Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Sep 14.
Published in final edited form as: J Sci Study Relig. 2009 Dec;48(4):794–804. doi: 10.1111/j.1468-5906.2009.01481.x

Serving God and Country? Religious Involvement and Military Service among Young Adult Men

Amy M Burdette 1, Glen H Elder 2, Janel Benson 3, Victor Wang 4, Terrance Hill 5
PMCID: PMC3442261  NIHMSID: NIHMS376398  PMID: 22984296

Abstract

Despite important connections between religion and military action throughout world history, scholars have seldom explored the association between religiosity and military enlistment. Using data from the National Longitudinal Study of Adolescent Health (Add Health), we used a person-oriented analysis to categorize young men according to patterns of adolescent religious involvement. Youth indentified as “highly religious evangelical” are more likely to enlist in the military compared to their “highly religious non-evangelical” and “non-religious” counterparts; however, these findings hold only for those young men without college experience. These findings are discussed along with study limitations and promising directions for future research.


Despite important connections between religion and military action throughout world history (e.g., the Christian crusades from 1095–1291), scholars have seldom explored the association between individual religiosity and military enlistment. Scholars have tended to focus heavily on economic incentives for military enlistment (Elder 1986; Johnson and Kaplan 1991; MacLean and Elder 2007; Sampson and Laub 1996), while other reasons for joining the armed forces (including religious motivations) remain understudied. The question of whether religion influences military enlistment is particularly important in an era of the all-volunteer force. Indeed, the global “war on terror” is the first large-scale military engagement in which the United States has not recruited soldiers using a draft, instead mobilizing units of the National Guard and reserves (MacLean and Elder 2007).

Why might early religious experiences influence military enlistment? Religious conservatives tend to view America as a “covenant society,” with a unique calling to fulfill God's purpose on earth. Part of this obligation employs the advantaged economic position of the United States to expose all nations to Christianity, as well as to protect America's religious heritage (Wuthnow 1988: 247–48). By contrast, the version of civil religion espoused by mainline Christians and Catholics focuses on America's international role as a provider of resources to fight hunger and other social problems, rather than acting as a defender of Christian theology (Wuthnow 1988: 250). In addition, those raised within mainline religious traditions may be exposed to messages actively discouraging military enlistment, as members of these churches have taken a prominent role in peace-keeping organizations such as Witness for People (Kurtz and Fulton 2002: 367–369). With these distinctions in mind, the conservative view of civil religion, which is reinforced by evangelical worldviews, represents a likely source of encouragement for young adults to join the military rather than to pursue other career choices. Indeed, limited evidence indicates that military personnel are disproportionately members of conservative religious affiliations, such as the Church of God and Seventh-Day Adventists1 (Segal and Segal 2004).

Religion-specific explanations aside, there is reason to believe that religious conservatism might also encourage military enlistment by contributing to conservative social attitudes and political orientations. It is widely documented that religious conservatives tend to be politically and socially conservative (e.g. Hayes 1995; Hill, Mouton, and Burdette 2004; Layman 1997; Roof and McKinney 1987; Shortell 1997). Studies also link political and social conservatism with military enlistment (Bachman et al. 2000; Dunivin 1994; Franke 2001; Ricks 1997).

Using data from Waves I and III of the National Longitudinal Study of Adolescent Health (Add Health), we investigate the influence of religiosity on military enlistment among young men. The data come from a nationally representative, school-based sample of young people in grades 7–12 (n= 6,328). Despite the complex, multidimensional nature of religion (e.g., Levin, Taylor, and Chatters 1995; Stark and Glock 1968), studies have neglected to use a holistic approach in studying the impact of religiosity on social outcomes. Instead of focusing on the impact of individual variables, we investigate how multiple aspects of religion come together to influence military enlistment. By employing a person-centered approach, we explore individual religious profiles across patterns of religious factors (Crosnoe and Elder 2002; Magnusson and Cairns 1996). We then use these religious profiles to predict military enlistment.

