silent (1925-1942), Boomers (1943-1960; some argue 1946-1964), Of these, religious preference, strength of religious affili- GenX (1961-1979 or 1965-1979), Millennials (1980-1994; ation, religious service attendance, and confidence in reli- for reviews, see Strauss & Howe, 1991; Twenge, 2014), and gious institutions are public religious variables, and belief in iGen (1995-2012). These birth year cutoffs are arbitrary and an afterlife, believing the Bible is literal, frequency of pray- are not necessarily justified by empirical evidence, but are ing, belief in God, identification as a religious person, and useful shorthand labels for those born in certain eras. identification as a spiritual person are private religious Differences among those of the same generation certainly variables. occur, and these can be seen in the tables and figures; we use these labels merely for convenience. Possible Moderators Results We analyzed moderation by gender (men vs. women), race (White vs. Black, the only racial groups measured in all sur- Trends in Religious Orientation vey years), education level (high school graduate and below American adults in the 2010s were less religious than those vs. attended some college and above), and U.S. region in previous decades, based on religious service attendance (Northeast, Midwest, South, and West). and more private religious expressions such as belief in God, praying, identifying as a religious person, and believing the Procedure Bible is the word of God (see Table 1 and Figure 1). These findings held when restricted to 18- to 29-year-olds Data collected over time can be analyzed in many ways, (see Table 2 and Figure 2), demonstrating that Millennials including grouping by 20-year generation blocks, by decades, 5 Table 1. Religious Orientation Among All Adult Americans, 1972-2014, General Social Survey. d (2006 d (earliest Survey items n 72-74 75-79 80-84 85-89 90-94 95-99 00-04 2006 2008 2010 2012 2014 vs. 2014) vs. 2014) Religion “none” (%) 58,893 6% (0.24) 7% (0.26) 7% (0.26) 7% (0.26) 8% (0.27) 13% (0.33) 14% (0.35) 16% (0.36) 17% (0.37) 18% (0.38) 20% (0.40) 21% (0.40) .13*** .51*** Strength of religious 53,938 3.18 (0.86) 3.14 (0.86) 3.19 (0.87) 3.12 (0.89) 3.08 (0.92) 2.98 (1.03) 2.98 (1.03) 2.92 (1.04) 2.93 (1.06) 2.91 (1.08) 2.89 (1.11) 2.87 (1.12) −.05 −.30*** affiliation (1-4) Attend religious services 58,347 4.13 (2.63) 3.94 (2.68) 4.05 (2.66) 3.99 (2.67) 3.89 (2.69) 3.67 (2.71) 3.66 (2.72) 3.57 (2.79) 3.56 (2.77) 3.53 (2.78) 3.49 (2.82) 3.33 (2.83) −.09** −.30*** (0-8) Never attend religious 58,347 11% (0.32) 14% (0.35) 13% (0.33) 14% (0.35) 15% (0.36) 17% (0.38) 19% (0.39) 22% (0.42) 22% (0.42) 23% (0.42) 25% (0.44) 26% (0.44) .09** .41*** services (%) Confidence in organized 38,281 2.27 (0.69) 2.15 (0.70) 2.15 (0.70) 2.01 (0.70) 2.01 (0.69) 2.07 (0.68) 2.04 (0.68) 2.03 (0.69) 1.95 (0.68) 1.96 (0.67) 1.96 (0.67) 1.96 (0.66) −.10** −.46*** religion (1-3) Praying (1-6) 30,333 — — 4.24 (1.55) 4.24 (1.49) 4.19 (1.52) 4.24 (1.52) 4.28 (1.58) 4.26 (1.72) 4.16 (1.75) 4.18 (1.75) 4.20 (1.77) 4.12 (1.80) −.08** −.07** Never pray (%) 30,333 — — 3% (0.16) 1% (0.08) 1% (0.10) 2% (0.14) 4% (0.19) 11% (0.30) 12% (0.32) 13% (0.34) 14% (0.35) 15% (0.36) .12*** .48*** Believe in God (1-6) 19,542 — — — 5.26 (1.22) 5.23 (1.31) 5.11 (1.43) 5.21 (1.34) 5.18 (1.32) 5.11 (1.41) 5.02 (1.45) 5.03 (1.45) 4.98 (1.47) −.14*** −.20*** Do not believe in God (%) 19,542 — — — 13% (0.33) 15% (0.35) 18% (0.39) 15% (0.35) 16% (0.37) 18% (0.39) 20% (0.40) 20% (0.40) 22% (0.41) .15*** .24*** Bible literal (1-3) 30,084 — — 2.25 (0.68) 2.20 (0.68) 2.19 (0.67) 2.15 (0.68) 2.18 (0.68) 2.17 (0.70) 2.13 (0.71) 2.14 (0.73) 2.10 (0.73) 2.10 (0.73) −.10** −.21*** Bible = fables (%) 30.084 — — 14% (0.34) 15% (0.36) 15% (0.36) 17% (0.37) 16% (0.37) 17% (0.38) 20% (0.41) 21% (0.41) 22% (0.42) 22% (0.42) .13*** .20*** Believe in afterlife (%) 35,391 76% (0.43) 76% (0.42) 79% (0.41) 80% (0.40) 82% (0.40) 82% (0.38) 82% (0.38) 83% (0.38) 81% (0.39) 81% (0.39) 81% (0.39) 79% (0.41) −.10** .07* Religious person (1-4) 12,924 — — — — — 2.65 (0.95) — 2.68 (0.94) 2.62 (0.96) 2.57 (0.97) 2.57 (1.01) 2.51 (1.00) −.18*** −.14*** Not religious at all (%) 12,924 — — — — — 15% (0.36) — 14% (0.35) 16% (0.37) 18% (0.39) 20% (0.40) 20% (0.40) .16*** .13*** Spiritual person (1-4) 12,862 — — — — — 2.72 (0.93) — 2.89 (0.92) 2.84 (0.92) 2.84 (0.92) 2.85 (0.97) 2.82 (0.97) −.07** .10** Not at all spiritual (%) 12,862 — — — — — 12% (0.32) — 9% (0.29) 10% (0.30) 11% (0.31) 11% (0.31) 11% (0.31) .07** −.03 Difference between service 37,419 0.19 (1.19) 0.18 (1.26) 0.17 (1.24) 0.11 (1.24) 0.08 (1.24) −0.03 (1.26) −0.05 (1.23) −0.10 (1.26) −0.08 (1.24) −0.06 (1.27) −0.08 (1.25) −0.10 (1.26) .00 .24*** attendance and belief in afterlife (z score) Never attend services but 37,419 7% (0.26) 8% (0.26) 7% (0.25) 9% (0.29) 9% (0.28) 12% (0.32) 12% (0.32) 15% (0.36) 13% (0.34) 13% (0.34) 15% (0.36) 15% (0.36) .00 .25*** believe in afterlife (%) Note. — = question was not asked in those years. Standard deviations are in parentheses. d = difference in standard deviations. *p < .05. **p < .01. ***p < .001, based on a t test. 6 SAGE Open Figure 1. Percentage of all American adults with no religious affiliation, who never attend services, never pray, do not believe in God, are not religious at all, and are not spiritual at all. are less religious than previous generations were at the same the less certain response choices such as “I don’t believe in age. While religious affiliation and service attendance have God”; “I don’t know whether there is a God and I don’t been declining since the 1990s, the decrease in more private believe there is any way to find out”; or “I don’t believe in a religious expressions began fairly recently, becoming pro- personal God, but do believe in a Higher Power of some nounced only after 2006 (see Figures 1 and 2). Effect sizes kind”; these responses were combined into “Do not believe ranged from moderate (around d = .50; Cohen, 1988) to in God” in Tables 1 and 2). By 2014, however, 22% expressed small (around d = .20). The increase in never praying among doubts, a 69% increase. Among 18- to 29-year-olds, 30% had 18- to 29-year-olds was d = .80, equaling the guideline for a serious doubts by 2014, more than twice as many as in the large effect. late 1980s (12%). As found in previous research, fewer Americans now Americans have also become less likely to believe that the affiliate with a religion. Although the majority of Americans Bible is the word of God. In 1984, 14% of Americans are still religious, three times as many in 2014 (vs. the early believed the Bible “is an ancient book of fables, legends, his- 1970s) have no religious affiliation, and twice as many never tory, and moral precepts recorded by men” rather than the attend religious services. Fewer have confidence in orga- word of God; by 2014, 22% of Americans believed this, a nized religion; the number who said they had “hardly any” 57% increase. Among 18- to 29-year-olds, 29% believed this confidence went from 14% in the early 1970s to 24% in by 2014, nearly twice as many as in the late 1980s (15%). 2014, a 71% increase, and those who said they had “a great Fewer Americans identify as religious; 62% said they deal” of confidence was cut in half (from 41% to 20%). were “moderately” or “very” religious in 1998, compared By 2014, the declines in religious orientation extended to with 54% in 2014, a 13% decrease. Among 18- to 29-year- more personal and private religious beliefs. Five times as olds, 49% said they were moderately or very religious in many Americans in 2014 (vs. the late 1980s) never prayed 1998, compared with 38% in 2014, a 22% decrease. Similarly, (eight times more among those ages 18-29). Slightly more those who said they were “not religious at all” increased Americans in 2014 (vs. the 1980s) said they prayed “several from 15% in 1998 to 20% in 2014 among all adults (a 33% times a day” (28%, up from 26%), but the 20% who prayed increase), and from 23% in 1998 to 28% in 2014 among 18- “less than once a week” in the 1980s became only 11% in to 29-year-olds (a 22% increase). 2014, apparently moving to “never” praying (3% in the Has religiosity been replaced with spirituality? It does not 1980s vs. 15% in 2014). appear so. Identifying as a spiritual person increased between Americans in 2014 were less likely to say they believed in 1998 and 2006, but then declined between 2006 and 2014 God. In the late 1980s, only 13% of U.S. adults expressed (see Table 1 and Figure 1). In all, 62% identified as moder- serious doubts about the existence of God (choosing one of ately or strongly spiritual in 1998, compared with 70% in 7 Table 2. Religious Orientation Among American 18- to 29-Year-Olds, 1972-2014, General Social Survey. d (2006 d (earliest Survey items n 72-74 75-79 80-84 85-89 90-94 95-99 00-04 2006 2008 2010 2012 2014 vs. 2014) vs. 2014) Religion “none” 13,724 12% (0.32) 12% (0.33) 12% (0.32) 12% (0.32) 12% (0.33) 22% (0.41) 22% (0.41) 23% (0.42) 28% (0.45) 26% (0.44) 31% (0.46) 31% (0.46) .18** .53*** Strength of religious affiliation 12,495 2.97 (0.91) 2.92 (0.92) 3.02 (0.94) 2.93 (0.94) 2.91 (0.96) 2.72 (1.10) 2.73 (1.10) 2.69 (1.10) 2.65 (1.16) 2.67 (1.16) 2.56 (1.18) 2.57 (1.17) −.11 −.38*** (1-4) Attend religious services 13,631 3.60 (2.51) 3.39 (2.56) 3.56 (2.51) 3.48 (2.51) 3.52 (2.50) 2.96 (2.44) 3.18 (2.52) 2.99 (2.51) 2.86 (2.59) 3.11 (2.68) 2.71 (2.52) 2.68 (2.61) −.12* −.36*** (0-8) Never attend religious 13,631 14% (0.35) 17% (0.38) 14% (0.34) 17% (0.37) 15% (0.35) 21% (0.41) 21% (0.41) 28% (0.45) 29% (0.46) 28% (0.45) 31% (0.46) 33% (0.47) .11* .31*** services Confidence in organized 9,158 2.17 (0.68) 2.10 (0.69) 2.09 (0.69) 1.99 (0.70) 2.02 (0.69) 2.07 (0.69) 2.05 (0.71) 2.02 (0.73) 1.99 (0.70) 1.99 (0.73) 1.97 (0.67) 1.88 (0.64) −.20* −.43*** religion (1-3) Praying (1-6) 6,818 — — 3.78 (1.54) 3.79 (1.43) 3.76 (1.54) 3.75 (1.54) 3.87 (1.60) 3.76 (1.81) 3.62 (1.78) 3.67 (1.87) 3.44 (1.85) 3.55 (1.88) −.11 −.14** Never pray (%) 6,818 — — 3% (0.17) .04% (0.07) 1.4% (0.13) 3% (0.18) 5% (0.21) 19% (0.39) 17% (0.38) 22% (0.41) 25% (0.43) 24% (0.43) .12* .80*** Believe in God (1-6) 4,126 — — — 5.15 (1.21) 5.12 (1.39) 4.88 (1.54) 4.94 (1.45) 4.90 (1.48) 4.86 (1.54) 4.71 (1.57) 4.61 (1.56) 4.61 (1.63) −.19** −.38*** Do not believe in God (%) 4,126 — — — 12% (0.33) 16% (0.36) 25% (0.43) 21% (0.40) 22% (0.42) 24% (0.42) 26% (0.44) 29% (0.46) 30% (0.46) .19** .45*** Bible literal (1-3) 6,500 — — 2.24 (0.69) 2.15 (0.68) 2.18 (0.68) 2.08 (0.69) 2.11 (0.69) 2.12 (0.69) 2.06 (0.70) 2.08 (0.74) 1.95 (0.72) 1.97 (0.74) −.21** −.37*** Bible = fables (%) 6,500 — — 15% (0.36) 17% (0.38) 16% (0.36) 20% (0.40) 19% (0.39) 21% (0.41) 21% (0.41) 24% (0.43) 29% (0.45) 29% (0.45) .19** .33*** Believe in afterlife (%) 8,775 73% (0.44) 72% (0.45) 79% (0.41) 79% (0.41) 78% (0.42) 81% (0.39) 82% (0.38) 82% (0.38) 79% (0.41) 82% (0.39) 79% (0.41) 80% (0.40) −.05 .17* Religious person (1-4) 2,632 — — — — — 2.39 (0.97) — 2.42 (0.96) 2.33 (0.99) 2.31 (0.96) 2.25 (0.99) 2.20 (0.97) −.23*** −.20*** Not religious at all (%) 2,632 — — — — — 23% (0.42) — 21% (0.41) 25% (0.44) 25% (0.44) 29% (0.45) 28% (0.45) .17** .11 Spiritual person (1-4) 2,628 — — — — — 2.51 (0.93) — 2.64 (0.93) 2.61 (0.98) 2.50 (0.98) 2.48 (0.99) 2.48 (1.02) −.17** −.03 Not at all spiritual (%) 2,628 — — — — — 14% (0.35) — 13% (0.34) 16% (0.36) 18% (0.39) 18% (0.39) 19% (0.39) .17** .13 Difference between service 8,734 0.03 (1.26) 0.08 (1.33) −0.01 (1.27) −0.06 (1.29) 0.00 (1.25) −0.25 (1.27) −0.23 (1.25) −0.34 (1.17) −0.27 (1.24) −0.20 (1.25) −0.35 (1.28) −0.38 (1.23) −.03 −.33*** attendance and belief in afterlife (z score) Never attend services but 7,596 11% (0.32) 10% (0.31) 8% (0.27) 11% (0.32) 8% (0.27) 15% (0.36) 14% (0.35) 19% (0.39) 18% (0.39) 15% (0.36) 22% (0.42) 21% (0.41) .05 .27*** believe in afterlife (%) 370 Note. — = question was not asked in those years. Standard deviations are in parentheses. d = difference in standard deviations. *p < .05. **p < .01. ***p < .001, based on a t test. 8 SAGE Open Figure 2. Percentage of 18- to 29-year-old Americans with no religious affiliation, who never attend services, never pray, do not believe in God, are not religious at all, and are not spiritual at all. 2006 and 65% in 2014; thus, identification as a spiritual per- person and spiritual person variables were not asked in son increased 5% between 1998 and 2014, a small increase enough years to be included, so we limited this analysis to compared to the larger declines in religious belief and prac- the other eight variables.) We used the continuous form of tice. In addition, the percentage of 18- to 29-year-olds iden- six variables (strength of religious affiliation, religious ser- tifying as moderately or strongly spiritual declined 6%, from vice attendance, frequency of prayer, belief in God, belief in 50% in 1998 to 47% in 2014. In 1998, 14% of 18- to 29-year- the Bible as literal, confidence in religious institutions), with olds said they were not spiritual at all, rising to 19% by 2014, religious affiliation (none vs. affiliated) and belief in an a 36% increase (see Table 2 and Figure 2). Thus, there is afterlife (yes vs. no) dichotomous. We included only respon- some suggestion that young people were less spiritual in dents who completed at least four of the eight items. Horn’s 2014 versus 1998, though the decline was not statistically (1965) parallel analysis of n = 8,513 cases with no missing significant. In 2014, fewer 18- to 29-year-olds (Millennials) values indicated that only a one-component solution had an identified as spiritual (47%) than those 50 and above (72%). Eigenvalue better than chance levels. Moreover, all variables This suggests that identification as a spiritual person may loaded highly onto a single principal component explaining continue to decline. 