The Liberalization of Young Evangelicals: A Research Note

Authors


Correspondence should be addressed to Buster G. Smith at the Institute for Studies of Religion, Baylor University, One Bear Place #97236, Waco, TX 76798. E-mail: Buster_Smith@Baylor.edu

Abstract

Media outlets and observers of American religion suggest that young evangelicals are retreating from the ranks of the “Christian right” and are embracing more liberal positions on controversial social issues. We test this hypothesis using the Baylor Religion Survey. We examine two separate measures of evangelical identity as well as a wide variety of political identifications and attitudes. Our study indicates that young evangelicals (1) are significantly more likely than older evangelicals to think that more should be done to protect the environment; (2) hold views similar to older evangelicals regarding abortion, same-sex marriage, stem cell research, marijuana use, government welfare spending, spending on the nation's health, and the war in Iraq; and (3) remain significantly more conservative than nonevangelicals on these same social issues. We find no strong evidence to support the notion that young evangelicals are retreating from traditional positions or increasingly adopting more liberal positions on hot-button or controversial social issues.

Introduction

In recent years considerable media attention has emphasized the dissatisfaction of young evangelicals toward organized religion and the conservative political views so often associated with those on the religious right. Indeed, a number of observers, especially during the 2008 U.S. presidential race, indicated an increasing percentage of young evangelicals were becoming more liberal than their parents on a range of political issues (e.g., Dokoupil and Miller 2009; Edwards 2008; Wehner 2007; Zogby 2009a). For example, an ABC World News Tonight news story (2008) discussed this issue under the title “Are Young Evangelicals Skewing More Liberal?” Similarly, the PBS series Generation Next ran a story entitled “Are Young Evangelicals Leaning Left?” (Mendenhall 2006). Pieces like these in the popular news media tend to conclude young evangelicals, unlike older evangelicals, are becoming increasingly alienated from positions held by religious conservatives on various social issues. Since evangelicals may comprise as much as one-third of the U.S. population (Dougherty, Johnson, and Polson 2007; Kosmin and Keysar 2009), it is not hyperbole to suggest a shift among young evangelicals from a more conservative to a more liberal political position on various issues would be consequential for American politics.

So convincing has been the argument for the shift of young evangelicals to the left, that one pollster has even concluded that Republicans are currently in the process of losing an entire generation (Zogby 2009b). But are we really observing a sea change as some suggest? What tends to be missing from these claims is empirical evidence documenting a significant shift to the left among young evangelicals on either specific issues or political ideology.

This article examines the differences between young and old evangelicals in the United States on various political views. We build on previous speculation on this topic through two complementary social scientific methods. First, we use established methods of categorizing evangelicals. Previous work on this topic has either failed to mention how evangelical status is assigned or has almost always used a single measure. Instead, we test two established methods within the sociology of religion using both affiliation and religious tradition. This emphasis on examining several measures of latent concepts is extended to the dependent variables of interest as well. For example, we simultaneously examine several measures of political liberalism and conservatism. While these terms are often used to refer to monolithic groups within American political society, important nuances exist. This study looks more precisely at identification with particular political parties and political stances. In addition, we examine a variety of specific political issues. By using a broad array of relevant issues, we find it is possible to see not only whether young evangelicals are different, but in what particular ways. Our research also differs from previous journalistic approaches by going beyond descriptive statistics. We apply inferential statistics to each of the measures of how young and old evangelicals differ to determine if the differences that may be present are statistically meaningful. Many of the previous statistical analyses do not do this, which has led to much speculation on these matters.

Data and Methods

In this study we use the Baylor Religion Survey (BRS) from 2007. The 2007 BRS is the second wave of a national random sample of the United States that used a combination of initial phone interviews followed by mail surveys. For more information about the methodology of the BRS, see Bader, Froese, and Mencken (2007). For the purposes of this study, young was defined as any respondent between the ages of 18 and 29. Similar studies have used this age range when dealing with the issue of whether young evangelical Christians are more liberal (Cox 2007). Thus, to allow comparisons of results, we used the same age range.