METHODS

Data come from the National Longitudinal Study of Adolescent Health (Add Health). Add Health is a nationally representative, school-based sample of 20,745 adolescents in grades 7–12 surveyed during the 1994–1995 academic year. The sampling frame consisted of all high schools in the United States. The adolescents were interviewed three times during a 7-year period in 1994–1995, 1995–1996, and 2001–2002. The overall sample is representative of schools with respect to region of the country, urbanicity, and school type and size. Members of ethnic minorities were over-sampled. Further details regarding the sampling design and weights are available at: http://www.cpc.unc.edu/projects/adhealth/. Military enlistment, the dependent variable, is measured at Wave III of the survey.

All independent variables, including indicators of religious involvement, are measured at Wave I. Data from this time period was used in order to capture the influence of religious experiences during the most significant years of the life course; when adolescents are making decisions that will impact their initial educational and careers transitions. Measuring religious involvement at later waves may introduce uncertainty concerning casual order. In addition, 187 respondents were not interviewed at Wave III due to being active-duty military. Although religious involvement may have changed between waves of data collection, our current approach is preferable in that it allows us to retain our military sample.

We have chosen to limit our sample to males for two primary reasons. First, less than 3% of young women in Add Health had enlisted by Wave III (compared to a 10% enlistment rate for young men), leading us to be less confident in our findings for females than for males. Second, given the gendered nature of many religious institutions (Bartkowski 2001; Grasmick, Wilcox and Bird 1990), there is reason to believe that men may feel a particular obligation to enlist in military service.

Our sample of young men is drawn from 15,170 cases successfully interviewed during Wave III, as well as the active-duty military sample noted above. In addition to females (n=8,045), 119 Native Americans were excluded because of their small sample size and differing religious views and12 cases were dropped because they did not provide information on military enlistment. We also deleted a small number of respondents (n=191) who were disabled, as well as those respondents who had not graduated from high school (n=51) by Wave III. Finally, 424 cases were deleted due to missing information on the independent variables or lack of sample weights. Thus, the sample consists of 6,328 young men.

There are three major advantages of using Add Health in a study of religious involvement and military enlistment. First, the Add Health data cover a relatively long period following high school graduation (the respondents' ages range from 18 to 27 in Wave III). Previous studies have tended to focus only on enlistment behavior right after high school graduation. Add Health's wider age coverage enables us to generalize our finding to a broader population. Second, Add Health includes multiple measures of religious involvement. Instead of simply including a measure of religious service attendance or religious affiliation, the data provides information on religious identity, beliefs and behaviors. Finally, Add Health includes an array of measures not available in previous studies of enlistment. This enables us to examine the association between religion and military enlistment, net of a number of important sociodemographic influences.

Dependent Variable: Military Enlistment

Military enlistment is a binary indicator composed of three questions about the respondent's current and previous military experience. Respondents were asked whether they were currently serving in the armed forces, as well as if they had previously been full-time active duty military. Respondents were also asked if they had ever served in the military reserves. Respondents were coded as having military experience if they answered “yes” to any of these three questions. Of the military sample, 83% were currently or had previously been active-duty military, while 17% served in the military reserves.

Religious Profile Measures

We created religious profiles with a cluster-analytic approach to capture dimensions of religious identity. Our measures capture five key aspects of religious involvement, namely evangelical religious identity (i.e., born-again Christian), biblical views, religious salience, public religious participation and prayer. Scholars have long noted that having a personal conversion experience that involves the repenting of sin and accepting Jesus Christ as a personal savior (i.e., being “born again”) is the core element of an evangelical Christian identity (Smith 1998; Wilcox 1996). Youth were coded as being a born-again Christian if they indicated that they thought of themselves in this manner. Biblical views were measured with responses to the question, “Do you agree or disagree that the sacred scriptures of your religion are the word of God and are completely without any mistakes?” Those who indicated that they agreed with this statement were coded as holding conservative biblical views. Those respondents who did not agree, or indicated that their religion did not have a sacred text acted as the reference category. Religious salience is measured using the question, “How important is religion to you?” Response categories ranged from (1) “not important at all” to (4) “very important.” Public religious participation is measured by a summed index of two items (r=0.60). Respondents were first asked “In the past 12 months, how often did you attend religious services?” Then, they were asked, “Many churches, synagogues, and other places of worship have special activities for teenagers, such as youth groups, Bible classes, or choir. In the past 12 months, how often did you attend such youth activities?” Response categories for both items ranged from (1) “never” to (4) “once a week or more.” Finally, respondents were asked how often they prayed. Response categories for this item ranged from (1) “never” to (5) “at least once a day.” Although each of these measures captures a unique aspect of religious involvement, these items are significantly correlated with one another (see Appendix A).