46% of the variance, with a model fit of .94 (on a 0-1 scale). One increase in religious belief did emerge: Slightly more Using the omega function available in the {psych} package Americans believe in life after death (see Tables 1 and 2). in R (Revelle, 2015) indicated that 50% of the common vari- Thus, more Americans believe in life after death even as ance in the item scores could be accounted for by a general fewer belong to a religion, fewer attend religious services, factor of religious orientation. In addition, the omega coeffi- and fewer pray. In the 1970s, only about 7% of Americans cient, which is the best estimator of single factor saturation never attended religious services but nevertheless believed in (see Zinbarg, Revelle, Yovel, & Li, 2005), was .70 suggest- life after death; by 2014, twice as many (15%) showed this ing that a single factor accounted for much of the variability disconnect between behavior and belief, and 21% among in these items. The principal components analysis indicated young people. a single principal component across the decades of data col- lection. Therefore, all variables were z scored, and a compos- ite religious orientation variable was formed (n = 52,497, Mixed-Effects Analyses to Separate Time Period, M = 0.01, SD = 0.69, α = .83). Generation, and Age Next, we performed mixed-effects analyses to separate First, we performed a principal components analysis to the effects of time period, generation, and age on the com- determine whether the religion variables could be combined posite variable. The SD in intercepts for period (survey into a composite variable for use in the mixed-effects APC year) was .12 [.09, .16] and for cohorts was .03 [.00, .04], model analyses; combining these variables into an index suggesting that almost none of the variability in religious ori- increases internal reliability over single items. (The religious entation was due to cohorts. There was also a statistically Twenge et al. 9 Figure 3. Adult Americans’ religious orientation by generation (cohort/birth year) and time period (survey year), in mixed-effects analyses separating time period, generation/cohort, and age. significant effect for age (b = .011 [.010, .014]) indicating private religious practice and belief was smaller (d = −.18 that older individuals were higher on religious orientation between 1984 and 2014) and began later (with a consistent (thus, religious orientation increases about d = .01 with each decline beginning around 2006-2008; d = −.12 of the change year of age). There was a weaker quadratic effect of age occurred between 2006 and 2014). (b = −.00005 [−.00008, −.00002]) indicating that the linear effect is not as strong at older ages. Moderators of the Decline in Religious Overall, there was a marked time period effect when gen- Orientation eration and age were controlled (see Figure 3). Religious ori- entation declined d = −.38 from 1973 to 2014, and d = −.15 We next analyzed whether the time period and cohort between 2006 and 2014. The generational effect was weaker, decrease in religious orientation (controlled for each other with religious orientation declining the most between those and age) differed based on race, U.S. region, sex, and educa- born in the 1930s and the Millennials born in the 1980s-1990s tion level. The trends were moderated by race, with no (d = −.06). change in religious orientation for Black Americans (d = .00) Although religious orientation formed a single factor, we and a large decrease among White Americans (d = −.48). In also examined whether the pattern of change was different the early 1970s, Whites and Blacks differed little in religious for public (affiliation, strength of affiliation, service atten- orientation (d = .15, 1973-1974, with Blacks higher), but by dance, confidence in religious institutions) and private (belief 2014, there was a marked racial difference, with Blacks in the afterlife, belief that the Bible is literal, praying, belief higher (d = .67). Cohort effects were weak for both Whites in God) religious practice. Similar to the analyses with one and Blacks. combined variable, time period explained more of the change The effects also differed by U.S. region, with the decline than birth cohort for both public and private religious in religious orientation largest in the West (d = −.42), Northeast practice. However, the pattern of change and its size differed (d = −.27), South (d = −.10), and Midwest (d = −.07). However, (see Figure 4). The decline in public religious practice was Midwesterners showed a pronounced cohort effect from larger (d = −.50 between 1972 and 2014, and d = −.42 those born in the 1880s to those born in the 1990s (d = −1.15), between 1984 and 2014) and began sooner (with the consis- compared with the nonexistent cohort effects in the other tent decline beginning around 1991-1993). The decline in three regions. In the early 1970s, Southern residents were 10 SAGE Open Figure 4. Time period changes in adult Americans’ public religious practice (left) and private religious belief/practice (right), in mixed- effects analyses separating time period, generation/cohort, and age. only somewhat more religious than those in the Northeast (d the Bible was the word of God than they were in previous = .23), but by the 2010s, Southerners were moderately higher decades. Thus, the decline in religious affiliation found in in religious orientation compared to Northeasterners (d = previous research has now extended to religious service atten- .40). The West was the least religious region in both eras, dance and, by 2008 and afterward, to personal religious belief with Westerners lower than Southerners in 1972 (d = −.47) and practice. The only exceptions were an increase in belief but even more so in 2014 (d = −.78). in the afterlife and a small increase in identifying as spiritual An intriguing pattern appeared when examining men and between 1998 and 2006 limited to those above 30. The women separately: The time period difference was some- declines in religious orientation were particularly striking what larger for women (d = −.28) than for men (d = −.12), between the early 2000s and 2014 and among those 18 to 29 but men showed a pronounced cohort decline in religious years old. Nearly a third of Millennials are not just religiously orientation (d = −.93), while women showed virtually no unaffiliated, but secular in other ways (doubting the existence effect for cohort (d = −.02). of God, believing the Bible is a book of fables, not attending Similarly, the time period decline in religious orientation religious services, describing oneself as “not religious at all,” was somewhat larger among those who had not attended col- never praying), and one out of five also say they are “not spir- lege (d = −.28) compared with those who attended at least some itual at all.” Although religious orientation is often concep- college (d = −.15). However, there was a moderate cohort tualized as a multidimensional concept (e.g., Cornwall, decline in religious orientation among those who attended col- Albrecht, Cunningham, & Pitcher, 1986; Idler et al., 2003), lege (d = −.39) and virtually none for those who did not attend the present data indicated that declines in religious affiliation college (d = −.02). Overall, gender, race, education, and extended across various measures of religious participation regional differences in religious commitment grew larger and commitment. The decline in religious affiliation and par- between the 1970s and the 2010s or between cohorts born in ticipation has now extended to private practices and beliefs, the late 19th century and those born in the late 20th century. though the decline in private religious practice and belief is smaller and began later than the decline in public religious practice. Discussion Mixed-effects analyses demonstrated that these trends By 2014, American adults were less likely to pray, believe in were primarily due to time period. Millennials were less reli- God, identify as religious, attend religious services, or believe gious than their Boomer and Generation X predecessors were Twenge et al. 11 at the same age, demonstrating that their lower religious com- Principal component and omega analyses demonstrated mitment is not solely due to their developmental stage of that a single factor captured the eight religious orientation young adulthood. However, this appears to be due to a time variables. Although religiosity is usually conceptualized as period effect in which all generations are growing less reli- multidimensional (Cornwall et al., 1986; Idler et al., 2003), gious over time. This suggests support for the idea that grow- in this data set, the majority of variation in religious orienta- ing individualism has been accompanied by less religion on a tion was determined by a single factor. We tried to strike a larger cultural basis, with a larger linear cohort decline in balance between internal reliability and diversity among some groups (men, Midwesterners, the college educated). individual items by presenting analyses of single items in the These findings contradict popular culture notions of genera- tables and focusing the APC analyses on the composite mea- tions cycling back and forth with, for example, a less religious sure and on the public and private practice measures. generation being followed by a more religious one. For exam- Our focus here was on individuals in the United States, so ple, generational theorists Howe and Strauss (2000), who cross-cultural studies should examine temporal trends in reli- adhere to the theory that generations come in cycles, proposed gious orientation in other countries. Religious groups may that Millennials would be more religious than GenX’ers. also differ in how they manage the discrepancy between reli- However, these data strongly suggest that the opposite is true. gious participation and afterlife beliefs, based on teachings Men and women, Blacks and Whites, the college edu- about the afterlife and whether (and how) the afterlife is con- cated and not college educated, and the South versus the nected with choices or behaviors during this life. Northeast are becoming more polarized in their religious ori- entation: While differences in religious commitment between Conclusion these groups were small during the 1970s, they have grown larger in recent years and with recent cohorts. The decline in The 2010s are a time of tremendous change in the religious religious commitment was most pronounced among men, landscape of the United States. Although the majority of Whites, and those in the Midwest, Northeast, and West, and Americans are still religious, the declines in public religious was nearly absent among Black Americans and small in the affiliation observed in previous research have, by 2014, South. It appears that groups with relatively high social extended to private religious belief and action (such as power are less likely to see themselves as having a signifi- prayer, belief in God, and identifying as religious). This cant need for religion or God in recent years. decline was not replaced by a substantial increase in those In comparison with those from earlier years and genera- identifying as spiritual. The slight increases in afterlife belief tions, American adults in