We utilize two separate categorizations of evangelical Christians. First, we use the religious tradition typology (RELTRAD) suggested by Steensland et al. (2000). This relies primarily on the denomination with which a respondent affiliates. Second, we use a measure of whether the respondent self-identifies as evangelical, which is available in the BRS. This is based on the survey question, “How well do the following terms describe your religious identity?” Evangelical was one of the terms listed. The response categories “somewhat well” and “very well” were coded as evangelical, while “not at all” and “not very well” were coded as not evangelical. This distinction is important because a substantial portion of respondents who are affiliated with an evangelical denomination do not self-identify as evangelical (42 percent); in addition, some respondents who affiliate with other religious traditions do self-identify as evangelicals (17 percent). By using two separate measures, we will be better able to test what sort, if any, of American evangelical Christians shows differences in political views based on age.

We use Goodman and Kruskal's gamma as the measure of association (Goodman and Kruskal 1954). This is the preferred measure because each of the subsequent cross-tabulations includes an ordinal variable. The gamma score allows us both to determine statistical significance, as well as being interpretable as a proportional reduction in error (PRE) value.

Are Young Evangelicals More Liberal?

Religious leaders on the left, like Jim Wallis and Tony Campolo, have been very active regarding the role of religion in public life, and both have spent a great deal of time trying to educate evangelicals, particularly young evangelicals, on how to think more progressively on matters of social justice rather than merely adopting monolithic positions traditionally championed by religious conservatives (Campolo 2006, 2008; Wallis 2006). As previously stated, at least the perception is surfacing that young evangelicals have been trending away from conservative positions on social issues heretofore supported by the religious right. Until now this discussion has largely consisted of anecdotal accounts and media stories without the benefit of empirical scrutiny.

  • Hypothesis 1: Young evangelicals are less likely to associate with the Republican Party than older evangelicals.

To test this hypothesis, we will use the BRS, which allows for several measures of what it means to be an evangelical Christian. Table 1 presents distributions of both young and old evangelicals, as well as other young and old respondents, regarding their political party identification. While there do appear to be some differences between young and old evangelical Christians in regard to political party identification, the association is very weak. Evangelical identity has a gamma value of −.15 with a p-value of .052, while RELTRAD affiliation has a gamma value of −.09 and a p-value of .312.1 If we examine the standardized adjusted residuals (not shown), the only difference between old and young evangelicals with a value above 2.0 is that young evangelicals are more likely to be independent.

Table 1.  Cross-tabulations of young evangelicals and political identification by identity and affiliation (row percentages with number of cases)
Identity MeasureStrong Republican % (N)Moderate Republican % (N)Leaning Republican % (N)Independent % (N)Leaning Democrat % (N)Moderate Democrat % (N)Strong Democrat % (N)Gamma Scorep-Value
Young evangelical17 (12)24 (17)14 (10)24 (17)6 (4)13 (9)1 (1)−.09.312
Old evangelical22 (72)31 (100)10 (32)15 (49) 5 (15)10 (33) 7 (24)  
Young nonevangelical6 (8)9 (12)6 (8)26 (36)16 (22)20 (27)18 (25)−.15.013
Old nonevangelical 8 (60)17 (128) 7 (56) 24 (185) 9 (66) 21 (162) 13 (102)  
Affiliation MeasureStrong Republican % (N)Moderate Republican % (N)Leaning Republican % (N)Independent % (N)Leaning Democrat % (N)Moderate Democrat % (N)Strong Democrat % (N)Gamma Scorep-Value
  1. Note: Data from BRS 2007. Weighted by wgt.