Socio-demographic Controls

Previous research establishes a number of individual-level sociodemographic characteristics as correlates or predictors of young adult occupational roles, including military enlistment (Elder 1986; Johnson and Kaplan 1991; MacLean and Elder 2007; Sampson and Laub 1996). We can only be confident of our conclusions regarding possible religious variations in military enlistment if we include statistical adjustments for these potentially confounding factors. Therefore, all models include controls for the following variables: age (in years); race/ethnicity (0 = non-Hispanic white, 1 = all other categories, including African American, Asian and Hispanic); log household income, parents education (0= parents have less than a high school education, 1= all other categories including parents have a high school education and parents have at least a college education); family structure (0= biologically-intact family, 1= all other categories including single parent, biological parent and a step-parent, and other family structure); region of the country (1= South or Midwest 0= other region)2 and high school grade point average (GPA). In addition to the above predictors, we also divide the sample according to college enrollment in the final models. Individuals were assigned to the college enrollment category if they reported at least some college, whether it be a two-year or four-year college. Descriptive statistics for all of our measures are presented in Table 1.

TABLE 1.

Descriptive statistics for add health variables (n=6,328).

Mean/Proportion S.D.
Dependent Variable
 Military Enlistment .08 -
Profile Religious Items
 Born-Again Experience .26 -
 Conservative Biblical Views .67 -
 Religious Salience 2.97 1.08
 Public Religious Activity 4.73 2.18
 Prayer 3.43 1.58
Religious Profile Types
 Highly Religious Evangelical .25 -
 Highly Religious Non-Evangelical .26 -
 Moderately Religious Non-Evangelical .26 -
 Non-Religious .22 -
Sociodemographic Controls
 African American .20 -
 Asian .08 -
 Hispanic .17 -
 Non-Hispanic White .54 -
 Age 16.24 1.70
 Biologically-intact family .55 -
 Respondent lives in the South or Midwest .62 -
 Log Household Income 46.08 38.99
 Parents' have less than a high school education .12 -
 Parents' have a high school education .33 -
 Parents' have at least a college education .55 -
 Biological Parent& Step-parent .15 -
  Single Parent .23 -
 Other Family Structure .07 -
 GPA at Wave I 2.45 .75
 College Enrollment .57 -

Analysis

In our first step, we employed cluster analytic techniques to generate religious profiles based on similarity/dissimilarity across five dimensions of religiosity. In order to achieve uniform variability, each of the five cluster variables was first standardized. To create our profiles, kmeans cluster analyses in STATA were employed. K means analysis is a non-hierarchical partition clustering method that assigns observations to the appropriate clusters based on our five religion items.3 This is the most appropriate technique when there is a theoretical rational for a specific number of clusters (Hair et al. 2000).

Partitioning the data into a number of groups or clusters using the k-means cluster analysis is also an iterative process. First, the number of clusters is specified and initial group centers are generated. Then, STATA assigns observations to the nearest center based on Euclidean distances. The mean of the observations assigned to each of the groups is then computed. Reassignment of cases and re-calculation of the means continues until no observations change groups. Informed by literature and theory, we experimented with different numbers of groups (four to eight) in order to obtain the optimal number of profiles or clusters that represent people with different religiosity. The four-cluster solution was finally chosen as it provided a manageable number of profiles while maintaining important conceptual diversity. Sensitivity analysis was performed to test its robustness. We were able to replicate the four-cluster solution on random sub-samples with only slight variations in the means of our five religiosity items. Therefore, we are confident about the consistency of our four-cluster solution both statistically and theoretically.

The next step predicts military enlistment with the four religious profiles. We did so using binary logistic regression. The results of this analysis are displayed in Table 3. We first estimated a base model, including only our four religious profiles. Model 2 incorporates a number of individual and family predictors of military enlistment. In order to test whether religious variations in military enlistment are simply the result of limited educational opportunities, a measure of grade point average in high school is included in Model 3. The relationship between higher education and military enlistment is complex. Military enlistment may be an alternative to a college education, a means to a college education, a career choice following college graduation, or a path taken following an unsuccessful attempt at a college education. Therefore, Models 4 and 5 further investigate the relationship between religious involvement and postsecondary education by dividing the sample according to whether or not the respondent enrolled in college in conjunction with military enlistment.