recent years and generations were represent a potentially important exception to this pattern. slightly more likely to believe in an afterlife. Combined Overall, the data suggest a pervasive decline in religious par- with the decline in religious participation and belief, this ticipation and belief among Americans, with a burgeoning might seem paradoxical. One plausible, though speculative, minority becoming decidedly nonreligious. explanation is that this is another example of the rise in entitlement—expecting special privileges without effort Declaration of Conflicting Interests (Campbell, Bonacci, Shelton, Exline, & Bushman, 2004; The author(s) declared no potential conflicts of interest with respect Twenge & Foster, 2010). Entitlement appears in religious to the research, authorship, and/or publication of this article. and spiritual domains when people see themselves as deserv- ing spiritual rewards or blessings due to their special status. Funding Entitlement centered on afterlife beliefs could be seen as a The author(s) received no financial support for the research and/or modern rendition of Pascal’s wager, in which the individual authorship of this article. observes that believing in God and a positive afterlife has few downsides, but not believing has the major possible Notes downside of condemnation to eternal suffering (Hájek, 1. Birth cohort refers to everyone born in a given year, and gen- 2003). However, the current data make it difficult to deter- eration to those born within a specified period. Both refer to mine the cause of rising belief in the afterlife. the effects of being born during a certain era and thus are thus somewhat interchangeable; we will use the term generation Limitations and Future Directions most of the time but will use birth cohort when we are spe- cifically referring to birth year. Generational labels (such as Using the GSS data set has several major advantages, includ- Boomers and Millennials) use arbitrary birth year cutoffs; we ing the ability to examine trends among carefully sampled use these labels only for ease of presentation. U.S. adults over long periods of time. Nonetheless, this form 2. In the 2014 survey year, the 18- and 19-year-olds were born of research also has its limitations. Responses are limited to after 1995 and thus are iGen instead of Millennials. The n of self-report, and measures must be brief. As such, the GSS 18- to 19-year-olds was too small to justify a separate analysis does not provide the opportunity for nuanced or in-depth (e.g., n = 51 in 2014). As a proxy, we examined 18- to 22-year- measurement of specific ideas of interest over time. olds (n = 153 in 2014; total n 1972-2014 = 4,927), which in 12 SAGE Open 2014 includes those born 1992-1996 (and thus, those at the psychology of religion and spirituality (2nd ed., pp. 380-398). cusp between Millennials and iGen). In most cases, the decline New York, NY: Guilford. in religious orientation was even more dramatic among 18- to Finke, R., & Stark, R. (1988). Religious economies and sacred 22-year-olds than among 18- to 29-year-olds. For example, the canopies: Religious mobilization in American cities, 1906. percentage of 18- to 22-year-olds who reported no religious American Sociological Review, 53, 41-49. affiliation rose from 11% in 1972-1974 to 36% in 2014; the Finke, R., & Stark, R. (2005). The churching of America, 1776- percentage who reported they never prayed rose from 4% in 2005: Winners and losers in our religious economy. New 1980-1984 to 28% in 2014; the percentage who said they were Brunswick, NJ: Rutgers University Press. “not spiritual at all” rose from 13% in 2006 to 25% in 2014. Fukuyama, F. (1999). The great disruption: Human nature and the Belief in God declined d = -.54 (1988-2014), being a spiritual reconstitution of social order. New York, NY: Free Press. person d = -.21 (1998-2014), and attendance at religious ser- Fuller, R. C. (2001). Spiritual, but not religious: Understanding vices declined d = -.48 (1972-2014). This suggests that iGen unchurched America. Oxford: Oxford University Press. will continue the decrease in religious orientation rather than Graham, J., & Haidt, J. (2010). Beyond beliefs: Religions bind reversing it, even in spirituality. individuals into moral communities. Personality and Social 3. Some controversy has surrounded the issue of which intervals Psychology Review, 14, 140-150. to use in APC models (Bell & Jones, 2013, 2014). We analyzed Hájek, A. (2003). Waging war on Pascal’s wager. Philosophical the data in 2-year, 5-year, and 10-year intervals, and found that Review, 112, 27-56. they all produced very similar results. Hill, P. C., & Pargament, K. I. (2003). Advances in the concep- tualization and measurement of religion and spirituality. Implications for physical and mental health research. The References American Psychologist, 58, 64-74. Bell, A., & Jones, K. (2013). The impossibility of separating age, Hill, P. C., Pargament, K. I., Hood, R. W., McCullough, M. E., period and cohort effects. Social Science & Medicine, 93, Swyers, J. P., Larson, D. B., & Zinnbauer, B. J. (2000). 163-165. Conceptualizing religion and spirituality: Points of common- Bell, A., & Jones, K. (2014). Another “futile quest?” A simula- ality, points of departure. Journal for the Theory of Social tion study of Yang and Land’s hierarchical age-period-cohort Behaviour, 30, 51-77. model. Demographic Research, 30, 333-360. Horn, J. (1965). A rationale and test for the number of factors in Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, factor analysis. Psychometrika, 30, 179-185. S. M. (1985). Habits of the heart: Individualism and commit- Hout, M., & Fischer, C. S. (2002). Why more Americans have ment in American life. Berkeley: University of California Press. no religious preference: Politics and generations. American Berger, P. L. (Ed.). (1999). The desecularization of the world: Sociological Review, 67, 165-187. Resurgent religion and world politics. Grand Rapids, MI: Wm. Howe, N., & Strauss, W. (2000). Millennials rising. New York, B. Eerdmans. NY: Random House. Berger, P. L. (2011). The sacred canopy: Elements of a sociological Idler, E. L., Musick, M. A., Ellison, C. G., George, L. K., Krause, theory of religion. New York, NY: Open Road Media. N., Ory, M. G., . . . Williams, D. R. (2003). Measuring mul- Berger, P. L., Davie, G., & Fokas, E. (2008). Religious America, tiple dimensions of religion and spirituality for health research. secular Europe? A theme and variation. Burlington, VT: Research on Aging, 25, 327-365. Ashgate. Kesebir, P., & Kesebir, S. (2012). The cultural salience of moral Blaine, B., & Crocker, J. (1995). Religiousness, race, and psycho- character and virtue declined in twentieth century America. logical well-being: Exploring social psychological mediators. Journal of Positive Psychology, 7, 471-480. Personality and Social Psychology Bulletin, 21, 1031-1041. Konrath, S. H., O’Brien, E. H., & Hsing, C. (2011). Changes in Campbell, W. K., Bonacci, A. M., Shelton, J., Exline, J. J., & dispositional empathy in American college students over time: Bushman, B. J. (2004). Psychological entitlement: Interpersonal A meta-analysis. Personality and Social Psychology Review, consequences and validation of a new self-report measure. 15, 180-198. Journal of Personality Assessment, 83, 29-45. Lim, C., MacGregor, C. A., & Putnam, R. D. (2010). Secular and Chaves, M. (2011). American religion: Contemporary trends. liminal: Discovering heterogeneity among religious nones. Princeton, NJ: Princeton University Press. Journal for the Scientific Study of Religion, 49, 596-618. Cornwall, M., Albrecht, S. L., Cunningham, P. H., & Pitcher, B. L. Markus, H. R., & Kitayama, S. (2010). Cultures and selves: A cycle (1986). The dimensions of religiosity: A conceptual model with of mutual constitution. Perspectives on Psychological Science, an empirical test. Review of Religious Research, 27, 226-244. 5, 420-430. Dougherty, K. D., Johnson, B. R., & Polson, E. C. (2007). Myers, D. G. (2000). The American paradox: Spiritual hunger in an Recovering the lost: Remeasuring U.S. religious affiliation. age of plenty. New Haven, CT: Yale University Press. Journal for the Scientific Study of Religion, 46, 483-499. Pargament, K. I. (1999). The psychology of religion and spiritual- Exline, J. J., Pargament, K. I., Grubbs, J. B., & Yali, A. M. (2014). ity? Yes and no. International Journal for the Psychology of The Religious and Spiritual Struggles Scale: Development and Religion, 9, 3-16. initial validation. Psychology of Religion and Spirituality, 6, Park, N. S., Lee, B. S., Sun, F., Klemmack, D. L., Roff, L. L., & 208-222. Koenig, H. G. (2013). Typologies of religiousness/spirituality: Exline, J. J., & Rose, E. D. (2013). Religious and spiritual strug- Implications for health and well-being. Journal of Religion & gles. In R. F. Paloutzian & C. L. Park (Eds.), Handbook of the Health, 52, 828-839. Twenge et al. 13 Pew Research Center. (2015). Religious landscape study. Retrieved 1960-2008. Journal of Cross-Cultural Psychology, 44, from http://www.pewforum.org/religious-landscape-study/ 406-415. Pichon, I., Boccato, G., & Saroglou, V. (2007). Nonconscious influ- Twenge, J. M., & Foster, J. D. (2010). Birth cohort increases in nar- ences of religion on prosociality: A priming study. European cissistic personality traits among American college students, Journal of Social Psychology, 37, 1032-1045. 1982-2009. Social Psychological & Personality Science, 1, Piff, P. K., Kraus, M. W., Cote, S., Cheng, B. H., & Keltner, D. 99-106. (2010). Having less, giving more: The influence of social Vandello, J. A., & Cohen, D. (1999). Patterns of individualism and class on prosocial behavior. Journal of Personality and Social collectivism across the United States. Journal of Personality Psychology, 99, 771-784. and Social Psychology, 77, 279-292. Plaut, V. C., Markus, H. R., & Lachman, M. E. (2002). Place mat- Wachholtz, A., & Sambamoorthi, U. (2011). National trends in prayer ters: Consensual features and regional variation in American use as a coping mechanism for health concerns: Changes from well-being and self. Journal of Personality and Social 2002 to 2007. Psychology of Religion and Spirituality, 3, 67-77. Psychology, 83, 160-184. Welzel, C. (2013). Freedom rising: Human empowerment and Presser, S., & Chaves, M. (2007). Is religious service attendance the quest for emancipation. Cambridge, UK: Cambridge declining? Journal for the Scientific Study of Religion, 46, University Press. 417-423. Yang, Y. (2008). Social inequalities in happiness in the United Reither, E. N., Masters, R. K., Yang, Y. C., Powers, D. A., Zheng, States, 1972 to 2004: An age-period-cohort analysis. American H., & Land, K. C. (2015). Should age-period-cohort studies Sociological Review, 73, 204-226. return to the methodologies of the 1970s? Social Science & Yang, Y., & Land, K. C. (2013). Age-period-cohort analysis: New Medicine, 128, 356-365. models, methods, and empirical applications. Boca Raton, FL: Revelle, W. (2015). psych: Procedures for personality and psycho- CRC Press. logical research (Version 1.5.1). Retrieved from https://cran.r- Ysseldyk, R., Matheson, K., & Anisman, H. (2010). Religiosity as project.org/web/packages/psych/citation.html identity: Toward an understanding of religion from a social Saucier, G., & Skrzypińska, K. (2006). Spiritual but not reli- identity perspective. Personality and Social Psychology gious? Evidence for two independent dispositions. Journal of Review, 14, 60-71. Personality, 74, 1257-1292. Zinbarg, R. E., Revelle, W., Yovel, I., & Li, W. (2005). Cronbach’s Schwadel, P. (2011). Age, period, and cohort effects on religious Alpha, Revelle’s Beta, McDonald’s Omega: Their relations activities and beliefs. Social Science Research, 40, 181-192. with each and two alternative conceptualizations of reliability. Shariff, A. F., & Norenzayan, A. (2007). God is watching you: Psychometrika, 70, 123-133. Priming God concepts increases prosocial behavior in an anon- Zinnbauer, B. J., Pargament, K. I., Cole, B., Rye, M. S., Butter, E. ymous economic game. Psychological Science, 18, 803-809. M., Belavich, T. G., . . . Kadar, J. L. (1997). Religion and spiri- Smith, C., Denton, M. L., Faris, R., & Regnerus, M. (2002). tuality: Unfuzzying the fuzzy. Journal for the Scientific Study Mapping American adolescent religious participation. Journal of Religion, 36, 549-564. for the Scientific Study of Religion, 41, 597-612. Smith, C., & Snell, P. (2009). Souls in transition: The religious Author Biographies & spiritual