Young evangelical17 (16)22 (21)9 (9)29 (28)2 (2)11 (11)10 (10)−.15.052
Old evangelical21 (86)26 (109)14 (57)19 (79) 5 (22) 8 (33) 7 (27)  
Young nonevangelical5 (8)10 (14) 7 (10)25 (36)17 (25)20 (30)16 (24)−.11.054
Old nonevangelical 8 (70)18 (156) 7 (58) 22 (191) 8 (73) 22 (190) 15 (134)  

While political party identification is important, people often are more interested in the political beliefs and attitudes of a respondent and use this party identification only as a proxy. Thus, it is important to examine how young evangelicals differ on their political stance.

  • Hypothesis 2: Young evangelical Christians are more likely to self-identify as liberal than older evangelicals.

On the BRS, respondents were asked, “How would you describe yourself politically?” with seven response categories ranging from “extremely conservative” to “extremely liberal.”Table 2 shows that no statistically significant difference exists between young and old evangelicals in regard to their identification with these labels. This is the case for both measures of evangelical, that is, by affiliation or by self-identification. All Christian evangelicals are much more likely to be conservative than other Americans and less likely to be liberal. Surprisingly, among nonevangelicals, age is an important predictor of political stance: younger respondents tend to be more liberal. Thus, not only are young evangelicals not more liberal than other evangelicals, but something about their religious affiliations also counteracts the typically more liberal views of younger respondents.

Table 2.  Cross-tabulations of young evangelicals and political attitudes by identity and affiliation (row percentages with number of cases)
Identity MeasureExtremely Conservative % (N)Conservative % (N)Leaning Conservative % (N)Moderate % (N)Leaning Liberal % (N)Liberal % (N)Extremely Liberal % (N)Gamma Scorep-Value
Young evangelical19 (13)33 (22)18 (12)19 (13)5 (3)5 (3)2 (1)−.06.586
Old evangelical15 (50)46 (151)13 (43)19 (62)3 (10)4 (12)1 (3)  
Young nonevangelical1 (2)13 (18)5 (7)19 (41)19 (26)26 (36)7 (10)−.27.001
Old nonevangelical3 (22)22 (170)10 (80)31 (239)11 (84)17 (130)5 (41)  
Affiliation MeasureExtremely Conservative % (N)Conservative % (N)Leaning Conservative % (N)Moderate % (N)Leaning Liberal % (N)Liberal % (N)Extremely Liberal % (N)Gamma Scorep-Value
  1. Note: Data from BRS 2007. Weighted by wgt.

Young evangelical16 (15)34 (33)8 (8)24 (23)6 (6)8 (8)3 (3)−.04.638
Old evangelical13 (52)38 (159)13 (55)24 (100)4 (18)5 (22)2 (9)  
Young nonevangelical5 (7)13 (19)9 (14)28 (43)18 (27)22 (34)5 (8)−.22.001
Old nonevangelical4 (31)25 (222)11 (94)31 (279)10 (91)15 (132)4 (39)  

Again, we can look more deeply into what it means to be conservative and liberal by shifting our focus. Using the BRS, we can examine attitudes toward a variety of politically sensitive issues, including abortion, same-sex marriage, and the environment. Examining these particular issues is important because young evangelicals may identify with the Republican Party and see themselves as conservative, but they might not necessarily share the same political views as people who typically fall into these categories. Table 3 shows the views of younger and older evangelical Christians on a variety of politically sensitive issues, including two abortion scenarios, the use of marijuana, embryonic stem cell research, same-sex marriages, civil unions, welfare, public health spending, and the environment. For eight of these variables, young respondents who are affiliated with an evangelical Christian denomination are not significantly different in their responses than older respondents from these denominations.2 Regarding abortion, both older and younger evangelical Christians are most likely to be in opposition, not only if the family cannot afford the child, but even if the pregnancy is the result of rape. A majority of both groups of evangelicals see the use of marijuana as always wrong, and the majority sees embryonic stem cell research as always wrong or almost always wrong. They also strongly oppose same-sex marriages and have almost as much disapproval of civil unions. Most evangelicals think more money should be spent on the nation's health, while less should be spent on welfare. The only divisive issue between these two age groups is over how much money should be spent protecting the environment. Younger evangelicals tend to see a need for more spending (57 percent vs. 43 percent). A similar difference exists between young and old nonevangelicals (77 percent vs. 68 percent).