TABLE 3.

Odds ratios for determinants of military enlistment, ADD Health.

Model 1 Model 2 Model 3 Model 4 College Model 5 Non-college
Religious Profiles
 Highly Religious Non-Evangelical .686* .757 .758 1.045 .628
 Moderately Religious Non- Evangelical .878 .898 .990 .821 .961
 Non-Religiousa .627* .624* .625* .936 .480**
Sociodemographics/ Controls
 African Americanb 1.100 1.103 1.188 1.031
 Asian .646 .646 .668 .604
 Hispanic 1.321 1.322 1.428 1.116
 Age 1.187* 1.187*** 1.382*** 1.050
 South or Midwest 1.337 1.338 1.360 1.304
 Log Household Income .887 0.887 .643 *** 1.365
 Parents' Educationc (high school) 2.433** 2.432** 1.566 2.994**
 Parents' Education (college & above) 2.544** 2.539** 1.738 3.594**
 Biological Parent& Step-parent 1.585** 1.588** 1.406 1.513
 Single Parent 1.018 1.020 .684 1.259
 Other Family Structure 1.334 1.337 .830 1.524
 GPA 1.008 .694** 1.731***

F 2.81* 4.86*** 4.50*** 5.29*** 3.15***
n 6328 6328 6328 3624 2704

p<.10

*

p < .05

**

p < .01

***

p<.001

a

Reference is Highly Religious Evangelical

b

Reference is non-Hispanic white

c

Reference is parents have less than a high school education

d

Reference is biologically-intact family structure

Results

Religious Profiles

Our cluster analysis is based on data from Wave I of the Add Health. The clusters represent four religious profiles, which are presented graphically in Figure 1 (with standardized values, mean=0). Below we describe these profiles more fully, on the basis of the five profile factors as well as the unadjusted means displayed in Table 2.

Figure 1.

Figure 1

Religious profiles

TABLE 2.

Unadjusted means for religious variations in military enlistment and sociodemographic characteristics (n=6,328).

Unadjusted Means Highly Religious Evangelical (n=1609) Highly Religious Non-Evangelical (n=1631) Moderately Religious Non-Evangelical (n=1671) Non-Religious (n=1417)
Military Enlistment .10bd .07a .09 .07a
African American .31bcd .22acd .12ab .13ab
Asian .05bcd .11acd .08ab .08ab
Hispanic .09bcd .19acd .23abd .15abc
Non-Hispanic White .53bd .46acd .56bd .60abc
Age 22.43cd 22.45cd 22.81ab 22.82ab
South or Midwest .76bcd .58a .55a .58a
Log Household Income 42.57 47.35a 47.86a 46.51a
Parents' have a high school education .33b .30acd .34b .36b
Parents' have at least a college education .58cd .57cd .53ab .51ab
Biological Parent & Step-parent .14d .12cd .16bd .19abc
Single Parent .22d .20d .22d .26abc
Other Family Structure .08 .07 .07 .07
Biologically-intact family .56bd .61acd .55bd .48abc
GPA 2.53cd 2.53cd 2.39ab 2.35ab
College Enrollment .61cd .64cd .54abd .49abc

Note: Age is reported at Wave III

a

Indicates significant differences from Highly Religious Evangelical (a= p < .05)

b

Indicates significant differences from Highly Religious Non-Evangelical (b= p < .05)

c

Indicates significant differences from Moderately Religious Non-Evangelical (c= p < .05)

d

Indicates significant differences from Non-Religious (d= p < .05)

Highly Religious Evangelical (n=1,609)

The young men who fit this profile exhibit high levels of all forms of religious activity, in combination with a conservative religious ideology. These young men report attending church and praying quite often, as well as high levels of religious salience. Further, these respondents believe the Bible to be the word of God and report having had a born-again experience. In terms of other characteristics, they are disproportionately African American, from the South or Mid-West, and from biologically-intact families with fairly well-educated parents. The young men who fit this profile also report relatively high levels of college enrollment.