lives of emerging adults. New York, NY: Oxford University Press. Jean M. Twenge is a professor of Psychology at San Diego State Stewart, A. J., & Healy, J. M. (1989). Linking individual develop- University and the author of Generation Me: Why Today’s Young ment and social changes. American Psychologist, 44, 30-42. Americans Are More Confident, Assertive, Entitled – and More Strauss, W., & Howe, N. (1991). Generations: The history of Miserable Than Ever Before (Atria Books, 2014). Her research America’s future, 1584 to 2069. New York, NY: William examines generational differences in values, behaviors, traits, and Morrow. work attitudes. Taylor, P. (2014). The next America: Boomers, Millennials, and Ryne A. Sherman is an assistant professor of Psychology at Florida the looming generational showdown. New York, NY: Public Atlantic University. He received his PhD in personality and social Affairs. psychology from the University of California, Riverside in 2011. Taylor, R., Chatters, L. M., Jayakody, R., & Levin, J. S. (1996). He is broadly interested in personality and situational influences on Black and White differences in religious participation: A real-world human behavior. multisample comparison. Journal for the Scientific Study of Religion, 35, 403-410. Julie J. Exline is a professor of Psychological Sciences at Case Twenge, J. M. (2014). Generation me: Why today’s young Americans Western Reserve University in Cleveland, Ohio. Her primary are more confident, assertive, entitled—And more miserable research interests focus on the psychology of religion and than ever before (2nd ed.). New York, NY: Atria Books. spirituality. Twenge, J. M., Campbell, W. K., & Carter, N. T. (2014). Declines in trust in others and confidence in institutions among American Joshua B. Grubbs is a currently a doctoral candidate in clinical adults and late adolescents, 1972-2012. Psychological Science, psychology at Case Western Reserve University and a pre-doctoral 25, 1914-1923. intern in professional psychology at the Louis Stokes Cleveland VA Twenge, J. M., Campbell, W. K., & Gentile, B. (2013). Changes in Medical Center. His areas of expertise are the psychology of reli- pronoun use in American books and the rise of individualism, gion and spirituality, human sexuality, and narcissism. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png SAGE Open SAGE

Declines in American Adults’ Religious Participation and Beliefs, 1972-2014:

Download PDF
13 pages

Loading next page...
 
/lp/sage/declines-in-american-adults-religious-participation-and-beliefs-1972-vy0Bbb0s39

References (66)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Inc, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
ISSN
2158-2440
eISSN
2158-2440
DOI
10.1177/2158244016638133
Publisher site
See Article on Publisher Site

Abstract

Previous research found declines in Americans’ religious affiliation but few changes in religious beliefs and practices. By 2014, however, markedly fewer Americans participated in religious activities or embraced religious beliefs, with especially striking declines between 2006 and 2014 and among 18- to 29-year-olds in data from the nationally representative General Social Survey (N = 58,893, 1972-2014). In recent years, fewer Americans prayed, believed in God, took the Bible literally, attended religious services, identified as religious, affiliated with a religion, or had confidence in religious institutions. Only slightly more identified as spiritual since 1998, and then only those above age 30. Nearly a third of Millennials were secular not merely in religious affiliation but also in belief in God, religiosity, and religious service attendance, many more than Boomers and Generation X’ers at the same age. Eight times more 18- to 29-year-olds never prayed in 2014 versus the early 1980s. However, Americans have become slightly more likely to believe in an afterlife. In hierarchical linear modeling analyses, the decline in religious commitment was primarily due to time period rather than generation/birth cohort, with the decline in public religious practice larger (d = −.50) and beginning sooner (early 1990s) than the smaller (d = −.18) decline in private religious practice and belief (primarily after 2006). Differences in religious commitment due to gender, race, education, and region grew larger, suggesting a more religiously polarized nation. Keywords religion, cultural change, prayer, religious affiliation, generations Are Americans less religious than they used to be? In previ- these sentiments. Finke and Stark (1988, 2005) contended that ous research, the answer depended on how religious commit- the overall religiousness of the American public has remained ment was measured. Most studies agree that religious relatively constant as a whole, although fluctuations in affilia- affiliation has declined in the United States since the 1970s; tion and expression have occurred. Similarly, Berger’s (1999) for example, more Americans in recent years chose “none” work explored the overall constancy of American religious when asked to identify their religion (e.g., Hout & Fischer, affiliation over time, with a particular focus on how religious- 2002; Lim, MacGregor, & Putnam, 2010; Pew Research ness moved back to the forefront of political and economic Center, 2015). However, several recent studies have con- discourse in recent decades—sentiments also echoed in cluded that religious service attendance, belief in God, and many other seminal works (Berger, 2011; Berger, Davie, & prayer have not changed or have even increased in recent Fokas, 2008). years (e.g., Dougherty, Johnson, & Polson, 2007; Presser & Thus, at least up to the mid- to late-2000s, research sug- Chaves, 2007; Smith & Snell, 2009; P. Taylor, 2014; gests that Americans’ private religious practice and beliefs Wachholtz & Sambamoorthi, 2011). Based on data up to (e.g., those religious practices, disciplines, and beliefs that 2008, Chaves (2011) concluded that belief in God and fre- may be conducted alone or without explicit religious affilia- quency of prayer did not change in the General Social Survey tion) and religious service attendance remained unchanged (GSS) since the 1980s. Examining 18- to 24-year-olds in the GSS 1972-2006, Smith and Snell (2009) found only small San Diego State University, CA, USA changes in religious affiliation and service attendance, and Florida Atlantic University, Boca Raton, USA no changes in frequency of prayer and belief in God. They Case Western Reserve University, Cleveland, OH, USA concluded that emerging adults “have not since 1972 become Corresponding Author: dramatically less religious or more secular . . . if such a trend Jean M. Twenge, Department of Psychology, San Diego State University, is indeed perceptible, it would seem to be weak and slight” 5500 Campanile Drive, San Diego, CA 92182-4611, USA. (pp. 99-100). Other sociologists of religion have echoed Email: jtwenge@mail.sdsu.edu Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License (http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open even as more did not affiliate with a particular religious tradi- Millennial generation (born approximately 1980-1994) is tion. Another possibility is that religious belief has been less religious because they are young or because of genera- replaced by spirituality (Fuller, 2001; Saucier & Skrzypińska, tional or time period change. That is, are Millennials less 2006). In other words, the prevailing conclusion has been religious than Generation X (born 1961-1979) and Boomers that Americans have remained just as religious and/or spiri- (born 1943-1960) were when they were 18- to 29-year-olds? tual in a private or personal sense, but less religious in a pub- This data may also provide an early look at iGen (born 1995- lic sense. This may be due to a more general disassociation 2012) and their religious attitudes. from large groups—for example, Americans have become Changes over time and generations in attitudes, values, significantly less confident in virtually all large institutions and personality traits are rooted in cultural change (Stewart & from government to medicine (Twenge, Campbell, & Carter, Healy, 1989; Twenge, 2014), with cultures and individuals 2014). Such an explanation would also be consistent with mutually influencing and constituting one another (Markus & many popular conceptions of religion as a socially organiz- Kitayama, 2010). One cultural change relevant for religious ing institution (e.g., Ysseldyk, Matheson, & Anisman, 2010) orientation is the rise in individualism, a cultural system plac- that transmits cultural values, mores, and rules (Graham & ing more emphasis on the self and less on social rules (e.g., Haidt, 2010). As societal norms have shifted away from Bellah, Madsen, Sullivan, Swidler, & Tipton, 1985; institutional identification to individualism, one would Fukuyama, 1999; Myers, 2000; Twenge, 2014). Several stud- expect commitment to religion, a ubiquitous social institu- ies have documented increases in focus on the self (Twenge, tion, to similarly decline. However, it is unclear whether Campbell, & Gentile, 2013; Twenge & Foster, 2010) and such decreases in external commitment would also be asso- declines in focus on institutions, empathy for others, and ciated with decreases in personal religious involvement or moral rules (Kesebir & Kesebir, 2012; Konrath, O’Brien, & practice. Hsing, 2011; Twenge et al., 2014). There are several reasons Despite popular conceptions that public religious involve- we would expect religion to decline with greater individual- ment has decreased while private expressions of religion and ism. First, religiosity implies some level of commitment to a spirituality have stayed about the same, stark distinctions larger group or organization. As Welzel (2013) suggests, the between religion and spirituality may be more theoretical trend in Western societies has been toward more freedom and than practical. Although religion and spirituality are known less commitment to groups. Second, belonging to a religious to be distinct constructs (i.e., religion comprises social and group may require assent with the group’s beliefs, opinions, ritualized aspects of personal belief, whereas spirituality and practices. This can create tension when differences in includes the search for meaning or transcendence in daily opinion arise between an individual and an organization (e.g., life; Pargament, 1999), these two constructs often overlap, Exline, Pargament, Grubbs, & Yali, 2014; Exline & Rose, and highly religious individuals often identify as being 2013). Third, religiosity usually involves some rule-follow- highly spiritual as well (for a review, see Hill & Pargament, ing and submission to authority (e.g., Graham & Haidt, 2010), 2003). Moreover, although some individuals certainly do another characteristic that goes against emancipation and identify as spiritual but not religious (e.g., Saucier & individualism. Fourth, religion often focuses on concerns out- Skrzypińska, 2006), a much larger proportion of individuals side of the self, such as helping others and serving God (e.g., identify as both religious and spiritual (Pargament, 1999), Pichon, Boccato, & Saroglou, 2007; Shariff & Norenzayan, and many people have difficulty substantively differentiating 2007). Thus, the increasing individualism of American cul- between the two on an individual level (Hill et al., 2000; ture may have produced decreased religiosity among more Zinnbauer et al., 1997). Therefore, as religious commitment recent time periods and generations. has decreased, one may also expect decreases in private reli- Based on previous research and cultural changes, we gious practice and individual spirituality. expect a decline in religious affiliation. We also predict In this article, we seek to examine whether Americans’ declines in religious service attendance; while religiously religious service attendance, religious practice, religious unaffiliated Americans may attend services for a time, they beliefs, religiosity, spirituality, confidence in religious insti- may become less likely to do so as they feel more disassoci- tutions, and religious affiliation have changed since the ated from religion. Most crucially, we predict declines in 1970s, with a particular focus on the years since 2006 and on more private expressions of religious belief and practice, 18- to 29-year-olds. We take the additional step of calculat- such as prayer, religiosity, and belief in God, with the declines ing effect sizes and performing statistical