Table 3.  Cross-tabulations of evangelical Christians by age on political attitudes (row percentages with number of cases)
How Do You Feel About the Morality of the Following …Always Wrong % (N)Almost Always Wrong % (N)Only Wrong Sometimes % (N)Not Wrong at All % (N)Gamma Valuep- Value
Abortion If the Pregnancy Is the Result of Rape?
 Young45 (44)16 (16)9 (9)30 (29) .15.09
 Old37 (155)13 (56)14 (59)36 (152)  
Abortion If the Family Cannot Afford the Child?
 Young69 (68)13 (13)9 (9)9 (9)−.01.96
 Old69 (289)13 (55)9 (37)9 (38)  
The Use of Marijuana?
 Young55 (54)17 (17)20 (20)8 (8) .06.53
 Old51 (214)22 (91)14 (57)14 (57)  
Embryonic Stem Cell Research?
 Young41 (40)15 (15)21 (21)22 (22) .09.26
 Old39 (160)12 (50)18 (75)31 (127)  
 Strongly Disagree % (N)Disagree % (N)Agree % (N)Strongly Agree % (N)Gamma Valuep- Value
  1. Note: Data from BRS. Weighted by wgt.

Homosexuals Should be Allowed to Marry?
 Young70 (64)15 (14)10 (9)5 (5) .13.22
 Old63 (233)20 (74)7 (26)10 (37)  
Homosexuals Should be Allowed Civil Unions?
 Young52 (50)9 (9)29 (28)9 (9) .03.74
 Old49 (181)15 (57)24 (88)12 (45)  
Government Spending on … Improving and Protecting the Environment?Too muchJust about rightToo little   
 Young14 (12)29 (25)57 (48) −.22.03
 Old18 (68)39 (145)43 (159)   
Improving and Protecting the Nation's Health?
 Young13 (11)24 (21)63 (54) −.01.93
 Old10 (39)29 (111)61 (235)   
Welfare?
 Young52 (44)24 (20)25 (21) −.18.09
 Old61 (229)23 (86)16 (62)   

Table 4 takes a more detailed look at environmental issues. For two of the three variables, young evangelicals are significantly more pessimistic about the future of the environment than older evangelical Christians. It is important to note that the same relationship exists for nonevangelicals. For example, regarding global climate change having disastrous effects, 17 percentage point shift separates young and old evangelicals on strongly agreeing, while a 13 point shift separates young and old nonevangelicals. Similarly, among evangelicals there is a 13 point increase from old to young in concern about exhausting natural resources, and a 14 point increase from old to young among nonevangelical respondents. Even with this similar trend based on age, there is a marked difference between the overall levels of concern about these issues among evangelicals. For respondents of all ages, there is less agreement on these issues among evangelicals than other Americans. For example, young evangelicals are less likely to strongly agree that climate change will have negative effects (40 percent) than older nonevangelicals (42 percent). Thus, while there may be important differences based on age within these groups, there are still much more substantial differences that exist across the groups in regard to environmental concerns.

Table 4.  Cross-tabulations of evangelical Christians by age on environmental attitudes (row percentages with number of cases)
Indicate Your Level of Agreement with the Following Statements …Strongly Disagree % (N)Disagree % (N)Agree % (N)Strongly Agree % (N)Gamma Valuep- Value
  1. Note: Data from BRS. Weighted by wgt.