Highly Religious Non-Evangelical (n=1,631)

The young men who fit this profile are somewhat similar to respondents in the first religious profile; however, they do not hold a conservative religious ideology. Although these individuals report relatively high levels of public religious participation, prayer, and religious salience, they do not report having had a “born-again” experience, indicating that they are not part of the evangelical community. With regard to other characteristics, these young men are the most likely among the various profiles to come from a biologically-intact family and report relatively high levels of college enrollment.

Moderately Religious Non-Evangelical (n=1,671)

The young men who fit this profile hold the Bible in high regard, pray occasionally, and report moderate levels of religious salience, but they also report relatively low-levels of public religious activity. Further, these respondents have not had a born-again experience, indicating that they are non-evangelical. Table 2 also indicates that these individuals are disproportionately Hispanic and are less likely to be from the South or Mid-West than other respondents.

Non-Religious (n=1, 417)4

In stark contrast to the first profile, these respondents report the lowest levels of religious involvement, as well as the lowest levels of military enlistment. With regard to other characteristics, these young men are disproportionally non-Hispanic whites and have parents with relatively low levels of education. Further, these respondents tend to come from single parent households and report relatively low levels of college enrollment.

Multivariate Results

Now that we have described the four religious profiles, we turn to the examination of how these profiles predict military enlistment. Table 3 displays the estimated net effects of the religious profiles and covariates on the odds of enlisting in the military. First, Model 1 reveals that those young men categorized as “highly religious evangelical” are significantly more likely to have enlisted in the military compared to their “non-religious” and “highly religious non-evangelical” counterparts. However, there is no significant difference between those categorized as “highly religious evangelical” and those labeled as “moderately religious non-evangelical.” The young men in the “non-religious” profile show an approximate 37% reduction in the odds of joining the military (OR= 0.627, p <.05) compared to those categorized as “highly religious evangelical.” Similarly, those classified as “highly religious non-evangelical” exhibit an approximate 31% decrease in the odds of military enlistment (OR= 0.686, p <.05).

Model 2 adds a number of sociodemographic controls for correlates of military enlistment, namely race/ethnicity, age, region of residence, adolescent household income, parental education, and family structure. Although the incorporation of these variables does little to change the association between those categorized as “non-religious” and military enlistment, the coefficient for “highly religious non-evangelical” is no longer significant in Model 2. This indicates that differences in military enlistment between those young men who fit the “highly religious non-evangelical” profile and those categorized as “highly religiously evangelical” are largely the result of variations in other social and family characteristics. Model 3 includes a control for high school grade point average. The inclusion of GPA is neither significant nor alters the associations between our religious profiles and the odds of military enlistment. In order to further investigate the relationship between religious involvement and postsecondary educational opportunities, we divide our sample according to college enrollment in Models 4 and 5. Model 4 shows no religious differentials in military enlistment among those young men with college experience. In contrast, Model 5 suggests notable religious variations in military enlistment for those respondents without college experience. For example, young men categorized as “highly religious non-evangelical” exhibit an approximate 37% reduction in the odds of military enlistment (OR= 0.628, p <.10) compared to their evangelical counterparts. Similarly, among those without college experience, young men fitting the “non-religious” profile exhibit roughly a 52% reduction in the odds of joining the military (OR= 0.480, p <.01) in contrast to those classified as “highly religious evangelical.”

Summary And Conclusion

Despite important historical connections between religion and military action, scholars have not fully explored the association between individual religiosity and military enlistment. Additionally, few studies have taken a holistic approach to studying the impact of religiosity on social outcomes. Employing this person-centered approach, we investigate how patterns of religious involvement form a religious profile, which in turn predicts military enlistment. Using data from Waves I and III of the National Longitudinal Study of Adolescent Health (Add Health), we find clear evidence of an association between religious identity and joining the military.

We find that those young men classified as “highly religious evangelical” are more likely to join the military compared to their “highly religious non-evangelical” counterparts; however, this finding only holds for those without college experience (p <.10). Conversely, those classified as “moderately religious non-evangelical” exhibit similar levels of military enlistment to their “highly religious evangelical” peers. Those who are “non- religious” consistently exhibit lower odds of enlistment in comparison to their “highly religious evangelical” counterparts (with the exception of the college sub-sample). Non-religious youth also exhibit the lowest levels of college enrollment. Although our presentation emphasizes military enlistment, adolescent religious participation may allow youth to take on various roles within the church as well as learn a variety of skills that increase success in secular environments. Indeed, evidence suggests that religious involvement may increase adolescent social capital, which in turn enhances educational achievement (Glanville, Sikkink and Hernandez 2008; Muller and Ellison 2001). Although it is beyond the scope of the current study, future research in this area should examine additional mechanisms by which adolescent religious involvement enhances occupational success among young adults.