significance testing especially evident among young people. Belonging to a reli- to quantify the size of the changes. We draw from the nation- gion and more privately believing in its tenets are tradition- ally representative GSS of U.S. adults conducted 1972-2014. ally linked (e.g., Park et al., 2013; Smith, Denton, Faris, & Because this survey draws from a multiage sample above 42 Regnerus, 2002); as more Americans are unaffiliated with years, it can isolate the effects of age from those of time religion, a greater proportion may become not just unaffili- period and generation. Thus, unlike some surveys conducted ated but secular in their beliefs and practices. These declines over a shorter period of time (e.g., 7 years: Pew Research may be especially evident in recent years and among 18- to Center, 2015), this data set can determine, for example, if the 29-year-olds, given the generational and cultural trends Twenge et al. 3 toward emphasizing social rules less and individual freedom sizes; (b) to examine whether these changes are due to genera- more (known as “Generation Me”: Twenge, 2014; or “eman- tion or time period; and (c) to examine whether the trends dif- cipative values”: Welzel, 2013). Moving away from social fer by gender, race, education, or U.S. region. institutions and community engagement would likely detract from one of the key facets of religion as a whole—that is, Method community involvement and social value transmission. A secondary question is whether changes in religious ori- Sample entation over time are caused by time period or generational We drew from the GSS, 1972-2014, a nationally representa- (cohort) effects. If successive generations are less religious tive survey of U.S. residents over 18. Depending on the item, (forming their religious orientation while young and not ns range between 12,862 and 58,893. As suggested by the changing), any decline would be due to generation. If people GSS administrators, we weight the analyses by the weight of all ages have become less religious during certain times, variable WTSSALL to make the sample nationally represen- any decline would be due to time period. New hierarchical tative of individuals rather than households and correct for linear modeling techniques (called APC or age-period- other sampling biases. However, these weighted analyses cohort analyses) attempt to separate the effects of age, gen- differ only very slightly from unweighted analyses. Also as eration, and time period (Yang, 2008; Yang & Land, 2013). suggested by the administrators, we excluded the black Some have argued that these techniques do not resolve the oversamples collected in 1982 and 1987. identification problem that has long plagued simultaneous analysis of age, period, and cohort effects (e.g., Bell & Jones, 2013, 2014); however, these criticisms appear to Items largely rest on untenable assumptions that are not consistent with basic APC models (Reither et al., 2015). In addition, We identified and analyzed all items on respondents’ own APC techniques have become widely used. For example, religious orientation asked in at least six administrations of Schwadel (2011) performed an APC analysis on some of the the GSS. They were as follows: GSS religion variables up to 2006. However, at that time the data included only a handful of Millennials, a generation 1. Religious preference: “What is your religious prefer- purported to be less religious; by 2014, however, Millennials ence? Is it Protestant, Catholic, Jewish, some other were the entirety of 20- to 29-year-olds. In addition, further religion, or no religion?” We analyzed the percentage time period change may have occurred in the 8 years of data of respondents who chose “no religion.” Asked available since 2006. Thus, we perform APC analyses to 1972-2014. examine whether shifts in Americans’ religious orientation 2. Strength of religious affiliation: “Would you call are due to generational or time period effects. yourself a strong (Christian, Jew, etc.) or not a very In addition, we examine possible moderators of change strong (Christian, Jew, etc.)?” Response choices were over time in religious orientation. Trends may differ among “not very strong,” “somewhat strong,” and “strong.” men and women, Blacks and Whites, education levels, and Asked 1974-2014. U.S. regions, as these groups differ in their levels of religios- 3. Religious service attendance: “How often do you ity and cultural focus (Blaine & Crocker, 1995; Piff, Kraus, attend religious services?” Response choices were Cote, Cheng, & Keltner, 2010; Plaut, Markus, & Lachman, “never,” “less than once a year,” “about once a year,” 2002; R. Taylor, Chatters, Jayakody, & Levin, 1996; Vandello “about once or twice a year,” “Several times a year,” & Cohen, 1999). We theorize that the decline in religious “about once a month,” “2-3 times a month,” “nearly orientation will be larger among demographic groups and every week,” and “every week.” Asked 1972-2014. regions with higher social power and more individualism, 4. Belief in the afterlife: “Do you believe there is a life including Whites, men, those with a college education, and after death?” Response choices were “yes” and “no.” living in the Northwest and West, and lower or nonexistent Asked 1973-2014. among groups with lower social power and less individual- 5. Believing the Bible is literal: “Which of these state- ism, including Blacks, women, those without a college edu- ments comes closest to describing your feelings cation, and the Midwest and South (e.g., Piff et al., 2010; about the Bible?” Response choices were “The Bible Vandello & Cohen, 1999). Groups with relatively high social is an ancient book of fables, legends, history, and power might not see themselves as having a significant need moral precepts recorded by men”; “The Bible is the for religion or God, so these groups might pioneer the move- inspired word of God but not everything in it should ment toward less religiosity. be taken literally, word for word”; and “The Bible is Thus, we have three goals in this article: (a) to perform a the actual word of God and is to be taken literally, comprehensive examination of American adults’ religious ori- word for word.” Asked 1984-2014. entation from 1972 through 2014, with a particular emphasis 6. Frequency of praying: “About how often do you on 2006-2014 and 18- to 29-year-olds, and including effect pray?” Response choices were “never,” “less than 4 SAGE Open once a week,” “once a week,” “several times a week,” or by individual year. Given our focus on both overall change “once a day,” and “several times a day.” Asked since the 1970s and change since 2006, we separated the data 1983-2014. into 5-year intervals from 1972-2004 and reported data by 7. Belief in God: “Please look at this card and tell me individual year from 2006 to 2014. We report the effect sizes which of the statements comes closest to expressing (d, or difference in terms of standard deviations) and p values what you believe about God.” Response choices were for t tests comparing 1972-1974 with 2014 and 2006 and “I don’t believe in God”; “I don’t know whether there 2014. We also include two figures with all of the year-by- is a God and I don’t believe there is any way to find year data for some variables. We report both continuous vari- out”; “I don’t believe in a personal God, but do ables (e.g., the 0-8 scale for religious service attendance) and believe in a Higher Power of some kind”; “I find dichotomous variables (e.g., the percentage who never attend myself believing in God some of the time, but not at religious services). We use the tables for means and report others”; “While I have doubts, I feel that I do believe percentage changes in the text. in God”; and “I know God really exists and I have no For the APC models, we estimated random coefficient doubts about it.” Asked 1988-2014. models allowing intercepts to vary across time periods 8. Confidence in religious institutions: “I am going to (years) and generations (cohorts). Thus, effectively, an inter- name some institutions in this country. As far as the cept (mean religious orientation) score is calculated (using people running these institutions are concerned, empirical Bayes) for each cohort and each survey year. In would you say that you have a great deal of confi- addition, a fixed intercept (grand mean) is estimated along dence, quite a lot of confidence, only some confi- with a fixed regression coefficient for age and age squared. dence, or very little in them?” One of the items is This model has three variance components: One for variabil- “organized religion.” Response choices were “hardly ity in intercepts due to cohorts (τ ), one for variability in u0 any confidence at all,” “only some confidence,” or “a intercepts due to period (τ ), and a residual term containing v0 great deal of confidence.” We excluded “don’t know” unmodeled variance within cohorts and periods. Variance in and “refused” responses. Asked 1973-2014. the intercepts across time periods and cohorts indicates 9. Identification as a religious person: “To what extent period and cohort differences, respectively (Yang & Land, do you consider yourself a religious person?” 2013). Thus, the technique allows for a separation of the Response choices of “not religious at all,” “slightly effects of generation/cohort, time period, and age. Weighting religious,” “moderately religious,” and “very reli- could not be used for the mixed-effects analyses because gious.” Asked 1998 and 2006-2014. proper probability weighting for variance component estima- 10. Identification as a spiritual person: “To what extent tion requires taking into account pairwise selection probabil- do you consider yourself a spiritual person?” ities, which is not possible in current statistical software. Response choices of “not spiritual at all,” “slightly In describing the trends in the text and tables, we will spiritual,” “moderately spiritual,” and “very spiri- sometimes use common labels for the generations such as the tual.” Asked 1998 and 2006-2014. G.I. or “Greatest” generation (born 1900-1924), Silent (1925-1942), Boomers (1943-1960; some argue 1946-1964), Of these, religious preference, strength of religious affili- GenX (1961-1979 or 1965-1979), Millennials (1980-1994; ation, religious service attendance, and confidence in reli- for reviews, see Strauss & Howe, 1991; Twenge, 2014), and gious institutions are public religious variables, and belief in iGen (1995-2012). These birth year cutoffs are arbitrary and an afterlife, believing the Bible is literal, frequency of pray- are not necessarily justified by empirical evidence, but are ing, belief in God, identification as a religious person, and useful shorthand labels for those born in certain eras. identification as a spiritual person are private religious Differences among those of the same generation certainly variables. occur, and these can be seen in the tables and figures; we use these labels merely for convenience. Possible Moderators Results We analyzed moderation by gender (men vs. women), race (White vs. Black, the only racial groups measured in all sur- Trends in Religious Orientation vey years), education level (high school graduate and below American adults in the 2010s were less religious than those vs. attended some college and above), and U.S. region in previous decades, based on religious service attendance (Northeast, Midwest, South, and West). and more private religious expressions such as belief in God, praying, identifying as a religious person, and believing the Procedure Bible is the word of God (see Table 1 and Figure 1). These findings held when restricted to 18- to 29-year-olds Data collected over time can be analyzed in many ways, (see Table 2 and Figure 2), demonstrating that Millennials including grouping by 20-year generation blocks, by decades, 5 Table 1. Religious Orientation Among All Adult Americans, 1972-2014, General Social Survey. d (2006 d (earliest Survey items n 72-74 75-79 80-84 85-89 90-94 95-99 00-04 2006 2008 2010 2012 2014 vs. 2014) vs. 2014) Religion “none” (%) 58,893 6% (0.24) 7% (0.26) 7% (0.26) 7% (0.26) 8% (0.27) 13% (0.33) 14% (0.35) 16% (0.36) 17% (0.37) 18% (0.38) 20% (0.40) 21% (0.40) .13*** .51*** Strength of religious 53,938 3.18 (0.86) 3.14 (0.86) 3.19 (0.87) 3.12 (0.89) 3.08 (0.92) 2.98 (1.03) 2.98 (1.03) 2.92 (1.04) 2.93 (1.06) 2.91 (1.08) 2.89 (1.11) 2.87 (1.12) −.05 −.30*** affiliation (1-4) Attend religious services 58,347 4.13 (2.63) 3.94 (2.68) 4.05 (2.66) 3.99 (2.67) 3.89 (2.69) 3.67 (2.71) 3.66 (2.72) 3.57 (2.79) 