Global Climate Change Will Have Disastrous Effects
 Young evangelical15 (13)12 (10)33 (28)40 (34)−.29.002
 Old evangelical22 (82)20 (73)36 (135)23 (84)  
 Young nonevangelical4 (6)16 (23)25 (36)55 (79)−.17.024
 Old nonevangelical8 (65)12 (100)38 (311)42 (342)  
We Will Exhaust the Earth's Supply of Coal, Oil, and Natural Gas
 Young evangelical15 (14)12 (11)42 (40)32 (30)−.20.033
 Old evangelical16 (63)18 (69)46 (178)19 (74)  
 Young nonevangelical2 (3)8 (11)40 (59)50 (74)−.31.001
 Old nonevangelical5 (39)18 (144)42 (342)36 (291)  
We Will Destroy Most of the Plant and Animal Life on Earth
 Young evangelical19 (17)23 (20)23 (20)36 (32)−.17.083
 Old evangelical18 (67)29 (111)35 (134)18 (67)  
 Young nonevangelical9 (12)23 (32)38 (54)31 (44)−.08.248
 Old nonevangelical9 (75)27 (222)37 (306)27 (225)  

One recent and particularly contentious political issue is the war in Iraq. Table 5 examines two separate questions about this conflict. First, respondents were asked whether “Going to war in Iraq was the right decision.” Second, they were asked whether “We should bring American troops home from Iraq immediately.” We found no difference based on age regarding views on the war in Iraq among evangelical Christians. However, both groups of evangelicals (those who affiliate and those who self-identify) hold quite diverse views on the issue with slight majorities thinking that going to war was the right decision and that troops should not be brought home immediately. We found a significant age effect among those who self-identify as an evangelical concerning bringing the troops home. Contrary to the idea that younger evangelicals have more liberal views, however, it is the older evangelicals who are more likely to either agree or strongly agree that the troops should be brought home (33 percent vs. 28 percent). Indeed, 48 percent of young self-identified evangelicals strongly disagree that the troops should be brought home immediately.

Table 5.  Cross-tabulations of evangelical Christians by age on attitudes toward the war in Iraq (row percentages with number of cases)
Please Indicate Your Level of Agreement with the Following StatementsStrongly DisagreeDisagreeAgreeStrongly AgreeGammap-
About World Events:% (N)% (N)% (N)% (N)ValueValue
  1. Note: Data from BRS. Weighted by wgt.

Going to War in Iraq Was the Right Decision
Self-Identified Evangelical
 Young17 (11)22 (14)31 (20)31 (20)−.01.95
 Old17 (51)20 (62)35 (108)28 (87)  
Evangelical Affiliation
 Young21 (20)19 (18)37 (35)23 (22) .14.11
 Old15 (55)21 (76)33 (121)32 (118)  
We Should Bring American Troops Home from Iraq Immediately
Self-Identified Evangelical
 Young48 (30)25 (16)18 (11)10 (6) .25.02
 Old30 (92)37 (116)17 (53)16 (51)  
Evangelical Affiliation
 Young33 (32)27 (26)28 (27)12 (11)−.04.70
 Old31 (115)39 (142)16 (58)15 (54)  

To test the robustness of these findings and assure they were not an anomaly of the BRS, we replicated the analysis, as far as possible, with the 2008 General Social Survey. In general, all substantive findings led to similar conclusions about young evangelicals and their political attitudes. In particular, using the RELTRAD typology we found no statistical difference between young and old evangelicals in regard to political party affiliation, attitudes toward abortion, welfare, and same-sex marriage. The statistically significant differences that do exist include the view of younger evangelicals that more money should be spent on the environment. The primary difference from BRS findings is that on the GSS there is no statistically significant difference among evangelicals concerning their political party identification. Tables of these results are available from the authors upon request.