Several important limitations of this research underscore the need for (a) caution in interpreting these findings and (b) further research into the links between religion and enlistment in the military. First, and perhaps most importantly, Add Health currently lacks information on the timing of the transitions into both the military and college. Although the high proportion with college experience among those fitting the “highly religious evangelical” profile suggest that this group is not educationally disadvantaged as asserted by previous research (e.g., Darnell and Sherkat 1997), we cannot answer more complex questions about the relationship between religion, college, and military enlistment. Future research in this area should include more precise measures of the timing of military enlistment. Related, our measure of higher education only captures college enrollment, rather than college graduation. Although research has yet to explore how religion may affect dropping out of college, there is reason to believe that religious profiles play an important role in college achievement.

In sum, despite the limitations of this study, we have found several notable and robust associations between religious identity and military enlistment. Given the intrinsic importance of this issue, as well as the changing patterns of military enlistment in the US population, the possible role of religious factors warrants careful investigation. Future work along the lines sketched above promises to clarify and extend our current knowledge about religion and military service in the lives of young people.

Supplementary Material

Supp App S1

Footnotes

1

The number of Seventh-Day Adventists currently serving in the military is particularly interesting given that historically this group has regarded military combat as a violation of both the Sixth Commandment, “Thou shalt not kill,” and of the nonviolent teachings of Jesus. However, over time this group has taken a more accommodating stance toward military service (Lawson 1996).

2

Residents of these regions are more likely to enlist in the military than residents of other regions of the country.

3

We chose a partition cluster approach, rather than a hierarchical technique, based on the large number of observations in our sample. Hierarchical methods are computationally intensive and cannot handle a sample of our size.

4

It is important to note that individuals that indicated that they had “no religion” were not administered subsequent religion questions. Consistent with other religion scholarship using Add Health data (e.g. Smith et al. 2002, 2003), we have coded these individuals in the lowest category on other religion questions.