3.56 (2.77) 3.53 (2.78) 3.49 (2.82) 3.33 (2.83) −.09** −.30*** (0-8) Never attend religious 58,347 11% (0.32) 14% (0.35) 13% (0.33) 14% (0.35) 15% (0.36) 17% (0.38) 19% (0.39) 22% (0.42) 22% (0.42) 23% (0.42) 25% (0.44) 26% (0.44) .09** .41*** services (%) Confidence in organized 38,281 2.27 (0.69) 2.15 (0.70) 2.15 (0.70) 2.01 (0.70) 2.01 (0.69) 2.07 (0.68) 2.04 (0.68) 2.03 (0.69) 1.95 (0.68) 1.96 (0.67) 1.96 (0.67) 1.96 (0.66) −.10** −.46*** religion (1-3) Praying (1-6) 30,333 — — 4.24 (1.55) 4.24 (1.49) 4.19 (1.52) 4.24 (1.52) 4.28 (1.58) 4.26 (1.72) 4.16 (1.75) 4.18 (1.75) 4.20 (1.77) 4.12 (1.80) −.08** −.07** Never pray (%) 30,333 — — 3% (0.16) 1% (0.08) 1% (0.10) 2% (0.14) 4% (0.19) 11% (0.30) 12% (0.32) 13% (0.34) 14% (0.35) 15% (0.36) .12*** .48*** Believe in God (1-6) 19,542 — — — 5.26 (1.22) 5.23 (1.31) 5.11 (1.43) 5.21 (1.34) 5.18 (1.32) 5.11 (1.41) 5.02 (1.45) 5.03 (1.45) 4.98 (1.47) −.14*** −.20*** Do not believe in God (%) 19,542 — — — 13% (0.33) 15% (0.35) 18% (0.39) 15% (0.35) 16% (0.37) 18% (0.39) 20% (0.40) 20% (0.40) 22% (0.41) .15*** .24*** Bible literal (1-3) 30,084 — — 2.25 (0.68) 2.20 (0.68) 2.19 (0.67) 2.15 (0.68) 2.18 (0.68) 2.17 (0.70) 2.13 (0.71) 2.14 (0.73) 2.10 (0.73) 2.10 (0.73) −.10** −.21*** Bible = fables (%) 30.084 — — 14% (0.34) 15% (0.36) 15% (0.36) 17% (0.37) 16% (0.37) 17% (0.38) 20% (0.41) 21% (0.41) 22% (0.42) 22% (0.42) .13*** .20*** Believe in afterlife (%) 35,391 76% (0.43) 76% (0.42) 79% (0.41) 80% (0.40) 82% (0.40) 82% (0.38) 82% (0.38) 83% (0.38) 81% (0.39) 81% (0.39) 81% (0.39) 79% (0.41) −.10** .07* Religious person (1-4) 12,924 — — — — — 2.65 (0.95) — 2.68 (0.94) 2.62 (0.96) 2.57 (0.97) 2.57 (1.01) 2.51 (1.00) −.18*** −.14*** Not religious at all (%) 12,924 — — — — — 15% (0.36) — 14% (0.35) 16% (0.37) 18% (0.39) 20% (0.40) 20% (0.40) .16*** .13*** Spiritual person (1-4) 12,862 — — — — — 2.72 (0.93) — 2.89 (0.92) 2.84 (0.92) 2.84 (0.92) 2.85 (0.97) 2.82 (0.97) −.07** .10** Not at all spiritual (%) 12,862 — — — — — 12% (0.32) — 9% (0.29) 10% (0.30) 11% (0.31) 11% (0.31) 11% (0.31) .07** −.03 Difference between service 37,419 0.19 (1.19) 0.18 (1.26) 0.17 (1.24) 0.11 (1.24) 0.08 (1.24) −0.03 (1.26) −0.05 (1.23) −0.10 (1.26) −0.08 (1.24) −0.06 (1.27) −0.08 (1.25) −0.10 (1.26) .00 .24*** attendance and belief in afterlife (z score) Never attend services but 37,419 7% (0.26) 8% (0.26) 7% (0.25) 9% (0.29) 9% (0.28) 12% (0.32) 12% (0.32) 15% (0.36) 13% (0.34) 13% (0.34) 15% (0.36) 15% (0.36) .00 .25*** believe in afterlife (%) Note. — = question was not asked in those years. Standard deviations are in parentheses. d = difference in standard deviations. *p < .05. **p < .01. ***p < .001, based on a t test. 6 SAGE Open Figure 1. Percentage of all American adults with no religious affiliation, who never attend services, never pray, do not believe in God, are not religious at all, and are not spiritual at all. are less religious than previous generations were at the same the less certain response choices such as “I don’t believe in age. While religious affiliation and service attendance have God”; “I don’t know whether there is a God and I don’t been declining since the 1990s, the decrease in more private believe there is any way to find out”; or “I don’t believe in a religious expressions began fairly recently, becoming pro- personal God, but do believe in a Higher Power of some nounced only after 2006 (see Figures 1 and 2). Effect sizes kind”; these responses were combined into “Do not believe ranged from moderate (around d = .50; Cohen, 1988) to in God” in Tables 1 and 2). By 2014, however, 22% expressed small (around d = .20). The increase in never praying among doubts, a 69% increase. Among 18- to 29-year-olds, 30% had 18- to 29-year-olds was d = .80, equaling the guideline for a serious doubts by 2014, more than twice as many as in the large effect. late 1980s (12%). As found in previous research, fewer Americans now Americans have also become less likely to believe that the affiliate with a religion. Although the majority of Americans Bible is the word of God. In 1984, 14% of Americans are still religious, three times as many in 2014 (vs. the early believed the Bible “is an ancient book of fables, legends, his- 1970s) have no religious affiliation, and twice as many never tory, and moral precepts recorded by men” rather than the attend religious services. Fewer have confidence in orga- word of God; by 2014, 22% of Americans believed this, a nized religion; the number who said they had “hardly any” 57% increase. Among 18- to 29-year-olds, 29% believed this confidence went from 14% in the early 1970s to 24% in by 2014, nearly twice as many as in the late 1980s (15%). 2014, a 71% increase, and those who said they had “a great Fewer Americans identify as religious; 62% said they deal” of confidence was cut in half (from 41% to 20%). were “moderately” or “very” religious in 1998, compared By 2014, the declines in religious orientation extended to with 54% in 2014, a 13% decrease. Among 18- to 29-year- more personal and private religious beliefs. Five times as olds, 49% said they were moderately or very religious in many Americans in 2014 (vs. the late 1980s) never prayed 1998, compared with 38% in 2014, a 22% decrease. Similarly, (eight times more among those ages 18-29). Slightly more those who said they were “not religious at all” increased Americans in 2014 (vs. the 1980s) said they prayed “several from 15% in 1998 to 20% in 2014 among all adults (a 33% times a day” (28%, up from 26%), but the 20% who prayed increase), and from 23% in 1998 to 28% in 2014 among 18- “less than once a week” in the 1980s became only 11% in to 29-year-olds (a 22% increase). 2014, apparently moving to “never” praying (3% in the Has religiosity been replaced with spirituality? It does not 1980s vs. 15% in 2014). appear so. Identifying as a spiritual person increased between Americans in 2014 were less likely to say they believed in 1998 and 2006, but then declined between 2006 and 2014 God. In the late 1980s, only 13% of U.S. adults expressed (see Table 1 and Figure 1). In all, 62% identified as moder- serious doubts about the existence of God (choosing one of ately or strongly spiritual in 1998, compared with 70% in 7 Table 2. Religious Orientation Among American 18- to 29-Year-Olds, 1972-2014, General Social Survey. d (2006 d (earliest Survey items n 72-74 75-79 80-84 85-89 90-94 95-99 00-04 2006 2008 2010 2012 2014 vs. 2014) vs. 2014) Religion “none” 13,724 12% (0.32) 12% (0.33) 12% (0.32) 12% (0.32) 12% (0.33) 22% (0.41) 22% (0.41) 23% (0.42) 28% (0.45) 26% (0.44) 31% (0.46) 31% (0.46) .18** .53*** Strength of religious affiliation 12,495 2.97 (0.91) 2.92 (0.92) 3.02 (0.94) 2.93 (0.94) 2.91 (0.96) 2.72 (1.10) 2.73 (1.10) 2.69 (1.10) 2.65 (1.16) 2.67 (1.16) 2.56 (1.18) 2.57 (1.17) −.11 −.38*** (1-4) Attend religious services 13,631 3.60 (2.51) 3.39 (2.56) 3.56 (2.51) 3.48 (2.51) 3.52 (2.50) 2.96 (2.44) 3.18 (2.52) 2.99 (2.51) 2.86 (2.59) 3.11 (2.68) 2.71 (2.52) 2.68 (2.61) −.12* −.36*** (0-8) Never attend religious 13,631 14% (0.35) 17% (0.38) 14% (0.34) 17% (0.37) 15% (0.35) 21% (0.41) 21% (0.41) 28% (0.45) 29% (0.46) 28% (0.45) 31% (0.46) 33% (0.47) .11* .31*** services Confidence in organized 9,158 2.17 (0.68) 2.10 (0.69) 2.09 (0.69) 1.99 (0.70) 2.02 (0.69) 2.07 (0.69) 2.05 (0.71) 2.02 (0.73) 1.99 (0.70) 1.99 (0.73) 1.97 (0.67) 1.88 (0.64) −.20* −.43*** religion (1-3) Praying (1-6) 6,818 — — 3.78 (1.54) 3.79 (1.43) 3.76 (1.54) 3.75 (1.54) 3.87 (1.60) 3.76 (1.81) 3.62 (1.78) 3.67 (1.87) 3.44 (1.85) 3.55 (1.88) −.11 −.14** Never pray (%) 6,818 — — 3% (0.17) .04% (0.07) 1.4% (0.13) 3% (0.18) 5% (0.21) 19% (0.39) 17% (0.38) 22% (0.41) 25% (0.43) 24% (0.43) .12* .80*** Believe in God (1-6) 4,126 — — — 5.15 (1.21) 5.12 (1.39) 4.88 (1.54) 4.94 (1.45) 4.90 (1.48) 4.86 (1.54) 4.71 (1.57) 4.61 (1.56) 4.61 (1.63) −.19** −.38*** Do not believe in God (%) 4,126 — — — 12% (0.33) 16% (0.36) 25% (0.43) 21% (0.40) 22% (0.42) 24% (0.42) 26% (0.44) 29% (0.46) 30% (0.46) .19** .45*** Bible literal (1-3) 6,500 — — 2.24 (0.69) 2.15 (0.68) 2.18 (0.68) 2.08 (0.69) 2.11 (0.69) 2.12 (0.69) 2.06 (0.70) 2.08 (0.74) 1.95 (0.72) 1.97 (0.74) −.21** −.37*** Bible = fables (%) 6,500 — — 15% (0.36) 17% (0.38) 16% (0.36) 20% (0.40) 19% (0.39) 21% (0.41) 21% (0.41) 24% (0.43) 29% (0.45) 29% (0.45) .19** .33*** Believe in afterlife (%) 8,775 73% (0.44) 72% (0.45) 79% (0.41) 79% (0.41) 78% (0.42) 81% (0.39) 82% (0.38) 82% (0.38) 79% (0.41) 82% (0.39) 79% (0.41) 80% (0.40) −.05 .17* Religious person (1-4) 2,632 — — — — — 2.39 (0.97) — 2.42 (0.96) 2.33 (0.99) 2.31 (0.96) 2.25 (0.99) 2.20 (0.97) −.23*** −.20*** Not religious at all (%) 2,632 — — — — — 23% (0.42) — 21% (0.41) 25% (0.44) 25% (0.44) 29% (0.45) 28% (0.45) .17** .11 Spiritual person (1-4) 2,628 — — — — — 2.51 (0.93) — 2.64 (0.93) 2.61 (0.98) 2.50 (0.98) 2.48 (0.99) 2.48 (1.02) −.17** −.03 Not at all spiritual (%) 2,628 — — — — — 14% (0.35) — 13% (0.34) 16% (0.36) 18% (0.39) 18% (0.39) 19% (0.39) .17** .13 Difference between service 8,734 0.03 (1.26) 0.08 (1.33) −0.01 (1.27) −0.06 (1.29) 0.00 (1.25) −0.25 (1.27) −0.23 (1.25) −0.34 (1.17) −0.27 (1.24) −0.20 (1.25) −0.35 (1.28) −0.38 (1.23) −.03 −.33*** attendance and belief in afterlife (z score) Never attend services but 7,596 11% (0.32) 10% (0.31) 8% (0.27) 11% (0.32) 8% (0.27) 15% (0.36) 14% (0.35) 19% (0.39) 18% (0.39) 15% (0.36) 22% (0.42) 21% (0.41) .05 .27*** believe in afterlife (%) 370 Note. — = question was not asked in those years. Standard deviations are in parentheses. d = difference in standard deviations. *p < .05. **p < .01. ***p < .001, based on a t test. 8 SAGE Open Figure 2. Percentage of 18- to 29-year-old Americans with no religious affiliation, who never attend services, never pray, do not believe in God, are not religious at all, and are not spiritual at all. 2006 and 65% in 2014; thus, identification as a spiritual per- person and spiritual person variables were not asked in son increased 5% between 1998 and 2014, a small increase enough years to be included, so we limited this analysis to compared to the larger declines in religious belief and prac- the other eight variables.) We used the continuous form of tice. In addition, the percentage of 18- to 29-year-olds iden- six variables (strength of religious affiliation, religious ser- tifying as moderately or strongly spiritual declined 6%, from vice attendance, frequency of prayer, belief in God, belief in 50% in 1998 to 47% in 2014. In 1998, 14% of 18- to 29-year- the Bible as literal, confidence in religious institutions), with olds said they were not spiritual at all, rising to 19% by 2014, religious affiliation (none vs. affiliated) and belief in an a 36% increase (see Table 2 and Figure 2). Thus, there is afterlife (yes vs. no) dichotomous. We included only respon- some suggestion that young people were less spiritual in dents who completed at least four of the eight items. Horn’s 2014 versus 1998, though the decline was not statistically (1965) parallel analysis of n = 8,513 cases with no missing significant. In 2014, fewer 18- to 29-year-olds (Millennials) values indicated that only a one-component solution had an identified as spiritual (47%) than those 50 and above (72%). Eigenvalue better than chance levels. Moreover, all variables This suggests that identification as a spiritual person may loaded highly onto a single principal component explaining continue to decline. 