Conclusion

Findings from this research suggest that shifting political views and affiliations of young Christian evangelicals in the United States may be exaggerated. While minor differences do exist, especially in political party affiliation and environmental attitudes, they are the exception. Young and old evangelicals have strikingly similar views on abortion, homosexuality, drug use, stem cell research, the war in Iraq, and their own political stance as it relates to conservatism and liberalism. Indeed, the most pronounced result in all these analyses is precisely that young evangelicals are as conservative as they are, given that on a wide variety of measures younger respondents tend to be much more liberal. Carroll (2002) argues that such findings should not be unexpected. She suggests that many young adults are increasingly embracing a more orthodox form of Christianity. These observations seem to support the conclusions reached by Stark and Bainbridge (1987), who theorized that religious groups with the clearest boundaries between themselves and the prevailing culture—“high tension groups”—would grow more dramatically and that “low tension groups” would dissolve into secularism and retreat from religion. In this line of thinking, the growth of religious nones may be an artifact of a shift of the nominally religious or left-leaning religious to the religious none category.

The one major exception to these findings—environmentalism—deserves to be studied much more intensely. This perceived divide over environmentalism as a liberal issue has been noted repeatedly by commentators who suggest that it demonstrates a shift toward liberalism by young evangelicals. While environmentalism may be a divisive issue among evangelicals as it is elsewhere, it does not appear to portend other liberalizing trends among this segment of the American population. Indeed, this divisiveness suggests that it may be more meaningful to understand environmentalism as a developing political/moral issue, in which separating lines have not yet formed. While it may be the case that Democrats and liberal organizations have led the way on certain environmental actions, this is not evidence in and of itself that opposition to environmentalism is a cornerstone of the Republican Party or among conservatives. Rather, this issue may well fall outside the exclusive territory of any political affiliation.

Led by Arnett (2000, 2004), a new wave of scholarship has begun to focus on “emerging adulthood”—a distinct category of young people generally falling within the same age range utilized in the current study. According to Arnett (2000), what people experience during the years from age 18 to 29 has changed significantly in Western or developed societies. Among these young people, a very clear trend has emerged to postpone marriage and parenting until at least their late 20s, while their late teens and mid-to-late 20s are very much self-focused in the areas of love and work. This period of time represents a new developmental stage between adolescence and young adulthood (Arnett, 2004). Research on emerging adulthood also holds considerable implications for those interested in the scientific study of religion and spirituality. Focusing on young evangelicals, for example, is but one of many potentially fruitful areas for social scientific studies of religion (see, e.g., Smith and Denton 2005; Smith and Snell 2009).

This study also points to the benefits of applying sociological methods to test findings within the popular press and from pundits. While previous news stories reporting that young evangelicals are becoming more liberal have been fairly unanimous in their findings and conclusions, they share several weaknesses. First, they lack established criteria for creating their typologies. The issue of defining evangelicals has been an area of much research within the sociology of religion and as such we are able to draw on two separate categorizations, one of self-identification and the other of religious affiliation. This more precise approach avoids the vague and undefined label of evangelical that clouds many of the media accounts about this issue. Second, a vast majority of statistics provided on this topic by journalists are entirely descriptive with no inferential analysis. By using significance tests on a variety of dependent variables we were able to show where there are meaningful differences, as well as where only superficial differences exist. This type of methodologically rigorous study can help validate or refute popular perceptions of the changes that are occurring within society.

Footnotes

  • 1

    These analyses were also run with the categories collapsed into the three broad political categories of Republican, Independent, and Democrat. In this case there was a statistically significant difference between young and old evangelicals based on affiliation with a gamma value of −.21 and a p-value of .031. Overall, young evangelicals are less likely than older evangelicals to be Republicans (47 percent vs. 61 percent), but more likely to be politically independent (29 percent vs. 19 percent). There is no statistically significant difference between the rates of Democrat identity (23 percent vs. 20 percent). When examining evangelical identity this difference does not exist. In this case the gamma value is −.08 and the p-value is .472.

  • 2

    We also ran the same analyses for evangelical identity with the same substantive findings. For none of these eight variables were older evangelical respondents statistically different than their younger counterparts.

Acknowledgments

Acknowledgments:  We would like to thank the three anonymous reviewers for their constructive feedback.

Ancillary