References

  1. Bachman Jerald G., Segal David R., Freedman-Doan Peter, O'Malley Patrick M. Who chooses military service? Correlates of propensity and enlistment in the U.S. armed forces. Military Psychology. 2000;12(1):1–30. [Google Scholar]
  2. Bartkowski John. P. Remaking the godly marriage: Gender negotiation in Evangelical families. Rutgers University Press; Piscataway, NJ: 2001. [Google Scholar]
  3. Crosnoe Robert, Elder Glen H., Jr. Successful adaptation in later years: A life course approach to aging. Social Psychology Quarterly. 2002;65(4):309–28. [Google Scholar]
  4. Darnell Alfred, Sherkat Darren E. The impact of Protestant fundamentalism on educational attainment. American Sociological Review. 1997;62(2):306–15. [Google Scholar]
  5. Dunivin Karen O. Military culture: Change and continuity. Armed Forces and Society. 1994;20(4):531–47. [Google Scholar]
  6. Elder Glen H., Jr. Military times and turning points in men's lives. Developmental Psychology. 1986;22(2):233–45. [Google Scholar]
  7. Franke Volker C. Generation x and the military: A comparison of attitudes and values between West Point cadets and college students. Journal of Political and Military Sociology. 2001;29(1):92–119. [Google Scholar]
  8. Glanville Jennifer L., Sikkink David, Hernandez Edwin I. Religious involvement and educational outcomes: The role of social capital and extracurricular participation. The Sociological Quarterly. 2008;49(1):105–37. [Google Scholar]
  9. Grasmick Harold G., Wilcox Linda P., Bird Sharon K. The effects of religious fundamentalism and religiosity on preference for traditional family norms. Sociological Inquiry. 1990;60(4):352–69. [Google Scholar]
  10. Hair Joseph F., Jr., Black William C., Grimm Laurence G., Yarnold Paul R. Reading and understanding MORE multivariate statistics. American Psychological Association; Washington, DC: 2000. Cluster analysis; pp. 147–205. [Google Scholar]
  11. Hayes Bernadette. The impact of religious identification on political attitudes: An international comparison. Sociology of Religion. 1995;56(2):177–95. [Google Scholar]
  12. Hill Terrence, Moulton Benjamin, Burdette Amy M. Conservative Protestantism and attitudes toward homosexuality: Does political orientation mediate this relationship? Sociological Focus. 2004;37(1):59–70. [Google Scholar]
  13. Johnson Robert J., Kaplan Howard B. Psychosocial predictors of enlistment in the all-voluntary armed forces: A life-event-history analysis. Youth and Society. 1991;22(3):291–317. [Google Scholar]
  14. Kurtz Lester, Fulton Kelly Goran. Love your enemies? Protestants and United States foreign policy. In: Wuthnow Robert, Evans John H., editors. The quiet hand of God: Faith-based activism and the public role of mainline Protestantism. University of California Press; Berkley, CA: 2002. pp. 364–380. [Google Scholar]
  15. Lawson Ronald. Onward Christian soldiers? Seventh-day adventists and the issue of military service. Review of Religious Research. 1996;37(3):193–218. [Google Scholar]
  16. Layman George. Religion and political behavior in the United States: The impact of beliefs, affiliations, and commitment from 1980 to 1994. Public Opinion Quarterly. 1997;61(2):288–316. [Google Scholar]
  17. Levin Jeffrey S., Taylor Robert J., Chatters Linda M. A multidimensional measure of religious involvement for African Americans. The Sociological Quarterly. 1995;36(1):157–73. [Google Scholar]
  18. MacLean Alair, Elder Glen H., Jr. Military service in the life course. Annual Review of Sociology. 2007;33(1):175–96. [Google Scholar]
  19. Magnusson David, Cairns Robert B. Developmental science: Toward a unified framework. In: Cairns Robert B., Elder Glen H., Jr., Costello EJ., editors. Developmental science. Cambridge University Press; New York: 1996. pp. 7–30. [Google Scholar]
  20. Muller Chandra, Ellison Christopher G. Religious involvement, social capital, and adolescents academic progress: Evidence from the national education longitudinal study of 1988. Sociological Focus. 2001;34(2):155–83. [Google Scholar]
  21. Ricks Thomas E. Making the Corps. Scribner; New York: 1997. [Google Scholar]
  22. Roof Wade, McKinney William. American mainline religion: Its changing shape and future. Rutgers University Press; New Brunswick, NJ: 1987. [Google Scholar]
  23. Sampson Robert J., Laub John H. Socioeconomic achievement in the life course of disadvantaged men: Military service as a turning point, circa 1940–1965. American Sociological Review. 1996;61(3):347–67. [Google Scholar]
  24. Segal David R., Segal Mady Wechsler. America's military population. Population Bulletin. 2004;59(1):1–42. [Google Scholar]
  25. Shortell Timothy. Religious affiliation, commitment and ideology among U.S. elites. In: Williams Rhys., editor. Cultural wars in American politics: Critical reviews of a popular myth. Aldine De Gruyter; Hawthorne, NY: 1997. pp. 101–17. [Google Scholar]
  26. Smith Christian. American Evangelicalism: Embattled and thriving. University of Chicago Press; Chicago, IL: 1998. [Google Scholar]
  27. Smith Christian, Denton Melinda Lundquist, Faris Robert, Regnerus Mark. Mapping adolescent religious participation. Journal for the Scientific Study of Religion. 2002;41(4):597–612. [Google Scholar]
  28. Smith Christian, Faris Robert, Denton Melinda Lundquist, Regnerus Mark. Mapping American adolescent subjective religiosity and attitudes of alienation toward religion: A research report. Sociology of Religion. 2003;64(1):111–33. [Google Scholar]
  29. Stark Rodney, Glock Charles. American piety. University of California Press; Berkeley, CA: 1968. [Google Scholar]
  30. Wilcox Clyde. Onward Christian soldiers? The religious right in American politics. Westview Press; Boulder, CO: 1996. [Google Scholar]
  31. Wuthnow Robert. The restructuring of American religion. Princeton University Press; Princeton, NJ: 1988. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp App S1

RESOURCES