46% of the variance, with a model fit of .94 (on a 0-1 scale). One increase in religious belief did emerge: Slightly more Using the omega function available in the {psych} package Americans believe in life after death (see Tables 1 and 2). in R (Revelle, 2015) indicated that 50% of the common vari- Thus, more Americans believe in life after death even as ance in the item scores could be accounted for by a general fewer belong to a religion, fewer attend religious services, factor of religious orientation. In addition, the omega coeffi- and fewer pray. In the 1970s, only about 7% of Americans cient, which is the best estimator of single factor saturation never attended religious services but nevertheless believed in (see Zinbarg, Revelle, Yovel, & Li, 2005), was .70 suggest- life after death; by 2014, twice as many (15%) showed this ing that a single factor accounted for much of the variability disconnect between behavior and belief, and 21% among in these items. The principal components analysis indicated young people. a single principal component across the decades of data col- lection. Therefore, all variables were z scored, and a compos- ite religious orientation variable was formed (n = 52,497, Mixed-Effects Analyses to Separate Time Period, M = 0.01, SD = 0.69, α = .83). Generation, and Age Next, we performed mixed-effects analyses to separate First, we performed a principal components analysis to the effects of time period, generation, and age on the com- determine whether the religion variables could be combined posite variable. The SD in intercepts for period (survey into a composite variable for use in the mixed-effects APC year) was .12 [.09, .16] and for cohorts was .03 [.00, .04], model analyses; combining these variables into an index suggesting that almost none of the variability in religious ori- increases internal reliability over single items. (The religious entation was due to cohorts. There was also a statistically Twenge et al. 9 Figure 3. Adult Americans’ religious orientation by generation (cohort/birth year) and time period (survey year), in mixed-effects analyses separating time period, generation/cohort, and age. significant effect for age (b = .011 [.010, .014]) indicating private religious practice and belief was smaller (d = −.18 that older individuals were higher on religious orientation between 1984 and 2014) and began later (with a consistent (thus, religious orientation increases about d = .01 with each decline beginning around 2006-2008; d = −.12 of the change year of age). There was a weaker quadratic effect of age occurred between 2006 and 2014). (b = −.00005 [−.00008, −.00002]) indicating that the linear effect is not as strong at older ages. Moderators of the Decline in Religious Overall, there was a marked time period effect when gen- Orientation eration and age were controlled (see Figure 3). Religious ori- entation declined d = −.38 from 1973 to 2014, and d = −.15 We next analyzed whether the time period and cohort between 2006 and 2014. The generational effect was weaker, decrease in religious orientation (controlled for each other with religious orientation declining the most between those and age) differed based on race, U.S. region, sex, and educa- born in the 1930s and the Millennials born in the 1980s-1990s tion level. The trends were moderated by race, with no (d = −.06). change in religious orientation for Black Americans (d = .00) Although religious orientation formed a single factor, we and a large decrease among White Americans (d = −.48). In also examined whether the pattern of change was different the early 1970s, Whites and Blacks differed little in religious for public (affiliation, strength of affiliation, service atten- orientation (d = .15, 1973-1974, with Blacks higher), but by dance, confidence in religious institutions) and private (belief 2014, there was a marked racial difference, with Blacks in the afterlife, belief that the Bible is literal, praying, belief higher (d = .67). Cohort effects were weak for both Whites in God) religious practice. Similar to the analyses with one and Blacks. combined variable, time period explained more of the change The effects also differed by U.S. region, with the decline than birth cohort for both public and private religious in religious orientation largest in the West (d = −.42), Northeast practice. However, the pattern of change and its size differed (d = −.27), South (d = −.10), and Midwest (d = −.07). However, (see Figure 4). The decline in public religious practice was Midwesterners showed a pronounced cohort effect from larger (d = −.50 between 1972 and 2014, and d = −.42 those born in the 1880s to those born in the 1990s (d = −1.15), between 1984 and 2014) and began sooner (with the consis- compared with the nonexistent cohort effects in the other tent decline beginning around 1991-1993). The decline in three regions. In the early 1970s, Southern residents were 10 SAGE Open Figure 4. Time period changes in adult Americans’ public religious practice (left) and private religious belief/practice (right), in mixed- effects analyses separating time period, generation/cohort, and age. only somewhat more religious than those in the Northeast (d the Bible was the word of God than they were in previous = .23), but by the 2010s, Southerners were moderately higher decades. Thus, the decline in religious affiliation found in in religious orientation compared to Northeasterners (d = previous research has now extended to religious service atten- .40). The West was the least religious region in both eras, dance and, by 2008 and afterward, to personal religious belief with Westerners lower than Southerners in 1972 (d = −.47) and practice. The only exceptions were an increase in belief but even more so in 2014 (d = −.78). in the afterlife and a small increase in identifying as spiritual An intriguing pattern appeared when examining men and between 1998 and 2006 limited to those above 30. The women separately: The time period difference was some- declines in religious orientation were particularly striking what larger for women (d = −.28) than for men (d = −.12), between the early 2000s and 2014 and among those 18 to 29 but men showed a pronounced cohort decline in religious years old. Nearly a third of Millennials are not just religiously orientation (d = −.93), while women showed virtually no unaffiliated, but secular in other ways (doubting the existence effect for cohort (d = −.02). of God, believing the Bible is a book of fables, not attending Similarly, the time period decline in religious orientation religious services, describing oneself as “not religious at all,” was somewhat larger among those who had not attended col- never praying), and one out of five also say they are “not spir- lege (d = −.28) compared with those who attended at least some itual at all.” Although religious orientation is often concep- college (d = −.15). However, there was a moderate cohort tualized as a multidimensional concept (e.g., Cornwall, decline in religious orientation among those who attended col- Albrecht, Cunningham, & Pitcher, 1986; Idler et al., 2003), lege (d = −.39) and virtually none for those who did not attend the present data indicated that declines in religious affiliation college (d = −.02). Overall, gender, race, education, and extended across various measures of religious participation regional differences in religious commitment grew larger and commitment. The decline in religious affiliation and par- between the 1970s and the 2010s or between cohorts born in ticipation has now extended to private practices and beliefs, the late 19th century and those born in the late 20th century. though the decline in private religious practice and belief is smaller and began later than the decline in public religious practice. Discussion Mixed-effects analyses demonstrated that these trends By 2014, American adults were less likely to pray, believe in were primarily due to time period. Millennials were less reli- God, identify as religious, attend religious services, or believe gious than their Boomer and Generation X predecessors were Twenge et al. 11 at the same age, demonstrating that their lower religious com- Principal component and omega analyses demonstrated mitment is not solely due to their developmental stage of that a single factor captured the eight religious orientation young adulthood. However, this appears to be due to a time variables. Although religiosity is usually conceptualized as period effect in which all generations are growing less reli- multidimensional (Cornwall et al., 1986; Idler et al., 2003), gious over time. This suggests support for the idea that grow- in this data set, the majority of variation in religious orienta- ing individualism has been accompanied by less religion on a tion was determined by a single factor. We tried to strike a larger cultural basis, with a larger linear cohort decline in balance between internal reliability and diversity among some groups (men, Midwesterners, the college educated). individual items by presenting analyses of single items in the These findings contradict popular culture notions of genera- tables and focusing the APC analyses on the composite mea- tions cycling back and forth with, for example, a less religious sure and on the public and private practice measures. generation being followed by a more religious one. For exam- Our focus here was on individuals in the United States, so ple, generational theorists Howe and Strauss (2000), who cross-cultural studies should examine temporal trends in reli- adhere to the theory that generations come in cycles, proposed gious orientation in other countries. Religious groups may that Millennials would be more religious than GenX’ers. also differ in how they manage the discrepancy between reli- However, these data strongly suggest that the opposite is true. gious participation and afterlife beliefs, based on teachings Men and women, Blacks and Whites, the college edu- about the afterlife and whether (and how) the afterlife is con- cated and not college educated, and the South versus the nected with choices or behaviors during this life. Northeast are becoming more polarized in their religious ori- entation: While differences in religious commitment between Conclusion these groups were small during the 1970s, they have grown larger in recent years and with recent cohorts. The decline in The 2010s are a time of tremendous change in the religious religious commitment was most pronounced among men, landscape of the United States. Although the majority of Whites, and those in the Midwest, Northeast, and West, and Americans are still religious, the declines in public religious was nearly absent among Black Americans and small in the affiliation observed in previous research have, by 2014, South. It appears that groups with relatively high social extended to private religious belief and action (such as power are less likely to see themselves as having a signifi- prayer, belief in God, and identifying as religious). This cant need for religion or God in recent years. decline was not replaced by a substantial increase in those In comparison with those from earlier years and genera- identifying as spiritual. The slight increases in afterlife belief tions, American adults in recent years and generations were represent a potentially important exception to this pattern. slightly more likely to believe in an afterlife. Combined Overall, the data suggest a pervasive decline in religious par- with the decline in religious participation and belief, this ticipation and belief among Americans, with a burgeoning might seem paradoxical. One plausible, though speculative, minority becoming decidedly nonreligious. explanation is that this is another example of the rise in entitlement—expecting special privileges without effort Declaration of Conflicting Interests (Campbell, Bonacci, Shelton, Exline, & Bushman, 2004; The author(s) declared no potential conflicts of interest with respect Twenge & Foster, 2010). Entitlement appears in religious to the research, authorship, and/or publication of this article. and spiritual domains when people see themselves as deserv- ing spiritual rewards or blessings due to their special status. Funding Entitlement centered on afterlife beliefs could be seen as a The author(s) received no financial support for the research and/or modern rendition of Pascal’s wager, in which the individual authorship of this article. observes that believing in God and a positive afterlife has few downsides, but not believing has the major possible Notes downside of condemnation to eternal suffering (Hájek, 1. Birth cohort refers to everyone born in a given year, and gen- 2003). However, the current data make it difficult to deter- eration to those born within a specified period. Both refer to mine the cause of rising belief in the afterlife. the effects of being born during a certain era and thus are thus somewhat interchangeable; we will use the term generation Limitations and Future Directions most of the time but will use birth cohort when we are spe- cifically referring to birth year. Generational labels (such as Using the GSS data set has several major advantages, includ- Boomers and Millennials) use arbitrary birth year cutoffs; we ing the ability to examine trends among carefully sampled use these labels only for ease of presentation. U.S. adults over long periods of time. Nonetheless, this form 2. In the 2014 survey year, the 18- and 19-year-olds were born of research also has its limitations. Responses are limited to after 1995 and thus are iGen instead of Millennials. The n of self-report, and measures must be brief. As such, the GSS 18- to 19-year-olds was too small to justify a separate analysis does not provide the opportunity for nuanced or in-depth (e.g., n = 51 in 2014). As a proxy, we examined 18- to 22-year- measurement of specific ideas of interest over time. olds (n = 153 in 2014; total n 1972-2014 = 4,927), which in 12 SAGE Open 2014 includes those born 1992-1996 (and thus, those at the psychology of religion and spirituality (2nd ed., pp. 380-398). cusp between Millennials and iGen). In most cases, the decline New York, NY: Guilford. in religious orientation was even more dramatic among 18- to Finke, R., & Stark, R. (1988). Religious economies and sacred 22-year-olds than among 18- to 29-year-olds. For example, the canopies: Religious mobilization in American cities, 1906. percentage of 18- to 22-year-olds who reported no religious American Sociological Review, 53, 41-49. affiliation rose from 11% in 1972-1974 to 36% in 2014; the Finke, R., & Stark, R. (2005). The churching of America, 1776- percentage who reported they never prayed rose from 4% in 2005: Winners and losers in our religious economy. New 1980-1984 to 28% in 2014; the percentage who said they were Brunswick, NJ: Rutgers University Press. “not spiritual at all” rose from 13% in 2006 to 25% in 2014. Fukuyama, F. (1999). The great disruption: Human nature and the Belief in God declined d = -.54 (1988-2014), being a spiritual reconstitution of social order. New York, NY: Free Press. person d = -.21 (1998-2014), and attendance at religious ser- Fuller, R. C. (2001). Spiritual, but not religious: Understanding vices declined d = -.48 (1972-2014). This suggests that iGen unchurched America. Oxford: Oxford University Press. will continue the decrease in religious orientation rather than Graham, J., & Haidt, J. (2010). Beyond beliefs: Religions bind reversing it, even in spirituality. individuals into moral communities. Personality and Social 3. Some controversy has surrounded the issue of which intervals Psychology Review, 14, 140-150. to use in APC models (Bell & Jones, 2013, 2014). We analyzed Hájek, A. (2003). Waging war on Pascal’s wager. Philosophical the data in 2-year, 5-year, and 10-year intervals, and found that Review, 112, 27-56. they all produced very similar results. Hill, P. C., & Pargament, K. I. (2003). Advances in the concep- tualization and measurement of religion and spirituality. Implications for physical and mental health research. The References American Psychologist, 58, 64-74. Bell, A., & Jones, K. (2013). The impossibility of separating age, Hill, P. C., Pargament, K. I., Hood, R. W., McCullough, M. E., period and cohort effects. Social Science & Medicine, 93, Swyers, J. P., Larson, D. B., & Zinnbauer, B. J. (2000). 163-165. Conceptualizing religion and spirituality: Points of common- Bell, A., & Jones, K. (2014). Another “futile quest?” A simula- ality, points of departure. Journal for the Theory of Social tion study of Yang and Land’s hierarchical age-period-cohort Behaviour, 30, 51-77. model. Demographic Research, 30, 333-360. Horn, J. (1965). A rationale and test for the number of factors in Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, factor analysis. Psychometrika, 30, 179-185. S. M. (1985). Habits of the heart: Individualism and commit- Hout, M., & Fischer, C. S. (2002). Why more Americans have ment in American life. Berkeley: University of California Press. no religious preference: Politics and generations. American Berger, P. L. (Ed.). (1999). The desecularization of the world: Sociological Review, 67, 165-187. Resurgent religion and world politics. Grand Rapids, MI: Wm. Howe, N., & Strauss, W. (2000). Millennials rising. New York, B. Eerdmans. NY: Random House. Berger, P. L. (2011). The sacred canopy: Elements of a sociological Idler, E. L., Musick, M. A., Ellison, C. G., George, L. K., Krause, theory of religion. New York, NY: Open Road Media. N., Ory, M. G., . . . Williams, D. R. (2003). Measuring mul- Berger, P. L., Davie, G., & Fokas, E. (2008). Religious America, tiple dimensions of religion and spirituality for health research. secular Europe? A theme and variation. Burlington, VT: Research on Aging, 25, 327-365. Ashgate. Kesebir, P., & Kesebir, S. (2012). The cultural salience of moral Blaine, B., & Crocker, J. (1995). Religiousness, race, and psycho- character and virtue declined in twentieth century America. logical well-being: Exploring social psychological mediators. Journal of Positive Psychology, 7, 471-480. Personality and Social Psychology Bulletin, 21, 1031-1041. Konrath, S. H., O’Brien, E. H., & Hsing, C. (2011). Changes in Campbell, W. K., Bonacci, A. M., Shelton, J., Exline, J. J., & dispositional empathy in American college students over time: Bushman, B. J. (2004). Psychological entitlement: Interpersonal A meta-analysis. Personality and Social Psychology Review, consequences and validation of a new self-report measure. 15, 180-198. Journal of Personality Assessment, 83, 29-45. Lim, C., MacGregor, C. A., & Putnam, R. D. (2010). Secular and Chaves, M. (2011). American religion: Contemporary trends. liminal: Discovering heterogeneity among religious nones. Princeton, NJ: Princeton University Press. Journal for the Scientific Study of Religion, 49, 596-618. Cornwall, M., Albrecht, S. L., Cunningham, P. H., & Pitcher, B. L. Markus, H. R., & Kitayama, S. (2010). Cultures and selves: A cycle (1986). The dimensions of religiosity: A conceptual model with of mutual constitution. Perspectives on Psychological Science, an empirical test. Review of Religious Research, 27, 226-244. 5, 420-430. Dougherty, K. D., Johnson, B. R., & Polson, E. C. (2007). Myers, D. G. (2000). The American paradox: Spiritual hunger in an Recovering the lost: Remeasuring U.S. religious affiliation. age of plenty. New Haven, CT: Yale University Press. Journal for the Scientific Study of Religion, 46, 483-499. Pargament, K. I. (1999). The psychology of religion and spiritual- Exline, J. J., Pargament, K. I., Grubbs, J. B., & Yali, A. M. (2014). ity? Yes and no. International Journal for the Psychology of The Religious and Spiritual Struggles Scale: Development and Religion, 9, 3-16. initial validation. Psychology of Religion and Spirituality, 6, Park, N. S., Lee, B. S., Sun, F., Klemmack, D. L., Roff, L. L., & 208-222. Koenig, H. G. (2013). Typologies of religiousness/spirituality: Exline, J. J., & Rose, E. D. (2013). Religious and spiritual strug- Implications for health and well-being. Journal of Religion & gles. In R. F. Paloutzian & C. L. Park (Eds.), Handbook of the Health, 52, 828-839. Twenge et al. 13 Pew Research Center. (2015). Religious landscape study. Retrieved 1960-2008. Journal of Cross-Cultural Psychology, 44, from http://www.pewforum.org/religious-landscape-study/ 406-415. Pichon, I., Boccato, G., & Saroglou, V. (2007). Nonconscious influ- Twenge, J. M., & Foster, J. D. (2010). Birth cohort increases in nar- ences of religion on prosociality: A priming study. European cissistic personality traits among American college students, Journal of Social Psychology, 37, 1032-1045. 1982-2009. Social Psychological & Personality Science, 1, Piff, P. K., Kraus, M. W., Cote, S., Cheng, B. H., & Keltner, D. 99-106. (2010). Having less, giving more: The influence of social Vandello, J. A., & Cohen, D. (1999). Patterns of individualism and class on prosocial behavior. Journal of Personality and Social collectivism across the United States. Journal of Personality Psychology, 99, 771-784. and Social Psychology, 77, 279-292. Plaut, V. C., Markus, H. R., & Lachman, M. E. (2002). Place mat- Wachholtz, A., & Sambamoorthi, U. (2011). National trends in prayer ters: Consensual features and regional variation in American use as a coping mechanism for health concerns: Changes from well-being and self. Journal of Personality and Social 2002 to 2007. Psychology of Religion and Spirituality, 3, 67-77. Psychology, 83, 160-184. Welzel, C. (2013). Freedom rising: Human empowerment and Presser, S., & Chaves, M. (2007). Is religious service attendance the quest for emancipation. Cambridge, UK: Cambridge declining? Journal for the Scientific Study of Religion, 46, University Press. 417-423. Yang, Y. (2008). Social inequalities in happiness in the United Reither, E. N., Masters, R. K., Yang, Y. C., Powers, D. A., Zheng, States, 1972 to 2004: An age-period-cohort analysis. American H., & Land, K. C. (2015). Should age-period-cohort studies Sociological Review, 73, 204-226. return to the methodologies of the 1970s? Social Science & Yang, Y., & Land, K. C. (2013). Age-period-cohort analysis: New Medicine, 128, 356-365. models, methods, and empirical applications. Boca Raton, FL: Revelle, W. (2015). psych: Procedures for personality and psycho- CRC Press. logical research (Version 1.5.1). Retrieved from https://cran.r- Ysseldyk, R., Matheson, K., & Anisman, H. (2010). Religiosity as project.org/web/packages/psych/citation.html identity: Toward an understanding of religion from a social Saucier, G., & Skrzypińska, K. (2006). Spiritual but not reli- identity perspective. Personality and Social Psychology gious? Evidence for two independent dispositions. Journal of Review, 14, 60-71. Personality, 74, 1257-1292. Zinbarg, R. E., Revelle, W., Yovel, I., & Li, W. (2005). Cronbach’s Schwadel, P. (2011). Age, period, and cohort effects on religious Alpha, Revelle’s Beta, McDonald’s Omega: Their relations activities and beliefs. Social Science Research, 40, 181-192. with each and two alternative conceptualizations of reliability. Shariff, A. F., & Norenzayan, A. (2007). God is watching you: Psychometrika, 70, 123-133. Priming God concepts increases prosocial behavior in an anon- Zinnbauer, B. J., Pargament, K. I., Cole, B., Rye, M. S., Butter, E. ymous economic game. Psychological Science, 18, 803-809. M., Belavich, T. G., . . . Kadar, J. L. (1997). Religion and spiri- Smith, C., Denton, M. L., Faris, R., & Regnerus, M. (2002). tuality: Unfuzzying the fuzzy. Journal for the Scientific Study Mapping American adolescent religious participation. Journal of Religion, 36, 549-564. for the Scientific Study of Religion, 41, 597-612. Smith, C., & Snell, P. (2009). Souls in transition: The religious Author Biographies & spiritual lives of emerging adults. New York, NY: Oxford University Press. Jean M. Twenge is a professor of Psychology at San Diego State Stewart, A. J., & Healy, J. M. (1989). Linking individual develop- University and the author of Generation Me: Why Today’s Young ment and social changes. American Psychologist, 44, 30-42. Americans Are More Confident, Assertive, Entitled – and More Strauss, W., & Howe, N. (1991). Generations: The history of Miserable Than Ever Before (Atria Books, 2014). Her research America’s future, 1584 to 2069. New York, NY: William examines generational differences in values, behaviors, traits, and Morrow. work attitudes. Taylor, P. (2014). The next America: Boomers, Millennials, and Ryne A. Sherman is an assistant professor of Psychology at Florida the looming generational showdown. New York, NY: Public Atlantic University. He received his PhD in personality and social Affairs. psychology from the University of California, Riverside in 2011. Taylor, R., Chatters, L. M., Jayakody, R., & Levin, J. S. (1996). He is broadly interested in personality and situational influences on Black and White differences in religious participation: A real-world human behavior. multisample comparison. Journal for the Scientific Study of Religion, 35, 403-410. Julie J. Exline is a professor of Psychological Sciences at Case Twenge, J. M. (2014). Generation me: Why today’s young Americans Western Reserve University in Cleveland, Ohio. Her primary are more confident, assertive, entitled—And more miserable research interests focus on the psychology of religion and than ever before (2nd ed.). New York, NY: Atria Books. spirituality. Twenge, J. M., Campbell, W. K., & Carter, N. T. (2014). Declines in trust in others and confidence in institutions among American Joshua B. Grubbs is a currently a doctoral candidate in clinical adults and late adolescents, 1972-2012. Psychological Science, psychology at Case Western Reserve University and a pre-doctoral 25, 1914-1923. intern in professional psychology at the Louis Stokes Cleveland VA Twenge, J. M., Campbell, W. K., & Gentile, B. (2013). Changes in Medical Center. His areas of expertise are the psychology of reli- pronoun use in American books and the rise of individualism, gion and spirituality, human sexuality, and narcissism.

Journal

SAGE Open – SAGE

Published: Mar 23, 2016

Keywords: religion; cultural change; prayer; religious affiliation; generations

There are no references for this article.